• Protip: Profile posts are public! Use Conversations to message other members privately. Everyone can see the content of a profile post.


19 October 2022
Columbus, Ohio
30 year owner of 92 w/25k miles that runs great, but have a likely self imposed lighting mess. About a year ago I crossed the subwoofer power leads.(with the key and radio on!) and blew the stereo which also blew the ignition switch and around that time lost right taillight, left turn signal, all side marker lights, dash illumination, glove box and rear license plate. Deleted the stereo and replaced the Ignition switch and still no lights. Since we are on the last gasp of parts, I decided to throw parts at it (for spares)and replaced the lighting combo switch/signal cancel switch/dash rebuilt by MITA/dash brightness controller below instr panel/dash brightness control on firewall, ICU in drivers footwell, lighting and taillight relays/both taillights and bulbs. Had our best local tech put in a few hours to no avail except headlights and neither taillight now work (argh). So I decided to start on the taillight circuit as power was not getting to #38 taillight fuse. So after testing the new relay of course, I jumped the taillight relay(underhood fuse box hot relay term to relay term going to fuse #38) and got the left taillight back which sort of led me to thinking of a switch circuit issue although all that is new except the hazard switch and the alarm control box. When I also jumped the headlight relay I did get headlights. So to my question: up to this point I’ve ignored the alarm system but after looking at the schematics it seems to tie into about everything, including even the stereo. It is still working properly but I have not done the full diagnostic yet. I was hoping to avoid this system to simplify my diagnosis, but do I need to get into this system or can I ignore it for this lighting issue? Any thoughts to improve my diagnosis greatly appreciated. Pic of my relay bypass attached. Thanks!


  • 20231203_034519.jpg
    146.1 KB · Views: 9
This is the headlight and tail light control circuit.

Headlight circuit.jpg

You said
When I also jumped the headlight relay I did get headlights.
I am going to guess that means that means that you jumpered the A and B terminals of the lighting relay together? If so, that indicates that everything on the fuse 52 and fuse 49 circuits is working. The better test is to jumper the C terminal of the lighting relay (blue/red wire) to ground. This should energize the relay and cause the headlights to illuminate (the headlight doors will not open). If the bulbs do not light up when C is grounded then the relay is damaged or you have relay socket damage.

The red line is the ground path for switching the lighting relay / headlights on.
Headlight ground path.jpg

Pretty simple. If the headlight relay and socket are OK then
- there is a broken wire between the relay and the lighting switch
- there is a broken wire in the lighting switch
- the switch part of the lighting switch has failed
-that diode on terminal 1 of the lighting switch has failed
-the G401 / G402 ground point is damaged.

The last 4 items on the list will also disable the tail light relay. Take a jumper wire and back probe terminal 10 (blue / red wire) on the lighting switch and connect the other end to ground. This should cause the lighting relay to power up and illuminate the headlight bulbs. If it doesn't then you have a problem with the wire harness between the headlight relay and the switch which is going to be a major pain to locate and repair. You will need the 1992 Electrical Trouble Shooting manual to do that. If you are lucky the headlight bulbs illuminate which means that the problem is in the switch or the ground connection G401/402. Money should fix that.

The security system is not the problem. The security system can turn the headlights on by applying a ground connection on the blue/red wire; but, it cannot prevent the headlights or tail lights from being turned on. If you need to convince yourself just unplug the security control unit to see if it resolves the problem. It is a lot of work to get at the security control unit so I would not bother.

Your 'best local tech' should probably be restricted to bolting and un bolting stuff because they clearly didn't bother to look at the schematic diagrams. The headlight control circuit is a dumb-ass simple circuit. Since you have done about 50% of the testing already by confirming that the headlight relay is getting power, once you have removed the pieces required to get access to the light switch to test at the connecting plug it should take about 5 minutes tops to locate the problem area. The repair may take longer - a lot longer if it is in the wiring harness.

I suspect that once you figure out what is amiss with the headlight control circuit you will also resolve the tail light problem.
Thanks so much Old Guy for your step by help. I would likely be totally screwed without your help. Progress report: Going thru your steps but hit a snag. First G401 and G402 showing continuity to ground(they’ve never been off). On bypassing the headlight relay I can’t really get to the C with the relay in so I jumped the three non-control relay terminals into the fuse box and jumped the C terminal to ground and got headlights telling me the issue is the switch circuit so I’m feeling good. So moving from fuse box toward the steering column I pulled the C246 connector in the pass footwell (I’ve pulled the whole dash at this point for easy access) and found continuity from the underhood relay box female “c” terminal to the Blue/red wire in the C246 connector. So next with that connector still pulled I connect that connector Blue/Red wire to ground G401 thinking I should again get both headlights, but instead I get the right headlight only and the left headlight goes up and down but no light. I did both of these tests twice. I was planning on next cont checking between C470(which connects to C246 and goes to the steering column) and the combo switch and then using a cheap “open finder” but this headlight response has me thrown as my takeaway from your note was that there should be no digital interference (from icu/etc) and that the combo switch simply grounds the circuit. Again combo/cancel switch/dimmer/icu controls all new. I did pull the back off the underhood fuse box and I see no visible issues. Thanks!
On bypassing the headlight relay I can’t really get to the C with the relay in so I jumped the three non-control relay terminals into the fuse box and jumped the C terminal to ground and got headlights telling me the issue is the switch circuit so I’m feeling good.

