Again I agree, I acknowledge this, and then it is in the process of tuning the car that you are accounting for these other variables (consciously or not) that aren't reflected when using fuel temp, i.e. efficiency of your setup, ic heatsoak, environment, etc.

Edit: I think you were getting at this now that I re-read your post. I'll restate anyway.

 

I have logged UICP temps vs Intake temp at the inlet (which Eric, I realise is what its for).

That on a good intercooler when it is moving air and not heatsoaking was within 9*.

Mychailo showed 64* inlet, 3.5" intercooler core, 26.6psi of boost, 73* UICP.

I showed very similar with my 3586 and 4", and fuel temp was in the middle. I did not have a true thermocouple in the intake pipe. I can do that I suppose but it defeats the purpose of why I do it the way I do other than for the knowledge. I just used a brand new infared on the surface. On Jeff's 1G racecar we do have pre and post to see how well the 6" intercooler works at 48psi. This is an extreme case, but at a 95* inlet it was 128* at 48psi in the UICP.

Jay,

Our MAF is good but not that good. Take a well tuned car and throw an Intake manifold on that works (i.e. Magnus) it will lean out. Why? Volume in thru the turbo at a given boost level doesnt have to go up. How the engine uses what it consumes does. I learned that 2 years ago when the Magnus first got released and went on Billy@Englishracing's landlord's car (Spaceballs I know).

 

DSMlink ignores it after a certain calculated Hz level yes. AEM ignores it under preset ranges or however you configure it for that matter. I have tested this in extremes, I am being 100% serious when I say 16*F to 106*F. AFR was 12.0 at both times on the same car. On 110 it made 647 both times with no changes and when we put Q16 in it and turned the boost up it ran 142mph in 90*+ heat without any tuning.

Anything is a band aid really. Its how big is the band aid vs the wound? I mean sometimes I have used band aids when I need stitches but I have gotten stitches when a butterfly would have been plenty.

I am agreeing with you, I am just trying to say there isnt a perfect system because entropy doesnt allow it. The UICP is the "best" solution but still has issues because it heatsoaks on sitting. Mine assumes a very good intercooler setup and relies on the heatsoaking to work perfectly. If you ran SD on an Ebay intercooler (like a lot of the hondas we see) it becomes obvious that it doesnt work that way.

 

I never did publicly thank Sean@FFTEC for telling me I didnt know what I was missing out on with SD.

Thanks dude

Aaron

 

Yep, Aaron, your method is just another way to skin the cat. If your true IAT temps post IC are very consistent and deviation from fuel temp is very consistent, then it will work without issue.

It's just not for me...I'm sort of a data junkie. I need the data to be right, for reference to other cars, other data, or even other logs from years past. I just feel better having the ECU know the true mass airflow when it counts.

But, there is more than one than way to get to a common goal. Your approach is just another. Thanks for sharing...I was never bashing it of course. Just making sure people understand what it's doing.

 

I don't completely agree with this, though.

The Mitsu maf based system does take VE into account. The reason you were leaning out when testing Magnus intakes on a MAF system was from other reasons, such as possibly the intake pipe being used, etc, etc.

The Mitsu MAF measures volumetric airflow. If the Magnus intake increased VE, the the volume of air the engine was using per time would increase, causing more air to be pulled through the MAF and measured. A properly functioning Mitsu MAF system will always account for changes in VE because of this. Any AFR changes you see are from completely different reasons altogether (burn rates, timing, tuning, other parts, etc).

SD on the other hand, doesn't take VE into account, since we are giving it the VE to calculate a mass airflow, and needs to be retuned for any mod that changes VE, such as an IM, cams, etc.

 

We are on the same team . I am posting this since recently there have been ALOT of questions about how we do SD, what is it if its not real SD, how does it compare. You guys have not let me down in making sure the proper labels have been applied and in your case the proper formulae.

In my car and Mychailo's it seems to be very consistent and his was on a stock appearing turbo which should be hotter for a given pressure ratio anyway depending on where a BBK runs at 26.6psi

 

That was constant since it was still on a MAF, it only changed the coupler to the turbo. The only change was the UICP was chopped to meet the angle of the Magnus. I will look at the thread to see what else might have been different but I am positive that was as back to back as we could get it. 441whp before at 28psi, 474whp at 28psi after.

 

I was still editing my post as you were responding, but there are various (too many too name) reasons as to why the AFR changed when testing the Magnus. But, the Mitsu MAF system definitely takes VE changes into account. That was the point I was trying to address without having to write a 2 page answer.

I simply stated something like an intake pipe, since that is one thing that could throw off a MAF. But, it can get much deeper, such as the new VE needing different timing advance, causing different burn rates, etc, etc, etc. Difference in timing does affect AFR, for example. It's obviously a complex system and gets into the true engineering of an engine.

I agree, Aaron. This was just a good thread and I jumped in. You're doing a great job. Just like discussing things like this.

 

Aphpa-N is fundamentally different then "speed density" in that it is more of a direct look up system. Alpha-N is meant to be the most responsive to driver input and can use all kinds of compensation on top of the TPS vs. RPM and still be alpha-N. It is also largely used because of unstable pressure signals that high overlap cams and ITBs give. Thus the idea is that at a given TPS and RPM it will make a % of what ever the maximum torque is. That maximum torque level is then based on things like IAT, plenum pressure, coolant temp, and other variables that can be more slowly updated without causing a problem. Thus the ECU can focus more energy into calculating changes due to driver inputs.

Boost Comp, VE Comp, Hyper Boost Comp, blah blah blah they are all versions of "speed density" just different ways to skin a cat.

Boost comp is mostly popular simply because it is very quick to setup a fuel map that is very close for all boost conditions. Should you over boost beyond where the motor is tuned for, it is likely fuel will be close to what is needed. On a straight IPW based map by contrast, if you haven't programmed the pulsewidth at that pressure and RPM point correctly, the fuel won't be any where close.

VE comp is similar to boost comp, but typically separates out VE tuning from AFR tuning and it automatically accounts for density changes due to plenum pressure and IAT. This IMO is my favorite system as it allows you to focus on tuning the AFR separately from engine VE so if you make a change to an engine part, it is pretty simple to remap the VE curve with just matching the logged AFR with the target AFR table. If I ever get around to writing my own 3D SD patch, this is how I will set it up.

You say Potato...

 

subscribed! thinking about going SD myself. just trying to get as much info as possible prior to making a decision.

 

very good information in this thread.

 

x10!

 

This is a great thread Aaron. Thank you for sharing with the community about SD. It is more complicated then a lot of people think. You don't just throw a 4 bar on and go to town.

My brain feels a bit mushy now...

Mikey

 

Would you guys like me to give that a shot?

I figure I can just use the rpm and load axis that the tephra rom is using should be trivial in that regard to get a 3d ve map just need to repoint the map and function to use the get byte map function instead.

If any rom was to get this done what version would you prefer? 9417 or the 9653 variants?