My Headlights Stay Up After Adding Headlight Relays

[drtidmore]

Hal,

GREAT to hear that you got things working. As for the headlight door partial retraction, I would compare the time it takes both to open as well as the rate that they close. If there is more than a fraction of a second difference, I would chase a mechanical binding issue first. Remember that the HDM shuts down power based on the fact that the motor presents a changing load to the HDM as it hits the mechanical stops, so any mechanical binding can trick the HDM. Curiously, if you manually crank it down somewhat (less than where it is partially stopping) with the headlights ON and THEN turn off the headlights, does it close as expected or stop at the same point?

David T

[2seater]

The headlight gave me a clue by starting to squeek, which it also never did before. This motor assembly had just been apart for new rollers and a check over. Interestingly, the rollers were almost perfect on this one. A small crack in two of them but all completely intact. I had done the right side some time ago and one was dust, one broken in three and one just mishapen. In any case, I had this one completely apart to check the brushes as well. I think when I had it apart, I had wiped down the shaft below the brushes to keep from contaminating them and was not lubed adequately when reassembled. I removed the endplay adjustment screw and nut and sprayed white lithium grease past the worm section on the shaft and reassembled. I also adjusted the endplay on both motors as I noticed they still had a fair amount of play and banged down when closing. Closing up the endplay in the shaft made a huge difference. The final result is the headlights now work perfectly and almost silently. The full travel has been restored, HDM is happy and the relay modification works perfectly. Oh happy day.

[wws944]

So the '90, and by extension, '91 HDM needs +12v on A to raise the doors and ground on C to lower the doors. Just like scheme #1 in MSG #33 on the Fiero thread said.

So what about '88-'89 Reattas? Looking at page 102-0 in the '89 FSM, the wiring is definitely different than '90/'91. That is, HDM input A is connected to the headlight door switch, and input C is connected to the headlight switch with a diode between the two. This seems to correlate with scheme #3 in the Fiero thread.

The other thing is why the Reatta Illustrations manual lists the same HDM part number for all years Reattas? Looking at the schematic, if a newer HDM is used in a '88-'89, input C is grounded through the headlight filaments. And that - of course - would detect the case where both filaments were bad! So does that mean that a '90-'91 HDM could be used in a '88-'89, but not the other way 'round?

[2seater]

I just tested three HDM's I received from DAVES89, one is from an '87 Firebird for certain, and the other two we believe are from the same source. I agreed to test them as long I was working in that area. I can report that none of them work on my early 1990. No identifying marks except one has a small unreadable paper tag that appears to say "replace, do not repair", the same tag that is on the back of my original module.

Maybe this should be in the Headlight module thread?

Edited September 8, 2013 by 2seater (see edit history)

[DAVES89]

Hal,

I guess you will have to bring them back and I will try them on both my '89 and completely stock '90. Not to knock your modification, but we probably should just know for sure.

[drtidmore]

So the '90, and by extension, '91 HDM needs +12v on A to raise the doors and ground on C to lower the doors. Just like scheme #1 in MSG #33 on the Fiero thread said.

So what about '88-'89 Reattas? Looking at page 102-0 in the '89 FSM, the wiring is definitely different than '90/'91. That is, HDM input A is connected to the headlight door switch, and input C is connected to the headlight switch with a diode between the two.

This is ONE in the SAME. The difference is the 90/91 did NOT have the separate headlight doors open switch on the console, but it appears that the HDM modules worked identically. THe 88/89 Headlight Door Diode provides the necessary +12 volts to the HDM "Headlight Doors Open" input (A) when the headlight switch powers up the headlights, BUT blocks the +12 v from powering up the headlights when the separate console switch is used to open the doors.

As for the grounding, in both 88/89 and 90/91, GM applied +12 to the HDM "Headlights On" input © when the headlights were powered (appears to be unnecessary and simply a matter of convenience and using the SAME wiring to both POWER the headlights as well as providing a path to ground for © when the headlights are turned off (headlights acting as a pull down resistor since the other side of the headlight is tied to ground)

What we found in the process of adding separate headlight power relays to remove all the high current off the factory headlight switch and factory wiring was that the HDM © input was internal to the HDM tied to a +5 pullup resistor which if applied to the added relay would HOLD it engaged for about 5-10 seconds AFTER the headlights were powered down. The HDM has a 5-10 second timeout on its operation so as to prevent it continuing to power the headlight door motor incase of a problem with the door mechanism. This was not all that bad a thing (ie the added relay holding for 5-10 seconds), but something that we found could be avoided entirely.

