What If? $8D (1227727 ECM) for TBI

[1project2many]

Reading the code, the value at $0293 is compared to bits 3 and 4 of a value obtained from $0032, which is a value based on a memcal read. With only bits 3 and 4 compared, this confirms the only values possible in $0293 are those previously listed; $00, $08, $10, and $18. There is a bit of additional code after this comparison but really, the only reason code 41 should set is because $0293 doesn't match those two bits in $0032. $0032 is transmitted on the ALDL stream as FMD Word 2 so you can probably view it in the ALDL and get a look at what the relevant bits are set to.

Dave, dumb question but is there any chance the code was left over from a previous test?

[dave w]

Reading the code, the value at $0293 is compared to bits 3 and 4 of a value obtained from $0032, which is a value based on a memcal read. With only bits 3 and 4 compared, this confirms the only values possible in $0293 are those previously listed; $00, $08, $10, and $18. There is a bit of additional code after this comparison but really, the only reason code 41 should set is because $0293 doesn't match those two bits in $0032. $0032 is transmitted on the ALDL stream as FMD Word 2 so you can probably view it in the ALDL and get a look at what the relevant bits are set to.

Dave, dumb question but is there any chance the code was left over from a previous test?

I used a 27SF512 chip for the testing, which was erased and flashed for each test. I used the same TBI Memcal with a Moates.net memory adapter board. I removed power from the ECM after each test for the chip swap. I verified no codes before each test.

Maybe in the near future we can regroup (build another harness) and test again?

dave w

[1project2many]

Love to do more testing. If I had time and space I'd set up my ecm bench and drive a few injectors here.

Any chance you recorded any $8D ALDL data? Maybe we could extract $0032 to see what was reported?

[dave w]

Love to do more testing. If I had time and space I'd set up my ecm bench and drive a few injectors here.

Any chance you recorded any $8D ALDL data? Maybe we could extract $0032 to see what was reported?

I think these files will have the information you are requesting?

dave w

[JeepsAndGuns]

I don't have a '7727 / '7730 harness for testing. I'll take you up on your offer to do some testing, it's greatly appreciated!

I ran out of time to do any further testing for a TBI '7727.

I used a memory adapter board with the '427 PCM TBI Memcal using a 27SF512 chip with the $8D .bin file. Resistors are not needed for testing when using Noid Lights or 4 TPI injectors on each bank. It should be Ok to not have injectors plugged in. The first test I'd like to have help with would be to set the $8D definition .bin file for cylinder hex value to the $18 Mark posted and see if the cylinder select error occurs? With my test, I got a cylinder select error with 0x293 set to hex $18.

Thanks,
dave w

Ok, to keep the thread from getting too mixed up, I will put this into this thread.

I hooked up a 7727, took a AUJP bin, and changed the 0x293 value to 18 for tbi. I then installed a stock un touched 7427 V8 tbi memcal and gave it a go. (using a moates G1 adapter) I too got the cylinder select error. I then tried my mpfi modded tbi memcal just for the hell of it, and got the cylinder select error too.
So it appears a one of a few things might be happeneing.
1. The 7427 tbi memcal is not giving the bin what it wants to see when its set to tbi (0x293 set to 18)
2. $8D can not run tbi, but someone added in the comments that it could if 18 was entered into 0x293
3. $8D can run tbi, but 18 is not the correct value
I'm thinking maybe the memcal is not giving the ecm what it wants to see.
Does anyone have one of those 7727 tbi memcals? Maybe measure what its has for resistance on the specific (unknown to me) pins and see if we could change a 7427 tbi memcal to the same resistance?

Or if you have any other things yould would like me to try, either make a bin and post it up/send to me, or tell me what to change and try.

[dave w]

I've done a bunch of reading and some testing. From the hardware perspective, here's the cyl select data. Bear with me if it seems repetitive. I'm trying to get this all together in one place.

Pin 56 of the memcal is the cylinder select pin. On some web pages it's referred to as cal56, on others it's VRFD, and in another it's called "pin 11" because it's connected to pin 11 of another chip. The voltage on this pin determines the fuel mode selected. Voltage is controlled by resistors connected between pins 56 and 58 and between pins 56 and 53 of the memcal. 58 is +5V and 53 is gnd. There are 5 modes available:

	MEM/CAL Mode Selection
Application Mode	Input Voltage (R29/R30 ratio)	Frequency Of Injection
1. Alternating TBI	2/3 VRFD + 10%	ALT. Every REF pulse
2. Single Injector TBI	VRFD + 0% - 10%	Every REF pulse
3. 4 Cyl. PFI	2/3 VRFD +/- 10%	Every 2nd REF pulse
4. 6 Cyl. PFI	1/3 VRFD + 10%	Every 3rd REF pulse
5. 8 Cyl. PFI	GND -0% + 10%	Every 4th REF pulse

Using this information and the schematic referenced earlier in the thread I went ahead and measured some memcals I have here.

