PumpEff adjustment via Turbonator wiki

[wheming]

I've been over the wiki a couple times now. I was working on getting a log with AFR to try the method of adjusting the pump efficiency table. With an automatic it might be a bit more of a challenge but i think i can get some useful logs.

Some observations after my log, -
i noticed my TPSvolts do not go over 4v on the log. On the DRBII i did show close to 5v at WOT.
- I had some recorded knock (and heard it! And backed right out.) With ECU knock retarding cyl 1, 3, 4
- i am lean in boost saw 13.8 @ 10psi
- i think i have some boost creep and hit overboost (my initial cal has overboost set around 13psi during break in)

If i input the .csv export from the rpm ranges where i was WOT, and then apply those to my cal for the suggested pump efficiency adjustments, my pump eff table will be at like 99% for some reason.
Does this actually indicate too low of base fuel pressure? I am scaled for 85#/hr right now (1000cc injectors at 35psi base press).
I've upped my fuel press a few psi, and i did alter my PumpEff table to be higher. This setup does move alot of air. I can tell this is going to be alot of fun when tuned!

[wheming]

Meant to add, for safety factor sake i did shift the curve a bit higher and out. Will try to get a log on the way home with this change in effect.
(Should have taken a pic before saving so you could see the significant shift in curve.)

[ShelGame]

You really have to make sure the data you are using is 1) WOT, and 2) Steady State. The O2 controller and the transient fueling will really mess with that tool. If either of those are active and you adjust the PEFTBL based on the fuel readings, it can really send you in the wrong direction.

Also, make sure the AFR settings in the AFR tool are right. There was an issue previously where the displacement wasn't saving correctly.

[wheming]

I still need to go through some new datalogs, but making the change i did above helped alot.

I had been taking it easy and hadn't done any WOT run before the other day when i found it lean. Part throttle seemed good.
So now, after the changes, the afr's are safer. I'd rather start on the richer side because it is easier to take away fuel than put a piston back together.
I took logs of the van running to an errand, so i need to find the WOT part and trim down the csv file for just that. I could post that file.

I'm not sure how you can ever get "steady state", as i was building boost during the run.
I am only importing the values when tps is at wot. (Oddly though MPScan shows a max TPSv as 4, not 5. The gauge settings are ranged 0-5v and the DRBII shows right near 5v at WOT)
I think i hit an overboost, is that what bit2 is? I've got a factory tach that doesn't seem to display accurate. I'll try to look at that when i have the cluster out for re-bulbing with better leds.
I saw some other turboflags, i'll have to see what those numbers were.

[Force Fed Mopar]

They way I always do it, is get the revs to 2500-3k and nail it, and watch the AFR's. If they are right when you spool but lean up top, then raise it in those rpms. If it's richer then you lower it. Basically you're just adjusting it so the AFR stays the same throughout the rpm range. Obviously it's better to start this at minimum boost

If it's lean everywhere, or rich, then you might need to rescale the injectors slightly, or scale your fuel tables up or down. If you right click a table, you'll get a menu box, in it is an option to scale the entire table on either axis. The Y axis is what you want to change.

[ShelGame]

They way I always do it, is get the revs to 2500-3k and nail it, and watch the AFR's. If they are right when you spool but lean up top, then raise it in those rpms. If it's richer then you lower it. Basically you're just adjusting it so the AFR stays the same throughout the rpm range. Obviously it's better to start this at minimum boost

If it's lean everywhere, or rich, then you might need to rescale the injectors slightly, or scale your fuel tables up or down. If you right click a table, you'll get a menu box, in it is an option to scale the entire table on either axis. The Y axis is what you want to change.

I would add that you need to setup the AFR tool for a constant AFR, and set your fuel tanks to match it before you time the PEFTBL. Otherwise, you run the risk of getting lost in changes.

The stock fuel tables do not call for a constant AFR in boost. They richen the AFR with increasing boost.

[wheming]

You meant "tables" not "tanks" right?
I did change the AFR tool to get to the 11.50 quicker since my minimum boost appears to be around 10psi. I wish i could have about 5 for the minimum.

Anyway, I'm probably not approaching this right, as i was following the wiki, and it seems some of the info in this thread is not there. I wonder who the closest turbonator cal tuner expert is to me.

I did a WOT run in 3rd, and when i look at the log i see i am getting a couple deg retard in 2 cylinders, even though the knock volts are less than the threshold. Also i seem to only get a max of 4V for a TPSv reading. I might need to look at some wiring now that i have the FSM diagnostic book.

Mini_WOTtest_5-4-15_PEtestC.csv:

https://docs.google.com/spreadsheets...p=docslist_api

[ShelGame]

You meant "tables" not "tanks" right?
I did change the AFR tool to get to the 11.50 quicker since my minimum boost appears to be around 10psi. I wish i could have about 5 for the minimum.

Anyway, I'm probably not approaching this right, as i was following the wiki, and it seems some of the info in this thread is not there. I wonder who the closest turbonator cal tuner expert is to me.

I did a WOT run in 3rd, and when i look at the log i see i am getting a couple deg retard in 2 cylinders, even though the knock volts are less than the threshold. Also i seem to only get a max of 4V for a TPSv reading. I might need to look at some wiring now that i have the FSM diagnostic book.

Mini_WOTtest_5-4-15_PEtestC.csv:

https://docs.google.com/spreadsheets...p=docslist_api

You have to remember that the ECU re-scans all of the data every 11ms or so. You'll never log that fast. So, it's possible to have a knock event that can't be logged. Especially if you are seeing just 2 deg of retard. That means the threshold was exceeded, and then knock went away almost instantly.

