: Timing Belt Tension - F9Q DCI Diesel Engine


tolsen
14th March 2009, 04:22 PM
Hi,

This is an idea that entered my head last night after having consumed 3 pints of London Pride ale.
According to the specification for the F9Q engine, the timing belt is to be tensioned such that it reads 63 Hz or 90 hz depending on model. Tension is measured using a special Renault tool which is not accessible to normal mortals. The tool uses a microphone which is placed within 10 - 20 mm from the back side of the belt. Belt frequency is measured while lightly tapping on the belt.
Has anyone attempted measuring belt frequency using a guitar or cello tuner?

C2 = 62.84 Hz and G2 = 98 Hz. Both notes are well within the tolerance specified for the engines.

Cheers,
TK :lager:

Azurael
14th March 2009, 04:59 PM
What... Just twang the belt, holding the tuner close enough that it hopefully picks up...? What an odd idea, it might just work if the belt resonates long/loud enough :d

The tuner app on my iPhone gives the current frequency in Hz, anyway, so you should be able to get it pretty much bang-on.

I don't have a dCi to play with though.

Pat.w
14th March 2009, 10:45 PM
Try a tuning fork ;)

and if that doesn't work, try dropping London Pride for a decent beer :)

Lagdti
15th March 2009, 12:49 AM
As I am in the 'trade' I can exclusively reveal that Jimi Hendrix has signed a deal with Snap-on to produce a range of guitars that produce a note in tune with manufacturers cambelt tension guides.

Mind you,Jimi plays the guitar upside down so what he calls number 1 string is infact the opposite,that would confuse a stupid person....

Needless to say,the guitars aren't cheap,but hey,just like the shampoo commercials,they're worth it.

Pat.w
15th March 2009, 12:58 AM
I take it that the Jimi Hendrix solution only applies to pre 1970 cars tho Lag ;)

planmeca
15th March 2009, 01:12 AM
As I am in the 'trade' I can exclusively reveal that Jimi Hendrix has signed a deal with Snap-on to produce a range of guitars that produce a note in tune with manufacturers cambelt tension guides.

Mind you,Jimi plays the guitar upside down so what he calls number 1 string is infact the opposite,that would confuse a stupid person....

Needless to say,the guitars aren't cheap,but hey,just like the shampoo commercials,they're worth it.


On this very subject. Whilst setting the tension on my belt earlier today I was unfortunate enough to break my G string does anyone know where I can get a new one from. PLease reply using a plain brown envelope. Many thanks:d

I appologise to anyone who knows me as the thought of me in a g string has probably created a horific picture in your mind Sleep well

Pat.w
15th March 2009, 01:21 AM
:rofl: :rofl:

I hear, that theres a couple of specialist shops on ebay :d

planmeca
15th March 2009, 01:27 AM
:rofl: :rofl:

I hear, that theres a couple of specialist shops on ebay :d

Yes there are. In fact I run one of them:d

tolsen
15th March 2009, 08:42 AM
Pretty sad response to a proposed solution to a common problem. Out of 7 replies only 1 is of any assistance to the OP. Thanks to Azurael.

Come on folks - there must be someone out there that possesses some enthusiasm and ingenuity.

Pat.w
15th March 2009, 11:39 AM
You can buy tuning meters for timing belts. Amonst others, a company called Breco make them. I've seen them used, but not for this application

Personally I believe that they are a bit of a gimmick and shouldn't be relied upon (especially when the belt isn't going to looked at for 5 years). The manufacture methods used in these belts is variable to say the least even through the same manufacturer, take samples from two or three manufacturers and I very much doubt that the frequency obtained would be the same if the belts were all tensioned exactly the same using conventional methods. I would have thought that very slight change in shore hardness of the rubber, or the thickness of the kevlar braiding would completely alter frequency of a belt and both of these have tolerences in the manufacture method. The braiding used isn't an exact science on pattern or placement, this could also have an effect.


Disclaimer: The above is my personal opinion based on my engineering knowledge rather than actual experience of using such devices

tolsen
15th March 2009, 02:54 PM
Thanks for your positive contribution Pat. W.
The problem with these specialist belt frequency meters is their high price. The Breco SM4 costs around GBP 600 plus VAT. Measuring the frequency of the belt is the best and most accurate method. The hardness and thickness of the rubber does not alter the frequency but dampens the oscillations pretty fast. A meter that reads reasonably quick is therefore required. I have already tried using my Fluke multimeter (has Hertz readout) using a microphone and amplifier. This did not work as the Fluke was too slow. I intend to try out a few guitar or cello tuners next and see if I can tune in to note C2 = 63 Hz.
Here is a link to a brochure for Breco Universial Belt Tension Meter SM4. (http://www.brecoflex.com/download/E207.PDF) It explains pretty well how the measurements are done.

Pat.w
15th March 2009, 05:16 PM
Thanks for your positive contribution Pat. W.
The problem with these specialist belt frequency meters is their high price. The Breco SM4 costs around GBP 600 plus VAT. Measuring the frequency of the belt is the best and most accurate method. The hardness and thickness of the rubber does not alter the frequency but dampens the oscillations pretty fast. A meter that reads reasonably quick is therefore required. I have already tried using my Fluke multimeter (has Hertz readout) using a microphone and amplifier. This did not work as the Fluke was too slow. I intend to try out a few guitar or cello tuners next and see if I can tune in to note C2 = 63 Hz.
Here is a link to a brochure for Breco Universial Belt Tension Meter SM4. (http://www.brecoflex.com/download/E207.PDF) It explains pretty well how the measurements are done.

I guess that we'll have to agree to disagree on those points Tolsen.

I'd much rather use a device that actually measures the tension rather than a musical note.

As for the construction, continue to think of your belt like the string on a musical instrument. If it was only the tension that changed the frequency then the strings would be identical, but set at different tensions

Whilst I'm all for 'enthusiasm and ingenuity' (you don’t get far in my business without it) I don't think that I'd want to trust something as critical and potentially expensive as timing belt tension to a guitar tuning aid

madnoel10
15th March 2009, 05:34 PM
Even if you could use musical instrument tuners to check the belt tension the problem would be to ensure the "pluck" was identical each time. Unlike musical strings timing belts aren't uniform in cross section and have lumpy bits on the inside. From a personal viewpoint I think the tension testers which measure a given amount of play under a known pressure are likely to be more consistent.:)
If you knew what the deflection should be it may be possible to rig up something using a spring balance or such. In other words if the deflection was supposed to be "X" mm under a load of "X" newtons, etc.

tolsen
15th March 2009, 07:13 PM
I guess that we'll have to agree to disagree on those points Tolsen.

I'd much rather use a device that actually measures the tension rather than a musical note.

As for the construction, continue to think of your belt like the string on a musical instrument. If it was only the tension that changed the frequency then the strings would be identical, but set at different tensions

Whilst I'm all for 'enthusiasm and ingenuity' (you dont get far in my business without it) I don't think that I'd want to trust something as critical and potentially expensive as timing belt tension to a guitar tuning aid
You are right Pat. W, mass of the string has also some effect and that is why the guitar strings are of different thickness.
See simplified equations for string vibration below taken from my school physics textbook:

f = v/λ
v = sqrt(T/)
f = sqrt(T/λλ)

Frequency (f)
Velocity of the wave (v)
Tension of the string (T)
Wavelength (λ)
Linear mass density of the string ()

I have however no control over the mass of the belt so need not worry about it. It shall be interesting to see if the guitar tuner will work. Of course I do intend to double check using a standard old fashioned belt tension gauge.
Using an instrument that displays frequency would be preferable, the problem is they are too expensive.

Think I did not make it sufficiently clear why I want to adjust tension of the timing belt by measuring its vibration frequency.
It is because the F9Q engine does not have a built in spring loaded tensioner. Renault specifies the vibration frequency for the belt, not the required tension or flex.

tolsen
15th March 2009, 08:22 PM
Even if you could use musical instrument tuners to check the belt tension the problem would be to ensure the "pluck" was identical each time. Unlike musical strings timing belts aren't uniform in cross section and have lumpy bits on the inside. From a personal viewpoint I think the tension testers which measure a given amount of play under a known pressure are likely to be more consistent.:)
If you knew what the deflection should be it may be possible to rig up something using a spring balance or such. In other words if the deflection was supposed to be "X" mm under a load of "X" newtons, etc.

Thanks for your contribution Madnoel. I don't think tension is that critical for the F9Q engine in my Kangoo where the specified belt frequency is 63 Hz. It is more critical I think, on the higher power engines like those fitted to the Laguna. The belt is exactly the same as mine but it is tensioned to a frequency of 90 Hz. If you use the simple string theory formulas I posted you will find that the Laguna timing belt is pre-tensioned 100% higher than the belt fitted on my Kangoo. This assumes the length of free vibration belt span is identical.
Cheers,
TK

PS. I have used a cheap Facom gauge like this one for years. I doubt if it will be OK for the Laguna:
http://i385.photobucket.com/albums/oo291/tkaald/FacomGauge.jpg

Pat.w
15th March 2009, 11:16 PM
I have however no control over the mass of the belt so need not worry about it. It shall be interesting to see if the guitar tuner will work. Of course I do intend to double check using a standard old fashioned belt tension gauge.
Using an instrument that displays frequency would be preferable, the problem is they are too expensive.

Think I did not make it sufficiently clear why I want to adjust tension of the timing belt by measuring its vibration frequency.
It is because the F9Q engine does not have a built in spring loaded tensioner. Renault specifies the vibration frequency for the belt, not the required tension or flex.


Surely you should be worried tho, if the mass of two belts is significantly different (which is very possible), then surely the adjustment could be out

tolsen
16th March 2009, 12:48 AM
Surely you should be worried tho, if the mass of two belts is significantly different (which is very possible), then surely the adjustment could be out
Should not be any problem Pat. w if you stick to an OEM timing belt or Gates.

Anyway, how would you check tension on a timing belt for a F9Q engine? Lagdti seems to know it all. Maybe he has the answer.

Waiting anxiously in anticipation.


Cheers,
TK :d

planmeca
16th March 2009, 12:55 AM
Why do you not seem to be able to construct a post that doesn't get personal?

Lagdti has helped many of us over the last few years with constructive advice and not sarcasm. I am sure there are many other forums that may appreciate your attitude but I doubt if many members on here do

This is a friendly forum and you seem to out to upset the smooth flow.:crazy:

Pat.w
16th March 2009, 01:19 AM
Should not be any problem Pat. w if you stick to an OEM timing belt or Gates.

Cheers,
TK :d

In my experience rubber moulding is far from an exact science, there are a lot of variables. Throw Kevlar into the mix, and consistency goes out the window. Outside of Aerospace, there are very few standards for composite materials (Kevlar) it's not inconceivable that the construction of belts from a single manufacturer will vary batch to batch, but two manufacturers :crazy:. Interesting project tho, but if you find another way to check the tension I would suggest you do it as a precautionary check.

I'll ignore the last bit

tolsen
17th March 2009, 08:07 PM
Hi there,

I have been busy experimenting with 3 guitar tuners attempting to measure vibration from the timing belt on my Citroen ZX 1.9 diesel engine (have it on my engine stand so easy access). The experiment was a total fiasco most likely due to the vibrations fading out to quickly for the tuner to pick up. I then read all the contributions made to this thread over a few times. Lagdti's input made me think of tuning my guitar to the required vibration frequency. Then armed with a medical stethoscope and a guitar I tried again. Using the guitar tuner I tuned a string to the required reference frequency. Bingo - it was dead easy. I found I could tune the belt in to C2 = 62.84 Hz and G2 = 98 Hz even without the stethoscope by holding my ear near to the belt. Madnoel raised an issue w.r.t. the pinging position. I was pinging in the middle of belt span to avoid harmonics. This is also what is recommended in Renaults workshop manual and by Breco for the SM4 instrument.