I am confused about "three non-control relay terminals". The lighting relay has four terminals with two of them being non control terminals. The dimmer relay has five terminals with three of them being non control terminals. Are you sure you were working with the correct relay? Or are you talking about shorting the terminals in the socket with the relay pulled out? The lighting relay socket might have 5 terminals with one of them unused. When you "got headlights" it was both the left and right low beams - just checking to confirm?

headlight relay.jpg

So next with that connector still pulled I connect that connector Blue/Red wire to ground G401 thinking I should again get both headlights, but instead I get the right headlight only and the left headlight goes up and down but no light. I did both of these tests twice.

You are doing the right thing, unfortunately you strayed over the fence in the way that you did the test. This is the diagram from the ETS showing the headlight wiring.

left headlight ground.jpg

The ground for the left low beam goes through C246. With C246 un plugged for the way you did the test you effectively disconnected the left low beam which is why only the right low beam came on. Plug C246 back in and then connect the blue/red wire to ground (any ground works) by back probing the blue/red wire in its connector plug. That should cause both low beams to come on. With C246 plugged in confirm you get both low beams. If so, I suggest the next step. I am not sure what is going on with the left headlight retractor; but, lets come back to that after the lighting relay control gets sorted out.

According to the ETS manual, C470 is not in the headlight circuit so I would not mess with it. The ETS implies that there is a direct connection of the blu/red wire from C246 to C416 at the turn signal / combination switch.
combination light switch.jpg


Back probe the blu/red wire at C416 and connect it to ground. If the previous test was positive this should also cause the right and left low beams to illuminate. If that works, then try operating the headlight switch to see if it illuminates both low beams. If it doesn't, that suggests that there is a problem with the switch. On page 23-177 of the service manual there is a function test for the lighting switch that you do from C416. Note that because of the diodes in the Lighting Switch, you have to make sure that the polarity of the tester is correct to get a correct result. The + terminal of the continuity tester must be connected to terminals 4 or 10 or 13 when doing the individual test otherwise the switch will always read open. If you get a 'fail' carefully examine the lighting switch wiring. There is only one plug for the switch assembly (C416); but, if you look at the above diagram you will see that there are two switch assemblies that have interconnecting wires. Maybe those wires got pinched and are damaged. If the problem is internal to the switch it may be possible to disassemble the switch to make a repair.
The 3-control comment was misleading, I do have the correct relay C935 and i do understand this standard 4 terminal operation(two large brass power/two small silver control). As i think you suggested I was wanting to ground the one "grounding" control terminal (Blue/Red), but didn't yet have the fuse box back off and didn't see an easy way to ground that terminal, so with the relay removed i jumped "three" of it's prongs back to the fuse box and took it's control terminal to ground. This resulted in the desired both headlights operation thus if i'm following correctly proving out the relay supply, the relay and the fuse box leads which were suspect. I will follow up on the C246 circuit and more in your note above. I'm learning. Thanks so much for your expertise!
Update: With C246 in pass foot well plugged BACK IN (I’ve gone “ over the fence” enough times my nads are scarred and likely got me into this jam to start with) and back probing the Blue/Red stripe at C246 I did get headlights. I then moved on to C417 at the column and back probed that Blue/Red and again got headlights, so I moved to the headlight switch test and it failed all tests. My Acura tech told me that he “back powered“ the taillights and that they were “good” from C246 on back, so I’m thinking he blew the diodes in by brand new $900 last on the planet headlight switch as the headlights were working when I went in and were not when It came out. So I bench tested the OE switch I took out and it tested good so I put it back on and I’m now back to where I started having headlights, left taillight, turn signals/brake lights but no markers and dash. Starting to stare endlessly at the schematics, but don’t’see how i can understand the ICU logic which appears to pertain to this lighting circuit. Will continue plugging away. I pulled the failed switch apart out of curiosity. Could I source the diode replacements? Assuming I develop the skills to solder under a microscope to see them! I don’t think BrianK does this work. So thankful for your input and attention and happy Holidays!
Haven't seen an ICU from the inside yet but if it's the same level of electronics as the rest of the car and there is no microcontroller involved it should be repairable without a magnifying glass.
The left and right front side marker lights are controlled by the ICU. The rear side marker lights are directly switched by the tail light relay - no ICU involvement. Is it just the front side marker lights that are the problem or are the rear marker lights also non operational? If it is just the fronts then that could be an ICU problem. Spoiler alert, the ICU also gets an 'on signal' from the tail light relay which could be the problem.

