I've run into this exact situation, but have yet to have time to check it out. My swap: 97 Civic LX with USDM 2000 Si B16A2 running a USDM P30 ECU OBD2a. Since it's easy enough to source the parts, I'm going to try to replace the under-hood fusebox which contains the ELD and hope that helps.

See Honda-ELD for a bit about how the ELD works.

 

 

 

So, I finally had time to check out mine, and as far as I can tell it's acting like it should and within spec despite the lights slightly dimming and interior blower fan slowing down a little. With my swap, I am using a brand new Honda alternator for OBD2a Del Sol VTEC, brand new Honda battery, brand new under-hood fuse box with ELD direct from Honda and an USDM OBD2a P30 ECU. My grounds and wiring are as good as factory as well. So I just have to accept that this is just how it works.

For example, if I turn on the A/C compressor, it kicks over to the high-charging system because the ELD senses the the car is pulling too many amps to use the low-charge system.

It's only happening between ~ 10-45mph which is apparently normal according to that link I posted above ^. Under 10, it's always using the high-charge system and if I accelerate past ~42mph, it will kick back to the high-charging system and my lights will get a little brighter and fan (if it's on full) will get a bit faster.

The idea is that if you're under very little load with both engine and or electrical system (like just cruising at 30mph in 4th or 5th) the ELD will inform the ECU that there is a low load and to kick the alternator into a low-charge state, which will improve fuel economy while delivering just enough juice to the battery to keep it charged. Under low-load, between 10-45mph with the a/c compressor off, you should see the alternator kick down to it's minimum, which that article says is 12.3v. If it's below that then it could be a problem with the alternator signal wiring or something else.

You can test it while driving with a bit of ingenuity. You can tape a multimeter backwards to your windshield so you can see the readout and run the leads to the battery to watch the voltage go up and down when the ELD talks to the ECU, which in turn talks to the Alternator. I've yet to do this, but will likely do it this weekend to test. Since the High/Low switching is normal, the light dimming just a tiny bit between 10-45mph doesn't really bother me too much for now.

 

 

I believe it would work.

 

Also check this paper. Super interesting and really breaks down how the dual-mode charging works: http://opensiuc.lib.siu.edu/cgi/view...text=auto_pres

 

 

Correct. having the A/C on will bypass the ELD and put the alternator into high-charge mode. So if you have A/C on, you should not see your lights bright and dim while pressing the gas.

From the paper I posted above are the conditions that have to be met to go into low-charging mode via the ELD:

What this is saying that once the engine is warmed up and lightly cruising with very low load, the computer will start lowering the alternator output, which will will reduce drag on the engine and increase fuel economy. This lowering of alternator output will cause your lights to dim slightly.

 

 

 

That's what so vexing about about this issue. I don't think mine did it before the swap but I honestly didn't pay all that much attention.

With my swap (2000 Si B16A2 / 1997 Civic LX), I've since installed new: alternator (OEM from Honda), new battery (OEM from Honda), new under-hood fuse block with ELD (OEM from Honda), and stock USDM OBD2a P30 ECU from 97 Del Sol VTEC. I've double checked all my grounds, which are all new OEM, and my wiring. It's all correct and within spec. So I have no choice at this point but to think this is just how the ECU is handling the electrical load under these narrow conditions. I have a spare USDM OBD2a P30 ECU that I checked as well. Same issue.

 

 

You could also just bypass the ELD. Search around as there are quite a few threads out there.

 

 

 

 

 

It sounds like there are a multitude of issues but the most credible seem to be related to the two master power / ground connectors under the intake manifold. One blue, one grey, which house shared power and ground. Sounds like there's a Honda TSB for wires rubbed down to the copper, causing CEL 20 for ELD. I don't have this code but I'll check out those connectors out this weekend.

See: https://sparkys-answers.com/2009/04/...use-blown.html

See also from that thread: https://honda-tech.com/forums/honda-.../#post46406917

As I've mentioned before, make sure you're not drawing amps directly from the battery for something like an amp, stereo, alarm, etc. The ECU needs to account for all used amps in order to switch the ELD correctly.

Lastly, with a multimeter, check Fuse #15 (under dash, 7.5a)

 

 

 

Doubtful. I have a few ECU's that i've tested. 2 stock USDM OBD2a P30s and a JDM ITR P73. They all do it.

 

 

Yeah, I mean, there's a lot to consider. I've seen some people talk simply about the wire gauge size being slightly different for the Alt-C wire between D and B series. It could be the B-seires alternators or engines pull or put out slightly different amps. We all take these B-series swaps for granted because it's mostly all plug and play but there really are lots of little differences between an OEM d-series car and a b-series car. Wire sizes, group power/grounds, shielded wires, etc. My guess it's it's somewhere in the Alt-C wire signal or common harness ground that's slightly off, just enough pull excess amps outside the ECU's knowledge.