[200q20v] Re: timing belt/dist rattle/cam chains

Bernie Benz b.m.benz at prodigy.net
Mon Feb 4 14:42:19 EST 2002


To answer your questions, let me expand on my "hypothesis", since apparently
there is some interest among the 20V lists.

An underdamped resonant system, electrical or mechanical, is a docile system
except when driven at or near its resonant frequency.  Driven at this
resonant frequency, the vibration amplitude increases sharply, dependant
upon, and limited only by the Q (quality factor) of the system.  (Q is an
inverse function of system friction.)

The infamous 20V distributor rattle apparently ocours, and is at its worst,
around 2200 RPM(?) and at a slightly negative engine torque condition, i.e.
no applied power and with some small engine braking.  (I haven't experienced
the problem, just guessing that this is the consenses. listening to the
descriptions of you rattlers.)  Under these conditions, timing belt static
running tension is at its minimum, inasmuch as there is a slightly negative
valve train decceleration force being transmitted which is just equal to the
positive force transmitted to overcome valve train friction.  2200 crank
RPM, 1100 cam RPM being the resonant angular frequency of the valve train
rotary system.  If there ever is to be a floppy T belt condition, this is

Remember, this resonant angular vibration is a cyclic speeding up and
slowing down of the cam system inertia superimposed upon its steady state
value, the 1100 cam RPM.  This resonant angular acceleration of the cam
system and the resultant rotary forces generated are transmitted by, or
coupled through the timing chain, the distributor drive gear system, and the
timing belt.  The distributor shaft inertia is forced to follow this
resonant angular vibration because it is tightly coupled to the cam by its
gear system.  IMO, it is this angular acceleration force required to drive
the distributor shaft inertia at this angular vibration frequency that
stresses the plastic gear to eventual failure, not just the steady state
torque required to turn the distributor shaft at half crank speed.

Timing belt spring rate is very low and nonlinear at low belt tensions.  At
some higher belt tension, when each tension cord is equally stressed and the
cover materials are fully compressed, the belt spring rate settles into a
much higher and linear value, the design opperating area.  Because of the
high valve train inertia of the 20V system, it apparently is critical that
the timing belt be run at some rather high tension to avoid this resonant
valve train syndrome.  I don't know what the correct tension is or how to
define over tension, but I'd elect to error on the high side.

John L., will you please further elaborate, from your 30 year "feel" base,
on 20V belt tension?  I would say that it should be considerably higher than
that required for a 10V because of much higher valve train inertia.  Will a
20V belt complain audibly if over tight, as apparently the 10V does?


> From: "Aaron Gjerde" <gjerdea at pconline.com>
> Date: Mon, 4 Feb 2002 11:24:56 -0600
> To: "Audi 20v List" <audi20v at rennlist.org>
> Cc: <b.m.benz at prodigy.net>
> Subject: re: timing belt/dist rattle/cam chains (was dubious distinction)
> Bernie - I'm intrigued.  Your hypothesis that "the tighter the belt, the
> higher the resonant frequency" seems very plausible.  I have a couple
> questions:
> 1.  At what point is the belt too tight?  I remember John has a "30 yrs of
> feel" criteria, which I certainly can't replicate.  Has anyone ever tried
> different belt tension settings and measured the rpms that the rattle occurs
> at and the *approximate* frequency of the rattle on some sort of scope?  Is
> that possible?  I know it isn't for me, due to time and tools required.
> 2.  Doesn't a tigher belt just increase the frequency, such that it make
> become inaudible (i.e. >20KHz)?  The amplitude of the wave may not
> necessarily be any less.  Does a higher resonant frequency have lower wear
> characteristics?  Or would we just be making the inevitable less audible?
> Finally, as food for thought, my car is an extremely early coupe (build date
> of 9/88) with the original dist at 150k miles.  The timing belt and water
> pump were replaced at 90k, and I have never experienced dist rattle.  I
> haven't done the timing cam tension test.  Is every part in the complete
> system (cams, chain, timing belt, tensioner, dist , dist gear) - all the
> same for all VINs?  My car is pre-production and sometimes has part nos.
> that no-one else's has.
> Based on this theory, we need to look beyond just the dist and its gear,
> since the harmonics and resonances are probably created by or at least
> perpetuated throughout the entire system.  Has anyone ever asked Audi about
> this?
> Interesting thread.  Glad we have some of the smart folks on this list that
> we do.
> - Aaron
> 89 CQ 20v (#197)
> Subject: Re: [200q20v] dubious distinction?
> From: Bernie Benz <b.m.benz at prodigy.net>
> Date: Sat, 02 Feb 2002 14:54:14 -0800
> X-Message-Number: 7
> -snip-
> IMO, both the distributor rattle and chain wear problems, as well as timing
> belt life variability, are caused by a resonant torsional vibration problem
> in the cam drive system.  The only varible among different 20V engines
> affecting this torsional vibration system is the input spring rate, the
> timing belt tension.  A low belt tension will both lower the system
> torsional resonant frequency and also allow for a much increased torsional
> vibration amplitude. This increased vibration amplitude and the associated
> higher angular acceleration forces, IMO, is the root cause of this 20V cam
> drive train problem. This problem is not evident in the 10V equivelent
> system because it has only half the cam inertia, thus doubling the torsional
> vibration frequency and pushing it out of the critical range, even with a
> loose belt.
> Thus, a tight timing belt should have a longer and more consistant life than
> a loose one, while also preventing the other cam drive system rattle and
> wear problems.
> Keep your belts tight, kids!
> Bernie

More information about the 200q20v mailing list