ECU - when has it "learned" enough?

[95lotus]

Well, I think I made the stupid mistake of having the battery disconnected last night, and then taking the S4S in for Arizona emissions today, which it promptly failed (2.59g/mile HC, and 40.92g/mile CO). At most, I put 50 miles on her (probably about 30) before going in for the test, and I've heard that the ECU tends to run a richer mixture while it's learning (please correct me if I am wrong).

I hooked up Any Whittaker's ALDL hardware and ran Freescan while dumping the raw communications data to disk. Looking at the data bits for "closed loop mode" and "learn control enable" from a 37 second run, I see both of them toggling on and off (although it's mostly in closed loop mode).

I'm wondering:
1) How do you know when the ECU has finished learning, or learned enough? Is it when it enters closed loop mode?

2) Does the ECU go into open loop mode when it gets to a part of the tables it hasn't "learned" yet, and then run in closed loop thereafter?

3) Is it likeley that I need to drive her some more (heaven forbid! ) under various conditions to fill in the tables?

4) Am I "done" when the ECU stops exiting closed-loop mode while driving around?

5) Is the documentation for FreeScan available for download? I didn't see it on Andy's site nor LEW.

thanks in advance,
Bill
'95 S4S

[paul c]

95lotus said:

...Is the documentation for FreeScan available for download? I didn't see it on Andy's site nor LEW. thanks in advance, Bill '95 S4S

The only documentation i know of is the LEW freescan manual. This can be purchased with or without the ADSL lead on LEW's E-shop: www.lotusespritworld.co.uk/Sale/shop.html

Well worth the 20 quid and then some, if you figure the time it will save you

Paul C

[arium]

I think you are failing emmissions due to the fact you are bouncing out of closed loop mode. When an engine is started cold it is in 'open loop' mode. That means that until certain sensors warm up and can give accurate infromation the computer takes a 'best guess' based on engine and outside temperature along with a couple other environmental factors. Once the engine (and sensors) heat up or a predetermined period of time elapses the system goes 'closed loop'. No more guessing anymore, everything goes by the sensors or it throws a code (usually).

Point is, once things heat up, you should go into closed loop or throw a code. If you are jumping off the bandwagon something is wrong. Map, coolant temp, and O2 sensors are the usual gm suspects here. If you are in closed loop and the integrator is fairly neutral (128ish) then it's time to look at the cat.

When you failed, did they give you a printout of hc, c0 and nox emmissions? Looking at those would probably clarify things further.

Steve

[Paula&Marcus]

95lotus said:

I'm wondering: 1) How do you know when the ECU has finished learning, or learned enough? Is it when it enters closed loop mode? 2) Does the ECU go into open loop mode when it gets to a part of the tables it hasn't "learned" yet, and then run in closed loop thereafter? 3) Is it likeley that I need to drive her some more (heaven forbid! ) under various conditions to fill in the tables? 4) Am I "done" when the ECU stops exiting closed-loop mode while driving around? 5) Is the documentation for FreeScan available for download? I didn't see it on Andy's site nor LEW. thanks in advance, Bill '95 S4S

1) It never stops learning ...
2) No, it goes into open loop with WOT and after start-up till the cat has warmed up
3) Yes Sir.
4) No, see 1) amd 2)
5) No, see other posting ...

Cheers
Marcus (www.PUKesprit.de)

[95lotus]

Thanks Paul, Steve and Marcus...I appreciate the responses.

I captured another run, this time to a CVS file, and she stayed in closed loop mode. I'm going to capture some more runs when the weather improves, it's actually cold and rainy today in Phoenix...not ideal Esprit weather.

Steve, the data from the emissions test is:
HC 2.59 (0.8 is the limit)
CO 40.52 (12)
NOX 1.82 (2)

BTW, the car was warmed up before I went in. I could almost feel the pain of my baby while they were testing it...between killing it a few times and the smell of the clutch burning, I was dying inside. The second driver finally got through it, but I wonder what effect it had on my results.

The integrator is varying in the range of 76-181 during the short run I did.
O2 was in the range of 0.036 - 0.955
MAP 0.14 - 1.15
coolant temp 71 - 74.8

I'm going to check the TPS next. It maxed out at only 1.33V during the run.

I bought the FreeScan manual which has helped with my understanding of the ECU. I'll read over it again and try to narrow things down.

When I get some free time, I'll look into what it would take to write some software to analyze the data.

thanks again everyone,
Bill
'95 S4S

[AZ88Turbo]

Billy One,

Check your email, I just emailed you.

TPS should be around there at fully closed I think, less perhaps (should be around 4.5v fully open), so that could be an issue.

Marcus can probably tell you better than I can.

Mark - 88 Turbo

[Paula&Marcus]

Yes, Mark is right, your MAP and TPS reading donot look right.
Please take your Freescan and recheck that. MAP should be around 800-1000 mbar (= 0.8-1 bar = around atmospheric pressure) when the ignition is ON and the engine NOT running. When the ingine is idling you will see a lower MAP (because of the N/A vacuum) and with boost you should see pressures more than 1 bar (between 1-2 bar ABS pressure). Please keep in m,ind that MAP is showing ABS pressure ! BOOST = MAP ABS - 1 !!!
TPS should vary from 0% to 100% when you press the pedal slowly.

Please report.

Cheers
Marcus (www.PUKesprit.de)

[95lotus]

Paula&Marcus said:

Yes, Mark is right, your MAP and TPS reading donot look right. Please take your Freescan and recheck that. MAP should be around 800-1000 mbar (= 0.8-1 bar = around atmospheric pressure) when the ignition is ON and the engine NOT running. When the ingine is idling you will see a lower MAP (because of the N/A vacuum) and with boost you should see pressures more than 1 bar (between 1-2 bar ABS pressure). Please keep in m,ind that MAP is showing ABS pressure ! BOOST = MAP ABS - 1 !!! TPS should vary from 0% to 100% when you press the pedal slowly. Please report. Cheers Marcus (www.PUKesprit.de)

The MAP is around 0.85 with the engine off, and the TPS varied during a run between 0 and 98 with a MAP max of 2.24.

I took it into the dealer, and it looks like the #3 cyclinder valve guides may be worn. I'm going to pick it up today, and get with AZ88Turbo to have a look. Mark's had his engine apart a couple of times, and this would be a good learning experience for me, so we're gonna have a go at it and make some improvements while we're in there.

Thanks again for all of the help here!
Bill
'95 S4S