When you mention no dash lights, I assume this means no dash cluster illumination lights and you are not talking about the dash warning lights? If so, the cluster illumination lights and the brightness control are not controlled by the ICU. However, all of this including a whole bunch of other stuff (right tail light, rear side markers, some stereo stuff (ominous connection there!) and the on signal to the ICU all look like they are on the fuse #38 circuit which is powered up by the tail light relay. Since the left tail light is working then the tail light relay is switching on (fuse #39 is good). Pull fuse #38 to make sure that it is not open circuit. With the fuse out and the lights switched on check to make sure that you are getting +12v at one of the terminals of the fuse #38 socket. If you are then all of your original problems may be originating somewhere on the #38 fuse circuit and depending on what you did with your stereo fiddling that might have been the cause. Depending on what you find then there are some simpler tests that can be carried out to confirm the problem is the #38 fuse circuit and where the problem might be.

Going to send you a PM re the light switch.
Things that DON’T work: All marker lights(front turn signal markers do flash), right taillight, dash illumination(indicators appear good), glove box, rear license plate. Fuse #38 with fuse out DOES have one socket hot. I was getting ahead of you on my ICU freakout comments. I put in a new ICU yet never tested the OE unit as wanting to have a spare. Thanks for your continued attention.
Here is the fuse layout diagram for fuse #38.

Taillights 3.jpg

All of the stuff you have identified as non working is on the #38 circuit and is circled in red. The license plate lights, right tail lights, rear side marker lights (and ICU) are all on the next page connected by E. So, given that the glove box light and the gauge assembly is dead, I think it is reasonable to deduce that the problem is at C450 or somewhere between C450 and the under hood fuse / relay box.

You didn't mention the status of the front parking lights. If the front parking lights are non functioning then that would seem to imply that the problem might be right at the under hood relay box at the fuse #38 terminal. The electrical trouble shooting manual provides no details on how the parking lights are connected to fuse #38. Normally it would call out a junction connector; but, it doesn't so I don't know whether the parking lights run individually to the fuse box. If the parking lights are non functioning I would be inclined to lift the relay box to have a look at the underside of fuse #38 to check for damage.

If the parking lights work, then the problem would seem to be between C246 and C450. C 246 is relatively accessible behind the right kick panel and you can test for fuse #38 voltage on the red black wire there.

C 246.jpg

If you have voltage at C246 then the next location would be to check C450. C 450 is a major junction connector. I think you said you already have your dash out? Even with the dash out getting at C450 and extracting it to do testing is going to be a treat.

C 450.jpg

Given the major hassle associated with getting at C450 I would be inclined to first check to make sure that the problem is not before C450. The dimmer control amplifier is moderately easy to access on the back bulkhead. You can pull C497 at the dimmer control and check for voltage on the #38 circuit there. If you have voltage at that point then I think you are doing the deep dive to extract C450 for testing.

C 497.jpg

If it were me, I would be hoping for damage right at the fuse / relay panel because that is probably the easiest thing to get to.
Making progress. I test for power at the Red/blk going into C246 with the headlights switched on and get nothing. Turning to fuse #38 headlights turned off I get no power out of #38. Headlight switch on I do get power out of 38 (so much for my engineering degree-I didn’t realize I needed the load to show power). So this proves out the fuse box connections I’m thinking. So I now do a continuity check from the fuse output to the unplugged C246 Red/blk and get nothing. So I’m now thinking I do have an open in the Red/blk between the underhood fuse #38 and C246. But before I cut into the harness, I’m thinking I should run a temp jumper from #38 output and tie into the red/blk just before it goes into C246 to prove out the “open here” assumption. Since I don’t have an extra connector pin I’m thinking that I will have to cut the red/blk a couple of inches upstream of the connector(ouch). I should then get all my stuff back and then I go into the harness with my $29 short finder or simply run a parallel line. Any red flags here for this test? Regarding the parking lights, I had Arclight LED’s on the front and did remove to take them out of the equation(perhaps the cause of all of this?) and lost the plugs to the small outer bulbs. I did test for power going to these loose wires and get none. Once I get things rolling with all OE, I’ll get into these LED’s later. Thanks!
Sorry to the list, i can't get the PM's to Neil to work, but i wanted to catch before a response as my conditions suddenly changed. In getting ready to jump from the fuse 38 to C246 to prove an open in that line, i now see i have intermittent loss of power to the fuse 38. My loss of lights is not intermittent. I see no signs of damage on the back of the underhood fuse box. I'm using both a snap-on multimeter and a snap-on digital safe test light. Thanks!
Subject your latest post and what you find might be going on with fuse #38.

