Pennock's Fiero Forum
  Technical Discussion & Questions - Archive
  some 3.4 tdc ecm confusion

T H I S   I S   A N   A R C H I V E D   T O P I C
  

Email This Page to Someone! | Printable Version


some 3.4 tdc ecm confusion by caddyrocket
Started on: 11-13-2005 08:26 PM
Replies: 16
Last post by: caddyrocket on 11-17-2005 09:48 AM
caddyrocket
Member
Posts: 651
From: OK
Registered: Nov 2003


Feedback score: N/A
Leave feedback

Rate this member

Report this Post11-13-2005 08:26 PM Click Here to See the Profile for caddyrocketSend a Private Message to caddyrocketDirect Link to This Post
I've got a couple issues I'm trying to work out. Hoping one of you guys can help. I've got a 91 tdc which came with a massively hacked wiring harness and an ecm with 93 Lumina 3.4 DOHC written on it. It's a 16149396 ecm and has what should be bin BCDY written on the same label. good to go right? Nope. The memcal cover has AXSJ written on it. Clearly not the same bin. According to TunerCat, AXSJ requires a $A2 code mask. 60degreev6.com, says it's for a 3.4 (not specificly DOHC) 91 W-Body and uses a $DF code mask. No matter. I just burned a 29C256 with the manual trans BCFA bin. Plopped that sucker on the memcal and installed it. All good except the engine won't run.

Some interesting notes: With the BCFA bin, the fuel pump will activate for the 2 second prime. With the 27256 that was in it with the AXSJ bin, you can hear the fuel pump relay click off after 2 seconds but the pump doesn't come on. Weird for sure. I'm pretty confident in the wiring as I used the Fiero mated to the DOHC connectors were applicable. I checked and rechecked all the pinouts before installing it.

The car cranks and will start but doesn't run. If it stays started, full throttle is required to keep it started. Stays are around 1000rpm. It appears to be getting steady and well timed spark. The cam timing is stock. It has a fresh cam belt, tensioner, and idler pullies.

So my first question is do I need to get an entirely different memcal to work with? (IE are they application specific?) Or can you plop any program on any memcal and it will work? I'm assuming there must be some differences between memcals for different applications evevn if it's using the same ecm? Does anyone have any recommendations?

IP: Logged
PFF
System Bot
TK
Member
Posts: 10013
From:
Registered: Aug 2002


Feedback score: (2)
Leave feedback





Total ratings: 200
Rate this member

Report this Post11-13-2005 09:21 PM Click Here to See the Profile for TKSend a Private Message to TKDirect Link to This Post
BCC year is 1991

BCC= BCDY1694 Scan id= 1421 Part number= 16181394
Release date= 06/05/92 Engine size= 3.4 Trans Type= Auto trans
ECM/PCM: ECM #16149396
Used in cars: Chevy Pontiac Oldsmobile
Possibly used in:
LUMINA LUMINA EUROSPORT LUMINA Z34
GRAND PRIX LE GRAND PRIX SE GRAND PRIX GT
GRAND PRIX STE CUTLASS SUPREME SUPREME INTERNAT'L
CUTLASS SUPREME SL CUTLASS SUPREME CONV

Options:
With LQ1 3.4L GAS 6 CYL (3.4X) SFI OHC V6 HO
With M13 AUTO 4 SPD (4T60-E)
With NA5 FEDERAL REQUIREMENTS
With QGW ALL P215/60 R/16 N BL R/PE ST TL HWY
With QXJ P225/60R16/N NL R/PE ST TL AL3


BCC year is 1991

BCC= AXSJ2368 Scan id= 2391 Part number= 16162364
Release date= 11/09/90 Engine size= 3.4 Trans Type= Auto trans
ECM/PCM: ECM #16149396
Used in cars: Chevy Pontiac Oldsmobile
Possibly used in:
LUMINA LUMINA EUROSPORT LUMINA Z34
GRAND PRIX LE GRAND PRIX SE GRAND PRIX GT
GRAND PRIX STE CUTLASS SUPREME SUPREME INTERNAT'L
CUTLASS SUPREME SL CUTLASS SUPREME CONV

Options:
With LQ1 3.4L GAS 6 CYL (3.4X) SFI OHC V6 HO
With M13 AUTO 4 SPD (4T60-E)
With NA5 FEDERAL REQUIREMENTS
With QGW ALL P215/60 R/16 N BL R/PE ST TL HWY
With QXJ P225/60R16/N NL R/PE ST TL AL3


BCC year is 1991

BCC= BCFA1705 Scan id= 1441 Part number= 16181397
Release date= 06/05/92 Engine size= 3.4 Trans Type= Manual trans
ECM/PCM: ECM #16149396
Used in cars: Chevy Pontiac Oldsmobile
Possibly used in:
LUMINA LUMINA EUROSPORT LUMINA Z34
GRAND PRIX LE GRAND PRIX SE GRAND PRIX GT
GRAND PRIX STE CUTLASS SUPREME SUPREME INTERNAT'L
CUTLASS SUPREME SL CUTLASS SUPREME CONV

Options:
With LQ1 3.4L GAS 6 CYL (3.4X) SFI OHC V6 HO
With M27 MAN 5 SPD GETRAG 84MM 2.46 1ST

[This message has been edited by TK (edited 11-13-2005).]

