Stock R32 Dyno - All-wheel Dyno Dynamics Dynamometer

[Prefekt]

Courtesy of GHL, my stock R32 was dyno'd today at Dynocomp in Scottsdale, AZ. Dynocomp uses an all-wheel Dyno Dynamics dynamometer to produce a true load based dyno plot, unlike a Dynojet.

Stock numbers are:

174 awhp @ ~6250RPM
180 awtq @ ~3150RPM

I'll have the sheet scanned in tonight and I'll edit this post with it. Tuesday GHL will be fitting their first R32 full exhaust to my car and a post dyno at Dynocomp will follow. Stay tuned for the results!

And for comparison sake, my buddies EVO8, which GHL is also working with, put down 212 awhp stock and 226 awhp with just straight exhaust at Dynocomp.

EDIT: Here are the graphs -

Horsepower and Torque

Horsepower and Air/Fuel


Modified by Perfekt at 4:26 PM 2-13-2004

[Prefekt]

Oh, and for those that are curious about how the new Haldex system works on the R32, my car spun all 4 rollers equally during the dyno. It responded just like a full time all-wheel drive car would which I thought was interesting since the previous Haldex systems act like FWD until there is slip then transfer power to the rear. My car twisted all four wheels 100% of the time.

[Thinman61]

Hmm.
27% loss HP
That seems high no?

[j-dub]

Did you try to test it in FWD mode by setting the hand brake as Daemon42 has sugested is possiable? I guess the TT is also capable of this. Just curious.

[carma]

Quote, originally posted by Perfekt »

Stock numbers are: 174 awhp @ ~6250RPM 180 awtq @ ~3150RPM

That's weird, because those numbers are the exact same as a stock MKIV 12v VR6.

[Prefekt]

Quote, originally posted by Thinman61 »

Hmm. 27% loss HP That seems high no?

Seems about right to me. STi's and EVO's are in the low to mid 30's on drivetrain loss.

Quote, originally posted by j-dub »

Did you try to test it in FWD mode by setting the hand brake as Daemon42 has sugested is possiable? I guess the TT is also capable of this. Just curious.

No thanks. I drive it without the e-brake on, I want it dyno'd the same. But seriously, I would rather know the real world all-wheel numbers than the 'bench' FWD numbers.

Quote, originally posted by carma »

That's weird, because those numbers are the exact same as a stock MKIV 12v VR6.

But stock 12v are FWD on a Dynojet most likely, and I've yet to see a non-feel good dyno give a stock 12v 174whp. If you mean that's the same as the stock 12v's crank numbers, then yes!


Modified by Perfekt at 1:49 PM 2-13-2004

[euro sport]

Quote, originally posted by JustinTheGIMP »

those are good numbers...now I really want one

you really think those numbers are good? Or are you being facetious?

I don't believe you should see anything below 200whp on the dyno. The numbers that Perfekt posted look like 2.8l VR6 numbers.

[Thinman61]

Well the Evo in the first post was showing a 23% loss no?

I'm confused.

[Prefekt]

Quote, originally posted by euro sport »

you really think those numbers are good? Or are you being facetious? I don't believe you should see anything below 200whp on the dyno. The numbers that Perfekt posted look like 2.8l VR6 numbers.

He's being serious. I think the numbers are right on target. Comparing it to Neuspeed's dyno is no good. HKS's dyno is a 'feel good' dyno and an inertia dyno at that.

[Prefekt]

Quote, originally posted by Thinman61 »

Well the Evo in the first post was showing a 23% loss no? I'm confused.

The stock number I posted was by word of mouth as I wasn't there to see it. The 'with exhaust' number I saw first hand today. I cannot vouch for the validity of the stock EVO8 number.

[euro sport]

Quote, originally posted by Perfekt »

He's being serious. I think the numbers are right on target. Comparing it to Neuspeed's dyno is no good. HKS's dyno is a 'feel good' dyno and an inertia dyno at that.

Okay... it's obvious that most dynos will produce differing numbers. Those just seemed extraordinarily low.