ECM	CAL	APPLICATION	RFD CHIP NO	RESISTANCE TO GND (PIN 53)	RESISTANCE TO VCC (PIN 58)	CALCULATED CYL SELECT VOLTAGE
1227165		89 VETTE TPI	16133420	7.5K	INFINITE	0
1227165		1987 VETTE TPI	16055376	7.5K	INFINITE	0
1227727		3.1 PFI ENGINE		7.5K	15K	1.7
1228706	AYZD	1992 3.3 V6		7.5K	15K	1.7
16197427		1994 4.3 CPI		7.5K	15K	1.7
1227730	ADPS	2.8 V6		7.5K	15K	1.7
1228253	AJRZ	3.8 SFI		7.5K	15K	1.7
16196396		94-95 TBI TRUCK MANUAL TRANS		15K	7.5K	3.3
1227749	AMDD	1988 QUAD 4		15K	7.5K	3.3
16197427	BJYA	1995 TBI TRUCK AUTO TRANS		15K	7.5K	3.3
1227128	ARZW	W BODY 3.1		15K	7.5K	3.3
1227730	BABM	92 BERETTA 3.1		15K	7.5K	3.3
1227749	AMDD(2)	1988 QUAD 4		15K	7.5K	3.3
16197427	BPKW	4.3 VIN W		15K	7.5K	3.3
1227165	ACKD	4 CYL TRUCK TBI		INFINITE	7.5K	5
1227807	ASDD	W BODY QUAD4		INFINITE	7.5K	5
1227165	ACWS	2.5 TBI		INFINITE	7.5K	5

According to the info in the first table, 5V is single injector TBI, 3.3V enables 4 cyl PFI or alternating TBI, 1.7V sets 6 cyl PFI, and 0V is 8 cylinder PFI mode. Almost all the measurements make sense except the Q4 memcal set for "single point TBI" and the cpi equipped 4.3 engines set for "alternating bank TBI or 4 cylinder PFI." I still need to understand the relationship between software controlled firing and the cyl select pins. I realized last night I don't even remember my logic gate diagrams anymore.

JeepsAndGuns,
Thanks for testing the '7727 with a '427 Memcal.

I'm thinking the TBI '7727 cylinder select is looking for 5 volts.

I'm thinking the '427 Memcal will need to be modified so the resistance to ground on pin 53 is infinite, not 15K. I'm thinking the infinite resistance on pin 53 of a '427 Memcal will send the 5 volts the '7727 is looking for when 0x293 is set to $18.

I have a spare, '427 Memcal, but I think the Limp Home is faulty. When this Memcal was in service, the TBI injectors were dumping lots of fuel for no apparent reason. I could snail mail you the spare Memcal, once I figure out how to modify pin 53 for infinite resistance.

dave w

[JeepsAndGuns]

I have a few spare V8 7427 tbi memcals that are all workig. It would be quicker and cheaper just to tell/show me exactly what pins to connect and I will add a jumper wire to one and give it another test. I dont think a faulty memcal would be good to test this with.
Just snap me a picture and draw arrows pointing to the exact pins that need connected together and I will do it.

[dave w]

I have a few spare V8 7427 tbi memcals that are all workig. It would be quicker and cheaper just to tell/show me exactly what pins to connect and I will add a jumper wire to one and give it another test. I dont think a faulty memcal would be good to test this with.
Just snap me a picture and draw arrows pointing to the exact pins that need connected together and I will do it.

I'm working on figuring out how to get pin 53 to 5.0 VDC. The pics below are from some of my research over at Thridgen.org. I think the challenging part is 2/3 VRFD (3.3 VDC) for alternating TBI, and have 3.3 VDC to the cylinder select without an error.

Post #39 at ThridGen.org http://www.thirdgen.org/techboard/di...-resistor.html

dave w

[dave w]

I have a few spare V8 7427 tbi memcals that are all workig. It would be quicker and cheaper just to tell/show me exactly what pins to connect and I will add a jumper wire to one and give it another test. I dont think a faulty memcal would be good to test this with.
Just snap me a picture and draw arrows pointing to the exact pins that need connected together and I will do it.