IP: Logged
caddyrocket
Member
Posts: 651
From: OK
Registered: Nov 2003


Feedback score: N/A
Leave feedback

Rate this member

Report this Post11-13-2005 09:52 PM Click Here to See the Profile for caddyrocketSend a Private Message to caddyrocketDirect Link to This Post
Thanks for the info! If I understand correctly, all the LQ1 memcals are the same? I verified the BCFA burn on the 29C256 twice. It's setting code 51 among other things. It also set 33 which is for the MAP but I verified the map wiring and replaced it with a known good map. It leads me to believe it's memcal or atleast ecm related. Is it normal for the tag on the ECM to have a different bin than the memcal?
IP: Logged
TK
Member
Posts: 10013
From:
Registered: Aug 2002


Feedback score: (2)
Leave feedback





Total ratings: 200
Rate this member

Report this Post11-13-2005 10:42 PM Click Here to See the Profile for TKSend a Private Message to TKDirect Link to This Post
Yes, all LQ1 mem-cals are the same for 91-93 and a different one for 94-95. I hope someone didn't take a 94-95 for 6401 ECM and add the 91-93 code although it should run but I seem to remember the knock filters were different. Frankly, I don't that that is your problem.

When GM builds the car, the ECM has a tag with the mem-cal BCC code (four characters) on it. If there is an update the mem-cal will have a newer BCC and so it won't match. BCFA is the last update from GM for the 91-93 LQ1 manual trans. No matching doesn't suprise me.

Since you are getting malf codes, the program in the mem-cal is working correctly. Obviously it could still be the wrong one.

I'm sure Darth Fiero will chime in and fix you up! You might want to PM or email him.

TK

Edit - Malf 51 is a checksum error. Make sure it is correct or the experimental byte is set. Also I don't recall ever seeing an A2 production code (mask) for a 91-93 DOHC. That doesn't mean it's not correct but in the many I'm worked on I haven't seen one. All have been DF unless very early on GM used a different program and cal tables.

[This message has been edited by TK (edited 11-13-2005).]

IP: Logged
caddyrocket
Member
Posts: 651
From: OK
Registered: Nov 2003


Feedback score: N/A
Leave feedback

Rate this member

Report this Post11-13-2005 11:49 PM Click Here to See the Profile for caddyrocketSend a Private Message to caddyrocketDirect Link to This Post
Terry - thanks again for the reply! I had hoped the mismatched BCCs would have provided an easy answer but there are often non in this work.

On the DTCs it spit out a bunch of them but some didn't make sense. It spit out several three digit codes (flashing the SES light). One of them was 7-14 another was 1-16. It had the 5-1 as noted but I'm questioning the authenticity of the codes based on the above. But I've never messed with one of these before. I've been using tunercat and HHD Hex Editor. BTW, where is the checksum located?

EDIT: Also wanted to mention, when I did manage to get the engine to run a little bit, the ses light flashed continuously seemingly indicating a the ecm seeing a serious fault (which was further evidenced by the barely running engine lol).

[This message has been edited by caddyrocket (edited 11-13-2005).]

IP: Logged
TK
Member
Posts: 10013
From:
Registered: Aug 2002


Feedback score: (2)
Leave feedback





Total ratings: 200
Rate this member

Report this Post11-14-2005 12:59 AM Click Here to See the Profile for TKSend a Private Message to TKDirect Link to This Post
A continuously flashing SES is a checksum error and the ECM is running in backup mode. I don't think the P4 ECM can flash the light without the code at least running but it will go all goofy either way.

The checksum for DF is two bytes at 8006 (8007) and is the sum of 8008 through FFFF mod FF.

The correct checksum for BCFA is 2F04. You can also just skip all that and set 8008 to AA. This makes the code ignore the checksum. It won't care what it is. I will do that while running an emulator and then let the program recalc the checksum when I am done.

TK

[This message has been edited by TK (edited 11-14-2005).]