Another possible method more suited to those without guitars and guitar tuners is to use the string vibration formulas to calculate the slack side belt pre-tension that corresponds to the specified vibration frequency.
Preliminary calculations indicate the following pre-tensions:
90 Hz - 324 N.
63 Hz - 159 N.
My gut feeling is that above slack side pre-tensions seem to be on the low side. Could someone please check the figures?

The next step is to work out force required to deflect the belt on slack side. Access to the belt is much better on top of the engine between diesel pump sprocket and camshaft sprocket. I have estimated the force required to deflect the belt 6 mm in the middle of the above span, again with slack belt:
Frequency - Slack belt pretension - Force required to deflect slack belt 6 mm.
90 Hz - 324 N - 10 N.
63 Hz - 159 N - 5 N.

I did not allow for any belt elasticity when estimating the deflection forces. The calculation was purely based on static force equilibrium assuming belt tension remaining constant when belt was deflected. Belt span assumed was 20 cm.

More work is obviously required to work out the set distance deflection forces. Looking for volunteers.

I am pretty satisfied with the frequency tuning method and intend to write a separate post on how to do this on the F9Q engine provided there is sufficient interest.

Cheers,
TK

tolsen
18th March 2009, 10:51 AM
Need to make a correcion. The specified vibration frequency for the F9Q Dci engine in the Kangoo is 67 Hz, not 63 Hz are previously stated. Source: Peter Russek Pocket Mechanic for Kangoo.

En electric piano is most likely the best suited tool for producing the reference notes:

Target 90 Hz (Laguna, Megane etc)
F2 = 87.3 Hz
F#2 = 92.5 Hz
G2 = 98.0 Hz

Target 67 Hz (Kangoo)
C2 = 65.4 Hz
C#2 = 69.3 Hz
D2 = 73.4 Hz

Maybe someone in the process of changing their timing belt could test this method?
TK :d

tolsen
19th March 2009, 04:06 PM
Those without guitars or pianos can check tension on your timing belts using the proper tool. Sealey has the VS099 on promotion for about 400 pounds. This tool measures and display the vibration frequency of the timing belt.
http://i385.photobucket.com/albums/oo291/tkaald/vs099.jpg
Cheapest supplier is Justoffbase, (http://www.justoffbase.co.uk/VS099-SEALEY) 388 pounds incl p&p.
Still rather expensive so I intend to use my special previously described method using an electric piano and a medical stethoscope next time.

Is there any interest for a more detailed procedure explaining how to tune the belt? I am thinking of checking tension on my Kangoo F9Q Dci engine soon. It might be possible to record the tuning steps and add a link to the recording in the post.
Cheers,
TK

rob4d
20th April 2009, 11:43 AM
Thanks for the excellent info on timing belt frequencies, just what I was looking for! Thanks for pointing me in the right direction. I Hope the following info is of help to you.

As an ex BT electronic exchange engineer this look like an oscilloscope problem to me. There are many PC oscilloscope software programs that allow the PC to be used as an oscilloscope using the PC sound card as an interface. Last night I tried twanging an elastic band in front of my pc and managed to produce around 100 Hz. (Obviously the sound card limits the frequency response at very low frequencies, near zero Hz due to the capacitor input.) Calibration could be checked with musical instrument or mains hum from a noisy transformer. (This would be at 100Hz as the frequency heard is double as there is no permanent magnet involved. Loudspeakers . earpieces etc have a permanent magnet so as to stop frequency doubling. Pushing and pulling of the air column is achieved with the permanent magnet, not just pulling of the diaphragm which would result in freq. doubling if a perm magnet was not fitted)

Cutting out the above blurb all one appears to need is the free software (soundcard oscilloscope ver 1.30), for private use only,(http://zeitnitz.de/Christian/Scope/Scope_en.html)and a microphone.
Gives good read out of frequencies and has trigger, run button for data capture etc. Good manual in english.

Useful formula periodic frequency (time of one cycle) is the reciprocal of the frequency. eg. mains at 50 Hz, time of one cycle is 1/50 sec = 0.2 sec = 20 ms (millisecs).

Other option would be to buy a scope meter at circa 100 off ebay, and making a mic interface, but then anybody reading this already has a computer.

Will be trying this out on mothers car as the local Renault dealer (Southern Ireland) want €750 (yes thats right) to supply and fit the cam belt which too expensive for her.

I was an ex Ford main dealer apprentice over 42 years ago on 105e and Mk4s and Mk 1 & 2 Cortinas etc, before working for BT.

dejan72
21st April 2009, 09:08 AM
from time to time there comes a gizmo like that on ebay.
I bought mine one-too diapaz (used by renault and opel) for 100EUR, There's also gates stt, a couple of rebranded diapaz (continental)- they all measure frequency and also cheaper mechanical ones that measure belt tension.
So in these times of crysys and foreclosures, we might expect some more on ebay. It's not that hard to get one. It's a bit harder to get the Hz values to make use of it.

tolsen
21st April 2009, 12:28 PM
Many thanks for your positive and constructive input Rob4d and Dejan72.
I have downloaded the soundcard oscilloscope software and is about to test it out on a timing belt.
The frequencies are not that hard to get. Just see below:

Engine type Fitting tension (Hertz)
D7D/D7F 145 5
K4J / K4M AUTOMATIC TENSION WHEEL
K7M 702,703,790 144 5
K7J / K7M 704,710,
720,744, 745,746 162 5
K9K AUTOMATIC TENSION WHEEL
F4P / F4R / F5R AUTOMATIC TENSION WHEEL
F9Q 650,718,750,751,
752,754,760,762,772,
774,800,820,826 T1 = 95 3, T2 = 90 3
F9Q 260,660,757,758,
759,800,804,808,812 T1 = 90 15, T2 = 80 15
F9Q 710,717,722,731,
736,744,770,780,782 AUTOMATIC TENSION WHEEL
F8Q/F9Q T1 = 68 3, T2 = 61 5
G9T / G9U AUTOMATIC TENSION WHEEL
L7X 106 4
S8W / S9W AUTOMATIC TENSION WHEEL
P9X AUTOMATIC TENSION WHEEL

Above info originates from Renault Technical Note 3786A Edition Anglaise issued OCT 2003.
Cheers,
TK

rob4d
21st April 2009, 02:38 PM
Lots of useful info TK.

Would be very interested to know how you get on.

I am more of a landrover man myself. (Timing chains on the TD 5) but need the info for my mothers renault sceneic 2003 1.9 diesel dci early model. Can you tell me how this engine is referred to please (short code, like TD on landrover or 300tdi).

Re PC oscilloscope software, the quality of the soundcard is the limiting factor. If one is using a pc mounted mic like on laptops then some noise is present from hard drive an internal fan. Just need to remember not to look at the noise.

To check calibration use a know source of signal and compare it to the one you are observing.

A bent piece of plastic waterpipe is good for conducting soundwaves from the source to the destination. ie cam belt to microphone. (I have talked very successfully over a 20 mm underground waterpipe over 30 metres and very clear signal. My fatherinlaw RN retired, says when he was in the service voice tubes were still used in emergency).

Done a bit more research on the web and Maplin do a very nice handheld oscilloscope at around 80, which also looks ideal for the job and would be useful for checking abs sensors etc. Have ordered one for myself and will post with my test results.

PS I like the scenic very much but not the cam belts as my son has just had a failure on his fiesta and bent one valve and cost €350 to fix ourselves.

Cheers Rob

tolsen
23rd April 2009, 09:45 AM
Lots of useful info TK.

Would be very interested to know how you get on.

I am more of a landrover man myself. (Timing chains on the TD 5) but need the info for my mothers renault sceneic 2003 1.9 diesel dci early model. Can you tell me how this engine is referred to please (short code, like TD on landrover or 300tdi).

Re PC oscilloscope software, the quality of the soundcard is the limiting factor. If one is using a pc mounted mic like on laptops then some noise is present from hard drive an internal fan. Just need to remember not to look at the noise.

Cheers Rob

I have tried measuring frequency using the Soundcard Scope but was experiencing trouble probably due to exessive noise and poor quality external microphones.

Engine code for your 1.9 Dci is shown on a nameplate riverted to the block and possibly a label stuck onto the "rocker" cover. On the first line there will be engine type (F9Q) followed engine approval letter and by a 3 digit number signifying engine rating (758). F9Q758 (this is just an example) is sufficient to identify required tension settings.

Cheers,
TK

rob4d
24th April 2009, 11:07 PM
I have tried measuring frequency using the Soundcard Scope but was experiencing trouble probably due to exessive noise and poor quality external microphones.

Cheers,
TK

TK

Was just about to post that yes you can tune a guitar low enough on the low E string when my hand held scope HPS 10 arrived from Maplin.

The scope is the bees knees for measuring frequency and can measure all the frequencies needed to tune a guitar or a belt. The pc oscilloscope software sound generator is measurable on the hand held oscilloscope down although the level is only just audible on the pc at this frequency. Will post some more hopefully tomorrow.

Meanwhile would be interested in details of your tension method and make of gauge used.

Regards Rob

rob4d
25th April 2009, 10:44 AM
Tk

Updated post as promised, stayed up late last night to test oscilloscope and it works.............! just like I thought. Thanks to everyone for posting their info it was a great help in putting me on the right track. Will try on the car soon, but I will try and find an old belt before doing the job and stretch that on a frame and play around with that first.

The below works brilliantly on the guitar. My only concern is that the belt amplitude might be a bit low, I have never had to twang one yet.

Can you actually here the note ok on the Renault? Holding my mike near the hole on the guitar the mike picks up audio that I cant hear but my ears are shot anyway. If low output on car may need to use a cheap preamp.

Equipment

Hand held oscilloscope HPS10 from Maplin 70 plus vat.

600 ohm balanced mike with balun to convert ot 50k ohm unbalanced which I already had. (so a 50k ohm
unbalanced coil mic should work fine note higher output voltage from 50k mic doesnt need amplification when used as described.

Method, I just put the batteries in and connected the probe to the mic unbalanced output. Turned on the
osc. and adjusted the screen contrast and pressed the auto button. Held mic in front of the guitar and
twanged the lower string. Did several times to get a nice sign wave and pressed memory button to hold
display. Then pressed marker button to adjust markers.

Got periodic time of 9msec. freq = 1/periodic time. Gave me 110Hz on the E string which aint bad seeing as the grandchildren have been playing with the guitar since my wife adjusted it. Manage to repeat with consistent results.

There is no noise on the screen so long as the mic is kept away from stray magnetic fields (like loudspeakers, transformers or people talking, and walking nearby)

Below is a description of how we adjusted the guitar. So now I got myself a tool to do the job more versatile than what is offered to do the job at a quarter of the price. Any more info wanted email me or post.

Playing around with the Pc oscilloscope software again. The sig generator is only usefull on my pc above 130 Hz, below that one might only be reading noise or harmonics. Tried on a piano. Sound card limitation problem.

Still it is useful for tuning a guitar and yes it is possible to retune a guitar to the lower freqs required for Renault cambelt tuning. (my wife is a dab hand at tuning guitars). The pc oscilloscope is really good for tuning and its free!

Useful info. Tuning standards vary, but using the equal tempered method and taking middle A as 440 HZ.
A below = 220 Hz and the one below that is 110hz. eg double the octave and double the freq, lower the
octave half the freq.

13 notes in a chromatic scale. The difference between each note is 12th root of 2.

(if fequency = 1 and is doubled to 2 and there are 12 steps then 1r^12=2.
r^12=2 and therefore r=12th root of 2 where r is the ration between the notes)

Therefore if A=440Hz, A#=440 x 12th root of 2=446.16Hz and G#=440/12th root 2=415.304 etc.

The problem with the PC scope software is reading a low enough value but it is useful for tuning the strings

on a guitar in conjuntion with its signal generator.

Using the signal generator, tune D string to 164.81Hz, G string to 220Hz, B to 277.18Hz and top E string

to 369.99Hz. The lower E and A string can be tuned using the fourth fret method by ear.

