sskooter90

New guy here. I have an interesting issue that I find intriguing and I need some help. First, the car. It's a frankenstein. Not a "build", like most of you think of, but a slapped-together-with-parts-laying-around type of build. I didn't do all the slapping, for I bought the car as-is for cheap from a former coworker.

-1992 Civic DX 1.5L 5 speed
-It has a ECU from a different car: 37820-P06-A52 (always has a code "19" for the automatic)
-I believe the cylinder head is from a CX. It only has two valves per cylinder (eight total). If my research is correct, the DX originally came with a 16V head.
-CEL #9 (CYP)

Questions (in red):
Below, in the Issues and Remedies, is a description of the problems I'm having. I have this first so readers don't get bored or lost before arriving to the questions. Please refer to the rest of the body for symptom questions. I've decided to try to test the ECU, but I'm not familiar with doing that. I know the basics of electricity and how it works (I've wired up my basement and garage), but sensors and ECUs confuse me, because I'm old school. What's the proper procedure for testing the ECU sensors? I'm testing the ECU connection of the CYP sensor. I have taped threading needles to the probes of my multi meter and "backprobed" the B11 and B12 pins. With the connector at the distributor connected, I get 420 ohms (350-700 is the range). The same for the other dist. sensors. Testing the B11 to ground yields a closed circuit. According to a diagram I found somewhere, if any of the dist. sensor wires have a continuity to a ground, there's a problem. But I check B11, B13, and B15 (CYP, TDC, CKP, respectfully), all of them have continuity to a ground. This doesn't make sense to me because all of them are "grounded" but I'm only getting a #9 CYP code. What do these readings tell me? Am I doing this wrong? Has anybody done this before? If I am doing this correctly, is this a sign the the ECU is bad, or perhaps the wires are pinched somewhere? Keep in mind, as stated below, the distributor is brand new, thus a new CYP sensor, which does read around 410 ohms at the dist. connection sensor side (pins 4 and 8). If there is a thread describing this procedure, please point me in that direction. I have a difficult time navigating in these forums.

All of the below is just to describe the issue. Please refer to this part of the body for symptom questions.
The issue:
At startup or shortly after, the car bogs down, hesitates, and jerks. If the throttle pedal finds its way to the floor, the car accelerates without issue. However, settling back down to a cruising state (2-4K RPMs) , the car continues its bucking. During this time, the throttle has to be pressed down further than normal just to keep it at a highway speed (55mph). After a few miles, sometimes longer, of continuous jerking, the car finally decides to operate smoothly. It's like night and day. One second the car is running terrible, the next it's accelerating because of the pedal being pressed. It's almost like a sensor starts/stops working, or the ECU finally compensates for whatever the problem is (if it can do that, I'm not sure). Also, this problem often doesn't present itself again until the ignition is shut off and turned back on. Sometimes it happens again though. Has anybody had an issue like this before? If so, was it fixed and what was the solution? Again, if there is already a thread with this issue, feel free to point me in that direction.
Hopeful Remedies:
I've replaced spark plugs, wires, cap, rotor, and coil. I've checked voltage to the TPS and MAP. I've checked voltage and listened to each injector. I've replaced air and fuel filters. Because of the CYP code (#9), I've replaced the whole distributor to no avail. I would reset the ECU by pulling the fuse, ensure the #9 is no longer present (#19 is always there), and during the short test run, the symptoms return and the CEL #9 returns, even with a new distributor. I have adjusted ignition timing multiple times to see if it's a timing issue. No matter how many degrees advanced or retarded the timing is, the car acts the same.

mk378

Test sensor resistance and shorts to ground at the ECU plug with the ECU unplugged.

You can move jumpers inside the ECU to reprogram an OBD1 auto ECU to manual.

CX-Adam

Great writeup, OP. Too few people put any effort into asking their questions. I'm heading out right now and won't have time to look into your questions today but I'll send you a PM in a minute that might help.

crazyhorse2011

I have a question in regards to this. My ECU is a very similar model number, 37820-P28-A52, and it is a 5 speed manual on a same gen car, a 95 civic with the d16z6. My car does NOT and never has thrown a code 19 for the auto solenoid that is missing. Why would this be? A52 is the designation for a auto ecu if i remember correctly. Would this lockup solenoid circuit be shorted with a resistor of some sort to alway trick the computer into thinking the car is in neutral or something?

mayfly

not to be rude, but start a new thread, dont hi-jack this guys thread so it doesnt get off topic

homercan2

I had similar symptoms, if I would had done a compression test in the beginning, I would had known valves were not properly adjusted (cylinders 2,3,4 were at 168 and 1 was at 60). Also my timing was WAY off. Wont hurt to do a leak down either.

sskooter90

@mk378 Thank you for the advice. I will try that today.

@CX-Adam Thank you, also, for the PM. Hopefully that will be of help.

@crazyhorse2011 I suppose, from what mk378 mentioned, that you're ECU has a jumper moved in order to remove the code 19, just as you suspect. But don't take my word for it. I don't know ECU's at all. Give me a carburetor (or a row of four like on my 1980 Honda motorcycle), and I'll have that baby screaming. This electronic stuff confuses me, and this is a Ninety's car...

@homercan2 I guess I failed to mention that in the original post. The compression is good and valve clearances are adjusted correctly.