IP: Logged
caddyrocket
Member
Posts: 651
From: OK
Registered: Nov 2003


Feedback score: N/A
Leave feedback

Rate this member

Report this Post11-14-2005 09:25 AM Click Here to See the Profile for caddyrocketSend a Private Message to caddyrocketDirect Link to This Post
One last thing that is confusing me. I've seen the disassembly for the BCFA. It also had the code mask referenced at 8008 but I was/still am confusing myself on that issue. My hex editor (indeed any that I've used) has got that data stored in 0008. In a 256k chip, how could there be an 8008? Wouldn't a 256k chip stop at 8000?

Also, if I am correct in the checksum being located at 0006 and 0007, the BCFA I had previously is bunk. It is the one that came of of 60degreev6 but the checksum is different. It's 26ec. I downloaded another bin with the correct checksum so I'll work from that. Interestingly enough, the ASXJ does have a 0008 set to a2 so I see where tunercat got that.

IP: Logged
TK
Member
Posts: 10013
From:
Registered: Aug 2002


Feedback score: (2)
Leave feedback





Total ratings: 200
Rate this member

Report this Post11-14-2005 03:58 PM Click Here to See the Profile for TKSend a Private Message to TKDirect Link to This Post
The physical location in the chip vs where the ECM looks is two different things. The ECM sees the calibration data starting at 8000 but it's physically in the chip at the start of 0000.

Since the address range grows from the bottom (FFFF backwards), the start address is dependent on the size of the chip. for a 256k eprom, the start address is 8000. The ECM wiring internally handles mapping the chip to the correct address.

I write my stuff to reflect the ECM address and not the chip address since GM references the ECM address. Minor detail though.

Terry

IP: Logged
caddyrocket
Member
Posts: 651
From: OK
Registered: Nov 2003


Feedback score: N/A
Leave feedback

Rate this member

Report this Post11-14-2005 04:22 PM Click Here to See the Profile for caddyrocketSend a Private Message to caddyrocketDirect Link to This Post
That's good stuff. I would have never guessed that in a million years. But it makes sense now because I was burning the fiero bins into the top range on the 256k chip. So ideally, I should take the corrected checksum chip and try it again. That should atleast correct the DTC 51. I don't know why the 60degreev6 version had a bad checksum though unless someone did it purposefully. When I did a compair, the checksum itself was the only thing different.

I'll give the fresh BCFA a try and see how it works and report back. Thanks again for all your help.

IP: Logged
TK
Member
Posts: 10013
From:
Registered: Aug 2002


Feedback score: (2)
Leave feedback





Total ratings: 200
Rate this member

Report this Post11-14-2005 09:25 PM Click Here to See the Profile for TKSend a Private Message to TKDirect Link to This Post
Check you email.
IP: Logged
Darth Fiero
Member
Posts: 5919
From: Waterloo, Indiana
Registered: Oct 2002


Feedback score: (5)
Leave feedback





Total ratings: 361
Rate this member

Report this Post11-15-2005 10:44 PM Click Here to See the Profile for Darth FieroClick Here to visit Darth Fiero's HomePageSend a Private Message to Darth FieroDirect Link to This Post
 
quote
Originally posted by caddyrocket:

Thanks for the info! If I understand correctly, all the LQ1 memcals are the same? I verified the BCFA burn on the 29C256 twice. It's setting code 51 among other things. It also set 33 which is for the MAP but I verified the map wiring and replaced it with a known good map. It leads me to believe it's memcal or atleast ecm related. Is it normal for the tag on the ECM to have a different bin than the memcal?

caddyrocket, I have had several problems when trying to burn chips for these 16149396 computers. They are extremely picky, and 99% of the time I try to erase and reprogram the stock chip on the mem-cal; the computer will flip out and not read it correctly, even though it verified every time I checked it.

What I had to do was either use a new 27c256 eprom or, better yet, use the AT29C256 flash memory chip and just unsolder/cut the stock chip off the mem-cal and solder on the new chip. The Atmel AT29C256 chips are nice because they don't require UV light to erase and are supported by most EPROM burners. You can also use the 27SF512 flash chips instead which are less expensive but you will have to change the "Start of Device" setting in your burner software so you load the .bin into the upper half of this larger chip so it will work properly. Also the 27SF512 chips are not supported in a lot of older burners.

------------------
power corrupts. absolute power corrupts absolutely.

Custom Computer Tuning | Engine Conversions | Turbocharging | www.gmtuners.com

[This message has been edited by Darth Fiero (edited 11-15-2005).]

