Using $6E Mask with 4+3 Trans?
 

Has anyone done this? My 88 Anniversary has a bad cold start switch that I cannot find a source to replace. Leading to excessively long crank times. I had great success uploading the 89 APYP bin that reduced cranking to around 3 seconds. Just hit me this morning though that the ECM controls the overdrive function in the Doug Nash unit and that a bin from a 6 speed manual has no such control.

Any guidance out there? I'd like to do away with the cold start injector system but maintain the original control over the overdrive.

Thanks!
Matt

I know with ebl flash you can actually program the overdrive to act with the TCC lock up in the bin file. If you have those tables available in the file you should be able to to set lock and unlock tps vs speed in a table. I'm not sure if it is exactly the same. I would think those features would be there though.

The crank really fueling isn't that much different between $32B and $6E.

$6E delays the crank fueling until the ecm sees 8 distributor reference pulses and the tuning is slightly different vs coolant temp. $6E does allow larger maximum crank fuel pw since the scalar is doubled, but this can be easily duplicated in $32B if needed for small injectors.

Try logging during startup to see the actual difference in the cranking pw between the two bins and tune accordingly. This can be difficult since the data connection may be lost during transition from run to crank. Sometimes it works, but not always. I've had better luck logging crank fueling with $32B than $6E, but I usually only run $6E temporarily for test purposes, so I don't have much of my own $6E data.

1-2 second cold start cranking time should be achievable with $32B and no csi under most conditions with proper tuning.

Here's an old log for example:
https://datazap.me/u/tequilaboy/log-...-60&mark=35-19

In this case took around 1.8 seconds to fire. This log is from 2013 and was exported from tunerpro. I don't recall who's car it is or injector size details. Just an example startup log.

If anybody has data to share, zip it and post it here.

Sounds good, will give that a try. I haven't played at all with the bins as I had been planning to keep the stock 88 tune, until the cold start issue sprung up. I have a 78 with a 16197427 ECM driving a TBI system that I've worked extensively. Lots of tables in the later model 7427 to really dial in!

Keeping the $32B would be ideal as I'm not certain there is any way to link the $6E to properly control the overdrive portion of the trans.

Just too much work to add all the 4+3 control logic to a $6E bin. Could be done, but its not worth my time.

Here's another start-up example log:

https://datazap.me/u/tequilaboy/cold...357&mark=26-49

This one's a blower car with 42 lb/hr injectors. In this case, I'm varying the crank pw multiplier as a function of drp, so you can see that the crank fuel pulse width is modulated over time to reduce risk of flooding with the larger injectors.

Started my datalogging this morning. Didn't have Tunerpro all the way setup as I think the autoprom was using the 88 min while the definition file was the 89. Looking at the export log shows some goofy things. Uploaded the 89 bin to the shooting and all logged properly, although it wasn't a true cold start. I

How long should I wait to do another trial and have it accurately represent a cold start? 4 hours? 8? 12?

Also, does anyone have a copy of this 32B Extended file that scorp508 posted? Here is the link to his original posting, but hosting site no longer active. Link

Here are a few $32B related files obtained from gearhead-efi and one of my old adx files which may be helpful:

Thanks! I'm a big fan of the gearhead site. Was a treasure trove of info when I fuel injected my 78!

I found an older thread where you were giving some detailed steps to improve the crank fueling: Backport 6e to 32b

Post #6 - I'm getting tripped up with adding the scalar to the XDF file. When I use the address 0x372, it defaults to a value of 25. I'm guessing I must be using the wrong address, wouldn't it default to 6554, that I would then double to 13108?

So for $32B, the address: 0x372 is correct. It is a 2 byte value with default setting of 0x199A or 6554 (decimal). Not sure why you would see 25. Check the conversion tab. Maybe you're dividing the 100 ms (converted value) by 4.

Found the issue, thanks to your note above. The default value was 1 byte. Changing it to 2 byte went great.

Here are my steps so far, mimicing the earlier thread from a few years ago:

1) Added the Max Crank PW Scalar to the $32B XDF file using 0x372 as address and increasing value from 6554 to 13108
2) Doubled the conversion calculation on the Crank PW vs Temp table from 0.39065*X to 0.781300*X
3) Copied the values in the Crank PW vs Temp table from the APYP $6E table into my ABTR table

Now get to wait until morning to do a good test, or at least later tonight. I was playing around earlier with some tables so not a good cold start example. However, I did upload the changes above into the AutoProm and the Vette fired right up, so at least my changes didn't do anything catastrophic!