Do you, by chance, have any dyno numbers from a 12v VR6 that was ran on that same dyno (preferably stock)? I'm just curious what they would produce.

[Blitz16v]

How much crank is that?

[GSB Enterprises.com]

Making 175+ at all four wheels is excellent. WRXs make like 150. AWD cars lose vast amounts of power through the driveline, many times close to 30%. The R32s are very fast in real life and will beat anything in the same power range in general because the huge, flat torque curve.

[Prefekt]

Quote, originally posted by Blitz16v »

How much crank is that?

VW rates the R32 as 240hp and 236ftlbs at the crank.

Quote, originally posted by euro sport »

Do you, by chance, have any dyno numbers from a 12v VR6 that was ran on that same dyno (preferably stock)? I'm just curious what they would produce.

I do not know of a stock 12v run at this dyno. But if I find one I will definately post it up.

Quote, originally posted by GSB Enterprises.com »

Making 175+ at all four wheels is excellent. WRXs make like 150. AWD cars lose vast amounts of power through the driveline, many times close to 30%. The R32s are very fast in real life and will beat anything in the same power range in general because the huge, flat torque curve.

Exactly my sentiments! Once I post the graph you'll really see what he's talking about with the torque curve. Its like a damn plateau starting at almost idle!!

[Daemon42]

Quote, originally posted by Thinman61 »

Hmm. 27% loss HP That seems high no?

Not really. It's one of the reasons I was very suspicious of Neuspeed's dyno as it shows
much lower drivetrain loss and higher whp than I'd expect for an AWD dyno.

Perfekt: A better comparison than the Evo would be to know what a stock WRX puts down
on that same dyno. Do they know? If so, I want to know. There's hundreds of WRX dynos out there
to compare to so they make a good baseline. A typical stock WRX puts down 165-169whp
which is a similar 27% drivetrain loss. My *guess* is that a WRX on that particular dyno
probably is a bit lower than that, as a WRX has center diff and the R32 does not
so its AWD drivetrain loss should be somewhat less. But maybe they're closer
than I think in AWD mode.

As for it spinning the rollers at the same speed, once it's moving that should be normal.
If you hammered it in first gear from a stop though, the front wheels should turn
about 15 degrees before the rears start to move, and by 45 degrees (angle of this V)
at full power the rears should be turning the same speed as the fronts. An AWD system
with a center diff will also turn them at the same speed (ala WRX/S4). An AWD system with
only a center viscous coupling and no diff (of which are are a number of cheap mini-SUV
examples and more exotics like Porsche 911TT and Diablo)
will always turn one axle a bit slower than the other as the viscous fluid
never provides a full lock.
I have a strong suspicion that the Haldex system only needs a rotational difference
to apply the initial force on the clutch, but from then on, the torque
imbalance at the input shaft keeps it clamped. Think of it like your brakes.
You push the pedal down to engage the brakes, but then the pedal stops
and it holds the same force. No additional "pumping" is necessary to maintain
that force, thus no additional wheelspin will be detected. Haldex also
has an electric pump to prime the slave resevoir so it should be able to
initiate engagement even before wheelspin is detected, such as during launch
although I sure that it still needs that torque imbalance to provide full lock.

As for the numbers themselves..
Folks have to keep in mind that as soon as the R32 is out of first gear the Haldex
clutch will be disengaged most of the time and total drivetrain loss will decrease
significantly. The above numbers represent the worst case scenario.

Perfekt, did you get a chance to try the pulling the handbrake trick to disengage Haldex
just to see if it works?

ian


Modified by Daemon42 at 2:10 PM 2-13-2004

[Daemon42]

Quote, originally posted by Perfekt »

No thanks. I drive it without the e-brake on, I want it dyno'd the same. But seriously, I would rather know the real world all-wheel numbers than the 'bench' FWD numbers.

That logic if flawed. The R32 spends most of it's time in FWD mode, not AWD.
So in the "real world" the FWD numbers are probably a lot closer to what you're really putting to
the ground. Only sees AWD drivetrain loss during hard launches and hard cornering.

ian

[Thinman61]

Yeah your all are right.