Lower E comes out at 92.499Hz and A at 123.47 Hz.

For different lower freqs just build a spreadsheet in excel and shift the notes one or two notes ( or change
the reference frequency) to achieve exact target freq required. Eg. if one wants exactly 95Hz make the
lower E string 95Hz the next harmonic 190Hz, the next E 380Hz etc and recalculate the other notes. Play
a few cords on the guitar and check that it sounds ok although a lot lower tone.

Have checked guitar against my noisy 110v mains transformer (50 Hz freq doubled to 100Hz as per my
earlier post as no permanent magnet) and the guitar is a shade lower.

If one wants a 50Hz reference point, touch an audio input on a HiFi amp and listen to mains hum. (Or connect a loudspeaker to a suitable low voltage transformer)


I agree from the other posts that the composition of the belt will affect the ultimate tension of the belt if
relying on the frequency method. I would guess that as one is trying to reach the correct tension then it
probably would have been better if Renault had quoted belt tension in the first place. I also wonder how
accurate the commercial meters are and how often they need recalibration. For my money I would want to
calibrate against a known source such as mains frequency before each use.

On mothers car I hope to use hand held scope, check with guitar and plastic pipe as stethoscope, and tension method.

Have written a guitar string calculator in excel so as to be able to to achieve exact target frequency which I
will try to attach to my post if possible

Thanks TK for encouraging a load of ideas and getting everyone thinking.

Rob

rob4d
25th April 2009, 03:05 PM
I have tried measuring frequency using the Soundcard Scope but was experiencing trouble probably due to exessive noise and poor quality external microphones.

Cheers,
TK

Tk

Since reviewing hand held oscilloscope in my last post I just found an adapter to plug my coil mic into my laptop sound card. Guess what it works a treat. Mic deatails in my last post.

Will try to post pic

Rob

rob4d
25th April 2009, 03:30 PM
Hopefully the screen dump of the pc-oscilloscope will show ok illustrating the retuned guitar lower E string proving that both the software works and so does the guitar

My toshiba laptop works fine with the sound card at low frequencies so I could have managed with just my PC without buying a hand held scope.

Some sound cards may well not work. Alos mic needs to be a good quality moving coil. In my case I used a 600 ohm balanced mic and converted to 50K ohm unbalanced. Less chance of picking up mains hum.

Also all the mic connection need to be really clean and well made because of the low voltages we are trying to measure (millivolt range)

Sorry cant post the guitar tuning excel file cos its not an allowed format on this post. Any one want a copy drop me an email and I will email it back if you dont want to do the maths crunching yourself.

Regards Rob

tolsen
27th April 2009, 10:27 PM
Hi,

Thanks to the excellent input from Rob4d I now consider the problem of measuring vibration frequency in timing belts to be finally solved.

Below I shall attempt to describe the ultimate, cheapest and most reliable method for checking timing belt tension by measuring its vibration frequency.

Tools required:
1) A pc, preferably a laptop with a soundcard.
2) A microphone.
3) Soundcard Oscilloscope software downloaded free from this site: http://zeitnitz.de/Christian/Scope/Scope_en.html

You also need the specified vibration frequency setting for your engine. There is a list of these in one of my previous posts in this thread.

How the measurements are done.
1) Connect the microphone about 5 - 10 mm (required distance will vary depending on how good the microphone is) away from the belt at the specified position (you need to refer to Haynes or make a post in this forum), usually middle of longest belt span on slack side of belt.
2) Turn engine around at least two revolutions and stop at TDC.
3) Pretension the belt on slack side. Renault specifies a special tool, which is operated with a torque wrench. Torque applied is only 11Nm so these wrenches are expensive. There is no need for that special tool. Just apply a force by hand minimum 8kg to the middle of the belt where the microphone is located.
4) Start the oscilloscope software. See photos for approximate settings.
5) Twang the belt and read off vibration frequency.
6) Turn engine around another two revolutions and stop at TDC. You should confirm TDC by inserting the 8 mm diameter timing pin. Pretension belt on slack side and measure vibration frequency.

http://i385.photobucket.com/albums/oo291/tkaald/DSC01647.jpg
Above photo shows belt frequency measurements being carried out on a Citroen ZX 1.9D engine. This was the easiest way to test out the method as the ZX engine was on an engine stand in my shed.

http://i385.photobucket.com/albums/oo291/tkaald/DSC01648.jpg
I used a headset microphone and a second headset for listening to the sound.

Below photos show 3 screenshots of the measurements. Note that there is very little spread in the readings.
http://i385.photobucket.com/albums/oo291/tkaald/Scope1.jpg

http://i385.photobucket.com/albums/oo291/tkaald/Scope2.jpg

http://i385.photobucket.com/albums/oo291/tkaald/Scope3.jpg

Cheers,
TK :d

rob4d
28th April 2009, 11:28 AM
TK

Your new post looks absolutely brilliant. You put a lot of hard work (as the excellent pictures show) in on this as did I.

It does not look like there are any hamonics on the wave form of your screen shot and it is as I found with the guitar a very high degree of accuracy can be achieved.

I think that this is a quality tool far superior to the commercial ones available. (I very much doubt if it would be possible to tune a musical instrument with a commercial cam belt tool due to the harmonics). The hand held oscilloscope HPS 10 that I referred to in one of my previous posts is harder to use especially if harmonics are present. If TK can check for harmonics and none are present then the HPS 10 from Maplin would be a fine tool also.

For completeness perhaps you could post a picture of the frequency analysis screen.

This would prove conclusively that anyone doing this is not looking at a harmonic.

Below illustrates why it is necessary to check. (Guitar tuning is very akin to cam belts in theory)

It was not possible to tune the wifes guitar from the oscilloscope screen due to the harmonics which give the guitar its tone. The harmonics combine together and give misleading results.

For the guitar on the lower E string the frequency analysis screen displays at least two distinct harmonics and displays the lower one as the main frequency, even though it was the fundamental to which I was listening. (I think because the sub harmonic peaked higher this showed up as the main frequency which of course it was not).

As one tunes the guitar the A string showed about 4 distinct harmonics.

Method of tuning guitar is to roughly (reasonable accuracy but not too fussy) set the guitar using the signal generator.

Then using only the frequency analysis screen (use the log scale on the x axis) tune using the most convenient sub harmonic.

I chose to work from Lower E string up as I found this easier for me.

I used 164.8/2 as this showed as the main frequency. (When the sub harmonic of 82.4Hz was tuned then the E note became 164.8Hz. I only chose to tune the sub harmonic as a digital frequency read out was given. I could have just tuned to 164.8 if I had set the marker to this, but the digital readout would have been 82.4 as the spike peaked stronger altough obviously a lower power output). On some of the higher strings the read out alternated between two or three harmonics sometimes more (cos thats how guitars work, lots of harmonics) But it did not matter as I could set the marker on the frequency analysis screen to which frequency I chose to adjust. (all of the harmonics are shown simultaneously as spikes on the x scale). Adjusting the sub harmonic (to its calculated target) obviously alters the pitch of the note we are trying to achieve. (A harmonic is a multiple either above or below of the fundamental and can be odd or even multiples, the guitar appears to be even multiples but I have read that the clarinet is odd multiples)

On testing the guitar afterwards it sounds and feels far more alive and using the fret method you can see the unplucked note resonate (vibrate)

Of course the frequency analysis screen would be just as good (may be even better) for adjusting cam belts as one can set the marker on the screen for the target frequency and just adjust the belt until the marker lines up.

Moral of the story is use one ears as well when adjusting belts and both screens. (Oscilloscope and frequency analysis screen)

I dont think that there are any harmonics present on your screen shot but it would be just as well to check.

Excellent job you did with the post and the pictures, I think the problem is now well solved at little or nil expense for those with PCs.

Finally depending on the output of abs sensors someone might like to come up with an interface for the PC-Oscilloscope for testing these? (Hint audio amp and loudspeaker or interface wired to pc sound card with suitable protection)

Thanks again TK for all the emails between us and the practical work you did so well.

Regards Rob

Pat.w
28th April 2009, 12:38 PM
I have to say that I'm very impressed with the amount of work that you guys have put into this.

As I mentioned earlier in the thread I do have my doubts about this method of tensioning belts, but then I've only seen it used in industrial applications where checks are carried out more frequently and failure is less catastrophic. I'd still be interested to see if different belts give different tones, but that just the engineer in me being fussy :rolleyes:

Keep it up guys, nice work :)

tolsen
28th April 2009, 02:05 PM
We'll do Fourier analysis on the timing belt vibration waveform next to see what harmonics we have that might lead to errors and wrong readings. I've found a way of publishing the belt vibration sound via this forum so you can all download the sound "wav" file and test it on your own PCs. (Will be in my next post perhaps).

It is getting more and more common with modern engines that the manufacturers (not solely limited to Renault) specify belt tension settings by vibration frequencies.

As previously mentioned, the problem for the home mechanic is the extortionate price for the special instrument required for checking the vibration frequency.

The Robols method (first 3 letters of Rob4d and tolsen) is free and is suited for anyone that can lay his hands on a PC.

TK :d

rob4d
28th April 2009, 06:25 PM
[QUOTE=Pat.w;503209
I'd still be interested to see if different belts give different tones, but that just the engineer in me being fussy :rolleyes:

Keep it up guys, nice work :)[/QUOTE]

Pat.w

Thanks for that.

So would I. I have my doubts as do you but I guess the manufacturer has already thought of this.

Maybe Tk would like to try a spring balance and measure the force required to deflect the cam belt at its mid position a reasonable amount measured by a dial test indicator and suggest a standard for the rest of us to test to.

Forum members could replicate the test and post so as to compile a database of different belts and their results including the wav file if enough interest.

I suggest that all will tune to the same frequency ok, but the tension may vary a bit. But I also guess that adjusting to a frequency is probably more accurate a standard to adjust to than using dubiously calibrated tension gauges having thought about it more.

Its easy to calibrate ones pc with a known reference. (For example one could record a 440hz computer generated tone and then email (wav) it to a forum member to get him to check against his pc. Same with the cam belt wav.

It all hinges on the quality control of the belt manufacturer. Perhaps they should be brave and state the frequency to tune to for a given tension.

Out of interest I will test mothers cam belt before removal and see how far it has deviated from fitting, assuming it was fitted correctly.

Regards Rob4d

rob4d
29th April 2009, 10:56 AM
Hi all on forum

Just re-read all the posts on cam belt tension.

Surpising how now the problem is solved everything is quiet on the forum. I would be (and I am sure TK also) interested to hear what you think, what your test results are.

Does anyone want more info such as oscilloscope basics, what is displayed on the screen etc.

All the original posts, light banter and comments encourgaged finding a solution.

Thanks all for your original input.

Regards Rob4d

Pat.w
29th April 2009, 10:59 AM
Perhaps everyone is waiting for the results of a physical test Rob :)

tolsen
29th April 2009, 11:15 AM
Perhaps everyone is waiting for the results of a physical test Rob :)
The physical test is forthcoming within an hour or two.
Intend to check tension using 3 methods:
1) Vibration frequency measurements and Fourier analysis of wave form using the unique Rodols method.
2) Universial timing belt tension gauge.
3) Twist of belt method.

tolsen
29th April 2009, 03:30 PM
Hi there,

Have checked the waveform using Fourier analysis. Conclusion: There is very little harmonics in the signal as can be seen from below photo.
http://i385.photobucket.com/albums/oo291/tkaald/Fourier1.jpg

I have made some recordings of the belt. You can download these here: http://cid-12c7d40e53cb9a1d.skydrive.live.com/browse.aspx/.Public The frequencies of these sounds are approximately 72, 108 and 123 Herz. It is best using the Sound Recorder in Windows to play these. Select "waive out" as input to the Soundcard Oscilloscope - do this under the "Extra" tab in Soundcard Oscilloscope. The above sounds are rather poor recordings but the quality is OK for measuring vibration frequency using the Soundcard Oscilloscope software. There will be more harmonics in the recordings than in the live sounds I used due to the analogue to digital conversion.

