Ahh, back from a few days at Watkins Glen. I really, really love that track, it is a perfect turbo car track. We had perfect weather and no incidences. This is like crack, I'm back and looking for my next fix.

Anyway, in reading this thread, I'm sorry that it has turned the way it has and I'm sorry that some of you are unhappy with the results and are questioning the accuracy of the dyno or the operator.

A couple of comments; The wide band O2 sensor is a BOSCH LSU 4 which is a 5 wire heated wide band sensor that is about 300ms fast. It uses the control sourced from a MoTec PLM. I have a MoTec PLM in my car and have compared numbers which are very similar. Clamping the sensor in the tailpipe is not as accurate as having it properly placed in the exhaust pipe like you would do in a tuning session. How far off it would be, I don't know.

As Adrial pointed out, there is not correction factor similar to that used when running a car on a Dynojet dyno. The rear rato is calculated by multiplying the R&P ratio by the gear used, in this case 4th gear to get the proper ratio. It is then verifyed by doing a hold test which is performed at 3000rpm to make sure the dyno is in synch with the car. This is extreemly important since the dyno controls the car. A preset rpm limit is entered into the dyno software say for argumet 6500. The dyno will NOT let the car rev beyond that limit. It also has the ability to hold the car at any rpm, so when doing the 3000rpm test, the car is floored, boost is building, but the dyno is holding the car steady at 3000rpm by the use of hydraulics. I feel confident that we entered the ratios properly since we verified every one of these.

As for detonating, there were two cars we heard detonating on, I believe the white one which we increased the fuel pressure to 65psi and then the red one everyone is discussing. It was very audible if you were near the front of the car.

The red car was indeed showing 10psi on his fuel pressure gauge (the same gauge and pressure regulator that was in the white car). Fuel injectors do not work very well below 2bar of pressure (29psi) and will not give you a proper spray pattern. Typically a 3bar base fuel pressure is used (vacuum line disconnected) and will drop to about 2.5bar at idle when the vacuum line is connected. This will give you 4bar of fuel pressure at 1bar of boost, 4.5bar of fuel pressure at 1.5bar of boost, etc.

Anyway, I really, really need to go to the track again.

 

Geoffrey, please don't take MY comments as either criticism of DeMan or you personally-I've already commented upon the professionalism and overall good vibes I get from you guys.

That said, there seems to be something not right about how this dyno measures 968's-and I'd like to get to the bottom of that. If the only evidence I have is that the torque peak as measured on the Dynopack is almost 1000 rpms away from where it should be-well, that's enough to tell me something is awry. I figure the truth is out there somewhere!

 

I definately don't take it personally. I too try to resolve inconsistencies so that I can find out what is actual. As a reference, the June issue of Turbo magazine tested 6 dynos on the same day with a Nissan 350Z. For what its worth, here were the results.

Dynapack 5500 - (249.5, 249.2, 249.5, 249.4hp) (242.7, 243.1, 242.6,242.8torque)
Dynamic Test Systems (DTS) - (257.2, 259.6, 256.1, 257.6hp) (552.2, 555.7, 561.7, 556.5torque)
Dynojet 248 (windows) - (235.1, 235.3, 237.0, 235.8hp) (226.4, 227.1, 230.0, 227.8torque)
Dynojet 248 (DOS) - 240.2, 244.4, 246.7, 243.7hp) (233.1, 238.5, 240.0, 237.2torque)
Clayton VTT (265.3, 266.2, 265.7, 265.7hp) (240.2, 241.5, 240.3, 240.6torque)
Super Flow SF-1853 - (228.9, 228.9hp) (226.7, 226.5torque)

Unforunately the graphs are more telling than the raw numbers. They were the most interesting part of the article.

 

But is there any way we can explain this major error in the location of the torque peak of my car? It is consistent on all the 968's you've dynoed-surely something is skewed, which MAY explain the low hp numbers.

As Sherlock Holmes would say, 'The game is afoot, Watson!'

 

I found the turbo magazine article online:
<a href="http://www.turbomagazine.com/tech/0306tur_dynodash/index.html" target="_blank">http://www.turbomagazine.com/tech/0306tur_dynodash/index.html</a>

They draw no conclusions at the end of the article.

I'll draw one very obvious one: The dynapack's runs were more consistent than those on the dynojet running either windows or DOS software.

Geoffrey,
I so need to get to the track again! Even autocross would do it at this point. I haven't had a good hit (hehe) of the track crackpipe in all too long! I will hopefully be at the Glen sometime in the next 2 months.

 

Maybe this was already posted and I missed it in this thread...

But were all the 968 6-speeds? Any Tips?

Karl.

 

Karl,
You didn't miss it, it was never mentioned. All the 968s were 6 speeds.

 

Perhaps I'll put my all stock 91 clubsport on the dyno the next time I'm there. It should show similar results as a 968 but with less HP & torque. We were going to do a 90 cab but the clutch was slipping too much.

Again, dyno results do vary from dyno to dyno, from mfg to mfg, from day to day. It is a relative comparison tool.

I personally really like the Dynapack dyno because it is a very effective, very reliable, very consistent tuning tool. All of these are important when working with any type of racing application or when tuning EFI systems whether aftermarket, chipped OEM, or piggyback. Additionally, the dyno is portable which makes it perfect for a racing team.

 

Geoffrey, what's a 91 Clubsport?

 

 

Was that a US Porsche offering? I know the RoW 968 CS's weren't available here.

 

I'll try to scan and post my graph tomorrow.

EDIT: Crap - I forgot about this. Now it will have to wait until I'm back from Pocono on Monday.

 

I'm even more confused now! If the 'gain' value is the ring and pinion value multiplied by the fourth gear ratio then my car should have a gain of 4.2 rather than the 4.5 on the chart. (3.778x1.111). The 944T gain is correct at 3.375x1.034=3.49. Does this have something to do with my suspect reading? And I have a chart that Rick gave me of another 968 that was dyno tuned, and the gain on that was 3.6--but that also had a peak torque at almost 5000rpm instead of 4100.

 

<a href="http://dan.type944.com/dandyno1.gif" target="_blank">http://dan.type944.com/dandyno1.gif</a>
<a href="http://dan.type944.com/dandyno2.gif" target="_blank">http://dan.type944.com/dandyno2.gif</a>
<a href="http://dan.type944.com/dandyno3.gif" target="_blank">http://dan.type944.com/dandyno3.gif</a>

 

OK, better late than never. Here is the graph for my stock '94 968 coupe:

<img src="http://boards.rennlist.com/upload/mike_graph.jpg" alt=" - " />