Hey, do you happen to know if the dyno averages front and rear? I'd be curious to know what the power distribution was front/back.

[Prefekt]

Quote, originally posted by Daemon42 »

That logic if flawed. The R32 spends most of it's time in FWD mode, not AWD. So in the "real world" the FWD numbers are probably a lot closer to what you're really putting to the ground. Only sees AWD drivetrain loss during hard launches and hard cornering. ian

There was no 'hard launch' on the dyno in the 4th gear pull and all four wheels spun the same, so I would think your logic is flawed. I agree that previous Haldex systems drove that way and I had assumed the new system in the R would as well. But actual execution proved otherwise. For comparison with that, Dynocomp dyno'd an '01 Audi TT and had to pull the 3 fuses to disable the Haldex system on it as the car was not transferring power to the rear during a smooth acceleration. My R had no issues what so ever.

I did not try the e-brake trick.

[Daemon42]

Quote, originally posted by Thinman61 »

Yeah your all are right. Hey, do you happen to know if the dyno averages front and rear? I'd be curious to know what the power distribution was front/back.

Depends on the dyno. This one is load based, so if it applied the same load to
each roller, then the torque split was 50/50. Haldex can send nearly 100% of the
torque to the rear axle if that's where 100% of the load is, up to a couple thousand
Nm before the clutch starts slipping (it's rated at 1000Nm, but I've heard folks
say it's been tested up to 2000Nm). If you had an inertial dyno (like the dynojet)
then the torque split would be 50/50 if both rollers were the same sized, which is
not always a given. I've heard of AWD dynos with different sized rollers
so an AWD car with an open center diff would try to split the torque 50/50
and end up spinning the smaller roller faster. Anything with a viscous coupling
in addition to the center diff (like the WRX) would try spin them at close
to the same speed and more torque would go to the side with the larger roller.
A Torsen diff (ala A4, S4) might try to do the same. Same speed, but more torque to
the larger roller.

BTW, load based dynos (like the Mustang) in general show lower whp numbers than inertial dynos (like Dynojet).
Inertial dynos can be fooled by things like light or heavy wheels.
The advantage Dynojets have is that they require minimal if any calibration, so you can compare
the numbers from two different dynojet dynos and have reasonable confidence in their
consistency, if not complete trust in their accuracy.

ian

[Daemon42]

Quote, originally posted by Perfekt »

There was no 'hard launch' on the dyno in the 4th gear pull and all four wheels spun the same,

I already explained why they turn the same once the car is rolling, and yes obviously there is
no hard launch in 4th gear, which is why you wouldn't see any wheelspin necessary to
activate the system. It was activated a moment after the front tires started turning
and stayed active the entire time there was a torque imbalance between front and
rear.

Quote »

so I would think your logic is flawed. I agree that previous Haldex systems drove that way and I had assumed the new system in the R would as well. But actual execution proved otherwise. For comparison with that, Dynocomp dyno'd an '01 Audi TT and had to pull the 3 fuses to disable the Haldex system on it as the car was not transferring power to the rear during a smooth acceleration. My R had no issues what so ever.

I know for a fact though that once the torque imbalance goes away (out on
the highway) the clutch will disengage nearly 100%. That's the whole *point* of
Haldex. To give you AWD only when it's needed, and FWD most of the time
for fuel economy purposes. There have been piggyback boxes
made for the Haldex ECU which will keep it engaged all the time (Forge made one) , and just like
a 4WD truck with a locked center diff, it binds up during cornering, and mileage
is very very bad. Nothing about your dyno experience disproves what I'm saying.

There may well have been some tweaking between the older TT's Haldex system and
the new one, but there is no way that it remains in AWD mode on the highway.
Ask Haldex if you don't believe me.

ian

[Prefekt]

I see what you are saying now. Makes sense that the rear is disengaged during straight line driving.

[Thinman61]

Nice. Thanks for the explanation (again ).