Tests carried out:
1) Set timing belt to vibrate at 123 Herz.
2) Used my DM 16 gauge to check belt tension. Belt was too tight.
http://i385.photobucket.com/albums/oo291/tkaald/DSC01650.jpg
3) Tried twist of belt method. Could twist belt about 25 degrees, hence too tight.
4) Set belt to vibrate at 71 Herz, see below photo:
http://i385.photobucket.com/albums/oo291/tkaald/Scope7.jpg
5) Used my DM 16 gauge to check belt tension. Belt was OK.
http://i385.photobucket.com/albums/oo291/tkaald/DSC01656.jpg
6) Tried twist of belt method. Could twist belt about 45 degrees, hence considered OK.
http://i385.photobucket.com/albums/oo291/tkaald/DSC01655.jpg

Please note that frequency readout in the Soundcard Oscilloscope is continuous. The vibration frequency of the belt is displayed for a few seconds before it changes to display the backgound noise frequency.

You can also measure frequency on the display by setting time markers, see photo below 4) above. Displayed frequency is 70.6 Herz. Time markers are set at 0.3 msec and 78.5 msec. There are 5.5 waives between the markers. Time for the 5.5 waives is 78.2 msec. Measured frequency becomes 5.5/0.0782 = 70.4 Herz.

Maybe someone with time to spare can have a go at it.

Cheers,

TK

rob4d
29th April 2009, 09:14 PM
TK

Excellent test and pics. Hope someone else tries it out now.

Downloaded your 72Hz file and couldnt manage to directly input it directly into Pc-oscilloscope, so did it via my mic off of my laptop speakers which isnt a good idea and the results are shown in osc.jpg. Lots of rubbish shown on the screen between the markers set at 72.9 Hz (discussed below) (used media player and set the trig to single, as below so as to get a trace without fiddling with the run stop button). But any one at home trying this for themselves on a cam belt will get a clean signal.

However it was a good test because the freq analysis.jpg below shows a signal peaking at 75.5Hz

It emphasises what I have said previously, use both screens and check for harmonics and use ones ears!

TK I can pick out a tone of 75.5 and therefore if you are telling me that your original tone read on your machine was approx 72Hz without harmonics then I would agree with you that the belt was adjusted to the said frequency of approx 72Hz.

The frequency analysis read from 75.05, 116.95, 132.52, 170.17, 192.84, 229.01, 267.75Hz etc.

I dont think that these freq in red are harmonics as I cant see any pattern in the progression. Not odd or even harmonics anyway. I think that they are spurious signals (computer fan, background noise, noise and distortion due to the conversion of analogue to digital and analogue, loudspeakers to mic and analogue to digital again).
(Perhaps someone out there knows more and might contribute to the post)

The vertical dotted line on the freq analyses.jpg is the marker controlled by the mouse used to read of frequency at that point. I like to use the trigger on single shot so as to freeze the display on short duration pulses automatically.

Finally I have since reading the original post retuned my wifes guitar enough times to know that consistent, accurate results are easily obtainable at low frequencies using the Robols method.

Thanks TK for the test post and the further info on belt deflection.

Regards

Rob4d.

tolsen
29th April 2009, 11:22 PM
Thanks Rob,

My PC has produces lots of background noise, especially the fan.

To send the recording to the Soundcard Scope, all you need do is this:

1) Start Sound Recorder which comes with Windows XP. You find it under Programs, Accessories, Entertainment, Sound Recorder.
2) Open the recording in Sound Recorder.
3) In Soundcard Scope, go to Extras tab, click input under Windows Sound Parameters and select Wave Out. The oscilloscope will now get the sounds directly from Sound Recorder when you press the play button.

I have played a bit more with the Soundcard Oscilloscope and also partly read the manual.

You might have seen at the end of my last post, I tried to explain how to measure frequency graphically. I did it the hard way. Tried within 2 seconds to set the markers and hit print screen to save the info. Managed after lots of attempts but there is a much easier way: Just use the Run/ Stop button, see below:
http://i385.photobucket.com/albums/oo291/tkaald/Scope8-1.jpg
The screenshot shows the 108 Hz belt frequency sound frozen. I have then placed the markers. At bottom of screen you can see dT = 73.909 msec and f = 13.53 Hz.
Actual frequency becomes 13.53 x [number of wave cycles between markers] = 13.53 x 8 = 108.2 Hz.
This is the most accurate method, however the difference is not that great, only 0.35 Hz!

Cheers,

TK

Pat.w
30th April 2009, 01:18 AM
Nice work guys :)

I guess it does raise a question though.
If the 45 degree twist is ‘deemed acceptable’, and a 50 belt tension gauge does the job, we need a laptop, microphone, possible soundcard upgrade and I guess purchase the sound card oscilloscope because?

tolsen
30th April 2009, 08:03 AM
Nice work guys :)

I guess it does raise a question though.
If the 45 degree twist is ‘deemed acceptable’, and a 50 belt tension gauge does the job, we need a laptop, microphone, possible soundcard upgrade and I guess purchase the sound card oscilloscope because?
Thanks Pat W.
We should seek answers to those questions you have raised.

The main problem with both the tension gauge and the 45 degree twist method is their crudeness.

I have rather big hands and strong fingers. My 45 degree twist check results in a much tighter belt than that checked by a weaker person.

The tension gauge is too big to fit the belt at the point specified by Renault. Its readout is very inaccurate as visual indication of correct setting is a ring on a spring loaded cylinder or piston that only moves 5 mm.
The manual for the gauge does not explain whether to measure on the tight or slack side of the belt.

The Robols method for checking vibration frequency of the belt is free - there is no need for any external hardware except a cheap PC microphone and headphones (Soundcard Oscilloscope software is free, see my post no 32 for source). A laptop is advisable but no requirement, just easier to move around than a desktop PC.

We are now at a stage where the method is more or less ready to be rolled out. I'll do some calibration checks of the accuracy of the frequency readouts and investigate the effect of belt pretension a bit further before writing the final operation guide.

Cheers,

TK :d

rob4d
30th April 2009, 06:32 PM
Sorry may have sent a blank post.

Agree with all the comments and TKs excellent post. I dont have a tension gauge and dont want to buy one if I have the gear at home, although I did buy the HPS 10 from Maplin (which is more used to me than a gauge)

Rarely is a quality tool for free, as most have access to a PC. I have tried two laptops one Toshiba and one Advent cheapy used for mobile use. Both work fine on the Robols method when tuning guitars and therefore would be fine for cambelts.

Have spent the day in private message to TK trying to answer what appears to be a simple question obtaining a 100Hz signal for calibration and other technical scope stuff. Dont want to clog this forum up with too much detail.

On my Toshiba a simple patch cord from the mic to loud speaker socket cuts out background noise for those who want to download his recordings and check his results. Got far cleaner results but they were the same.

Regards Rob

tolsen
1st May 2009, 12:51 AM
Have carried out two sets of tests.

First Test.
Calibrated against mains voltage frequency using a 12 Volt iron core transformer driving a speaker connected in series to a resistor. As you all may know the mains frequency should be 50 Hz. Had my Fluke 112 multimeter connected as well to check mains frequency. Result:
Fluke: 50.02 Hz
Soundcard Scope 50.04 Hz
Plans to calibrate against rectified mains voltage frequency (100 Hz) was curtailed as the speaker coil overheated and got knackered.

Second Set of Tests.
Used the Soundcard Scope built in sound generator. Checked frequency against my Fluke 112. The Fluke was hooked up to the PC trough the headphone output. Generated a pure sine wave. Did the checks starting at 50 Hz increasing frequency in 25 Hz increments to 150 Hz. Soundcard Oscilloscope and Soundcard Frequency Analyser reported the frequencies spot on without any deviation. The Fluke reported a minor deviation, see below:

Scope Fluke
50 49.97
75 74.95
100 99.94
125 124.9
150 149.9
Unit for above frequencies is Herz.

Conclusion: Accuracy of frequency readout is more than adequate for the intended application and much better than the commercial instruments that only boast an accuracy within +/- 5%.

tolsen
1st May 2009, 10:56 PM
I stated in post 32 that pretensioning beyond approx 8 kg had little effect on belt vibration. An experiment carried out this afternoon proved however beyond any doubt that my previous statement was wrong. Correct pretensioning is extremely important.

The experiment.
I did it on my Citroen ZX 1.9D. The engine is on an engine stand, hence very convenient for testing purposes.

Preloaded the belt using a bathroom scale with a wooden stick screwed on to its base, thereby making it easy to apply a lateral force to the belt halfway between camshaft sprocket and injector pump sprocket. This is the longest belt span on this particular engine. The point where the lateral force was applied is also where belt vibration frequencies were measured. The engine was rotated two turns and stopped at TDC position. The lateral pretensioning load was applied in increments of 2.5 kg. Started at 0 kg pretensioning. Measured belt vibration frequency at each step using the unique and free Robols method, see discussions from post 32 and on.

Test result:
http://i385.photobucket.com/albums/oo291/tkaald/Pretension.jpg

The frequencies measured appear to level out up to 7.5 kg lateral pretensioning but then drops rather fast on the next steps, most likely because the sprockets start rotating slightly.

Renault garages use a special tool (price in excess of 50 pounds) for applying the pretenstion. The tool is torqued to 11 Nm which results in a lateral force being applied to the belt at the largest span on slack side of belt. While Renault specifies a tight tolerance for the measured vibration frequency, there is no specified tolerance for the applied pretensioning torque. This does not make much sense and torque wrenches are not that accurate.

Conclusion.
It is essential that the timing belt is correctly and accurately preloaded in order to achieve the specified vibration frequency and right belt tension.

Further work.
Pretensioning of the belt can be applied using a spring balance. Torque = force x arm. The required spring balance force can therefore be easily calculated if someone could measure and post the length of the arm on the special Renault pretensioning tool (need tool identification as well).
http://i385.photobucket.com/albums/oo291/tkaald/4068.jpg

The position where the force is to be applied is given by the length of the arm of the special tool.

rob4d
2nd May 2009, 11:15 AM
I stated in post 32 that pretensioning beyond approx 8 kg had little effect on belt vibration. An experiment carried out this afternoon proved however beyond any doubt that my previous statement was wrong. Correct pretensioning is extremely important.
The position where the force is to be applied is given by the length of the arm of the special tool.

My next question would be what happens if you dont preload, other than a rough starting preload enough to produce a musical note but not excessive tension.

Would it not be true to say that a certain frequency is only produced at one specific tension and that it is unique to that tension, and that particular belt.

My thought is back to the guitar. If I measured the tension of the string and slackened it off and then readjusted it back to the exact same tension then should I not get the same note. Preload would not be important providing the preload was not excessive. If I then changed the string to an new string manufactured to the same spec, and adjusted it to my previously measured tension should I not get a similar note within just a few hertz. (I have tension gauges that measure as low as a gram, used for adjusting PO 3000 type relays, and high speed 3/412 relays, if any one really wants me to try it).
This would be a check that the string was manufactured to a certain tolerance, and would just need tweaking a few hertz. Tension is difficult to measure accurately, but frequency is not.

Maybe this is the principal that we have been looking for as to why it is done this way. Tension is hard to calibrate but frequency is not. Springs tend to loose there springyness but 50Hz mains is always very close to 50Hz and scopes can be checked against this standard. A kg is hard to check accurately but how do we know it accurate. Tension is a rough check without specialised equipement, frequency is almost exact.
Back to the cambelt.

1. If x preload tension = y frequency
surely y frequency = x preload tension?

Have you tried working back the other way adjusting to the frequency and then measuring belt tension with the scales to see if the same chart can be reproduced.

If my statement in (1.) is correct then why would anything other than a rough pretension as a starting point be needed?

