Mark_ZZ3

Been tuning a 305TPI at WOT using TunerproRT and Autoprom for data logging.

What factors affect the Spark advance? I did two runs 0-60 and the spark curve in the data log is quite difference. One run 1, the spark from 2000-5000 was a steady climb 31-39 (which saw some retard show up). In the other run, Spark was flat in the 30-31 range for the 2000-5000 RPM. In both case, spark during the 2-3 shift is pretty flat at 28.

I'm trying to understand what affects it. It's the spark table, plus any from PE or temp correction. I was expecting the requested spark to be the same from run to run or in first or second given the same conditions.

I'm not seeing the bigger picture here ... or maybe it doesn't matter and I just set the numbers for best and go?

Mark.

1MeanZ

iTrader: (9)

Usually my timing is repeatable. Did IAT or coolant temp change from run to run?

Mark_ZZ3

If I recall temp was 217 or 213 for one and 203 for the other. I thought I removed the retard at that point to try and get consistent data. The IAT would have been different as well ... is there a table for spark vs IAT?

It must be the case that the ecm will average a number between two cell values?

Rather impressed with the 25'plus year old code.

Mark.

Mark_ZZ3

dumped the data from Tunerpro to Excel. Two runs. Same 0-60mph test.

Run 1
- TPS 3.94
- Temp was 217.
- Fans on
- Battery was 13.7 and 13.0 at fuel pump
- Average Spark at TDC from 3600-4850 was 37 (8 data points)
- LV8 averaged 207 over same data.
- MAF averaged 184.
- O2 averaged 964
- 1 knock at 4850 with 1.7 degrees retard


Run 2
- TPS was 4.08
- Temp was 203.
- Fans off
- Battery was 13.6 and 13.5 at fuel pump
- Average spark at TDC from 3600-4875 was 30.5 (7 data points)
- LV8 averaged 217 over same data.
- MAF averaged 196
- O2 averaged 955
- 3 knock detected at 4875 with 3.4 degrees retard

The spark advance table for 192 and 208 load between 3200 and 4800 RPM are almost the same except I pulled on degree out in one spot I had a knock. But otherwise the same as I am trying to get data to respond to.

I am puzzled at why one run had so much higher spark, and the other had less (and more knock). I'm not seeing a variable.

- I assume both runs are in Power Enrich?
- The higher load in the second run had more air flow (TPS open more). Spark should be the same given the tables. The Injector base was higher overall in the second run, in response to air flow.

Any ideas on what to check?

- I do need to see why the fuel pump voltage dropped out when fans on.

Mark.

Mark_ZZ3

I'm not making much headway.

I gave up on why the two runs are different and back on getting rid of the knock at 4800. Seem to always start at 4800-up ... consistent. So I removed some spark before and at 4800. Still knocks. I'm down to 24.96 at 4400 and 22.50 at 4800 with PE at 5.98. So that would be 31 and 28 at TDC?

REading the data in Tunerpro ... here is what I am seeing.

MAF goes up along with RPM, and then drops out when the knock is detected. So knock is real, not fake.
O2 sits at 940-ish the whole way BUT with a steady slow down to 900 at the end of the 0-60 run.

Here is a weird one ... the Injector base pulse width goes up until 4100 (9.48) and starts to go down to 8.50 by 4800! Why in the heck would it go down at WOT? I know it is based on LV8 and O2, but seems like a big drop.

Seems like knock occurs when fuel drops down ... so I am thinking it is a lean condition at 4800 ... but no idea how to add it there? PE or AE?

arg!
Mark.

1MeanZ

iTrader: (9)

Additional PE should fatten it up above 4800. I'm fighting a very similar issue right now as well. With only 28* I get knock detected above 4500, the ECM pulls timing and the car lays down. Narrow band is 950ish the whole time. I'm putting a wideband on the car to actually read what the AFR is. I suspect it's false knock in my case, which I'll verify with the wide band. The plugs look good and the engine runs great everywhere else so I suspect its actually leaning out or it's false knock.

In your case I don't know how you are sure it's NOT false knock just because the MAF drops out when knock is detected. In my situation if I can prove the AFR is good at WOT, I'm going back up to 30-32* and I'll kill the attack rate of the knock sensor. I'll watch the plugs and if they look good, and if I don't hear or feel any detonation, I'll roll with it.

1MeanZ

iTrader: (9)

by the way if you're clear down to 24-22*, that's really low. I can't fathom why you'd have knock there if the base timing is set correctly. I suspect you've got a lean condition or you're getting false knock.

Orr89RocZ

iTrader: (20)

Is maf maxing out by 4800? Inj pw may go down if maf is maxed and there isnt enough pe enrichment or it just could need more g/s for that area of maf table or pe enrich vs rpm needs increased. Sounds like it leans out there but a 940's mv is usually very rich

I usually dont mess with maf tables 3-5 for wot. I play with pe enrich vs rpm

Mark_ZZ3

Hmm ... all good info and advice. Thanks.