So that led to the decision to isolate the HDM © input and tie it to BOTH 87a contacts on the added respective high and low beam power relays. Allowing the HDM © to float when the headlights are up and running is not a problem. By tying it to BOTH 87a contacts, when the headlight are powered down, the added relays disengage and the HDM © input has a path to ground through BOTH headlight beams which along with the HDM (A) going to zero volts allows the headlight doors to close.

We know from the operation of the 88/89 HDM that simply grounding the HDM © does NOT, I repeat, does NOT close the headlight doors as that is exactly the situation that the factory setup presents to the HDM when the console door switch is activated (ie +12 to HDM (A) and HDM © tied to ground through the headlight filaments.

So once again, it is NOT the grounding of HDM © that closes the doors, but rather that BOTH HDM © has to have a ground path AND HDM (A) must be at zero volts for the doors to close. The HDM © is NOT the cause of the door closure, but rather a path for some of the internal HDM circuitry to be grounded.

David T

Edited September 8, 2013 by drtidmore (see edit history)

[drtidmore]

Finally the hot temps here in Dallas are breaking a bit so I have started my headlight relay mod. I decided to forgo my original intent to use 4 relays to control each headlight beam separately (ie just using 2, one for low beam, one for high beam) and since GM used 14g wire in the headlight wiring, I am going to continue using the factory wiring that feeds the passenger side headlight (I never installed higher wattage bulbs, so no chance of previous overheating) as it already is present at the drivers side and therefore does not require me running new wires over to the passenger side. My new HID Low / Halogen High setup are the same wattage as the factory bulbs and 14g wire is completely adequate for the run from the new relays on the drivers side over to the passenger side headlight PLUS it is already in place. I mounted my relays between the air cleaner housing and the fender on the air cleaner housing bracket (they just fit). With the air cleaner lid in place, they are completely hidden from sight, yet easily accessible for replacement.

I used Bosch/Tyco 30amp SPDT automotive relay (found them at Fry's for $5.99 each).

I did get a bit of a "you know better" reminder about the need to install a clamping diode across each relay coil (cathode/banded to +12 side of relay coil, anode directly to ground). I was testing for a solid grounding point in the intended relay area and rather than using my ohmmeter was just using the relay coil temporarily connected to +12V (battery) and before I connected the clamping diodes. I had crimped both relay coil ground wires into a spade connector and was holding it with my fingers. After finding a good ground, I released the temporary connection and as the relay coil magnetic field collapsed, I got a git of a nasty shock. Nothing serious, but far more than enough to surprise me. Fortunately there was NOTHING else connected that might be damaged by the reverse voltage surge. The clamping diodes I used (Fry's) are rated at 1000V, 100a, 70ns with a 1.1v forward drop. Definitely overkill, but Fry's didn't have a huge selection of high current diodes and it certainly does not hurt that what I installed is overrated for the job. I was looking for a FAST diode so as to ground that reverse voltage immediately on the release of the coil hold in voltage.

I also did the combo high/low beam that I had mentioned by simply putting a different diode (100v, 3A, .5v forward voltage drop) between the low and high beam wires from the factory headlight circuit (cathode toward the low beam side). Works perfectly. The important thing with this diode is that it can handle the initial surge current of the relay coil and has a small forward voltage drop to ensure that the low beam relay is fed as close to the +12 as possible whenever the high beam voltage from the factory high/low relay is powering it.

While my original intent was to NOT modify the factory wiring other than the HDM © input, after getting into the project, I realized that cutting the driver's side factory headlight socket off and splicing was a better solution from a failure point perspective. I really NEVER intend to return this to original but I will keep the removed headlight socket regardless.

FINAL Testing:

Work was slack, so I finished up the final stuff this morning. Testing went as expected. Low beams->Off, beams immediately off, door retract; High beams & Low Beams work at the same time; High/Low beams -> Off, beams immediately off, door retract.