Could the Renault pretension tool be just as a precuation to check that the belt is within specification?

Perhaps the preload is a test on the belts quality and if it is within spec, then the belt is just fine tuned to the exact tension by the frequency method. Does that make sense? That would overcome torque wrench accuracy problems.

It was suprising to see the the highest tension gave a lower frequency, I would have thought it would have been the other way around. I would have expected 0 kg to give 0 Hz and a higher tension a higher Hz. (was the preload really at 0 in the chart? a totally slack belt)

Nothing wrong with doing it as you suggest though (which is probably best) and the Robols method is still the best value for money.

Regards Rob4d

tolsen
2nd May 2009, 01:19 PM
Ref posts 47 & 48.

It appears that I did not explain sufficiently well the purpose for pretensioning the belt. When the engine runs, there is a slack side and a tight side in the belt transmission. The vibration measurements are carried out on slack side on the belt span between crankshaft sprocket and adjustable tensioner. This is also where the pretensioning is applied. Pretensioning results in the belt becoming slacker at this location and tighter between cam shaft sprocket - injector pump sprocket - crankshaft sprocket.

Too high a pretension leads to slacker belt at the measuring location which in turn must be compensated for by tightening the tensioner. The end result will be an overall tighter belt.

Renault Timing Belt Tensioning Procedure F9Q engine.
The F9Q engines require:
- A pretensioning torque to be applied (using the special tool for the engine type) to the belt section to be measured.
- A pre-stress T1 slightly greater than the final fitting tension T2 to be applied.
The purpose of these two operations is to stabilise the belt's internal tension, to make a reliable tension measurement possible.

Procedure.
Engine cold, ambient temperature.
Fit the new belt, with the timing at the setting point (Top Dead Centre).
Set the tension wheel tight against the belt by screwing a bolt into the lower timing cover.
Apply the pretensioning torque using tool (Mot.1543) with its various covers, using a torque wrench set to a torque of 1.1 daNm, to the belt section to be measured.
Measure the tension using tool (Mot. 1505), (Mot. 1715) or the unique Robols method, then adjust the tension using the screw on the tension wheel until the pre-stress T1 value is obtained.
Tighten the tension wheel.
Turn the crankshaft four revolutions and set the timing at Top Dead Centre.
Apply the pretensioning torque using tool (Mot. 1543) with its various covers, using a torque wrench set to a torque of 1.1 daNm, to the belt section to be measured.
Measure the tension using tool (Mot. 1505), (Mot. 1715) or the unique Robols method, then adjust the tension using the bolt on the tension wheel until the tension value T2 is obtained.

F9Q 650,718,750,751,752,754,760,762,772, 774,800,820,826: T1 = 95 3 Hz, T2 = 90 3 Hz
F9Q 260,660,757,758,759,800,804,808,812: T1 = 90 15 Hz, T2 = 80 15 Hz

http://i385.photobucket.com/albums/oo291/tkaald/F9Q_TimingBelt.jpg
(1) Pretensioning torque tool (Mot. 1543)
(M) Measuring point

Mot. 1543 looks like this:
http://i385.photobucket.com/albums/oo291/tkaald/4068.jpg

1.1 daNm = 11 Nm.

rob4d
4th May 2009, 09:59 AM
Re post #49

Thanks TK

Have carefully reread fitting instructions in manual. I had completely overlooked the fact of a slack side of the belt compared to the driven side and the reason for loading the belt.

Yes you are right in post #49 and your pretension/frequency chart makes full sense.

Hopefully some kind person out there will provide the tool details.

Regards Rob4d

BMboomer
5th September 2009, 01:43 AM
Hi
I am new to this forum, and this is my first post.
I run a 2006 Traffic DCI 100 with 1.9 diesel engine, for work in QLD Australia, (emigrated from England in 1989).
Last week I changed the timing belt, as the van has just past 100,000 kms.
I haven't got access to the Renault tools, so I used a spring balance as suggested in the Peter Russek manual.
I turned the engine clockwise until the camshaft mark appeared in the little timing cover window, put the 8mm rod in the engine lock hole, took the outside crankshaft pulley and cover off, measured the old belt, halfway between the crankshaft, and tensioner puylleys, (where the pretension tool would go).
I found that at 15lbs tension on the spring balance, the belt would deflect 2 tenths of an inch.
With that duly noted I marked the old belt where the timing marks were on the crankshaft, and camshaft pulleys. I couldn't find a mark on the fuel pump pulley, but I don't think that matters on a common rail diesel, but I marked the belt, pulley, and pump body anyway.
Next, I loosened the nut in the center of the tensioner pulley, moved the pulley away from the belt, and removed the belt.
I compared the old belt, with the new belt, (Dayco 94939) which came with 2 marks already on it. My marks on the old belt corresponded with the marks on the new belt.
I put the new belt on, made sure all the marks were lined up, pushed the tensioner onto the belt as tight as I could by hand, nipped the nut up, removed the 8mm rod, turned the engine clockwise a couple of turns, and a couple more for luck, back in with the 8mm rod, and checked the belt with my spring balance.
I hadn't read this thread at that stage, only the manual, so I didn't really appreciate what the pretensioning thing was all about.
My new belt was deflecting a bit more than the old one, so I loosened the tensioner nut, and levered the pulley towards the belt a bit more, nipped it back up, and rechecked it until it was the same. Finally I tightened it right up and put everything back together.
First day back to work after the weekend when I did the belt, I noticed that the engine noise seems to be a bit louder when I am driving. It seems to sound the same as before, just louder. I've done a week now, and about 750 kms, and it still hasn't settled down.
I had a bit of a Google around to see if anybody else has had that sort of thing, and that's when I found this excellent forum.
Do any of you think that my engine would seem a bit louder if I've got the belt tension a bit off?
I have downloaded the Scope program, and I might get under the Van again and give it a bit of a check with that.
Concerning the pretensioning. If you were using the official tools, am I right in thinking that you first bolt the lever to the cranksaft belt pulley, (so that the lever and pulley are as one) then take out the 8mm rod and apply the 1.1 KGM of torque. If this is correct, doing so would turn the crankshaft a little clockwise.
I imagine, that you would then remove the tool, before tuning the belt.
Would this not tend to turn the crank anticlockwise, and put tension onto the slack side of the belt, cancelling out the slackness that you have put there by the pretensioning?
Also Tolsen mentions "screwing a bolt into the lower timing cover to set the tension wheel tight against the belt", and "adjust the tension using the bolt on the tension wheel".
Am I missing something here? Is there another way of setting the tension wheel, other than loosening the center nut, and levering the wheel towards the belt and then tightening it back up, as I have done?
Hope some of you can shed some light on my engine noise, and clarify the pretensioning procedure a bit more for me.
Cheers

tolsen
5th September 2009, 06:50 AM
Hi BMBoomer,

That was a long post but I'll try answering your queries to the best of my abilities.

1) Peter Russek's spring balance method. I don't think this will work unless you first measure on a correctly tensioned new belt . Your old belt will be too slack.

2) There is no need on a common rail engine to time the pump. That is why it does not have any timing mark.

3) You compared the old belt, with the new belt, (Dayco 94939) which came with 2 marks already on it. Did you confirm number of teeth is same?
Have you confirmed engine timing marks still line up, i.e. when you turn engine to TDC, do camshaft timing marks line up?

4) Do any of you think that my engine would seem a bit louder if I've got the belt tension a bit off? I don't think there would be any more noise but could be wrong. A slack belt will eventually show up on the diagnostics. You'll get at fault saying something about lack of correlation between crankshaft and camshaft position sensors.

5) Noticed you have downloaded the Scope program. Suggest you also download my recorded belt sounds as well. You need to go back in this thread to find the link and instructions. This will allow you to play around with the software and measure frequencies before you try on your own engine.

6) Pretensioning belt using the official tools. Depending on which F9Q engine you have, the pretensioning tool is fitted on the crankshaft sprocket or on the idler wheel between crankshaft sprocket and tensioner. The tool does not turn the crankshaft sprocket or the idler wheel, it only pretensions the belt on the slack side where you measure belt vibration frequency. Applying a lateral force (and measuring with a spring balance) to the belt will have the same effect as using the pretensioning tool and an expensive torque wrench.

7) Am I missing something here? Is there another way of setting the tension wheel, other than loosening the center nut, and levering the wheel towards the belt and then tightening it back up, as I have done? The standard method on the F9Q is to use a long M6 set screw. This screws into the inner timing belt cover and touches the base bracket of the tensioner wheel. You may have to unbolt a clip for diesel pipes to find the treaded hole where you insert the M6 bolt. The procedure is explained in Peter Russek's manual and there is also one or two line drawings showing how it is done.

8) Hope some of you can shed some light on my engine noise, and clarify the pretensioning procedure a bit more for me. Could you describe the noise please? No metallic clatter I hope.
Correctly setting the pre tension on the belt is a rather complex and involved procedure if you are to do it correctly. The tools are expensive, especially the frequency meter and small torque wrench and then you also need the pretensioning tool.
The method I have described in this thread is free and pretty accurate, possibly more accurate for measuring belt vibration frequency than the expensive meters. You still have to pretension the belt. This can be done using a spring balance, see post no 49. My Irish friend was going to buy the pretensioning tool and supply dimensional details of it, the only essential measurement required is the length of the arm. It is then easy to work out required lateral force for the spring balance.
I wish you will go ahead and try out the Robols method, you are probably the first person to try it out as I had a rather poor response from other forum members.
Could you kindly advise which F9Q engine you have? Do you have the specification for required belt vibration frequency?

Cheers,
TK

tolsen
5th September 2009, 09:39 AM
The arrangement of your timing belt will be either as shown below or as shown in post no 49.
Work out required lateral force for belt pretensioning when using a spring balance instead of applying torque with tool Mot 1543 as follows.
1) Mark middle of free vibration span of belt at position M.
2) Measure distance from centre of crank shaft or centre of idler wheel to M. This length, hereafter named “L”, is for all practical purposes equivalent or very near to the length of the arm on pretensioning tool Mot 1543.
3) Renault specified pretensioning torque to be applied to belt using tool Mot 1543 is 1.1 daNm = 11 Nm = 1.12 kgm.
4) Required spring balance pretensioning force “F” equals Pretensioning torque devided by “L”, i.e.
“F” = 1.12/ L [Kg].
6) Example: Assume that L is measured to 17 cm. Converted to meters 17 cm is 0.17 m. Required spring balance force becomes F = 1.12/ 0.17 [kg] = 6.6 kg.

Belt Tensioning and Measuring Procedure:
A) You fit the belt ensuring timing is correct.
B) Tension the belt by turning the M6 set screw thereby pushing tensioner against belt. Tighten the tensioner bolt or nut located in its centre.
C) Turn engine 4 revolutions stopping at TDC. Easiest way to ensure you stop at TDC is by placing a timing mark for camshaft sprocket at the back cover. Renault cleverly only placed the mark on the outer cover which is of no use at this point because it has been removed.
D) Pretension belt using the spring balance at position M.
E) Measure timing belt vibration frequency. Adjust belt tension using the M6 screw on the tensioner wheel until pre-stress frequency T1 value is obtained.
F) Tighten the tensioner wheel.
G) Turn engine another 4 revolutions stopping at TDC.
H) Pretension belt using the spring balance at position M.
I) Measure timing belt vibration frequency. Adjust belt tension using the M6 screw on the tensioner wheel until the pre-stress frequency T2 value is obtained.
J) Tighten the tensioner wheel.

For the engine with timing belt layout as shown below, T1 = 68 3 and T2 = 61 5 Herz. For other layout see post 49.