I'm digging back into my INJ constants and the default pulse width. Moving to 22lb injectors over 19lb may not have been as smooth as I thought. At first to get the idle to work, I had to move the MIN BLM down to 100, and then moved the INJ constant up to 28! Seemed odd ... but it idled and BLM when running was better. A small change to the first couple MAF tables and the BLM was running at 125-128. Now I want to bring the INJ constant down as I suspect that number is used in other calculations.

What I figured out now is there was more to set with the larger injectors. A19 lb injector runs a default pulse width of 1.602 and a 22lb has 1.495. As well the cold start pulse is different. I'm going to try a new tune with setting the injector constant back down to 22 and see along with the default and min pulse width. If my mistake was setting the injector constant too high, then a lean condition at WOT makes sense. I am taking a shot in the dark here ... but one run being way too rich should prove the lean condition issue. I do recall the 4800-up issue only showed up when I got the BLM at idle up to 128.

The knock or false knock. If I read the MAF data as a "result", then it appears that the MAF false off dramatically when the knock counts go up. Now is that due to the spark being pulled out ... I'm going to say no as there are other places in the data log where I saw knock (minor accel run) and the knock was 1, spark removed was .9 and the MAF was smooth. So it might be the MAF drop is due to knock. Seems if it was false knock, at 4800 RPM under load, the engine should still be pulling air in ... real knock would create odd cylinder pressures.

Id have to say ... this has been an interesting tune. Learning lots.

Mark.

Orr89RocZ

iTrader: (20)

Yeah i never fudge the injector constant much, usually keep it around what the injectors are rated for. Use maf tables and pe mode enrich and open loop based tables to dial in fuel

1MeanZ

iTrader: (9)

Instead of fudging the injector constant, some tuners adjust the battery volts vs pulsewidth table.

Mark_ZZ3

Got some test done today. Set Inj constant back to 22. Set default PW for a 22lb injector. Set cold start PW for 22lb injectors. MAF tables 1-3 are still tweaked down.

It runs rich as expected. BLM's are in the 100-105 range at idle ... but not my issue today. Trying to solve WOT.

Did a test run ... felt better. Not as flat at 4800. So I put back in spark up to 27 on the table (which is around 32-33 with base timing in). Still conservative. It ran good. Still showed the same knock above 4800, but I did feel a "pop" at one time. So knock may be genuine.

Here is an odd observation. Two runs. Engine temp same at 203. One run saw requested spark (not TDC) from from 23 up to 38!!! whcih is 43 at TDC! Knock detected at 4900 and 3 degrees pulled out.

Did another run 30 seconds later, engine was not turned off, and requested spark was 23-27 which is 33 at TDC ... WHICH MATCHES THE TABLE!

So how in the heck did the first run add in 10 degrees??? And actually it ran quite well in the 3500-4500 with that aggressive curve.

I've dumped the data log to Excel and have gone through the values. Nothing stands out as different. It's like there is an extra Power Adder setting turned on in the first one ... or like highway mode was stuck on and adding in spark. Any ideas of where to look?

The one difference I can find is Air Ciontrol was reading as "diverted" for the first one, and on the second run is was "Not Diverted" up to 1100RPM.

Also, the first run has ZERO knock counts stored. And by the second run, there was 13 listed. So does the ECM take that into account when adding in spark? Prior knock counts = less aggressive?

This is weird why I get two different results ... Now the road I am on one way is slightly down hill (which was the higher spark). The other way is more flat.

This is so not like the old mechanical advance ... lol

RBob

iTrader: (1)

Oh, forced knock test, I've forgotten about that one. I hated that thing and was the first piece of code I removed (many years ago). Check this out:

This one:

Set it to 150* C.

Look at the parameter at LC239, up to 25* of advance can be added while the ECM is looking for knock to be reported (ouch).

RBob.

Mark_ZZ3

THATS IT! I recall there was a test to find Knock.

In the $6E (expanded) is that the ESC Diag Enable? Just uncheck it? Or is there a value?

Mark_ZZ3

In TunerproRT ... how do I access LC239? I can do the Hex Edit ...

I use to be a programmer many years ago, Assembler too. Not sure how to use TunerproRT to make the changes. Looks like I want to up the temp so the test is skipped. I got that. Now where do I put that?

Mark.

RBob

iTrader: (1)

Open BIN in TP, open hex editor (CTRL+H), go to location $237, type in FF, commit action, close hex editor.

Now, need to change any other parameter in the BIN via the normal way. So open something, type in the same value that is there (fake change), save parameter, save BIN.

After that I would exit TP, reopen it, go to the hex editor, and double check that the value at $237 is $FF.

RBob.

Mark_ZZ3

Thanks, I made the change, saved and verified!

I also tried to change an ECU file at that location. I have that element loaded now. I can see the old value was 193.5 for a 305 BIN and 203 for a 350 BIN. Now set to 304!

Was rather cool to add to the ECU file to make a XDF.

Now for the road test later today.

Mark.