The pic below is another pic from my research over at Thridgen.org.

dave w

[JeepsAndGuns]

All that my friend, is over my head....lol
When your figure out what pins need connected, let me know and we will give it a go.

[yelruhde]

I managed to run a tbi set up on a 7727 as a v6 mpfi using 8D and a vette tune by changing the values at both 0293 and 04F1 to 0C. It runs well enough to get driven everyday and does not throw error 41. Maybe changing both values to 18 with a tbi memcal will work.

[1project2many]

Dave,

Pin 53 is ground. That pin should remain at 0V. The voltage needs to be 5V on pin 56. Pin 56 of the memcal is directly connected to 8 of the RFID. Break this connection and you are free to solder resistors equal to GM's values between the remaining two pins. If you remove the leg at pin 8 then connect a 7.5K resistor between 8 and 10 you will provide 5V to the ecm at pin 56 which should enable "4 cylinder TBI mode." But now you may get a cylinder select code due to calibration set for 8 cyl and fuel mode set for 4 cyls. The easy answer to that would be to disable the error code. A more complicated answer would be to alter the code to allow this mismatch.

The 2/3 Vref +10% and -10% requirement is a fairly close tolerance value. Most resistors are 10% tolerance so you'll need 1% or 2% resistors, preferably temperature stable versions so the engine doesn't switch fueling modes in cold or hot weather. But with infinite resistance between GND and pin 56 you don't need to worry so much about the resistance to pin 58.

yelruhde, what was the memcal originally from? Was this on a six cylinder engine? Are you sure you changed 04F1 also? Why did you choose $0C?

Thanks.

[dave w]

Dave,

Pin 53 is ground. That pin should remain at 0V. The voltage needs to be 5V on pin 56. Pin 56 of the memcal is directly connected to 8 of the RFID. Break this connection and you are free to solder resistors equal to GM's values between the remaining two pins. If you remove the leg at pin 8 then connect a 7.5K resistor between 8 and 10 you will provide 5V to the ecm at pin 56 which should enable "4 cylinder TBI mode." But now you may get a cylinder select code due to calibration set for 8 cyl and fuel mode set for 4 cyls. The easy answer to that would be to disable the error code. A more complicated answer would be to alter the code to allow this mismatch.

The 2/3 Vref +10% and -10% requirement is a fairly close tolerance value. Most resistors are 10% tolerance so you'll need 1% or 2% resistors, preferably temperature stable versions so the engine doesn't switch fueling modes in cold or hot weather. But with infinite resistance between GND and pin 56 you don't need to worry so much about the resistance to pin 58.

1project2many,
I've been thinking about just disabling the cylinder select error code. I've been thinking, even with the cylinder select error code active the O'Scope showed alternating injector firing. What I have not been able to understand, what does the ECM do to fueling or spark when cylinder select error code is active. I've been thinking the cylinder select error was set up to detect / prevent a mismatch of Memcals on the assembly line? I've also been wondering, what error indication is given to the SES from the ECM when "Limp Home" is active. I'm thinking there is no indication "Limp Home" is active in the data stream? The only time I've ever encountered a cylinder select error was on my test bench, which if cylinder select causes "Limp Home" I did not see it. I'm thinking the cylinder select error is an "inert" error message and does not cause changes to the fuel or spark functions of the ECM? I'm thinking a cylinder select error does not set "Limp Home"?

dave w

[JeepsAndGuns]

Ya know, I too have wondered exactly what causes "limp home" Vs just a trouble code? I know you can get trouble codes and not go into limp home mode. So you make a good point, is there any way to know, (when using a test bench) if the ecm is in limp home mode?
When its running a engine you could probably tell. I know in my tuning adventures, there has been a couple times I have either didnt get the chip installed correctly, or forgot to install it completely and when I started the engine the iac pegged wide open and I could not get aldl connection. Not to mention the spas attack check engine light.
When running the 7727 on my test bench, I have noticed, when I changed it to tbi and got the cylinder select error, all the data appeaered to be the same as it was without the error. Iac counts, injector bpw, etc..

[RobertISaar]

from what i've seen on my bench(no scope), even with no MEMCAL attached(just a PROM) and setting a LOT of error codes, it still seemed to fire based on the setting in the BIN. i'm pretty certain it's only used in limp-home mode for fueling, since limp-home spark is done via the EST module without bypass voltage applied.

EDIT: easiest way to determine limp-home: ALDL will not function, since the PROM is entirely bypassed.