I don’t know the solution or root cause yet, but I am sure that I will in time. Car is an ‘88 5.3 L. A few years ago I installed Hella headlamps with 60w H4 bulbs; at the same time I added headlamp relays to reduce voltage drop- the relays are supplied by the left firewall buss post through a fusible link. I also added a supplemental ground from the engine to the upper radiator support using the same bolt that the right headlight grounds are secured with. All connections are clean and secure.

Recently I noticed that after a cold overnight soak, when the cold engine lock out prevented the HVAC blowers from coming on, when I turned on the headlights I could hear the a/ccompressor clutch engage. With the cold weather (40 days that the temperature never got above 32f), this happened pretty consistently, and compressor would run for maybe 30 seconds and then shut off. After driving a mile or so, blowers would come on normally. Delanaire system has always worked normally. Tried another Delanaire controller with no change.

So, yesterday I decided to upgrade the headlights and installed LED H4 replacements. Drove the car last night on an errand and noticed the the blowers came on at a low speed when the engine warmed. When I reached my destination, I turned off the headlights and the blower speeds doubled. Lights on blower low, lights off blower high. No change in system voltage either onfactory gauge or cigarette lighter monitor. Shut car off for 15 minutes, did same thing on restart. This morning, 30 degrees out, compressor on for 40 seconds when lights turned on, blowers came on as engine warmed but low speed when lights are on, high speed when lights were off.

There is no electrical connection between the two systems, especially when you consider the relays further isolate the two. Even the Delanaire panel is fiber optic illumination. I access and cleaned the right instrument panel ground (where the right blower and Delanaire system gets its ground; this was before the LED lamp installation.

I believe that the LEDs are somehow messing with the pulse width modulation that controls the blower speeds; may be that the higher draw of the 60 w H4 was causing the initial compressor malady, and the higher draw of draw LED are exacerbating it, what say you?

 

Gremlins, thats it.

Or, could be a coincidental electrical issue.

Remove the relay, see if the same thing happens. Turn lights on via the switch even though the relay won't turn them on. See what happens

If it's not coincidental, you may have a feedback issue of some sort, which can be fun to solve.

I doubt the LEDs are causing dirty power, but I believe a capacitor in line would smooth things out.

Doug

 

Why did you add another headlight relay, as there is already one in the system?
There is an OEM diode controlled system that allows the radiator electric fan to keep running if it is engaged when the engine is switched off, until the water pump inlet thermostatic switch cuts it off. Additionally the radiator electric fan on some models is automatically activated whenever the aircon comprressor comes on, so if you have backfed this area (even somehow via the OEM radiator fan relay wiring, as that is close to the OEM headlight relay) it could make the aircon cut in via current being backfed to it.
So maybe your extra wiring has somehow energised this system via earth feedback, causing the HVAC to activate.

 

That was my original suspicion also.
The relays that I added feed the headlamp fuses directly from the buss; so that the additional current of 60w bulbs no longer has to go through the headlight switch. The only current through the headlamp switch is what the relay coil requires. I am aware of the air conditioning relay and how it functions, and the fact that it and the left headlights (along with the condenser/radiator fan) ground at the same point (G2). I have ohmed out and cleaned the grounds and they measure as low a resistance as my Fluke meter will read (.1-0 ohms).

 

If the headlights are being fed directly from the firewall post, via a new relay, and the OEM relay wiring is intact, I think (not sure as understanding wiring diagrams are not my strong suit) I think that the current would backfeed to the OM relay.
Do you have the wiring diagram, if not I can post it.

 

[img alt="I don’t see how it would backfeed anything- basically what I did was take the top wires in the diagram out of the fuse box and used them to power the relay coils. I then connected the commonof the relays to the buss terminal via a fuse link. Than I connected the NO relay terminal to the fuse box where I had removed the wires. Headlight switch and relay now only carry the supplemental relay coil current.
I have installed two halogen seal beam lamps; on the only cold start so far, everything ok."]https://cimg1.ibsrv.net/gimg/www.jaguarforums.com-vbulletin/2000x1504/image_c19e21c3e28d3446b3ff41b6b554447727cf5d68.jpg[/img]

 

This image is at the bulb with an oscilloscope on the bulb terminals fed from a well filtered supply.
That is about 250mv at somewhere near 50kHz; the amplitude would vary with the supply impedance. The frequency seemed consistent but the amplitude varied with the supply voltage (above was at 13.5V). There is a cooling fan on the bulb, but its rpm remained constant from 10-14.5V. Also the bulb brightness was constant from 11-14.5V, which makes me believe the switching supply built in the bulb uses PWM switching to regulate LED current.

The analog ammeter on the supply indicated 1.5a throughout the voltage range. So I believe the hash above was interfering with the blower PWM circuitry.

 

I especially love a good challenge, but three issues stand out here that make help difficult until resolved . . .
. . .

1. these issues only arose with your modifications . . . is that correct?
2. with electrical / electronic schematics, show rather than tell by colour coding what is OEM and what is changed or added;
3. we are left to guess, but I sense a problem within your following remark . . . "I then connected the common of the relays to the buss terminal via a fuse link. Then I connected the NO relay terminal to the fuse box where I had removed the wires."

That emphasis is mine but useful because it indicates a dual feed anomaly via both the buss and powered fuse box because, IIRC, those positions where you have removed wiring and inserted new circuit(s) are internally powered . . . no problem when the NO is open (ie lights off) but, once the COMMON closes to the NO, (ie lights on), you are linking these otherwise separate power sources.

Still guessing, but I suspect Greg is spot on about feedback, but . . .
perhaps not feedback via grounds . . . more like backfeed of power.
If so, you have linked originally separate circuits . . . producing a dual feed anomaly.

Going forward, I like the idea of testing with your relay pulled.
Or . . . retest with all your mods removed,

Cheers and best wishes,

 

After a few days of cold soak starts there has been no further abnormalities, the only change was removing the LED bulbs, so I believe the switching supply internal to the LED bulbs is the root cause. The OE configuration has the headlight fuses fed from the OE headlight relay via the headlamp switch; my revision has these feeds only carrying relay coil current. My added buss source feeds the added relay common (high and low beam have individual relays). There is no feedback path.