[Force Fed Mopar]

FWIW mine runs better at 10.6-11.0 AFR in boost higher than around 10-12 psi.

[wowzer]

You have to remember that the ECU re-scans all of the data every 11ms or so. You'll never log that fast. So, it's possible to have a knock event that can't be logged. Especially if you are seeing just 2 deg of retard. That means the threshold was exceeded, and then knock went away almost instantly.

your log shows you are getting a full sample about every 250 mSecs. i would have thought it would be faster than that unless you have the logging rate set to 250..... typically i can get a byte back from the ecu every .3 - .4 mSecs so a 20 byte sample should only take about 80-100 mSecs. i'ld be interested to see your actual log file :-)

[wheming]

Yes, i see the logging rate was set to 250mSecs from previously (when working on logging on Slugmobile and were seeing erroneous reading at times for presumed failed samples).
I will set the value down to 80, if you think I can get good returns on that sample rate. i will eliminate at least one of my gauges, since O2v is really useless using the wideband.
Here's the link for the log both in .mpl and .csv
https://drive.google.com/file/d/0B5l...ew?usp=sharing
https://drive.google.com/file/d/0B5l...ew?usp=sharing
(these are the original full log files, which have a different filename. They contain alot of unnecessary data, since I was logging the entire time I was driving around on an errand, so mostly vacuum. I think it was about at sample 1029 when the condensed log was started.)

The AFR setting, that you get to from the left menu settings, is that used as a target AFR in the cal? Or is it just there for purposes of trying to tune the PumpEff table? I did change the fuel pressure on those settings to 38, It seems the intake air temperature is not a very accurate guess at 140F, the intake is usually closer to 100 or less during these logs. I need to go ahead and get my intake temp output reading to the smec, but I don't want to be changing too much before getting this more stabilized. Here is a screen shot of the AFR settings:
https://drive.google.com/file/d/0B5l...ew?usp=sharing

[wheming]

Ok, i've taken a log today that was a WOT period, I changed the sampling rate to 80 mS as discussed earlier and changed the display to 400 mS.
I haven't even looked it over myself yet, but I did set up a google drive page with the MPTune .tpl file, the log data in .csv and .mpl, and even my dashboard layout file .mpc
I still would classify myself as a novice, so these are baby steps for me. I will try to answer any questions, and appreciate you all's input.
Here is the link, it should work and let you view and copy.
https://drive.google.com/folderview?...mc&usp=sharing

I'm going to check out the log now myself, again it was on a short trip to an errand, so there is alot of normal driving too.
I do notice alot of 'hunting" in the wideband when I am going at very low speed when slowing. I'll look for that specifically in the log to point it out, it may be something that fixes itself once other things are correct. Also the car's wideband AFR moves much more with the 0-5V output to the ecu as opposed to the stock O2 sensor or the 0-1V output. Maybe something in that transfer function is not tight enough?

Oh, I remember I got an overboost or maybe over rev in 2nd at one point. Will have to look at that too.
ok, off to look at things.

[wowzer]

looking at the log info recap it looks like it can get a 21 byte sample about every 104 mS so just set the logging rate to 100 and maybe set the display rate to 250 mS so its a little smoother.

[wheming]

I spent alot of time last night looking over the log, and my cal, and the help sections of MPTune and the older how to's.
I did discover that i don't have things properly set as a basis.

In MPTune the AFR tool used for injector scaling, that should be set up to match your actual injectors, correct? And i guess you could do it two ways.
I have trick flow 95#/hr @45 psi injectors. But i have them set up as 85#/hr since running at 35psi. The initial cal had the injector scaling set for 85#/hr but the rated fuel press and actual fuel press was set to 55psi. It was mentioned to make sure the information in the tool was correct, so i started to.
But, if i want to have the tool be accurate, i should select 95#/hr injectors, set rated fuel pressure to 45psi, and set actual fuel press to 35psi. Then, as i make adjustments to fuel pressure, should i update the information in the tool?
How sensitive is the adjustment made for the air charge temp in the tool? Default in my cal was 140F but i was going to lower it to a closer number to actual, like 110-120F.

So, i'll probably go through and reset all the AFR targets through the MAP range, and then set up my base pulsewidth table for WOT. I think with my AFR closer to 10.5 i hope to see those timing retard events go away.

I wish i could have lower boost for the PumpEff setup. I thought about disconneting the arm from the swingvalve, but don't think i want to since it is just so tight in there with this turbo and on the TU header.

Got a long way to go, but i think I'm capable of learning! The recipes were alot simpler 20 years ago!

[wheming]

Ok, tried to do some more testing on Mini in hopes to develop some improvements to the tune.
Problems that i see that need some tweaking:
-Cold starts are too rich. Runs like 10:1 when starting on a 60F day.
-when in boost ( my lowest direct line to actuator yeilds 10-13psi) i get some instances of knock retard, and the AFR is running a bit higher than i am targeting (runs like lower 11's and i want about 10.5 in boost above 10psi)

I know the PumpEffTbl needs work to help get everything else on the right track. Today i thought to pump my wastgate actuator full open, in hopes to get a WOT run to redline in 2nd with minimal/no boost.
I found that this turbo will creep even with the wastegate wide open. (Using a GT3076 with a .63A/R chrysler flanged turbine housing)
I think i should have gone with the external wastgate.

Anyway, my run yeilded some values up to the point i hit an overboost (I've got the overboost set low while tuning). I only made it to like 4600rpm.

I took the values and imported them into the wideband tool in MPTune and here is what got plotted:

Does that look right?

I have this for FuelFullThrottle :

[wheming]

Another similar datalog import into the tool showed this, which is similar.


But shifted the curve lower is going to have the effect of lowering the fuel, isn't it?