First off I would not go cutting into the harness just yet. Unless there is outside physical damage to the harness ( cut / chaffing / impact damage) it would be really unusual to have a wire failure inside an intact harness. If you look at that diagram from the ETS manual in post #10 the parking lights connect into the red/black wire off of fuse #38 somewhere. Neither the ETS manual or the service manual indicate how this connection is made; but, a junction like that would be a logical failure point since you tested for power at the parking lights and did not get any. I would be inclined to go the the left and right front parking lights (which probably share a harness with the signal lights) and see if you can trace both harnesses back to some common point which is where the connection to the supply from fuse #38 might be and where there might be damage. If I were going to open up the harness it would be at that point.
Sorry to the list, i can't get the PM's to Neil to work, but i wanted to catch before a response as my conditions suddenly changed. In getting ready to jump from the fuse 38 to C246 to prove an open in that line, i now see i have intermittent loss of power to the fuse 38. My loss of lights is not intermittent. I see no signs of damage on the back of the underhood fuse box. I'm using both a snap-on multimeter and a snap-on digital safe test light. Thanks!

Have a careful look at the inside of the jaws on the socket that grab the blades on the fuse. Look for damage. Is the fuse really easy to extract? If so the jaws may be loose and making poor / intermittent contact with the blades. Finally, replace the fuse. Even though it does not appear blown there may be an intermittent failure in the fuse where the fuse element is soldered on to the internal blades.
Got all my lights back since Old guy steered me towards my front park lights. In wanting to increase my street visibility a while ago I installed ArcLight front LED park lights set up as DRL’s. These lights use the OE turn signal wires as triggers but get their power direct from added taps to the hot/wiper fuse box terminals thru external fuses. This makes the OE park wires superfluous so they got snipped(ouch). In doing so I deleted a crucial loop line to a whole bunch of other lights in the cabin/rear of the car. Since I rarely drive at night I was not able to put the Arclight install together with my loss of lights. For analysis, i removed the led’s and went back to OE. The front parking light wiring is not well shown so I had to pull the nose and split harnesses to chase this wiring which did show the problem. One would think that the underhood taillight fuse #38 hot (Red/Blk) would go directly into the cabin and then back out to the front parks, but instead it FIRST goes to the front left park light and then goes across the radiator to the right front park light and then another Red/Blk takes that power back into the front bulkhead C246 and then to C450 etc. Lesson here is when the manual shows a “dot” depicting a wire junction, they may have instead made that connection at a component which is likely a better practice yet not properly made it to the schematic folks. After re-connecting this loop per OE and testing grounds(as each front park has two separate grounds that I didn’t chase -argh) everything came back(Duh & Whoopie!) Lesson of course is to take notes when deleting/removing any electrical component so you can go back and analyze the situation without taking half of your car apart. I purchased the last front and rear Arclight Led’s which I will re-install as DRL’s once i regain my confidence. The dash lighting circuit is surprisingly complicated. While I personally don’t need most of the OE stuff for now i’ll keep it as intended. Whats the old saying that “don’t mess with mother nature?” that brings to mind “mess with mother Honda at your own risk!” But my boy will be roaring back soon better than ever. Thanks so much for Old Guy’s direction and list support.
Happy to hear that you got it resolved. Sounds like you might have a little harness re taping to do? If so, I recommend silicone self fusing tape for the repair. Home Depot and others sell Nashua, Gardner Bender and various brands. If you search around you can find larger quantities at better prices.

When stretched and wrapped with over laps it forms a continuous seal which looks good, has high temperature and oil resistance and does not unwrap. Don't use it if you think you might ever need to get back into the harness because the unwrap factor is real. If you need to remove it you need to score the jacket with a razor knife to peel it off. It does not stick to the wires; but, it does stick to itself - permanently.
Already repaired the harness prior to this post with 3m 33+and 3m 88 1 1/2" wide which seemed to be similar to what Honda used. My repairs were mostly only 12" long or less so except for across the radiator where this silicone would likely have been great. I noticed that Honda used a Tesa like product across the under dash wiring which i didn't have to dive into. I did buy some of that yet have not yet found a need yet. I am a mod guy and may have to go back into anything i do(God forbid!). I have been intrigued with the silicone tape but the "stretch" requirement seemed so critical it scared me, however with your experienced endorsement i'll put it back into my realm. Thanks!