sharkster

Hi marty.. look at the GIAC dyno and it's smooth. Look at the PSI one at 4500 rpms and see the "kinks" and dip? It should be a smooth flat curve and ramped up by that point. Clearly something isn't right... and either the AFR's are too high and it's pulling timing and killing power (would be good to see the AFRs) or it's what Stephen said earlier. It's hard to say why it's doing that but it's a software/programing issue....

TB993tt

Motronically controlled engine dyno curves are very often NOT smooth, have a look at these Cargraphic curves for the 996tt, they are taken from engine dyno sheets from RS Tuning:

http://www.cargraphic.de/stepone/dat..._P96Turbo2.pdf

I don't believe much of what is shown on chassis dyno curves, but what would be the explanation of why the chassis should show smooth runs and engine bumpy and (presumably) following knock ?
Edit
Having looked at Stephen's curve, it finishes at ~6750rpm whereas the RS engine ones are fully mapped to over 7000rpm. Is it that maybe to accurately control the heat (knock) in the engine to allow power for another ~500rpm requires the constant accurate mapping which results in the bumpy knock following curve

PorschePhD

If you look at the 4700RPM mark you will see the chart drops down almost 50 ft lbs of TQ from the initial jump then come up some but never really recovers. On the HP side it drops as best as I can tell about 35HP then tries to nose back up. This indicates that the car has pulled back something because it is not happy. It is the tuning because it is not present in the baseline. So the dyno can be ruled out. Octane can be an issue, but as seen on my dyno this is also on 91 and no dips are present. Overshooting or over requesting the frequency valve will send the system in to a protect mode hence pulling timing, boost or shutting the throttle body down.

PorschePhD

TB, ours stop at 6750 because that is what the base line was done at. The comparison was done exactly like what the baseline was done as. The curve can be smooth if the request is sent in before that targeted RPM. This is how we also tune the EFI cars but also steady state. The key in tuning is knowing what to request BEFORE it happens. Not to be misleading, the type of dyno will also cause overshoot. The inertia of the Mustang and Maha and so on will not create this situation.

TB993tt

Clear as mud

So are you saying that you don't map the ECUs over 6750rpm ?


I understood the key was setting parameters according to what was happening under real time under your set conditions ie set revs, cooling etc

Why overshoot if you are doing "steady state" tuning/mapping - is this not how the Motronics are done -even more

PorschePhD

No. For the run shown I stopped where the baseline was for the owner. You asked why the stop at 6750. The software is tuned for 7300. I will not run them there on stock rods though. 7K is the normal stop.

Again, you are misunderstanding me. I do not set parameters to overshoot anything. You can see that from the graph as well. We could have an hour conversation on tuning alone. The idea is you don't make a change if it is needed at a specific RPM. So if you knock at 4500 RPM you don't add fuel at 4500RPM. You need to be making changes before then to combat combustion temps so on. You watch temperatures, knock, and TQ as well as AFR. Although by this point the AFR has been set and it is the timing you are adjusting along with the boost request.

Is that any better or still mud?

TB993tt

Mud is diluting

I presume your watching torque 'cos that is what you are trying to maximise ?

If you have set the parameters on fuel/timing/boost to be within your temp limits (determined by knowing when knock will start to occur) at 4250rpm, then surely you just do the same at 4500rpm and the torque falls where it will - then carry on in your ~250rpm incremenrs to 7000rpm -join all the torque numbers together and get your curve which may be smooth or like the RS ones may be bumpy according to those temp limits.

The way you describe it doesn't sound to me like the "steady state" tuning I (think) I'm reffering to ?

PorschePhD

Correct, you maximize TQ not HP.

Well sure. You step each cell you want to tune and check until you have completed the map. You steady state each cell. First AFR, then timing then boost and timing again. Then sweep to see what the curve looks like. You will create different heat cycles with steady state and sweeps. It is not about one type of tuning, but several. When you are done you take it off the dyno and drive it on the street. You then fill in the cells as you are driving to make smooth transistions on and off idle. Lastly you tune the cold start and start up parameters. It is a long cycle to complete.

So in answer, yes I pick every 100RPM and hold it, tune then step to the next 100RPM, tune and continue until all the cells are filled in. Watching TQ as my tuning factor ONLY after the AFR and fuel cells have been set. Max TQ will still dictate ideal conditions. Not HP. So when you do all of these things you have all the data to set accordingly.

Woodster

Thanks, I see the dip (sounds like it could be a variety of things causing it),
and I am starting to understand some of this...

MK

johnyg

Gentlemen,
I appreciate all the comments regarding tuning principles and dyno comparisons. I also appreciate the civility in which this post is ending. There are many posters and companies stating comments on here and some are informational and some may be marketing. How does word of mouth get out if one can't talk about it? I think one reason there are several posts about exhausts and the PSI 550 kit by 969 is they are one of only a few to be on the street with it so far. PSI cars are much fewer in number and they (969) are trying out what they wish to market and be able to discuss expectations of the product. I have a PSI 480 car with an EP2 quiet and it runs very well. I have not affiliation with PSI other than being a satisfied customer. Neil at Orton has had several informative conversations with me as has Sharkey (plug reccomendations). I guess what I am saying is this is a company that is trying to break in to a difficult market late to the party. I see how some would be skeptical about the sellers of the product being the ones to praise it. I do think that they are some of the few running the bigger kits though. Most everyone who seems to have done the 480 kit seems happy as I have been. I am glad to have acces too all of technical expertice on this board. Lets keep evaluating all products out there and lets be honest about what we like and don't like. This should include service issues also. If I had to discover all of the pitfalls you guys have kept us from on my own I would have gone broke and sold my car in the process. THanks again, lets make 'em keep getting faster. 996TT's rule!

sharkster

I forgot to add- that's an EXCELLENT and I mean EXCELLENT book. OK it's not "exciting" to read but man does one learn a lot from it. Why wait till summer- go out in the snow with Stephen

Kevin

Kirby, Speedaddy, Robyn, I think that a few of us have mentioned that there might be a problem with boost and or fuel/AFR with the car that ran the dyno sheets. With that said, can you post the data logging that came with that dyno run? Didn't the dyno operator post recently? Have him toss up the data sheet that goes with this car and it will answer all of our "question" we will be able to see the boost vs when the ECU started to pull back the boost, or check to see how the AFR's are doing??

sharkster

I'm not sure if the software from PSI is an overlay or and "over-write" as such? I mean is the stock knock supression still in tact? With the "overlay" type Flashes all that "good" stuff is retained to protect the car from detonation and such. I just hope that the AFR's are ok and that the tune was really done for 91 and not something else from Europe. I noticed that one of the service managers from the dealer said he had to replace the motor for one of the PSI480 cars?!? I'd feel a lot safer knowing what type of approach the PSI software uses. I feel very comfy with Garret's stuff because he just doesn't get knock etc... He retains the factory knock supression tables and works with it rather than against it like Unichip or others did. I'm absolutely no expert, although I've seen hex tables (dunno what to do with them) so I'm sorry if this comes across confusing. I'm hoping either PSI or our panel of experts can PMAITK (put my *** in the know).

Oak

looks like timing is pulling back ...could be heatsoak.

sharkster

nod, I hope it was pulled back assuming the AFR's were not safe?