[dhauff]

Piv - do you use Torque or Torque Pro app. During my alternator woes, I used the Torque Pro app with my cheap bluetooth ELM OBD2 reader and was able to monitor key attributes such as rpm, voltage, etc and then was able to log these and download into a csv file and import into excel. Over several rides/days I could see the voltage drop occasionally and determined it had to be the alternator, even though I put a new (off brand) alternator in a few months earlier. A new rebuilt Denso resolved that issue.

 

[John Clark]

Torque or Torque Pro won't be able to see the ELD data. You need an enhanced data scan tool for that.

Foxwell NT510 or NT520 is a good option on a budget.

 

[nitely2]

A lot of weird electrical problems can pop up due to a bad alternator. If your current alternator is showing voltage swing of 12 - 18, that is a bad alternator. I would change that first or rebuild your old denso. I rebuilt mine a couple of years back and its been performing great since. https://www.odyclub.com/forums/24-1999-2004-odyssey/261601-compressor-pulley-has-side-side-play.html

 

[Frankiev#9]

My negative return cable from the battery to the trans is brittle and needs replacing. I will be changing it hopefully Thursday or Friday when the rain stops. I have also replaced the positive terminal post and morning drive to work - no flickering of the lights. Will keep you posted. View attachment 142943

 

[Piv]

The dash lights that are flashing are the Door Lights (all of them), the Brake Light and the Battery Light. I think they all share a BLK/YEL wire, It happens totally at random. Definitely not associated with bumps or any electrical components being turned on/off or application of the brakes or turning, etc. The windshield washer does work when the problem shows itself (previous question). The only constant seems to be that the lights/alternator output problem won't happen when at idle. When it happens at speed and you stop the van, it may stay on for a brief moment but not long enough to get under the dash or hood to troubleshoot.

Today I was monitoring the Alternator Control Signal (ALTC), WHT/GRN wire on the alternator plug. When the alternator drops off and the lights are flashing, this voltage goes from the normal 2.3 volts to zero. Not sure if this could be a cause or simply the effect of the alternator output going to 12 volts.

I just might change the alternator again. I'd be glad to pay the dealer cost for an alternator if it would fix the problem. Just having a hard time convincing myself that two alternators could have the exact same issue. I may have to dig deep and just do it to know for sure.

What I'd really like to be able to do is confirm the inputs at the four pin connector are all good but since they are only giving faulty readings when the problem happens, I am having trouble telling if one of the four pin connector wires is the cause of the issue or they are giving bad readings because of the alternator output.

I'll also be looking for a Faxwell Scanner.

BIG THANKS for everyone's help cause I'm stuck - BIG TIME - on this one.
Piv

 

[John Clark]

Your comment about the 2.3 volts on the alternator control output circuit is interesting. I'm guessing that the 2.3 volts you're seeing is an average voltage and that the PCM is sending a pulse width modulated signal to the internal regulator in the alternator. So, what's really happening is a 12 volt square wave with a certain on and off time. You would need a scope or a fast graphing meter to see what's actually happening there. Your DVOM is too slow to see the actual activity on that line.

You said you replaced the ELD, but did you check the power, ground, to the ELD, and the signal coming out of the ELD? It should have a 12 volt input, a ground, and then an output to the PCM. You need to make sure the power, ground, and output from the ELD are all good. The output will change based on the amount of current flowing through the ELD to the rest of the car (basically the demand on the alternator.) That ELD signal is sent to the PCM and the PCM outputs a square wave to the alternator to vary the output of the alternator. It sounds like that control output to the alternator is telling the alternator to stop putting anything out.

I would be suspect of the 12 volt ignition feed to the ELD, also. That is also the 12 volt feed to the alternator's white/blue wire (L circuit) and black/yellow wire (IG circuit.) If that feed is going to 0V then the ELD would lose power and the ELD would stop sending a signal to the PCM which might explain the white/green wire going from the average 2.3 volts to 0 volts. That power feed goes through the ignition switch and (either the #6 or #4 fuse in the driver's underdash fuse box, depending on what wiring diagram you look at) and may be losing power. It may be as simple as a bad fuse or ignition switch.

