Honda Civic / Del Sol (1992 - 2000) EG/EH/EJ/EK/EM1 Discussion

1998 civic no spark-pin 20 problem?

 
Old 06-10-2019, 07:23 PM
  #1  
jvf
Honda-Tech Member
Thread Starter
 
Join Date: Jun 2016
Posts: 12
Default 1998 civic no spark-pin 20 problem?

OK,

Iím stumped. After overheating and a head replacement, project Civic wonít start. No spark. ICM testing per manual says itís good and coil test says itís good (ohms testing claimed to be unreliable but readings fall solidly within specs). But, nothing mentions input to ICM as part of the test. Why is that? Itís pin 20 on connector A. Manual says it should read battery voltage with ignition on and drop somewhat when running depending on RPM. This implies that pin 20 supplies voltage, not ground, to the ICM even though the ICM supplies ground to the coil on itís output. Whatís so complicated about that? If you turn the key on, you should have battery voltage on the input to the ICM but that is not mentioned anywhere in the manual or in searches.



But, I donít have battery voltage on pin 20, I have 6.4 volts. Worse, if I disconnect the ICM connector to the coil I donít have any voltage. So, I got another ECU but get the same results. So, without an internal schematic of the ECU I canít troubleshoot any farther. If someone can produce a schematic of a 37820-P2E-A71 then I can continue. Barring that, I have a question: Has anyone measured the voltage of Connector A pin 20 with ignition on and, if so, what is it? I have checked voltage and grounds on all power and ground pins on connector A and they are good.





Thanks,

jvf

jvf is offline  
Old 06-10-2019, 08:44 PM
  #2  
Honda-Tech Member
 
muellersfan's Avatar
 
Join Date: Apr 2018
Posts: 978
Default Re: 1998 civic no spark-pin 20 problem?

The ECU provides ground to voltage supplied by the ICM on the Yel/Grn wire.

I don’t have battery voltage on pin 20, I have 6.4 volts.
6.4V is probably fine, but compare the voltage to what you read at the ICM on the Yel/Grn wire pin. It should be about the same.


muellersfan is offline  
Old 06-11-2019, 11:09 AM
  #3  
jvf
Honda-Tech Member
Thread Starter
 
Join Date: Jun 2016
Posts: 12
Default Re: 1998 civic no spark-pin 20 problem?

Hi muellersfan,

Thanks for responding. Except for the A20 input, your diagram doesn’t show the remaining four connections to the ICM. Under dash fuse nine supplies battery voltage to the ICM via a BLK/YLW wire. I had no voltage here at first and discovered that this fuse had blown. I’m always suspicious as to why a fuse may have blown. At any rate, I replaced the fuse and now have power on the BLK/YLW (which also splices over to the coil). The WHT/BLU from ICM to the coil is where the ICM supplies ground to the coil upon receiving inputs on your aforementioned YEL/GRN wire coming from A20. There is a BLU wire going from the ICM over to the test tachometer connector. The ICM appears to be grounded through its case because there are five connections shown but only four terminals exiting the ICM.

So, we have

1) BLK/YLW -power to the ICM and coil from fuse nine

2) WHT/BLU (TEC distributor) ICM coil control

3) BLU-tach connection

4) YEL/GRN-ICM input from A20

It’s very frustrating not to have an ECU output connector list showing which pins supply ground (like the ICM does) and which pins supply voltage (either battery or Vref). But, From the description of A20 in the manual, it’s strongly implied that the ECU supplies battery voltage to the ICM and, of course when pulsing, the voltage will go down somewhat depending on RPM as described.

Therefore, with the key on (position II) I think that you should have 12v on YEL/GRN even if you unplugged the distributor connector. But, when I unplug my connector, even my paltry 6.4 volts disappears.

We have also done the test suggested by several you tube videos where we put the test light to 12v and held the probe on the disconnected ICM output which results in a pulsing test light when cranking. We thought we had some weak pulsing at first but subsequent testing with this method never showed any activity. Thinking that the overheating might have damaged the ICM causing the blown fuse we replaced the ICM before I got curious about the possible lack of ICM input. If the ICM had fried, it would also be logical to assume that it might have spiked the ECU and blown pin 20. But, none of this appears to be true. The new ICM acts like the old one and, as I’ve discovered, we had no voltage on pin 20 (even 6.4v) with the old ECU because we had the distributor connector unplugged.

I’m going to measure for any current flow on the A20 input to see if there is excessive resistance to ground that might be pulling the voltage down. But, this again assumes that pin 20 outputs voltage so I’m asking again for any documentation that can clear up the issue of whether pin 20 outputs voltage or supplies ground.

Thanks,

jvf
jvf is offline  
Old 06-11-2019, 01:45 PM
  #4  
Honda-Tech Member
 
muellersfan's Avatar
 
Join Date: Apr 2018
Posts: 978
Default Re: 1998 civic no spark-pin 20 problem?