Did this add to the confusion perhaps?

http://i385.photobucket.com/albums/oo291/tkaald/KangooTimingBelt.jpg

freddybabe
5th September 2009, 10:00 AM
The arrangement of your timing belt will be either as shown below or as shown in post no 49.
Work out required lateral force for belt pretensioning when using a spring balance instead of applying torque with tool Mot 1543 as follows.
1) Mark middle of free vibration span of belt at position M.
2) Measure distance from centre of crank shaft or centre of idler wheel to M. This length, hereafter named L, is for all practical purposes equivalent or very near to the length of the arm on pretensioning tool Mot 1543.
3) Renault specified pretensioning torque to be applied to belt using tool Mot 1543 is 1.1 daNm = 11 Nm = 1.12 kgm.
4) Required spring balance pretensioning force F equals Pretensioning torque devided by L, i.e.
F = 1.12/ L [Kg].
6) Example: Assume the L is measured to 17 cm. Converted to meters 17 cm is 0.17 m. Required spring balance force becomes F = 1.12/ 0.17 [kg] = 6.6 kg.

Belt Tensioning and Measuring Procedure:
A) You fit the belt ensuring timing is correct.
B) Tension the belt by turning the M6 set screw thereby pushing tensioner against belt. Tighten the tensioner bolt located in its centre.
C) Turn engine 4 revolutions stopping at TDC. Easiest way to ensure you stop at TDC is by placing a timing mark for camshaft sprocket at the back cover. Renault cleverly only placed the mark on the outer cover which is of no use at this point because it has been removed.
D) Pretension belt using the spring balance at position M.
E) Measure timing belt vibration frequency. Adjust belt tension using the M6 screw on the tensioner wheel until pre-stress frequency T1 value is obtained.
F) Tighten the tensioner wheel.
G) Turn engine another 4 revolutions stopping at TDC.
H) Pretension belt using the spring balance at position M.
I) Measure timing belt vibration frequency. Adjust belt tension using the M6 screw on the tensioner wheel until the pre-stress frequency T2 value is obtained.
J) Tighten the tensioner wheel.

For the engine with timing belt layout as shown below, T1 = 68 3 and T2 = 61 5 Herz. For other layout see post 49.

Did this add to the confusion perhaps?

http://i385.photobucket.com/albums/oo291/tkaald/KangooTimingBelt.jpg



Holy Moley Torsen, that is one hell of a post! Well done.

I wondered if you could answer me a question?

I have just serviced a 07 Ford Transit 2.2 TDCi Duratorque.

Changing the fuel filter is without doubt the easiest one I have ever come across to change.
However getting the thing primed is nigh on impossible unless you fill it with diesel.

Or have I missed summat?

Regards, Freddybabe.

tolsen
5th September 2009, 10:47 AM
Holy Moley Torsen, that is one hell of a post! Well done.

I wondered if you could answer me a question?

I have just serviced a 07 Ford Transit 2.2 TDCi Duratorque.

Changing the fuel filter is without doubt the easiest one I have ever come across to change.
However getting the thing primed is nigh on impossible unless you fill it with diesel.

Or have I missed summat?

Regards, Freddybabe.

A bit off topic but I'll answer anyway. Yes you fill the new filter with fresh diesel. More details here. (http://fordtransit.org/forum/viewtopic.php?f=54&t=29718)

Cheers,
TK

tolsen
5th September 2009, 05:17 PM
Just a timesaver for those that wish to give it a go:

Go to post 32 (http://www.renaultforums.co.uk/showpost.php?p=503117&postcount=32) for details on setup of the Soundcard Scope and link to external site for free download of the software.

Go to post 40 (http://www.renaultforums.co.uk/showpost.php?p=503681&postcount=40) to download pre-recorded sound files of timing belt vibration measurements. Instructions on how to play and direct these sounds to the Soundcard Scope are in same post.

Screenshot of a timing belt vibration reading is shown below:
http://i385.photobucket.com/albums/oo291/tkaald/Scope8-1.jpg

BMboomer
6th September 2009, 12:06 PM
Hi Tolsen
Thanks for your really helpful post.
I had another go at the van today , and thanks to all the good advice I've got here I think it's all sussed now.

I got the Scope program set up on the spare computer, with a cheap old microphone which I had. I got the old belt from last weekend set up on a bit of a frame, so that it was roughly the right tension & had a good play with the program.

I found some settingss under the frequency tab using the peak hold that gave me the right sort of sensitivity & consistency to work.

For the pretension, I converted the 1.1 KgM, mentioned in Peter Russeks book to inch pounds which when divided by the length of the Renault pretension tool lever (in inches) would give me the pressure in pounds which is what my spring balance is graduated in.

I jacked the van up abit better this time, removed the wheel & used stands to make it easier to see what I was doing.

As soon as I got the covers off, I could see that I had managed to set the belt one tooth out. I was sure I had got it right, but it was a bit of a struggle getting the new belt on and I didn't have much room to work, but that's no excuse.

No wonder it sounded a bit harsh. It ran fine though

I can see now how to use one of the timing cover bolts in the little threaded lug on the tensioner back plate to adjust the tensioner, thanks to your advice.

I checked last weeks efforts with the PC Scope & found that the belt tension was a bit low, just like you said, about 75 Hz. It probably would have been even less had I applied pretension.

When I compared the illustration of the pretension tool in place, in the manual with my van, I guessed it to be about 4" long in the lever, which equates to about 24 pounds using my inch pound calculations.

I found the methods that you have outlined here to work brilliantly in practise. The peaks that I was reading before & after pretension were very consistent with what you have said.

I can't find an engine identification number on my engine, so I decided to use the 85 Hz specified in the manual.

After lots of twanging, pretensioning & turning. I was getting peaks pretty much bang on 85 Hz every time.

I found that it didn't matter when pretensioning whether you pushed or pulled the belt.

Either way, it seemed to bring the slack to the tensioner side & gave similar readings.

I tried my rough & ready method of deflecting the belt with a spring balance & measuring it with a ruler on the properly tensioned belt & found that it was more like 1.5 tenths of an inch for 15 pounds of pull.

Also when twisting the belt, it seems about the same as our other vehicles that use spring loaded tensioners.

Another interesting thing that I found was that if I pretensioned the belt run between the camshaft & pump (which can be accessed by only removing the engine mounting). I got very similar readings, well within the +- 3Hz tolerance given in the manual.

This would be very useful if you just wanted to check the belt tension without removing everything.

Anyway, now my van is running beautifully & sounds just like it did before & I am much less confused by all the pretensioning & electronic tools that I have never come across before.
Once again, thank you all for a great forum, I'll know where to come next time.

Thanks a million

tolsen
6th September 2009, 02:21 PM
Many thanks for an excelent post BMBoomer.

I am really glad you got your engine sorted and that you managed to adjust your timing belt using the method descibed in this thread. I think you are the first person to have done it!

I totally agree that vibration frequency measurements are ideal for checking the timing belt in service.
Important you keep a record of the readings and that readings are done at same location and under similar conditions. The belt frequency measured between camshaft sprocket and high pressure pump sprocket is most likely going to be different to belt frequency measured at the Renault specified location but very easy to access in service. Belt tension will be the same everywhere in the belt before you preload it. Frequency measured at various spans depends on the free vibration span length.

The initial reading should be done once timing belt has been correctly tensioned, later it is very easy to check in service for piece of mind and to get an early warning thereby avoiding expensive repairs.

Belt vibration theory

f = v/λ

v = sqrt(T/)

f = sqrt(T/λ2)

T = f2λ2

S = λ/2

λ = 2S

T = 4f2S2

Frequency (f)
Velocity of the wave (v)
Tension of the belt (T)
Wavelength (λ)
Free vibration span (S)
Linear mass density of the belt ()

tolsen
7th September 2009, 12:29 AM
Many thanks for an excelent post BMBoomer.

I am really glad you got your engine sorted and that you managed to adjust your timing belt using the method descibed in this thread. I think you are the first person to have done it!

I totally agree that vibration frequency measurements are ideal for checking the timing belt in service.
Important you keep a record of the readings and that readings are done at same location and under similar conditions. The belt frequency measured between camshaft sprocket and high pressure pump sprocket is most likely going to be different to belt frequency measured at the Renault specified location but very easy to access in service. Belt tension will be the same everywhere in the belt before you preload it. Frequency measured at various spans depends on the free vibration span length.

The initial reading should be done once timing belt has been correctly tensioned, later it is very easy to check in service for piece of mind and to get an early warning thereby avoiding expensive repairs.

Belt vibration theory

f = v/λ

v = sqrt(T/)

f = sqrt(T/λ2)

T = f2λ2

S = λ/2

λ = 2S

T = 4f2S2

Frequency (f)
Velocity of the wave (v)
Tension of the belt (T)
Wavelength (λ)
Free vibration span (S)
Linear mass density of the belt ()

My apologies for the above error - most formulas are wrong. I pasted the belt string theory into my post but the exponents were shifted to the base line.

Here is a jpg image of the the belt theory which should make more sence:
http://i385.photobucket.com/albums/oo291/tkaald/Beltvibrationtheory.jpg

I am surprised none of you clever members did not spot it!

BMboomer
7th September 2009, 09:45 AM
Hi Tolsen
I am surprised that I am the first to use the method that you folks have perfected and made so easy, and cheap, for everyone, by spelling it out in this thread.

The way you have developed it from a London Pride fuelled idea, to an accessable and workable method is just a triumph of ingenuity.

The theory behind it looks complicated, but in practice, it is easy, accurate and repeatable.
I feel like such an idiot, bungling my first attempt, at the belt change, and thinking about it today, I realise that I had managed to get the 8mm rod in the wrong hole, despite the warning in the manual.

The good side of it is that if I'd made a reasonable job of it the first time, I wouldn't have gone looking for answers, and I wouldn,t have found this thread. I'd just be happily driving around with a slack belt.
Using my van today, it is so quiet and smooth, compared to last week when I had the timing a bit off. I think it seems even better than before I changed the belt.

Cheers

PS. those pints of English beer are one thing that i still miss

fazzy4u
22nd May 2010, 04:07 AM
Hi Tolson, I have Laguna 2003 1.9 DCI and I wanted to find out if you have a procedure with pictures to change the Timing belt and the water pump for it.

Would appreciate your imput.

Kind Regards

tolsen
25th May 2010, 08:01 PM
Hi Tolson, I have Laguna 2003 1.9 DCI and I wanted to find out if you have a procedure with pictures to change the Timing belt and the water pump for it.

Would appreciate your imput.

Kind RegardsI recommend a good workshop manual, Haynes or Peter Russek.

Alan Welch
28th July 2010, 12:31 PM
...
After lots of twanging, pretensioning & turning. I was getting peaks pretty much bang on 85 Hz every time.

I found that it didn't matter when pretensioning whether you pushed or pulled the belt.
...
Another interesting thing that I found was that if I pretensioned the belt run between the camshaft & pump (which can be accessed by only removing the engine mounting). I got very similar readings, well within the +- 3Hz tolerance given in the manual.
...



Hello,
Having read this and related posts, I am wondering just what the pretensioning is really desgned to do. Since it would seem that one ought to be able to tension a belt without pretension. Is it simply to be able to get an octave higher to enable the frequency readings to be measured. (ie 40hz is inaudable/unreadable , but 80hz is more easier to read)? Therefore, the pretension force is simply to hold the belt stationary at the half way point (as does a fret on a guitar stock), Therefore the force soesn't need to be too harse.

tolsen
28th July 2010, 01:12 PM
Hello,
Having read this and related posts, I am wondering just what the pretensioning is really desgned to do. Since it would seem that one ought to be able to tension a belt without pretension. Is it simply to be able to get an octave higher to enable the frequency readings to be measured. (ie 40hz is inaudable/unreadable , but 80hz is more easier to read)? Therefore, the pretension force is simply to hold the belt stationary at the half way point (as does a fret on a guitar stock), Therefore the force soesn't need to be too harse.
The purpose of pretensioning is to simoulate the operation of the belt. There is a slack side and a tight side. Measurements are done on the slack side. Pretensioning is done on slack side. Adjusting belt without pretensioning results in a slacker belt. Pretensioning ensures specified tension in belt on slack side when engine is running.

tolsen
24th October 2010, 11:29 AM
Details of timing belt pretensioning tool Mot. 1543 can be found in this thread. (http://www.renaultforums.co.uk/showthread.php?t=59623&page=2)

tolsen
11th November 2010, 08:17 AM
My engine and PC are now on a Russian forum:
http://www.renaultforums.co.uk/_User-folders/vBulletinImg/2012/1/{649FBB5D-2195-42B8-9981-D5A72569EA21}.jpeg
Link: http://www.lagunaclub.ru/forum/showthread.php?t=25184
:d

davelowe
20th November 2010, 10:44 PM
Renault Timing Belt Tensioning Procedure F9Q engine.
The F9Q engines require:
- A pretensioning torque to be applied (using the special tool for the engine type) to the belt section to be measured.
- A pre-stress T1 slightly greater than the final fitting tension T2 to be applied.
The purpose of these two operations is to stabilise the belt's internal tension, to make a reliable tension measurement possible.