 

[Piv]

John - Thanks for the information. I will check everything out as you suggested and let you all know what I find. I did not check the ELD power, ground or output when replacing but will this weekend. I've got my fingers crossed too.

 

[John Clark]

I've seen some discrepancies between wiring diagrams. Below is the one in my 1999-2004 Odyssey service manual that says it's for the 1999-2001 models.

 

[Piv]

Thanks John - I wanted to let you know that under normal operation (lights not flickering and output at 14 volts) I do see the Alt Ctl Signal voltage vary when a load is turned on somewhere. For example, I see a fluctuation when selecting the window switches up with the windows already up. This seems to put a sizable load on the alternator. Probably not a great thing to do for the longevity of the window motors though.

After your explanation of the ELD circuit, I see why the PCM outputs are labeled "Signals". Square waves rather than voltage. I think I get Control Signal but not sure what the "FR" signal actually does.

Thanks - Piv

 

[dhauff]

piv.. Here is a better explanation of the ELD and alternator inputs/outputs. How often does this issue happen? while driving, several times an hour, a day, week, etc??

https://www.autoserviceworld.com/carsmagazine/honda-charging-systems/

I think you have a short some place

 

[johnboy1313]

Today I was monitoring the Alternator Control Signal (ALTC), WHT/GRN wire on the alternator plug. When the alternator drops off and the lights are flashing, this voltage goes from the normal 2.3 volts to zero. Not sure if this could be a cause or simply the effect of the alternator output going to 12 volts.

According to the service manual, the ALTC wire should be outputting approximately 8V when the engine is fully warmed up and running. I've attached a pic of the page from the manual.

 

[oldskewel]

Lots of good info in this thread. Hope you can get this tough problem figured out.

I'm following along, but will just throw in my own data point regarding the ELD: I've got a '99 EX, original owner, 230k miles, running with no problems at all. A year or so ago, perhaps triggered by an especially wet winter, the engine green key security immobilizer system would prevent starting. I thought it was voltage related, so in troubleshooting that (troubleshooting ended when the problem fixed itself) I installed a digital voltage gauge in the dash. So now I can see voltage (tapped off a point behind the radio) any time I want. And I do keep an eye on it, before starting, and while running, just to check on things.

I do have an ELD - I've visually confirmed it. But my voltage basically never ever goes below 14.3 or above 14.6 while the engine is running, regardless of headlights, engine RPM, etc. - system voltage is rock solid all the time. So it looks to me like my ELD is in place but not working. BTW, the van has not complained to me about this, and I don't consider it a problem I need to put any effort into, but I am still curious about what is going on.

So perhaps bypassing the ELD at least as a troubleshooting step, and maybe even as a permanent workaround would be something to consider.

Interpreting the circuit diagram by John Clark , it looks like the ELD has a coil that senses the magnetic field created by current flowing there in the fuse box, is powered by the ignition circuit, grounded, and passes the sensed result to the PCM. Then over on the right side of the diagram, the PCM comes back into play, controlling the voltage regulator in the alternator (based on the ELD sensor input). So I'm just pointing out (in case it's not already obvious) that the PCM input and PCM output at different locations on the diagram is how the thing is all connected. And we don't know from that diagram what sorts of computations the PCM is doing related to voltage regulator control.

 

[Piv]

Interesting that it's supposed to be 8 volts. I'm definitely not seeing that so perhaps we're on to something here. I'm going to spend some time checking the entire ELD circuit out this weekend and will let you know what I find.
Thanks for the info.

 

[John Clark]

When you check out the ELD circuit know that I made a mistake in one of my posts above. There is 12v power and ground to the ELD but, as someone else posted about, the ELD takes a 5v reference voltage supplied from the PCM and pulls it to ground or partially to ground--it won't be a 12v square wave. So, with the ELD unplugged there should be 5v on that signal wire. When plugged in it will vary based on the current flow through the ELD. It's just a matter of figuring out where the voltages are being lost when the problem crops up.

