When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
DFI and ECMDiscuss all aspects of DFI (Digital Fuel Injection), ECMs (Electronic Control Module), scanners, and diagnostic equipment. Fine tune your Third Gen computer system for top performance.
Original motor 305 with carburetor and automatic trans. Now has 350 and 5-speed manual - same electronic QJet.
Car runs extremely rich. The QJet was rebuilt by Jet Hot (I think) many years ago. I took it apart recently, cleaned it, replaced the float, and used a normal rebuild kit. It came out fine - but car still runs way too rich.
So I next tried the ALDL port. Using WinALDL, I get no activity showing on the serial port. After some frustration, I got out an oscilloscope and checked every pin on the ALDL port for activity, and can see no activity.
Jumpering the A and B together flashes "12" on the check engine light. Seems correct, except no error codes reported - and I'm sure there must be some since the check engine light is always on.
The pinout on my ALDL port does not match any of the ones I can find online. That's one of the reasons I got the oscilloscope - check for activity on all pins...but no activity at all.
Anyone run into this before? Any help or suggestions?
I am starting to lean towards chucking the QJet. Besides the cost of a new carb, I'd also have to chuck the distributor because it too is under computer control. The cost is lousy, but I'd like to get the car running again. It's been a long time...
So I next tried the ALDL port. Using WinALDL, I get no activity showing on the serial port. After some frustration, I got out an oscilloscope and checked every pin on the ALDL port for activity, and can see no activity.
Place a 10K resistor between A & B on the ALDL port. Many (all?) of the CCC ECMs require that to transmit data. Only problem is, is that it also changes how the ECM runs. But at least can get data.
Jumpering the A and B together flashes "12" on the check engine light. Seems correct, except no error codes reported - and I'm sure there must be some since the check engine light is always on.
thanks
gene
That is good as it means the ECM is actually running.
OK, but if the duty cycle is bad, I still won't know why.
You can check for power and ground to the ECM. But once you get to the point that it has good power and ground circuits, it won't communicate, and you haven't got a good MCS signal - all you can do it tap on the ECM or open it up and try to flex the board, etc. At that point you have to call it and try another one.
I'll look in my 86 manual and try to get a pinout for you. Unless someone has rewired the ALDL connector somehow which would be a useless thing to do..... I still suspect bad ECM. Most of the wires to the ALDL go to the ECM so if it's bad.....
When my ALDL wouldn't communicate it was a bad ECM. They had a tendency toward circuit board cracks.
You can put that scope to work and check for the duty cycle signal to the MCS (Mixture Control Solenoid).
GD
I took your advice and tried it. The MCS is toggling at rate of about 100 mS (10/second) and duty cycle is 90%, if your counting the "HIGH" as "ON". So in other words, the pulse goes low for 10% of the time.
Is that the rich or lean condition? I'm guessing it's the rich condition.
If the MCS were not operating at all, that would be your RICH condition. Pulsing the MCS leans the mixture.
To read it, you connect an old school dwell meter to the MSC set on 6 cylinder. It is usually adjusted for a 50% dwell.
Last edited by NoEmissions84TA; 06-08-2017 at 10:26 PM.
The computer in this car is 1227169. The ALDL pinout does not match anything I've seen, as it has 5-pins populated out of the 12 available. A and B seem correct as GND and Service (or whatever they call the B). Shorting A&B does flash the codes. It has no pin installed in the E position, but does have one in the F position. C and D are also populated, but I don't know what they do.
I've tried using my oscilloscope to look for any sort of data, but there is none. Sooooo - I wonder if this computer has to be queried FIRST before it sends out data? If so, maybe that's why the pinout is different?
In the past, I have run into a few things on this car that were meant for 1987 car - so they probably were changing over to the new year part way through manufacturing.
The last pic of non-scan isn't relevant if you successfully got a code #12.
I never got ALDL data (with a laptop) from the 4bbl ECMs. None of the different ones. And I tried. You should quit focusing on it frankly. I had a dedicated scanner I bought used though. The data just isn't critical on the 4bbl cars.
