11sORbust

I have recently ran into an issue with a quad driver. The car I was working on was my 90 grand am (quad 4 engine). Figure it might make for a good discussion, considering all ECMs have them.

From what I have gathered, a quad driver is similar to a relay. The difference is the quad driver is faster acting and is not mechanical in nature, it's also directly on the ECM circut board. Quad driver's can "control" up to four output pins (per driver). Things like accentuators and soleniods are controlled by a quad driver.

I don't think the 165 or 730 has the diagnostic ability to determine if there is a problem in a quad driver circut. It would be interesting to find out exactly how the code senses a quad driver error. The ECM I was working on will throw a code 26 if there is a quad driver circut problem.


To make a long story short, the canister purge soleniod was bleeding 12 volts through to the negative side. This caused a code 26. There is two wires in the canister purge circut, one is a positive thats fed from battery volts. The negative side of the canister purge circut goes direcly to the ECM(from solenoid). When the ecm wants to purge the tank vapors, it'll ground that negative to activate the solenoid. Well, the ECM tried to ground out the canister purge but there was 12 volts present!! Boom, fried quad driver. So I trace the negative (that is now a positive) back to the solenoid. There is 12v on both solenoid wires, disconnect the solenoid connector and the 12v is gone on the negative side. Then check both of the purge soleniod pins, there is no resistance between the two. Bad canister purge soleniod.


I'm wondering how many quad drivers are in the 165/730.....

Siggy

Vey interesting...while playing with the fan circuits on my 88 LB9 (165 ecm) I was worried that I had lost the quad driver. This link has the outputs for most GM ecms...http://www.cruzers.com/~ludis/1227165schematic.html
The 165 appears to have two, U3 & U6 Delco part 16042368. I have not been able to source them and they look like a real pain to replace. I suspect most people in need would just replace the ecm board with a used one..??? Which is great if you live near a USA junkyard...I have a spare 165 which cost me $200.00 in Australia.
Correct fusing should save them we hope. Any info on where to source these Delco parts??

RBob

iTrader: (1)

The drivers used in the P4's, with the power tab style case, have a fault line. This output can be read by the code in order to discover the failure of a driver.

The C3 16 pin DIP drivers were sourced by Harris and Natioanl Semiconductor. Here is a link to the discontinued part:

http://www.national.com/pf/DS/DS3656.html

For the P4 drivers look for the 3 digit code. If it is 014 then it is a National Semi part.

RBob.

11sORbust

There is still a problem. THe replacement ECM made no difference. NOW, with the canister purge solenoid disconnected there is just a code 62 (Transaxle gear switch signal circuits ). Funny thing is the SES light still comes on exactly like before, just a code 62 instead of a code 26. With the test light hooked up to the cooling fan control pin, I can see the ecm ground out that pin right as the ses light comes on. So, every time the cooling fan comes on, there is a ses light at the exact same time. The ecm is not trying to ground any other pin when the ses light comes on. I disconnected the fan, relay and wire from the ECM connector. That made no difference. My past experience says that a ground circut is not totally connected. Checked the grounds and everything is fine. Totally clueless at this point

JP86SS

iTrader: (1)

If you were to ground the secondary side of the solenoid through a DVM and see the current, then do the same with the solenoid and DVM in series to the ECM, that may indicate in fact the grounding circuit is not being made completly.
You may be able to disconnect grounds to determine the one that is the actual one for the driver and improve it. Thats only if removing the ground does stop the current flow. If there is no current flow through the ECM, that will be more difficult to locate why.

11sORbust

I disconnected several pins(including gear switch signal and cooling fan). It's still throwing a SES light when the cooling fan control pin is grounded, even w/o the wire connected. I'm really thinking it's a ground problem. Not sure how because the ground seems to be secure.
On a side note, I just rebuilt my v6 camaro engine. Took it for a drive and it would die after a minute or so. Brought it home and figured out it would die and/or throw a code right as the cooling fan came on. Every single time. I started to look around and the ground strap/harness grounds was very loose. Fixed that and everything was fine.
It makes me wonder what the ECM is thinking when stuff like that happens. Does the electricity try to find a different ground path or what. There is nothing in the code to address faulty grounds and other weird problems. BUT there is always an effect from such problems. The ECM always reacts, it does so in ways that can be random. Sometimes there is no codes, other times there is "rolling" codes(depending on the problem). Maybe I'm a geek but "unprogrammed responses" are interesting....

I do know this problem has me stumped. Any ideas?

11sORbust

I don't exactly understand.

jwscab

it does sound as though some ground(s) are not right. I just had my truck act reallly funny and die due to a loose negative terminal on the battery.

one quick note to you though, i noticed you said you probed the canister solenoid and measured 12 volts on both wires of the solenoid, when not 'turned on'. You WILL see 12 volts on both sides of the coil when the driver isn't grounding the one leg. You have no current draw, therefore no voltage drop. Make sense? And the coil may only have a few ohms of resistance.

check the ground wires( I know a real pain), it sounds as though one of the many grounds to the ecm is gone. If you can figure out which ground the driver uses, trace that one out first.

good luck!

JP86SS

iTrader: (1)

Basically if you can see the amps (current) flow amount by just grounding the solenoid (on the ECM side), you could compare that amount of current to what you read when the ECM grounds the solenoid.
Might indicate that the ground path is there but not fully making a good connection.
Once that is established you could wiggle each ground until the amp reading is the same, thus locating the ground for that driver.
Hope that helps

cliffharris

I found the part with Delco part number 16042368. It's a Texas Instruments TPIC2404, which they call a Quad Low-Side Switch. It does have self-analysis circuitry inside and can sense various problems. These parts are obsolete but may be available from surplus electronic component dealers. Here's a link to the datasheet:

http://focus.ti.com/lit/ds/symlink/tpic2404.pdf

Ponty

Hi everybody,
I am out of ideas wot is wrong with mi Pontiac. i replaced the ignition coil and the distributor i have spark, injectors and fuel pressure are good. engine cranks but wont start. it worked fine with old distributor. now when i put the old one in it doesn't work nether. when the car was still drivebel i noticed that mi 4th gear never came on in overdrive fuel economy was a disaster. exhaust smells like rotten eggs. No error codes are stored, so mi conclusion now is busted quad drivers. so i ordered new ones from ST electronics. desodered the old ones will put new ones in. but if i don't find the problem in the electrical i will fry the new ones us well. i drew with a broken gearbox mount for a wile when i accelerated there was a terrible knock against the chassis maybe it can be the cause of 4th gear lose(automatic transmission) .i am not sure that mi problem lies in quad drivers as the testing procedure is difficult to understand. if somebody have any advise on mi situation i will be most grateful.

cliffharris

Quad drivers are pretty robust and they are hard to damage. They have current limiting built in so if there is a short they just turn off and nothing bad happens.

Ponty

Agreed, but why no error codes are stored ? and the 4th gear is controlled by QD as well as all of the fuel and ignition calculations. I have two ECU's and both give the same result

cliffharris

The program code in the 127165 ECM does not have the ability to detect quad driver faults, so it has no way to know that one exists or to set an error code.. That capability was added in later ECMs, though.

Ponty

yes this is a good thing i have one from a 3.1 but it is not similar to 2.8 any idea how to make it work on 2.8 ? the connectors fit but is the functionality of the pins the same any one experience with it? I think it is a different topic thew. for the quad driver issue it is possible to replace. Will post back when i test it.