This circuit really does require an enhanced data scan tool and an oscilloscope to diagnose and troubleshoot. An enhanced data scan tool will show you the ELD voltage and alternator command. That really helps speed up the process and lets you know what the computer is seeing.

 

[dhauff]

The intermittent nature of this problem makes it real hard to figure out, unless you can replicate it or keep a vom on the signal wires while driving. piv.. how often does this happen? or as John said, put on a scanner that can read all the necessary voltages and keep it on until the problem happens so you can see what is causing the alternator to not put put the necessary voltage.

 

[Piv]

I drive 35 miles to work and it happens at some point almost every time I make the trip.

Also, I was checking out the ALT FR SIG (WHT/RED wire) today while driving. It is showing approx 5 volts and is unchanged by the problem.

If I don't see anything obvious when checking out the ELD circuit this weekend, I think I may purchase one of those Foxwell NT520 scanners that John was suggesting and see what it can tell me. I've got a Cen-Tech scan tool but it doesn't show any data for the PCM or ELD.

 

[dhauff]

Here is another really good article on Honda Charging system
https://opensiuc.lib.siu.edu/cgi/vi...cgi?referer=https://www.google.com/&httpsredir=1&article=1011&context=auto_pres

Please noted here: " As long as the IG and B circuits are connected the charging system will function." YOu might start by monitoring the IG connection at the alternator. It should be battery voltage all the time. See if it drops to 0 when you have the problem. If it does, that wire is your problem.

Also note this comment: "The warning lamp will turn on if the system is charging lower than specified. However, there have been many cases where the charging system will pass a performance test and still have a charge warning lamp on. In most cases, design variations are the culprit and most technicians prefer to use OEM alternators to avoid erroneous charging light problems"

 

[dhauff]

Another useful comment from a forum discussing the charging system:

"Just FYI, the Alternator C signal is the only one you need to worry about for no/low charging problem. It's a steady 8V provided by the alternator to the PCM, which grounds that voltage when it wants to shut the alternator off for fuel savings. If you see around 8V on that white/green wire with no charging taking place, then you're good to put an alternator in.

The FR is supposed to be a 0-5V square wave signal which the PCM uses for idle speed control, and nothing else. Zero impact on charging rate. With the alternator unplugged, look for 5V on the white/red wire."

 

[Piv]

You're right - very good article with expected voltages at the ELD . I'm getting a real education while troubleshooting this problem. Appreciate you providing this. It will definitely be helpful.

 

[Piv]

Wanted to provide an update - was out of town last week but now am back and still working through this problem:

I have monitored some voltages:
All 3 ELD readings are as they should be (output varies 5V toward ground with different loads)

IG - BLK/YEL voltage is not a steady reading - my multimeter sometimes fluctuates (dropping off to very low readings then jumps back to 12 or 13V. It would be nice to have a meter that could read this voltage more precisely but I think this might take an oscilloscope. I plan to call around to see if there are any shops that could take this reading for me....unless someone knows of a economical meter. Not sure if a Foxwell NT520 scanner will give me all of these readings or not.

ALTC- WHT/RED voltage is steady 3V but I believe it should be nearer 8 V. The 3V does not vary much. Scratching my head on this one. Once I figure out the ALT IG wire issue, I plan to take a harder look at this.

What I have done recently:
Got a 4 lead plug from the junk yard and am providing only one input at a time. Currently have jumped the IG wire from original plug so it is the only one going to the ALT right now. Time will tell but the input IG and ALT output appears to still fluctuate a lot. However, no dash lights yet. Just did this recently so I'm going to monitor it for a while.

I also got an original DENSO Alt from the junk yard and plan to try it and also rebuild it according to the info on the forum.

My next step:
Will provide power to the ALT IG wire direct from the battery via a 15 amp in-line fuse. Hook it up prior to driving and unhook when finished. Validate a good voltage input and see what the Alt output and dash lights do.

More info later as I continue to troubleshoot - Thanks to everyone for the great help. I/we will eventually figure this out. Maybe then someone else can also benefit.

Thanks - Piv