Tim Murphy

Yep, If it works as planned I will send one to everyone.

Lagavulin

<--- That's Tim making sure the preload is juuuuuuust right.

You GO boy!

PorKen

BrianG said: "A dynamic tensioner with controlable preload !!!"

Now add hydraulic dampening and then we'll have a proper timing belt tensioner!

Gretch

"damper" as in: to damp. "dampener" as in: to 'pee upon'

Carl Fausett

Video available -

This Schwarz/Schwarz 1990 GT install (that this thread is about) can be seen running on the dyno here:
http://www.928motorsports.com/ourvideos.html

Right-click on "32v Install" and save to your PC before running it for best results.

worf928

This would support Donald's 'pressure wave resonance' theory.
That makes perfect sense as the theoretical maximum power a/f is 12.6.

Pardon the hack job on the quote.

The above is why, in my first post, I mentioned the measurement technique. Out of context of the DynoJet if you are tracking a pure sine wave signal and sample every 360 degrees (or 2*PI radians) you will 'measure' a straight line. That's very smooth compared to sampling a sine wave every 180 degrees which gives your a sawtooth.

It all comes down to how many samples the Dyno can record during a pull. The more the better - up to a point. The DynoJet needs only to measure velocity of the drum, engine RPM and time. I can only think of a few sources of inaccuracy - RPM measurement and tire slippage. The former may tie in with 'grounding the chassis' during a pull.

Understanding the number of samples taken in each pull is the key to determining if the squiggly lines are real or not along with any understanding of the accuracy of the RPM measurement. How accurate is inductively-measured RPM?

Answers from the DynoJet website:

RPM accuracy: +/- 1/10th RPM, timing 1 microsecond, drum speed 1/100th MPH.

That's pretty darn accurate and leaves only tire slippage as a source of measurement inaccuracy. (Unless, static or other electrical effects threw off the RPM measurement.) It also means that the Dynojet can take as many samples as it pleases. From eyeballing the original graph it looks like samples every 50 rpm. That's about 14 samples per second given a 7 second run from 2000 rpm to 7000 rpm. Child's play in terms of rate for the DynoJet.

It would be interesting to get the raw data (e.g. each individual measurement) from the pulls. How many samples? They should be stored on the DynoJet's computer and available in a tab-delimited or CSV format or - worse case - a printout.

I'm not convinced of anything yet - other than Tim's kit bolting on a seriously cost-effective amount of power - but I think those squiggly lines at 11:5 in 4th gear may be real.

Old & New

I believe the inertia of the drums would preempt such sudden variances in the curve...

If they were "real", you would have experienced a horrible fluttering vibration on the dyno.

C'mon guys.

worf928

First, let me re-state that I have no conclusion. I'm trying to elicit more information and more understanding. The squigglies could be real or a problem with measurement - or maybe a little of both.

The inertia of the drum is resistance to acceleration. The variances in the curve simply show that the acceleration of the drum varied - cyclicly with small periods of very high acceleration followed by merely high acceleration. Or IOW - all positive acceleration just sometimes more-wildly positive than otherwise.

The biggest up/down variation is - by eyeball - about 20 ft-lbs from 4.5k to 4.9k with three cycles. At that point with over 350 ft-lbs on tap, 20 ft-lbs is less than 6% variation. I do not think buttometers would notice a 6% variation that cycles three times in under one second and then is more-or-less steady.

What is the threshold sensitivity of the average buttometer?

If they are not real then from where do they come? All guesses and/or informed opinions accepted...

They are not the result of too-few samples. Quite the opposite. The 5+% squiggliness we see is far more than what one would expect given the stated-by-the-maker accuracy of the DynoJet. If they are in fact by-products of measurement - either induced by measurement or uncharacterized noise - then there is a way to characterize and quantify the effect. A 5+% measurement variation is pitiful at this macro scale.

worf928

Carl, just confirm: did you use The Hammer after the run to read-out accumulated knock codes? Or did you have it hooked up during the romp?

Carl Fausett

I'm going to email the dyno owner and see if he cannot email me the raw dyno data files for you...

worf928

That would be cool. I would put those together with three other sets of GT raw data for comparison and manipulation purposes.

Tim Murphy

Yep Carl, I just watched the video of the GT on the dyno and the belt squeal is definitely there on the first run. No question about it, there is 30+ RWHP to be had with a non-slipping belt. I just did a quick spin in my car and made 14lbs boost using the die spring on the tensioner. If it holds 14 psi it will certainly hold 8 psi. Once I receive the springs I will send one to each of the Murf928 SuperShark owners.


Just to add to what Carl has stated above:

It is our policy to be as forth coming with information as we possibly can with all of our customers. Those that have dealt with me personally know that to be the case (at least I hope you all do . When asked how we accomplish certain tasks we will tell you EXACTLY how we do it because we want to answer any and all questions you might have, no surprises when UPS drops it off. The cost of our kits’ include 100% brand new components with all the factory warranties intact. Although we don't recommend it, if you wanted to source some used parts on your own, such as the supercharger, SFMU, gauges, injectors, blow of valve..etc, we will be happy to deduct the common retail cost from the Murf928 kits. With that all being said, I have to amend an earlier statement that I made regarding permanent modifications.... I forgot that you have to break the bracket off the passenger fuel rail that holds the plug for one of the knock sensors or it interferes with the intercooler. This will not be required on the stage one kits. Sorry for the oversight.

Old & New

Worf,

I think the question of fluctuation is one for the dyno manufacturer. I'll buzz them tomorrow.

worf928

Tim, my '94's right side fuel rail does not have that bracket. I suspect that it may have been deleted in later years (after '91 perhaps) and that a rail without the bracket could be ordered thus removing one non-reversible modification. The PET shows the same part numbers for '88 - '95. Nevertheless...

worf928

I completely agree.
Cool. I hope they are forthcoming. It is my suspicion that the DJ might produce the squigglies if there's harmonic vibration on the drum (due to for instance a hopping wheel) and/or if the inductive pick-up misses a pulse from a plug wire every once in a while (<1% might be enough.)

Regardless, they should have some idea...