I recently purchased a 1988 honda civic CRX that wouldn't start, and I have be unable to figure out what the problem is. The previous owner bought the car in the same condition, and sold it to me in exasperation after making no headway.

The symptoms: The car will crank fine, there is a smell of unburned gas out the tail pipe, but there is no spark. I removed each of the 4 spark plug wires (one at a time), put a plug in the end of the wire, and grounded it to the body so I could see any spark, no matter how weak. Nothing. An error code 4 is sometimes displayed (this might mean crank angle sensor)

History: The previous owner tried to fix the problem by replacing the entire distributer (coil, sensors [including the crank angle sensor], igniter unit, points, cap), the spark plugs, spark plug wires, the main computer, and the battery and engine grounds. None of this made any difference.

History continued: When I bought it, my father found a problem with the main power relay and re sodered the connections to fix this problem (the relay now clicks at the proper 3 times). Unfortunetly, this didn't fix the starting problem, so I looked up this web board and found comments by "Texasgreasemonky" advising on how to replace the MAP (Manafold absolute pressure) sensor, as that had felled his own honda. I did this, but the car still won't start, all the proper voltages are present in the MAP sensor, but the symptoms are unchanged. Does anyone have any ideas what other sensors there are that could block the car from starting? :confused:

. Did you test all under-hood and under-dash fuses with a test light or multimeter? Are any fuses missing; e.g. ACG (S) (ALT)? Are all fuses in the correct slots?

. To me, it sounds like your distributor housing is still bad, since it's throwing a Code 4 (bad CKP Sensor). I suggest installing a genuine Honda distributor housing and reusing your ICM and coil. AutoZone will test the ICM and coil for free. Be sure to clean the main ECM ground on the thermostat housing.

Does the car pass the 3 basic diagnostic questions? If it does, you are likely on the verge of success, unless the engine has been swapped.

Quote:

---

Originally Posted by okmjui2000

I recently purchased a 1988 honda civic CRX that wouldn't start, and I have be unable to figure out what the problem is. The previous owner bought the car in the same condition, and sold it to me in exasperation after making no headway.

The symptoms: The car will crank fine, there is a smell of unburned gas out the tail pipe, but there is no spark. I removed each of the 4 spark plug wires (one at a time), put a plug in the end of the wire, and grounded it to the body so I could see any spark, no matter how weak. Nothing. An error code 4 is sometimes displayed (this might mean crank angle sensor)

History: The previous owner tried to fix the problem by replacing the entire distributer (coil, sensors [including the crank angle sensor], igniter unit, points, cap), the spark plugs, spark plug wires, the main computer, and the battery and engine grounds. None of this made any difference.

History continued: When I bought it, my father found a problem with the main power relay and re sodered the connections to fix this problem (the relay now clicks at the proper 3 times). Unfortunetly, this didn't fix the starting problem, so I looked up this web board and found comments by "Texasgreasemonky" advising on how to replace the the MAP (Manafold absolute pressure) sensor, as that had felled his own honda. I did this, but the car still won't start, all the proper voltages are present in the MAP sensor, but the symptoms are unchanged. Does anyone have any ideas what other sensors there are that could block the car from starting? :confused:

---

Unfortunetly, yes all the fuses under hood and dash are OK, and a check of the carfax report revealed that the distributer as a unit ( which contains the coil, igniter unit, Top dead center sensor, and crank angle sensor) has been professionally replaced 3 times! I also checked the engine grounds with a trouble light tester, and they checked out fine. Since my last post, I have tried unsuccessfully to find other sensors whose failure could sabotage the spark system. The error code 4 is flat out telling me that the crank angle sensor is faulty or unplugged, but there is only one of these sensors, and it has now been replaced 3 times! I called a Honda dealer, and they say there is no knock sensor, no other crank angle sensor, and that even with faulty timing the car should still spark, even if the timing was way off. The only thing that doesn't work right is the tachometer - it isn't showing a connection to the body ground as it should, but grounding it out to the body made no difference, and removing the tachometer wire from the igniter caused the car to not crank at all, and instead beep 4 times. With the entire spark system new, the Only thing that makes sense to me is that somewhere, there is a sensor that is shot and actively sabotaging the spark system. Does anyone know of any other sensors, gauges or parts whose failure will shut down the spark system?

As you already know, code 4 is the TDC sensor. Have you
Taken a good look inside the distributor? Any rust?
(Common)

Code 4 is a common failing item-related code. As long as
Your PGMFI relay is good, and the charging system is OK,
Afraid it's gonna' be a distributor replacement, and take
My word for it, don't go aftermarket! Replace with OE
ONLY!

This is from one of my pre-made answers, so go OEM only
Please, check the wiring for the TDC sensor going to the
Computer also.

Hello all, my thanks to txgreasemonkey for your help on the diagnosis. I learned a bit more about the problem recently: The reason the computer didn't give a code 4 (Crank angle sensor faulty) error every time was apparently due to a shorted hot wire. I removed a lighting fuse and that fixed the short so that now the computer gives error code 4 every time. Next I got hold of the factory shop manual (hondahookup.com) and found the test needed to confirm a good crank angle sensor (350 to 500 oms). The sensor checked out at the plug on the distributer, but where the two wires for the crank angle sensor (orange and white) should have been on the main plug that jacks into the cars computer, there was one missing wire, and an unrelated green wire. I checked the manual, and all the second generations CRX's put this sensor port in the same position on computer plug, regardless of trim level or year (1988, 89, 90 so far as I know).

So now I was scratching my head wondering how the heck the wiring harness could be wrong; It was obviously a untouched harness with no signs of tampering. A further consultation of the manual reveled something else that was strange: the distributer on the car is apparently (based on external appearance) designed to mount on the single overhead cam engine (which is correct for the cars engine) but the plug on the harness is missing the two cylinder sensor pins, and is in fact identical to the harness plug on a double overhead cam engine which lacks the cylinder sensor. The obvious conclusion is that the engine used to be a double overhead cam, and was swapped for a single overhead cam engine, then the connecting plug on the distributer was re attached to accept the new plug (the tampering is evident). The fix should be to open up the distributer, track down the cylinder sensor wires, and jumper them to the harness.

Except that the harness seems to be inherently wrong.
I tried to fix the problem with the crank angle sensor by bypassing the harness and running the correct wires to the correct place in the plug. A quick test showed the required 350 to 500 oms of resistance so I reasembled the computer and tried to start the car. No change in the symptoms whatever. And to and insult to injury, it still gives a code 4 crank angle error even with the proper wires in the right place.

This is getting to be exasperating: has anyone ever heard of the correct response to coming across a wiring harness, untouched from the factory, that seems to be just plain wrong? Missing pins and the wrong colors in the wrong places?

This one has me stumped for the moment, and any insights would be very much appreciated.

Sounds like you've tested nearly everything except for continuity of the wires going from your ignition system to the computer. Also, wiring diagrams are notorious for being wrong. If I was you, I would invest a bit of time in physically tracking down the wires indicated by a plug test diagram. You will probably find that the plugs for a SOHC and DOHC dizzy are wired in a different configuration for the other inputs and outputs. You also didn't say whether the coil has been replaced. Many of your Honda distributors don't come with a new coil. You may also want to check that the ICM, coil, and Crank Pos. Sensor are all wired correctly inside the distributor. The terminals are different shapes. Electrical parts are notorious for coming out of the box bad. I do have to agree that you should go OEM for this particular part. Also, I don't think you mentioned whether the car has a slush box or a stick. Ifr it's an auto, there are several codes that don't save past the initial occurrence that cause a no spark condition. One of the most common is the shift solenoids which cost about 300 american each. They normally short to ground and have caused me several sleepless nights thinking that I was dealing with an ignition problem. Hope this helps. There are several forums on hondata.com and hondatuning.com which may be helpful as well.

That is part of the problem: I did test the continuity of the orange and white crank angle sensor wires, and they check out fine for continuity, and, show the 350 oms of resistance that is what a good sensor should put out. I traced the continuity of the wires by testing for continuity between the white and orange wires, and everything else on the plug. The white wire showed continuity in two places, the orange, only one. Both were in the wrong place.

The car is a manual thank goodness, so I should be free from transmission problems. As I said before, the wires for the cylinder censor do not come out of the distributer, and are absent from the harness, because the harness belongs to a duel overhead cam engine that didn't have that sensor, but despite this, the computer is not calling for a cylinder sensor, its calling for a crank angel sensor.

Have you put the dizzy on a scope to see if the hall effect signal from your CKP is actually being sent? And if so, how many volts, and how far towards the ECU does it make it? If it makes it there, then have you checked that it makes it back to the coil? You should be able to use either a good analog multimeter or an oscilloscope (if you have access to one). Your digital meters won't even blip if there is a problem unless they have a trace function. Also, have you checked all the inputs and outputs on the main PGH-Fi relay under the dash? Both ends of the circuits, not just at the relay. What you are looking for here is both continuity of the wires, and correct voltage at both ends of the circuits. Also, grounds can look good and test good until you run a voltage drop test on them. Hondas depend on grounds more so than most cars. Even a two volt drop across a ground on a twelve volt signal can cause enough degradation to make the PCM input invalid. Most of your inputs from your sensors only carry around five volts. Don't forget about the obscure grounds, such as clutch lock out grounds (might want to check that input, too), Trans grounds, even starter grounds. Even with a new ground kit installed, there can be enough drop to disrupt signals. I've seen a ground in the brake light circuit being bad cause the PCM go into limp-home mode because it was a secondary ground for the trans. There was a massive voltage drop across the primary trans grounds, but together there was enough of a signal until the brake ground broke.