IP: Logged
PFF
System Bot
caddyrocket
Member
Posts: 651
From: OK
Registered: Nov 2003


Feedback score: N/A
Leave feedback

Rate this member

Report this Post11-16-2005 06:30 PM Click Here to See the Profile for caddyrocketSend a Private Message to caddyrocketDirect Link to This Post
Darth - Thanks for the reply. I've been trying to use the AT29C256. But I didn't have to unsolder anything to replace the chip. I just popped the stock chip out of the socket and replaced it with the flash prom. I seen that the bcfa I was using had the wrong checksum. I haven't tested it yet as I was in the ER until this morning. 3mm kidney stone. 3 morphine shots and some percisett fixes all pain lol. By the time I left the hospital, I could have been run over by a train and been ok with it!

IP: Logged
caddyrocket
Member
Posts: 651
From: OK
Registered: Nov 2003


Feedback score: N/A
Leave feedback

Rate this member

Report this Post11-16-2005 06:32 PM Click Here to See the Profile for caddyrocketSend a Private Message to caddyrocketDirect Link to This Post

caddyrocket

651 posts
Member since Nov 2003
 
quote
Originally posted by TK:

Check you email.


Hey Terry. Some of the attachments got blocked. Please try sending to bryan_es@yahoo.com

Again, thanks so much for sending that stuff!!!

IP: Logged
caddyrocket
Member
Posts: 651
From: OK
Registered: Nov 2003


Feedback score: N/A
Leave feedback

Rate this member

Report this Post11-16-2005 06:49 PM Click Here to See the Profile for caddyrocketSend a Private Message to caddyrocketDirect Link to This Post

caddyrocket

651 posts
Member since Nov 2003
BTW, I was going to ask.. Has anyone successfully used $8F code on the 9396 ecm?

[This message has been edited by caddyrocket (edited 11-16-2005).]

IP: Logged
Darth Fiero
Member
Posts: 5919
From: Waterloo, Indiana
Registered: Oct 2002


Feedback score: (5)
Leave feedback





Total ratings: 361
Rate this member

Report this Post11-16-2005 09:51 PM Click Here to See the Profile for Darth FieroClick Here to visit Darth Fiero's HomePageSend a Private Message to Darth FieroDirect Link to This Post
 
quote
Originally posted by caddyrocket:

BTW, I was going to ask.. Has anyone successfully used $8F code on the 9396 ecm?

I have used the $8F code in the 16149396 ECM, but it wasn't perfect. Once I got the injector constant and timing tables nailed down to what the 3.4 DOHC was looking for, the idle still wasn't completely stable. It is hard to describe, other than it was just a little unstable. I really didn't have that much time to work with it since I had other work to be done but I do know it works.

I have done some other chips using the $8F code in the 16149396 ECMs for other people but haven't heard back yet on how those are working out.

It is interesting that you said you didn't have to unsolder anything to get the stock EPROM chip out of the mem-cal. Every one I have seen has been cold-soldered in. They didn't use much solder in this process so it it hard to see but it is there. While these memcals are robust, they are not indestructable and you might have damaged one of the contacts in the mem-cal itself when you pulled the stock chip out. Again, this is assuming you are using the stock mem-cal and not one of those moates.net memory adapters.

-ryan

IP: Logged
ltlfrari
Member
Posts: 5356
From: Wake Forest,NC,USA
Registered: Jan 2002


Feedback score: (4)
Leave feedback





Total ratings: 127
Rate this member

Report this Post11-17-2005 08:15 AM Click Here to See the Profile for ltlfrariClick Here to visit ltlfrari's HomePageSend a Private Message to ltlfrariDirect Link to This Post
If you want to swap chips a lot the best device is Craig Moates little dohicky adapter which you can find here --> http://www.moates.net/product_info.php?cPath=25_36&products_id=32

------------------
Dave E

www.ltlfrari.com

IP: Logged
caddyrocket
Member
Posts: 651
From: OK
Registered: Nov 2003


Feedback score: N/A
Leave feedback

Rate this member

Report this Post11-17-2005 09:48 AM Click Here to See the Profile for caddyrocketSend a Private Message to caddyrocketDirect Link to This Post
Darth - I can see small cold solder joints on the odd white board that sits over the backup chips. But the chip didn't have these joints in place. I thought I had missed them but I looked at the chip and the memcal and there is no evidence of the joints. I should check the conductivity however. It is a stock memcal.
IP: Logged



All times are ET (US)

T H I S   I S   A N   A R C H I V E D   T O P I C
  

Contact Us | Back To Main Page

Advertizing on PFF | Fiero Parts Vendors
PFF Merchandise | Fiero Gallery | Ogre's Cave
Real-Time Chat | Fiero Related Auctions on eBay



Copyright (c) 1999, C. Pennock