Also I now have the FULL battery voltage AT the headlights, whereas before I was at least a volt or so below battery due to resistance in the headlight switch, factory high/low relay and the factory wiring. This is how GM SHOULD have done this originally.

With ALL the high current now off the factory headlight circuits, the voltage drop where I connected the new relay coils (at the factory driver's side headlight socket location) is insignificant so the relays see for all practical purposes, battery voltage regardless of engine running status.

I have taken some pics and will do a nicer schematic as I know that a drawing is a lot easier for many to follow than the text that has dominated this thread.

Now to just tweak on the aiming of the HID beams as they are bit too low, at least compared to my vette's HID pattern. Have to say that even with the less than optimal aiming, the HID are an enormous improvement and remember these are still just 35W beams.

David T

Edited September 12, 2013 by drtidmore
final testing (see edit history)

[2seater]

David, perhaps your car is different than mine but all of the headlight wiring in my '90 is equivalent to 16ga, listed in the manual as .8(mm?), with the cross reference chart in the FSM, and confirmed by my wire stripper. I agree completely if it is 14ga, I would have left as is even @ 100w, but mine was not.

[drtidmore]

And so we find another area where the 88/89 vs 90/91 look to differ.

In the '89 FSM the Yellow/Black wire (input A, Headlight Doors OPEN) to the HDM is listed as .8(mm), but the Yellow (input C Headlights ON) as well as the wires throughout the headlight circuits and to the headlight sockets and from the drivers to passenger side are listed as 1(mm).

I did find the same FSM conversion chart from metric wire sizes to AWG, but that chart does NOT jive with online or on hand conversion charts. My references show that .8mm equates to 20 gauge (actually .821mm) and 1mm equates to 18 gauge (actually 1.024mm) and even without a comparison I can say that the wires that I encountered during the mod were larger than 18 gauge wire. (FYI 1.628mm equates 14 gauge, 1.291mm equates to 16 gauge).

I did not actually take a micrometer to the wires, but I used 14g wire for all new current carrying wires with the exception of the new relays main +12 feed (used 10gauge as I intended to power high/low simultaneously), so I had a visual comparison against the factory wiring (ie actual stripped wire size, NOT outside insulation diameter as I will explain below).

GM appears, at least in my '89, to have used a teflon-based insulation which is thinner (ie smaller wiring bundles), handles heat without deforming/failing far better than typical wire insulation and does not become brittle with age. I have used this type of insulated wire over the years and it has a stiffer, more difficult to strip characteristic, which conforms with what I encountered on my '89. I could be wrong on the insulation, but it certainly appears that way to me.

Before cutting into any of the wires, I looked at the outside diameter (ie insulation) and was sure that I would discover wire gauge outside my comfort zone, but the insulation was misleading (again, appears to be teflon-based). It was ONLY after stripping back the insulation and finding wire that certainly looked on par with the new 14 gauge wiring that I was using in the mod that I decided that I could live with the short factory run from the drivers side headlight area over to the passengers side headlight area (at least on the '89, that run is DIRECTLY across the radiator area at the top).

Given the way the GM originally wired the headlights so that the overwhelming majority of the wiring carried current to BOTH sides , the total current loads of 6amps (3amps each low beam) or 11amps (5.5amps each high beam) is still well within the capacity of 16g wire (rated to carry roughly 20amps DC). The section of factory wiring from the drivers side over to the passenger side ONLY carries the load of the ONE headlight, so we are talking 3amps(low beam) and 5amps(high) for the factory 35w/65W halogen headlights. I personally would NEVER depend on 16g wire to handle 20amp for anything other than short runs (a few feet), but it does appear that GM was NOT low balling the wire size when they designed the wiring harness.

Since my HID lamps (ie the digital ballast) pull 3.2 amp (35W HID low bulbs) and the high beams pull 5.5amps (65W Halogen bulbs), basically the same as factory, I was NOT concerned using the crossover wiring to the passenger side once I saw the actual wiring (the drivers side is ALL new wiring due to short distance to new relays). With the relay mod, I have removed everything except the trivial current of the two relay coils from the critical areas of the headlight switch, high/low relay and all the splices/connectors.