You have a test term. for the MCS, you have a code 12, and you have the service manual. Got the tools to adjust things?
On the 6cyl. scale, 30deg. is 50%. More dwell is an attempt to hold the primary rods into the jets to lean the mix. Less dwell is the opposite. This is because the rods are sprung full rich until the MCS starts. I didn't catch whether you had MCS travel stops set to allow a correct total travel.
Damon has some great posts here re the E QJet. Read up.
Looks like the CCC carb computers have no serial data stream. They can't do live data . The diagram clearly shows no term. E and no serial data pin at the ECM. Only codes... and you have none.
If it's at 90% duty cycle then it's trying to lean it out. Without success apparently. I would be finding out if the MCS is working.
I am really disappointed that the ECM doesn't send out any live data because now I am flying blind. My hope was the computer would tell me what is going on with the motor...sensor readings, etc., and why it's setting the advance curve, or MCS duty cycle. So I get nothing from this thing? Dang.
I will have to physically check and measure the O2 sensor. Maybe it's broken, the car has been sitting for a VERY long time. If I make a gizmo that lets me vary the MCS duty cycle directly, and I monitor the O2 readings, I should be able to see if the carb is at least capable of making the correct A/F ratio.
Seems a simple MOSFET and pulse generator will do the trick... I can try that out. Maybe the carb needs more tweaking.
I did follow someones instructions on how to set the lean and rich stops - wish I wrote down where I got the info...
If the MCS is getting set to full on LEAN, then it seems you are right and the carb is just running way too rich.
Don't measure the O2 for now. Just work on getting 30deg dwell on 6cyl scale.
Read the posts on the basic tune of the E QJet, which involves idle screws, MCS solenoid travel, and idle air bleed.
We know you'd like to have an octo processor heads up display, self-tuning, quad O2 sensor, CCC QJet, but you're simply focused on the frosting before the cake is baked right.
The CCC system does support a slower updating data display. It's probably less hassle to source a used scanner from the 90s on eBay than to fiddle with getting an Aldl prof. running.
I have a vague recollection that the SES light output is used for the data stream. As noted in the above images of the FSM, there is some confusion as to whether terminal D or E is used for the serial data.
If there is no terminal E, then connect to terminal D. Be sure to use the 10K across A & B.
some progress after a long day in the garage. Disassembled the QJ...again. I set the lean stop a little lower (3.75 turns above bottom), as well the rich stop is now lower. The float was a little bit high, so fixed that as well. It's now sitting at 0.415". The accelerator pump wasn't working right - turned out I had to sand the bore with 1000 grit to make the pump glide correctly. I still think one of the squirters is blocked (both were before), but this is better. I didn't touch the mixture screws this time around. Oh, and the air bleed is set with 3 threads visible.
Car started right up - nice. The timing was advanced way too much, so I corrected that.
So....drum roll... the dwell at the carb is now about 32 degrees at idle.
some aweful stuff had to cook out of the motor...but it's better now after running for 20 minutes or so. Is seems rich at the tail pipe though - but at least it's not billowing crap.
I really want to get the ALDL stuff but it can wait for now.
Next, figure out why my radiator fan won't come on. Fix that - and then the bigger task... new camshaft. But I have to get the carb and fan working first.
We know you'd like to have an octo processor heads up display, self-tuning, quad O2 sensor, CCC QJet, but you're simply focused on the frosting before the cake is baked right.
I have a vague recollection that the SES light output is used for the data stream.
RBob.
You may have something there...When I connect the the 10K resistor from A to B a few things happen:
1) The MCS pulses like crazy, immediately
2) The SES light flickers like crazy.
If I put the scope on that SES signal, maybe you are correct, and it's the elusive data stream I'm looking for.
The clicking MCS is normal in that mode. This is the problem with the CCC ECMs, to get data need the 10K resistor. But that also changes how the engine is run.
I have some experience on two different Buick CCC systems. The '85 with the 3.8l engine (2-bbl carb) would also add spark advance in this mode. It is actually meant to force detonation, which indeed it did.
06-07-2017, 09:25 PM