Procedure.
Engine cold, ambient temperature.
Fit the new belt, with the timing at the setting point (Top Dead Centre).
Set the tension wheel tight against the belt by screwing a bolt into the lower timing cover.
Apply the pretensioning torque using tool (Mot.1543) with its various covers, using a torque wrench set to a torque of 1.1 daNm, to the belt section to be measured.
Measure the tension using tool (Mot. 1505), (Mot. 1715) or the unique Robols method, then adjust the tension using the screw on the tension wheel until the pre-stress T1 value is obtained.
Tighten the tension wheel.
Turn the crankshaft four revolutions and set the timing at Top Dead Centre.
Apply the pretensioning torque using tool (Mot. 1543) with its various covers, using a torque wrench set to a torque of 1.1 daNm, to the belt section to be measured.
Measure the tension using tool (Mot. 1505), (Mot. 1715) or the unique Robols method, then adjust the tension using the bolt on the tension wheel until the tension value T2 is obtained.

F9Q 650,718,750,751,752,754,760,762,772, 774,800,820,826: T1 = 95 3 Hz, T2 = 90 3 Hz
F9Q 260,660,757,758,759,800,804,808,812: T1 = 90 15 Hz, T2 = 80 15 Hz

http://i385.photobucket.com/albums/oo291/tkaald/F9Q_TimingBelt.jpg
(1) Pretensioning torque tool (Mot. 1543)
(M) Measuring point

Hi - my first post here (although not the first on a Renault forum - 1600 and counting elsewhere).

I've read the technical note 3786A, the Haynes, and also the MR364MEGANE1 official 'office' guide (p187 et seq.).

I don't have the tool MOT 1543 and I'm a little mystified as to its use. As I understand it, the 11Nm torque is applied via a first class lever of length 115mm (http://www.renaultforums.co.uk/showthread.php?t=59623&page=2). Therefore:

1). Given the equation F1xD1 = F2xD2 where the left hand side = 11Nm, I'm assuming that by re-arranging the equation to calculate F2 or D2 a given home made tool will do the job. Question 1 is then how to do this, or rather, how has anyone done this already? I have in mind a length of aluminium flat bar with a short dowel to bear on the belt. At the operators end I would push some scales against the flat. Assuming the bar was 115mm long, I would press until 11/9.81 kg registered.

2). This one is going to sound a bit daft: Pre-tensioning - whilst I understand the principle and have read the factory manual, is the pre-tensioning momentary, or held until the M6 screw takes up the slack?

Before I came across this, I had been independently developing a near identical frequency measuring system. Nice work on this thread!

Cheers,
Dave

tolsen
21st November 2010, 09:40 AM
Hi - my first post here (although not the first on a Renault forum - 1600 and counting elsewhere).

I've read the technical note 3786A, the Haynes, and also the MR364MEGANE1 official 'office' guide (p187 et seq.).

I don't have the tool MOT 1543 and I'm a little mystified as to its use. As I understand it, the 11Nm torque is applied via a first class lever of length 115mm (http://www.renaultforums.co.uk/showthread.php?t=59623&page=2). Therefore:

1). Given the equation F1xD1 = F2xD2 where the left hand side = 11Nm, I'm assuming that by re-arranging the equation to calculate F2 or D2 a given home made tool will do the job. Question 1 is then how to do this, or rather, how has anyone done this already? I have in mind a length of aluminium flat bar with a short dowel to bear on the belt. At the operators end I would push some scales against the flat. Assuming the bar was 115mm long, I would press until 11/9.81 kg registered.

2). This one is going to sound a bit daft: Pre-tensioning - whilst I understand the principle and have read the factory manual, is the pre-tensioning momentary, or held until the M6 screw takes up the slack?

Before I came across this, I had been independently developing a near identical frequency measuring system. Nice work on this thread!

Cheers,
Dave

Hi Dave,

I thought this thread had gone completely dead due to lack of interest.

1) Easiest way of pretensioning the belt is using a spring balance. F = T/ L = 1.12/ 0.115 [m] = 9.7 kgf.
The force should be applied laterally exactly at the same point where the pretensioning tool will touch the belt. Applying the force elsewhere may lead to unforeseen errors.
2) The M6 bolt is only used for adjusting tensioner. You need to tighten the tensioner fixation bolt before proceeding with pretensioning belt and measuring belt vibration. Pretension is applied momentarily.

Why pretension belt: Because Renault's procedure says so.
Reason: When engine runs, timing belt is tighter on the run between camshaft and pump and even tighter between pump and crank shaft pulley. Run between crank shaft pulley, tensioner and camshaft goes slacker. Pretensioning the timing belt on the slack side therefore simulates a running engine. Setting belt tension without pretensioning belt will result in a too low tension in the belt.

Hope this answers your queries Dave.
Cheers,
TK

davelowe
21st November 2010, 02:58 PM
TK,

Thanks for your reply.

I think I now understand. Loosely quoting from the factory manual and incorporating your technique:

1) Fit a new tensioning roller
2) Fit the new belt paying attention to the timing marks(!)
3) Press the tensioner roller against the belt by means of the M6 temporary set screw to a nominal tension
4) The factory manual does NOT say to tighten the tensioner fixation bolt at this point
5) Apply a spring balance at a point 115mm on an arc from the centreline of the crankshaft. Pull momentarily and laterally so as to close the gap between the tensioner and the water pump pulley using a force of 9.7 kg. Remove the balance.
6) Measure the frequency at the midpoint between crankshaft sprocket and tensioner (90Hz +/- 3 target). Adjust frequency by turning M6 set screw
7) Tighten the tensioner nut to 10Nm (from the manual)
8) Turn the crankshaft through 2 revolutions
9) Find TDC
10) Pretension again as per 5)
11) Check the tension value is 80Hz (+/-5) Re-adjust as required.
12) Tighten the tensioner to 45 Nm (from the manual)
13) Remove the M6 set screw

I'm going to be doing this job on my car soon (together with the accessories belt and water pump). It obviously needs doing correctly. If I've missed anything (or misunderstood) please let me know.

I shall take many photos and post them here.

Cheers,
Dave

tolsen
21st November 2010, 03:16 PM
Item 4 Dave:

This is only a common sense advice.
Tensioner may sit at a slight angle if you don't tighten its fixation nut. Just make sure it is tight, does not need to be torqued. You'll find it will be tricky to achieve required vibration frequency if you leave fixation nut slack.

I'll keep my fingers crosses and look forward to viewing your photos.

TK http://i385.photobucket.com/albums/oo291/tkaald/lager.gif

davelowe
3rd March 2011, 09:39 PM
I have done the job and posted the guide on the 'other forum'. Here is a link to it:

http://meganeownersclub.co.uk/forum/viewtopic.php?f=2&t=16274

The post cites this website and thread as a source.

Thanks to all involved for their prior work in developing this technique.

Gonescenile
3rd March 2011, 10:00 PM
Thanks for the link ....... but it's asking for login details so cannot be viewed.
Why not copy it over to here and we could maybe pop it up as a sticky :d

davelowe
7th March 2011, 10:39 PM
Thanks for the link ....... but it's asking for login details so cannot be viewed.
Why not copy it over to here and we could maybe pop it up as a sticky :d

Here is a guide that I have compiled from personal experience of how to change the timing belt, the aux belt, the tensioner, the coolant pump and the coolant on 2002 - 2005 1.9dci F9Q800 engines. Some of the procedure was derived from the factory manual, Haynes, and a technique for belt tensioning on here:

http://www.renaultforums.co.uk/showthread.php?t=58093&page=12

This job is not for the faint hearted. You will need various tools including but not limited to 10, 13, 16, 18mm spanners and sockets. A 1/4" drive ratchet driver for access to tight areas is essential. A laptop running XP or better and a cheap microphone is required. Also needed are two axle stands and a trolley jack.

The following account is what I did and how, described to the best of my ability. TAKE SENSIBLE CARE WHILE WORKING. Read the factory pdf file from start to finish before attempting this job.

1. Here is an image of the parts you need:

http://i210.photobucket.com/albums/bb288/davelowe1977/DSCF1964.jpg

From left to right: Gates Powergrip timing belt and tensioner kit (or equivalent) - source ebay. The M6 x 75mm set screw was purchased locally for 12p. A spring balance - source ebay (2), timing locking pin tool MOT 1541 (NOT REQUIRED), aux belt (ebay), coolant pump (ebay), gasket (Renault parts dept). Don't skimp on the gasket - those sold with the pump are not of a sufficient quality for the job.

2. Unlock the car and open the boot and bonnet. Disconnect the battery earth. Jack up the front of the car and remove the driver's side road wheel. Remove the under tray - there are 7 M10 bolts - see image below.

http://i210.photobucket.com/albums/bb288/davelowe1977/DSCF1973.jpg

3. Remove the clips securing the front section of the driver's side wheel arch cover. Remove this from the car. You will see:

http://i210.photobucket.com/albums/bb288/davelowe1977/DSCF1970.jpg

4. Under the bonnet remove the clips securing the trim from both sides and the front section revealing this:

http://i210.photobucket.com/albums/bb288/davelowe1977/DSCF1969.jpg

5. Remove the bumper. Be careful and have an assistant on hand to help. There are more plastic clips to take off at the top, plus you must use a small screwdriver to depress the locking clips at the top. There are four of them, and they break easily. If you have fog lights, disconnect the wiring loom when the bumper is loose (yellow connector - not shown) Proceed with caution:

http://i210.photobucket.com/albums/bb288/davelowe1977/DSCF1975.jpg

6. With the bumper removed, drain the coolant by opening the filler cap and removing the radiator bottom hose - save the coolant in a clean container if it is new, or flush it away for safety. It has a sweet smell and is toxic - animals like to drink it and then they die. There are two bleed screws behind the battery where pipes pass through the bulkhead to the internal heater. Open them (one turn counter-clockwise). The coolant comes out FAST. Direct the hose into a container. Allow 10 minutes for full drainage. Hose any paintwork that has been splashed immediately.

http://i210.photobucket.com/albums/bb288/davelowe1977/DSCF1987.jpg

7. Remove the bracket shown in step 3 to reveal:

http://i210.photobucket.com/albums/bb288/davelowe1977/DSCF1976.jpg

8. Support the engine with a block of wood (I used 2x4) and a trolley jack - proceed carefully - lift the engine only 5mm or so:

http://i210.photobucket.com/albums/bb288/davelowe1977/DSCF2006.jpg

9. Unbolt the engine mount through bolt (not shown clearly) (18mm spanner) and the engine mount bolts x2

http://i210.photobucket.com/albums/bb288/davelowe1977/DSCF1990.jpg

Note that the bolt will not come out.

10. Remove the link (16mm bolts):

http://i210.photobucket.com/albums/bb288/davelowe1977/DSCF2005.jpg

11. Use a 16mm spanner on the aux belt tensioner and turn anti-clockwise to relieve tension. Note the factory and Haynes manuals specify clockwise - this is INCORRECT. An assistant is useful to get the belt off.