Hal, even in your situation with 80/100w aftermarket halogen beams, you were still well within the limits of 16gauge wire, but it appears not within the limits of the headlight switch, and factory relays. I suspect that the melted headlight sockets you found were a result of direct filament heating from your higher wattage beams, NOT from pulling excess current through the wires, but the discolored high/low relay clearly indicated that it was not able to handle the added current without overheating. Your experience of two failed headlights switches is a bit of mystery as I think about more. Maybe I misunderstood how the '90/'91 factory ENABLE relay mod was connected, but I thought it basically moved the current load from the headlight switch and ran it through a relay, which should have protected your headlight switch. Since I don't have the '90 FSM for reference, I am going on what I have read on this forum regarding the factory ENABLE headlight mod.

I certainly don't blame anyone that wants to replace ALL the current carrying headlight wiring (a secondary reason for the relay mod), but the short run of the factory wiring between the drivers and passenger side headlights, does not seem to be problematic due to it short length, even at 16 gauge.

David T.

Edited October 4, 2013 by drtidmore
found I had HDM input A&C swapped, reads correct now (see edit history)

[2seater]

David,

If my '90 HAD the enable relay modification, the issues I experienced likely would not have happened, but mine is an early model that carried all of the current through the switch.

I am now less certain that the original relays actually were damaged. I found nothing inside and the discoloration on the external termnals appears to be common to all of the OE relays I have pulled. The melted headlight sockets, primarily the ground, was worse on the passengers side. No empirical data to back this up, but I guessed that the excessive voltage drop to the headlights caused a higher than normal current draw, causing extra heat??

[drtidmore]

Hal,

Thanks for the clarification. Now there is NO doubt that frying of two headlight switches WAS due to excessive current. Even on low beam, you were drawing 40W MORE (20W per headlight) than stock High beam.

On my '89, I was dropping between 2-3 volts due to internal resistance and that was with 35w low beams. As the headlights and all the factory circuits form a series resistance circuit with each point of resistance (headlight switch, high/low relay, spices, connectors) dropping a small voltage. The amount of actual current the entire circuit would pull would depend on the total resistance, regardless, you were running more current that GM ever anticipated.

The melting that you saw was more likely to have been just local heating from the higher wattage headlight filament (your low beam was 20W per bulb higher than the factory design high beam). Those contacts on the back side of the bulb are thermally conducting the heat of the filament directly to the socket. It could also have been due to a poor connection between the ground contact in the headlight socket and the bulb.

Glad that you were able to get everything up and running and no longer have to worry about stressing the factory circuits, I KNOW that I am glad to know that all that wiring behind the dash is no longer carrying the high current of the headlights.

David T

[wws944]

The right-angle headlight sockets I used, Dorman #84790, claim to be rated for higher than OE (55/65w) wattage bulbs. They don't explicitly state how much higher the sockets are rated for. But do claim temperature rating of 482 degrees F.

With H4 bulbs and housings, the housing itself can provide some heat sinking for the heat from the bulb. Makes a case for metal housings, such as the Hella, over plastic ones.

[Guest Mc_Reatta]

The right-angle headlight sockets I used, Dorman #84790, claim to be rated for higher than OE (55/65w) wattage bulbs. They don't explicitly state how much higher the sockets are rated for. But do claim temperature rating of 482 degrees F.

What are they made from?

Would think ceramic or some thermoplastic.

[wws944]

What are they made from?

Would think ceramic or some thermoplastic.

Says on the box "high temperature nylon".

[drtidmore]

Some plastics can withstand substantial heat, but NOTHING beats the ceramic sockets that are available. Just as important as the socket is the insulation of the wires leading into the socket. These typically are high temp thermoplastic, but can be teflon as well (teflon more expensive but better at handling the heat). It is usually a safe assumption that the wire insulation chosen is in line with what the socket itself can withstand.

In my HID/Halogen upgrade, the manufacturer's bulb fixture (holds both bulbs) terminated with high temp wiring that plugged into the supplied wiring harness, so I did not have the bulb heat as a point of concern. As supplied, the HID harness plugged directly into the existing headlight socket well away from the heat of the headlight. I kept this arrangement for the passenger side headlight, but due to the installation of the relays to bypass the headlight power away from the factory circuits, I wired the driver's side directly into the created relay harness (yes, I cut off the driver's side headlight socket in the process).

David T.