Kinda off topic. The shop around me uses Dynapack 5000's.
I read up on the system on their website
http://www.dynapackusa.com/tech.htm
seems like a good system. Anyone have experience with these?

[plf]

daemon42 ... will you father my children. I don't think I have ever run into a person on any forum who knows as much as you do. Thank you very much for the thurough and consice explanations.

[Prefekt]

Bump for graphs

[vr6gtispeed]

I had a 96 GTI VR6 12V an I ran it on a dyno around christmas. The car was stock but with an AMS K&N cone filter. THe number were

160.7 whp
156.6 ft/lbs

Here is a video and the dyno sheet.
http://mywebpages.comcast.net/vr6gti/dynorun.jpg
http://mywebpages.comcast.net/vr6gti/dynorun3.wmv

I have an R32 and it's night and day with the power from the 96 GTI.

[Integrale]

Quote, originally posted by Perfekt »

Courtesy of GHL, my stock R32 was dyno'd today at Dynocomp in Scottsdale, AZ. Dynocomp uses an all-wheel Dyno Dynamics dynamometer to produce a true load based dyno plot, unlike a Dynojet. Stock numbers are: 174 awhp @ ~6250RPM 180 awtq @ ~3150RPM

How many miles do you have on your car? I dare say it'll increase over some time.

[9VW23yrs]

Just for reference on a Dyno Dynamics Dyno at Vishnu
Stock: WRX=155whp, EVO=190whp, STI=218 whp


Modified by 9VW23yrs at 5:46 PM 2-13-2004

[Prefekt]

Quote, originally posted by Integrale »

How many miles do you have on your car? I dare say it'll increase over some time.

I rolled over 700 miles leaving the dyno today. Word is from the euro owners that this motor gets really strong between 5k & 10k on the clock. But by then I'll have already significantly modded the motor.

Quote, originally posted by 9VW23yrs »

Just for reference on a Dyno Dynamics Dyno at Vishnu Stock: WRX=155whp, EVO=190whp, STI=218 whp

Thanks! Good to know!

[Daemon42]

Quote, originally posted by 9VW23yrs »

Just for reference on a Dyno Dynamics Dyno at Vishnu Stock: WRX=155whp, EVO=190whp, STI=218 whp

That's what I suspected. I see stock WRX's pull mid 160's on other dynos, but
I figured it'd probably be lower on this one, being load based.

Using that as a loose benchmark..
155/227 = .68 for a 32% drivetrain loss.
If the R32 had the same drivetrain loss (which I know it doesn't)
The 174 above would be 174 / .68 = 255hp. (very doubtful this early in its life)
If we took VW's factory power rating as truth.. then
it's 174/240 = .725 27.5% loss
If the R32 really has 250hp, then it's 174/250 = .696 .. 30% loss
It *should* have less drivetrain loss than the WRX even in AWD mode
because it has no center diff or viscous coupling. Once the Haldex clutch
is locked, it's basically just a straight through connection from front to rear.

ian

[LongviewTx]

Daemon42:

While this is off the specific topic, I had hoped to ask your opinion on the affect of wheel weight on performance and related, what affect a wheel weight difference would make on a Dyno test.

First, in considering optional wheels (mainly driven by desire for aesthetics and individuality) I want to have a accurate idea of what wheel weight will do to performance. 18" BBS CH @ 23.3 #/wheel (second choice) and my desired wheel @ 27.5 #/wh (RS4 nine spoke). Will this realistically make a difference in acceleration & performance.

Do you know what weight the O.E. OZ wheels are for a base reference? I haven't seen this info posted before. I know the superlaggers' weigh 18.* #/wh.

Your insight and opinion would be appreciated.

[NOVA337]

That has to be one of the flattest Tq curves I've seen!

[Daemon42]

While I can't give you specific numbers I will say that a 27# 18" wheel is
heavy as hell and is going to have a measurable negative impact on performance.
The rotational inertia for a mass distributed around a hollow cylinder
(not a horrible aproximation of a wheel)
is I=M*R^2
That means that while rotational inertia goes up directly proportional
to the increase in weight for the same diameter, it goes up by the square
of the increase in radius. That means an 18" wheel
which is 6.6% larger than a 17" wheel, has roughly 12% higher rotational inertia.
When you're talking a total of 160-200 lbs worth of tires and wheels that that you've
got to spin up to speed, you're going to feel that weight bigtime.