Will keep you posted on results!
Matt

Looking at the $6E multiplier table makes me think that the resulting crank pulse width will still end up at nearly the same value as $32B (after 10+ reference pulses). So looks like you only get 2 larger pulses when the multiplier is 0.5. I would like to see a $6E cranking log to confirm.

;--------------------------------------------
;Crank Fuel PW Multiplier vs Ref Pulse
;
;Val = Multiplier / 0.00390625
;
;Mult Ref Pulse
;--------------------------------------------
LC38B: .db 0x00 ; 0 1
.db 0x00 ; 0 2
.db 0x00 ; 0 3
.db 0x00 ; 0 4
.db 0x00 ; 0 5
.db 0x00 ; 0 6
.db 0x00 ; 0 7
.db 0x00 ; 0 8
.db 0x80 ; .5 9
.db 0x80 ; .5 10
.db 0x40 ; .25 11
.db 0x40 ; .25 12
.db 0x40 ; .25 13
.db 0x40 ; .25 14
.db 0x40 ; .25 15
.db 0x40 ; .25 16
LC39B: .db 0x33 ; .20 17
.db 0x33 ; .20 18
.db 0x33 ; .20 19
.db 0x33 ; .20 20
.db 0x33 ; .20 21
.db 0x33 ; .20 22
.db 0x33 ; .20 23
.db 0x33 ; .20 24

Here is a $6E log I took this morning. It was after I had gotten the engine to start (poorly) using a hashbag of bin/xdf/adx files. Have to upload the attachment as a csv according to the site...

Does that help at all?

Here is a log from my cold start this morning. Pretty pleased with the results. The car started up in roughly the same amount of time as the stock 89 APYP bin.

Something is up with the engine speed, though. The log shows it idling around 1300 but dash showed 900. Will peek into that later.

Here's a graphical view your log:

https://datazap.me/u/tequilaboy/cold-start-matt-abtr-mod?log=0&data=1-56-64-73&solo=1-73&mark=25-6

That's a lot of fuel (14.79 ms pw) and approx 2.6 seconds cranking time. I think there is room for improvement, but you're on the right track. Unfortunately the drp-based rpm signal is not present in the adx, but you can still see where crank fueling begins and engine is running.

In the previous $6E log, you can't see any injector pw before the engine starts. I suspect this is due to the zero multiplier for the first 8 pulses. Maybe it started without any crank fuel, just from residual fuel in the cylinders?

Thanks for the graph! That datazap feature is pretty slick.

I had noticed from the graph you had earlier of your results the drp. I have been trying to figure out how to get it into the adx. Been a year since I added definitions i to my 78 file, having to brush off the mental cobwebs.

I think you are correct in the $6e. Wasn't a true cold start, would likely have been residual fuel to start.

Disregard what I mentioned in #13 about the RPM data. Was correct and mirrored the dash tach exactly.

Here is a cold start log from this morning, using the stock APYP bin. It appears that the injector base pulse with spikes at 5.26ms then drops to 4.2-4.4. Would that be a good range to target with my 88 bin? It was a little over 14ms, so plenty of room to drop it, if that is the case.

Make small adjustments to the crank pw vs coolant table?

I don't think the APYP log is capturing the actual crank fuel pw. This may be due to the zero multiplier as already mentioned. Maybe set the multiplier to a non-zero value like 0.25 for the first 8 reference pulses to see if the crank pulse can be logged with $6E.

The first 6.5 ms pulse that you see (sample 16) is when the engine is already running. This is a normally calculated injector pulse (I think). But how is it starting without an early injector pulse? Maybe some residual, leakage, or csi still providing some fuel? I dunno.

My gut feeling is that you want the crank fuel pw to be about double the normal pulse width when the engine is running. Once the engine is running, you can see the pw is something like 2.5-3 ms. I think a crank pw of about 6 ms would be a good fit at this temperature.

This will vary as a function of temperature, fuel and cranking rpm, etc, so many variables to consider.