Originally Posted by jvf View Post

4) YEL/GRN-ICM input from A20

But, From the description of A20 in the manual, it’s strongly implied that the ECU supplies battery voltage to the ICM and, of course when pulsing, the voltage will go down somewhat depending on RPM as described.
ECU pin A20 inputs ground to voltage supplied from the ICM on YEL/Grn wire. The ground signal received by the ICM signals it to tell the coil to release its high voltage that generates spark at the plug. That's how the circuit works. You can accept this fact or abandon it. It's your choice.

Therefore, with the key on (position II) I think that you should have 12v on YEL/GRN even if you unplugged the distributor connector.
The voltage at the ICM Yel/Grn wire pin varies from 12V to 5V, but 10V is most common.

But, when I unplug my connector, even my paltry 6.4 volts disappears.
Which connector are you referring to?

We have also done the test suggested by several you tube videos where we put the test light to 12v and held the probe on the disconnected ICM output which results in a pulsing test light when cranking. We thought we had some weak pulsing at first but subsequent testing with this method never showed any activity. Thinking that the overheating might have damaged the ICM causing the blown fuse we replaced the ICM before I got curious about the possible lack of ICM input. If the ICM had fried, it would also be logical to assume that it might have spiked the ECU and blown pin 20. But, none of this appears to be true. The new ICM acts like the old one and, as I’ve discovered, we had no voltage on pin 20 (even 6.4v) with the old ECU because we had the distributor connector unplugged.
Unplug the Yel/Grn wire from the ICM. What voltage do you read at the ICM pin. If you measure 6.5V there but none at ECU pin A20, then the Yel/Grn wire is broken somewhere between the ICM and ECU.
muellersfan is offline  
Old 06-11-2019, 04:05 PM
  #5  
jvf
Honda-Tech Member
Thread Starter
 
Join Date: Jun 2016
Posts: 12
Default Re: 1998 civic no spark-pin 20 problem?

Thanks for helping me to work through this. The connector I’m referring to is the one near the distributor but I don’t recall its number at the moment. After getting 12v to the ICM and still getting no spark, I performed the Ignition Control Module Input test per section 23-109 in our manual and it tested good so we did the coil test and, resistance wise at least, it tested good also. But, as stated, we didn’t see test light activity on the ICM coil control terminal when cranking so how could the ICM be good? Also, to restate, I was curious that 23-109 did not mention checking the A20 YLW/GRN connector for voltage (or ground for that matter) and thought that to be a serious oversight contributing to the confusion of whether pin 20 supplied ground or voltage. 23-109 specifies a continuity test between the distributor and the ECU for the pin 20 wire which it passed. It also wanted a check between ground and the A20 wire which also passed (no continuity).

During all of this I have not measured any voltage at the ICM YLW/GRN terminal which would of course be required if pin 20 was switching ground. I’m developing a head cold and am quite fuzzy now but I need to redo these voltage tests ASAP hoping that I have missed something simple. I will then post my results. To proceed I will accept your assertion that pin 20 switches ground but, as a Mueller fan, you should appreciate my skepticism of this assertion without providing any documentation.

jvf
jvf is offline  
Old 06-11-2019, 04:42 PM
  #6  
Honda-Tech Member
 
muellersfan's Avatar
 
Join Date: Apr 2018
Posts: 978
Default Re: 1998 civic no spark-pin 20 problem?

Originally Posted by jvf View Post
Thanks for helping me to work through this. The connector I’m referring to is the one near the distributor but I don’t recall its number at the moment.
Probably the distributor connector itself.



After getting 12v to the ICM and still getting no spark, I performed the Ignition Control Module Input test per section 23-109 in our manual and it tested good so we did the coil test and, resistance wise at least, it tested good also. But, as stated, we didn’t see test light activity on the ICM coil control terminal when cranking so how could the ICM be good?
Because you did not perform all required tests on the YEL/Grn wire. You need to test not only for continuity but also for a short to ground. Doing the two voltage drop tests I mentioned earlier are also likely to be informative.

as a Mueller fan, you should appreciate my skepticism of this assertion without providing any documentation.
That's on you, bud, not me.

We can discuss this topic here:

https://honda-tech.com/forums/genera...32%2A-3334713/
muellersfan is offline  
Old 06-12-2019, 04:27 PM
  #7  
jvf
Honda-Tech Member
Thread Starter
 
Join Date: Jun 2016
Posts: 12
Default Re: 1998 civic no spark-pin 20 problem?

Yes, that's the connector. Actually, in the the Honda manual, 23-109 item 6 specifies a test of the YEL/GRN to body ground and this I have done (no continuity). Thanks to the information you have provided me I have a good understanding of voltage and current flow of this circuit and how it operates. Several test procedures come to mind. For instance, as a static test, if power and ground are supplied to the ICM we should see a healthy voltage on the ICM's A20 terminal and this could even be done with the ICM out of the vehicle. I also like the test lamp idea of checking for activity on the ICM output when cranking by hooking the test lamp to battery voltage and touching the disconnected ICM output with the probe end. But, since the ECU also pulses a ground connection, it would seem that A20 could also be checked for output in the same manner by probing the A20 wire connector. If we get no light activity we would then check the crank and cam position sensors.