From a performance perspective the 18" wheels on the R32 and 20AE
are a mistake. They look good and fit over the brakes, but you pay for it
bigtime. When it comes to putting new summer wheels on my
R32, I'll drop down to the lightest 17's that'll fit over the brakes I
can find.

To bring this back on the topic of dynos. Rotational inertia is felt most
during rapid accelleration. In the real world it means it hurts you
most during 1st and 2nd gear when your accelleration is highest.
On an inertial dyno that calculates power by measuring how fast
it can spin up a large heavy spinning cylinder in a relatively short
period of time, heavy and large diameter wheels will have a significant
negative impact. You see it more on the dyno because it's accellerating
the wheels from say 20mph to 100mph in 3rd or 4th gear in about
the same time you'd normally cover 2nd gear. The faster the
wheel accellerates, the worse the effect of rotational inertia.

On a load based dyno like this Dynamics or Mustang it can control the rate
at which the dyno's road speed increases (even stopping it and holding at a specific
speed) and measures power directly, so if you're going to run large wheels,
decrease the rate of accelleration to minimize the effect of the wheels.

ian

[Daemon42]

Thought it might be interesting to expand the vertical scale a bit to
give a better comparision with other dynos we're used to seeing.

I think this curve is why people's butt dynos are confused by the R32
and to some people doesn't feel as fast as it really is.
It starts high and decreases slowly toward redline. When you
mash on the gas at any given rpm, the torque is instantaneous.
You don't feel a rapidly rising torque curve. It's just there.
Notice that the power curve (actual work done) doesn't drop until
the last few rpms before hitting the rev limiter.

Compare to a 1.8t dyno (Revo 3Bar chip) making comparable wheel hp (keeping in
mind that we've yet to see an R32 dyno run in its more normal 2wd mode with
lower drivetrain loss)

The first area I've highlighted in yellow on the left is the steeply
rising part of the curve that makes turbo cars so addictive. The
slope of the curve there is largely determined by how fast the
turbo can spool up. I repeated the highlight twice more to the right
to show what happens if you're just cruising along at some higher
rpm at neutral throttle and then mash on the gas. You can't get
the instant torque of a NA motor, so instead you feel that same steep
rising curve which gives that "hand of God push" feeling unique to turbos.

Which one is actually quicker and faster, remains to be seen.

ian

[darrenewest]

That blows, check signature.

[Tristan]

Quote, originally posted by Perfekt »

Oh, and for those that are curious about how the new Haldex system works on the R32, my car spun all 4 rollers equally during the dyno. It responded just like a full time all-wheel drive car would which I thought was interesting since the previous Haldex systems act like FWD until there is slip then transfer power to the rear. My car twisted all four wheels 100% of the time.

This has to happen. The way any awd system work is that ensures that all 4 wheels are spinning at the same speed (or very close to) at all times. So as soon as the fronts start moving the system thinks their is slip and transfers power to the rear.
Note: this is different from real world driving. In rear wheel driving as the front wheels pulled the car the car would move, thus turning the real wheels. Thus, unless there is wheel spin all 4 wheels move at the same speed and Haledex stays dis-engaged. On a dyno however, unless the two drums are connected (which I don't think they are) nothing about the front wheels spinning means the rear wheels have to spin. Up untill haldex sees this an sends power to the back.

As far as the e-brake thing, that won't work. It would just be like dynoing your car with the breaks on (duh )

The only way I've ever herd of to make an Awd car work like a 2wd car is to disconnect two of the drive shafts. Thus even though power is being transfered to the back those wheels would move. But this would still give you most of the drivtrain loss, so the only reason to do this would be if you wanted to dyno on a 2wd dyno.


Modified by Tristan at 4:37 PM 2-16-2004