12. Have an assistant apply the brakes with the car in 4th gear AND TWO ROAD WHEEL BOLTS TEMPORARILY RE-FITTED. Unbolt the pulley wheel and put it to one side. Refit the bolt and torque to 40Nm.

13. Unscrew 4 bolts holding the timing belt cover on. It is removed FROM BELOW. Twist it around - it isn't easy to get off.

14. Rotate the engine clockwise by turning the pulley bolt until #1 cylinder is at top dead centre. This image shows when this is the case:

http://i210.photobucket.com/albums/bb288/davelowe1977/DSCF1992.jpg

Check that you can see this:

http://i210.photobucket.com/albums/bb288/davelowe1977/DSCF1997.jpg

15. Mark the position of the sprockets (not the high pressure fuel pump - this is un-timed) to the bodywork with a white pen or paint. Note you do not need to used the timing pin tool.

16. Ease the belt off and remove it. Ensure neither camshaft or crankshaft is disturbed <-- VERY IMPORTANT

17. Compare the old and new belts. Check for length and number of teeth.

18. Remove the coolant pump:

http://i210.photobucket.com/albums/bb288/davelowe1977/DSCF1985.jpg

and clean the old gasket off the block. I used a 39p wooden spatula from Sainsbury's.

http://i210.photobucket.com/albums/bb288/davelowe1977/DSCF1988.jpg

19. Compare the old and new pumps for hole patterns. Refit using new gasket. Note the new unit on the right has a metal impeller:

http://i210.photobucket.com/albums/bb288/davelowe1977/DSCF1984.jpg

20. Unscrew the belt tensioner centre nut (NOT THE MOUNTING PLATE - THE FOLLOWING IMAGE IS MISLEADING) and replace the bearing. Fit the M6 screw from above:

http://i210.photobucket.com/albums/bb288/davelowe1977/DSCF1983.jpg

21. Fit the new belt. Note the arrow pointing clockwise. The white bands on the belt will correspond to the markings on the sprockets you have already made (note there are several bands in different places - you must identify the correct ones before fitting). Start at the crankshaft and work around the various cogs leaving the last bit to the tensioner. I found the belt was so tight to get on that I had to tie a piece of binder twine to it and have an assistant pull hard upwards from under the bonnet while I pushed it into place. Count 28 spaces (not teeth) between the camshaft and the fuel pump. Spin the pump around to align the band on the belt to the mark on its sprocket if you want to be tidy.

22. Remove the pulley bolt and REFIT the aux belt PULLEY and the bolt again. If you have a micrometer, check the bolt is no longer than 49.1mm from shoulder to tip. If longer, obtain a new one. Torque to 40Nm. You should be back to this image now:

http://i210.photobucket.com/albums/bb288/davelowe1977/DSCF1976.jpg

23. Use the software described in the first part of this post to check the belt tension (90 +/- 15Hz). Pretension with the spring balance thus (11Nm):

http://i210.photobucket.com/albums/bb288/davelowe1977/DSCF1996.jpg

Adjust the M6 bolt to achieve this value. Tighten the tensioner bolt when you have the correct reading:

http://i210.photobucket.com/albums/bb288/davelowe1977/DSCF2000.jpg

24. Turn the engine through a few revolutions using the pulley belt. Don't be alarmed that the marks on the belt appear to misalign. This is just a function of the number of teeth. Eventually you will be back to the marks matching up.

25. Repeat section 23. Re-measure the tension to obtain 80 +/- 15Hz)

26. Have an assistant closely monitor the belt under the bonnet. Start the engine (without coolant or aux belt). Run for 30 seconds or so. Expect lights on the dashboard. No harm will be caused. Diesel engines will easily stand stand this situation. My engine fired first time and ran perfectly well.

27. Refit the aux belt with an assistant to aid (the belt is tight and you need one person to hold the tensioner off). Check that the grooves run true (several more revolutions by hand).

28. Re-fill and bleed the coolant. Re-assemble in the reverse order as described.

29. Start the engine and let it idle until the radiator fan starts. Test drive.

I hope this was a useful guide. Please post any errors or omissions and I will revise it.

DL

tolsen
8th March 2011, 12:18 PM
Well done Davelowe!
Just a wee comment on pretensioning:

Required pretensioning torque when using torque wrench and timing belt pre tensioning tool Mot. 1543 (Mitsubishi tool MB996048, Laser Tools 4058) is 11 Nm or 1.1 dNm.
http://i385.photobucket.com/albums/oo291/tkaald/DSC01817.jpg

Required force if using a spring balance to pretension belt is 9.7 kgf or 95 Newtons,
http://i210.photobucket.com/albums/bb288/davelowe1977/DSCF1996.jpg
see this post for details. (http://www.renaultforums.co.uk/showthread.php?p=583365#post583365)
I see you pretensioned to 10 Newtons. Perhaps you should recheck vibration frequency and retension as required whenever you have spare time. It would be interesting to learn if belt tension is significantly lower than that specified.

tolsen
18th March 2011, 07:51 AM
Davelowe Some comments please.

davelowe
20th March 2011, 04:24 PM
Unfortunately I cannot re-check the tension as I have sold the car. I can say with confidence that the belt wasn't too tight (it didn't whine) nor to loose (it didn't flap about).

I must admit that I'm confused over the torque setting. The factory manual recommends 11Nm applied 115mm from the centre of the crankshaft pulley to the belt in an arc.

How did you arrive at the figure of 95 newtons? Surely mechanical advantage calculations aren't relevant here: 11Nm is the force exerted (by whatever means).

tolsen
20th March 2011, 05:55 PM
Unfortunately I cannot re-check the tension as I have sold the car. I can say with confidence that the belt wasn't too tight (it didn't whine) nor to loose (it didn't flap about).

I must admit that I'm confused over the torque setting. The factory manual recommends 11Nm applied 115mm from the centre of the crankshaft pulley to the belt in an arc.

How did you arrive at the figure of 95 newtons? Surely mechanical advantage calculations aren't relevant here: 11Nm is the force exerted (by whatever means).
Specified pretensioning torque = T = 1.1 dNm = 11Nm = 1.12 kgm
Length of arm of Mot. 1543 = L = 0.115 m
Required spring balance force = F
Torque = Force x Length of arm
T = F x L
F = T/ L = 1.12/ 0.115 [kgf] = 9.7 kg force.
or when using SI system of units:
F = T/ L = 11/ 0.115 [N] = 95.7 Newtons
QED

The force should be applied laterally exactly at the same point where the pretensioning tool will touch the belt.

tolsen
24th May 2011, 07:58 PM
This thread has now had 32,063 views but as far as I am aware only 3 persons have actually used the method described in this thread for checking tension on timing belts by measuring vibration frequency.
Perhaps members get put off by the number of posts?
A summary on how to set it all up is in post 56. (http://www.renaultforums.co.uk/showthread.php?t=58093&highlight=f9q+belt+tension&page=12)
http://i385.photobucket.com/albums/oo291/tkaald/DSC01647.jpg

Same method can also be used to check tension of auxiliary drive belts. (http://www.renaultforums.co.uk/showthread.php?t=84791)

DELTAONE
6th October 2011, 10:10 AM
hi there tolsen just have some diffculty trying to sort out the timing belt issue on my Lag it is now running could not find the locking pin hole in the crank so i TDC it through the glow plug and tesioned the belt engine fired up fine and seemed to run ok , very smooth on tickover ,please can you advise mean on the tension of the belt as the new belt is huming slightly can you tell me were i could get the mic for the pc scope set up many thanks DELTONE

tolsen
6th October 2011, 11:16 AM
hi there tolsen just have some diffculty trying to sort out the timing belt issue on my Lag it is now running could not find the locking pin hole in the crank so i TDC it through the glow plug and tesioned the belt engine fired up fine and seemed to run ok , very smooth on tickover ,please can you advise mean on the tension of the belt as the new belt is huming slightly can you tell me were i could get the mic for the pc scope set up many thanks DELTONE
I do not know what you mean by "advise mean on the tension of the belt". You need to read through this thread to work out how to set it all up and measure vibration frequency. Required vibration frequency for your engine may be in this post (http://www.renaultforums.co.uk/showpost.php?p=504657&postcount=49). Not that easy for me to be more specific as I do not know which engine you have.
I strongly recommend that you time engine using the timing pin. Seek assistance from a neighbour, friend or professional mechanic if you have difficulty engaging the pin. Two heads think better than one as they say. Perhaps you do not apply sufficient force on pin to engage into timing slot in crank shaft or hole through crank case is full of dirt etc and needs a clean.

DELTAONE
6th October 2011, 11:58 AM
Hi will have a go with locking pin again although i did speak with a garage owner i know and his ex renault mechanic said it was a lip on the crank that the pin locked up to . the question regard tension was to do with the mic were can i get one from and does anyone have a part number ect /what type i did see some were that Maplins may have the mic many thanks Deltanone

tolsen
6th October 2011, 12:53 PM
See figure 4 in Sealey pdf document that comes up with this link. (http://www.tooled-up.com/artwork/PrdPDF/SEAINSVS4760v2.pdf) Fig 4 shows how the locking pin engages in timing slot on crank shaft. A cheap PC mic is all you need. I use an old Compaq mic that came with a PC retired more than 10 years ago. What is important is that the mic is small and cheap.
http://i385.photobucket.com/albums/oo291/tkaald/DSC02120.jpg
Old Compaq microphone held on to clothes peg with rubber bands. Checking vibration frequency of auxiliary drive belt on my Kangoo Trekka 1.9 Dci.

http://i385.photobucket.com/albums/oo291/tkaald/DSC01647.jpg
Head set microphone used for measuring vibration frequency when developing this procedure. Head set is too big and won't fit when engine is in situ. I recommend earphones when measuring as you can then hear the belt vibrating.

http://i385.photobucket.com/albums/oo291/tkaald/DSC02056.jpg
Same old Compaq microphone attached to a non-slip clothes peg with elastic bands used for checking vibration frequency of auxiliary drive belt on my Smart Cabrio Cdi. Sony PC shown was recovered from a local rubbish tip. Works fine.

DELTAONE
6th October 2011, 02:18 PM
Hi Tolsen thanks for info , Lag runs ok does however seem slight bit noisey nothing bad , pulls like a train if the valve timing was out would the car pull like this . thanks for the info with regards to pc software i have downloaded it and seems to work ok on my newish pc just need the mic now and will follow your post you guys seem to have done a lot of work on the cambelt tension issue really good stuff well done .Just ordered spring balance from ebay so i can reset pre tension and make sure that the belt does last , many thanks Deltaone

tolsen
7th October 2011, 09:39 AM
In my humble opinion it is more important to get timing right than get belt tension spot on.

Timing pin can be inserted without the help of another person but require some clever thinking and setup - you need a way to maintain preload on timing pin whilst you are turning engine.

Best method is turning engine until timing mark on cam shaft pulley is a notch or two before TDC, then get a mate to keep pushing timing pin towards crank shaft whilst you gently turn engine a wee bit further. You have missed the timing slot if timing pin goes all the way in and engine can be turned more.

See figure 4 in Sealey pdf document that comes up with this link. (http://www.tooled-up.com/artwork/ProdPDF/SEAINSVS4760v2.pdf) Fig 4 shows how the locking pin engages in timing slot on crank shaft.

There can be some sealant residue in hole through block for timing pin. Just clean up by inserting an 8 mm drill and pick up the residue in the flutes.

Never use an 8 mm drill as a timing pin. A drill bit is too brittle and may snap off in many pieces dropping into the sump leaving you with another problem to sort out.

DELTAONE
7th October 2011, 04:24 PM
Cheers mate for all the advice and info about the F9Q DCI engine i was going to check the TDC /Crank locking pin today but i just had to much to do hopefully i will have sorted this out over the weekend . I will try my pc scope and mic out as well thanks for your input ,i will let you guys know how it all works out
Deltaone