Until my bro-in-law turned this thing into a Stanley Steamer it was running fine so you can imagine our disappointment after replacing the head and turning the key only to have it lay there like a dead fish. I'm still under the weather so it'll take a day or two so sort this out. Hopefully, I'll conclude with a final note of successful startup. If not, I'll provide as much data as possible so someone can check my work.

Since the Crankshaft speed fluctuation sensor had to be moved and re-installed it has been a slight cause of concern that I might have messed it up but I've read the codes after the failed startups and there are none which should indicate that everything is OK with this and the other sensors.
jvf is offline  
Old 06-18-2019, 11:02 PM
  #8  
jvf
Honda-Tech Member
Thread Starter
 
Join Date: Jun 2016
Posts: 12
Default Re: 1998 civic no spark-pin 20 problem?

Despite our best efforts we have failed to achieve startup to date. I remain stymied by lack of detailed information on ECM requirements to pulse the ICM and, it appears, the fuel injectors. So, for now, I cannot troubleshoot any further without some guidance. As things seem to get more complicated I still can’t help feeling I’m overlooking something simple.

After reviewing the ignition circuit, I retested all wiring for continuity and shorts to ground but found no fault in the wiring. Earlier testing must have been erroneous due to inexperience because, depending on installed ICM module, I had 6.4 volts or 11.9 volts on A20. But, A20 would still not pulse ground. So, I inspected all the rotational sensors. I don’t think the Crankshaft fluctuation sensor is evaluated on startup but I included it anyway. All sensors had ohms reading within specs so I got out my USB scope and, after some study on what they should look like, observed the waveforms of the Crankshaft, CKP/TDC/CYP sensors. Everything looked good but the CYP sensor trace was very fat, like someone had gone over it with the wide part of a magic marker. This indicated noise which is not a good thing. A search of noisy CYP output brought up another world of Honda owner angst with tales of code 9, failed distributors and other woes.

About this time, I reviewed a video of ScannerDanner testing no start conditions on an old Civic trying to use only a test light. Before looking at the distributor he unplugged an injection connector to look for pulsing on the injector explaining that the ECM used the same sensor information to pulse the injectors as well as the ICM module. Since the injector ground wire tickled the test light he knew he had input to the distributor and proceeded from there.

Similar to accepting muellerfan’s knowledge of A20 pulsing ground, I also accepted ScannerDanner’s contention of the ECM’s use of distributor signals and decided to test for injector pulsing. I had none (but did have proper voltage on the other wire) and so had to consider that, although I didn’t know how much aberration a CYP trace could have, I did have a sorry looking trace and possibly the distributor was failing. So, without a detailed description of the ECM startup process what else to do but replace the distributor which we did. The CYP output is now very clean (as are the others-tested ‘em all) but we still have no pulsing on A20 or any injector. I have rechecked all ground and power inputs to the ECM and all are good. When the key is turned on, the fuel pump runs for a couple of seconds, etc, so all seems well with the main relay.

Living on an island is paradise until you have to pay for shipping if the item won’t fit into a USPS flat rate box. So, since the $50 Chinese knockoffs (if this thing ever runs again, it’ll prob. be a 5000 miles/yr beater so who cares?) would cost $150 to ship, we paid $200 to get a NAPA rebuild locally. This has really hurt the pocket book and is starting to make me look (deservedly) foolish.

Where is a detailed description of the startup process and a fault tree? Is it in a hidden Honda dealer only tech manual? How did muellerfan know that A20 pulsed ground? How did ScannerDanner know the ECM used the same information for spark and injector pulsing? Generally that should be true but different systems could consider other things as well so how did he know it would match exactly? I have built many machines using computer control for product manufacturing, testing, and the actual retail product itself and had to write enough of a manual so someone could fix it later. Where is the real Honda ECM manual? This is a 20yr old four banger. How hard can it be to fix these things? Hard enough, I guess. I’m getting nowhere.

jvf
jvf is offline  
Old 06-19-2019, 07:52 PM
  #9  
Honda-Tech Member
 
muellersfan's Avatar
 
Join Date: Apr 2018
Posts: 978
Default Re: 1998 civic no spark-pin 20 problem?

Unplug ECU connector A.

In the free connector, do you measure continuity to body ground at pins A9 (LG1), A22 (LG2), A10 (PG1), and A23 (PG2)?

Turn Key to ON(II). Do you measure battery voltage at pins A11 (IGP1) and A24 (IGP2)?
muellersfan is offline  
Old 06-20-2019, 05:10 AM
  #10  
Honda-Tech Member
 
muellersfan's Avatar
 
Join Date: Apr 2018
Posts: 978
Default Re: 1998 civic no spark-pin 20 problem?


muellersfan is offline  
 


Thread Tools
Search this Thread

Contact Us - About Us - Archive - Advertising - Cookie Policy - Privacy Statement - Terms of Service

© 2019 MH Sub I, LLC dba Internet Brands

We are a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for us to earn fees by linking to Amazon.com and affiliated sites.
 
  • Ask a Question
    Get answers from community experts
Question Title:
Description:
Your question will be posted in: