What's new
What's new

OT: Basic electrical question about adding diodes to a hitch light

rimcanyon

Diamond
Joined
Sep 28, 2002
Location
Salinas, CA USA
Since some of the EE's hang out here, I hope you can help me with a simple problem. I bought a Harbor Freight LED hitch light for my Ford T350, which has a factory hitch and trailer wiring. The Ford computer doesn't much like the HF hitch light, which uses the 4-pin adapter and connects to pins 1 and 4. Since pin 3 is not loaded, the computer thinks a trailer is connected and it has a burned out bulb on the left side, so it blinks the left turn signals at double rate. That is the problem I want to fix.

I think that if I rewire the HF hitch light with two diodes, so that there is one diode from pin 3 to the light's hot lead, and another from pin 4 to the light's hot lead, then the light should work as follows: the hitch light will be on whenever either or both taillights are on. i.e. if the vehicle is turning right or left the hitch light will flash, and it will be on steady if the brakes are on or the taillights are on, and the computer will be happy, so both sides will flash at the same rate. Do I also need to add a resistor between each diode and the light's hot lead, so that the circuit is slightly more loaded?

Assuming the above is correct, is there any simple way to get the hitch light to also blink when the turn signal is on and the taillights are on? That is what it does currently when turning right with the lights on, but I think that feature will disappear once pin 3 is added to the circuit, as described above.

-Dave
 
You should have green yellow, white and brown wires, unless Ford is different from what I have seen on others.

Green and yellow are for "Stop and turn" signals, white is ground, and the brown is the running lights.

If you do a diode from green and from yellow, that should keep things happy because the single light will flash for either turn and stop. I don't know about the running lights, they are usually separate, and getting one bulb hooked up to be everything is a bit more complicated.

You probably can do it by having a resistor from the brown to the light, and hooking the diodes directly to the light, bypassing the resistor. That way the running light would be dimmer, and the turn signal would be brighter. I do not know what is a good resistance, that depends on the bulb you have.

Might be best to put a diode in the brown wire too, so as not to back feed it with the turn signals That might not be a problem, but I do not know what the computer would do... it might think there is a fault in the running light controls.

What is the reason for it all? I do not know if blinking with the turn signal is a good idea.

If you just wanted it to be on, no turn signals, the brown wire and white is your connection.
 
You should have green yellow, white and brown wires, unless Ford is different from what I have seen on others.

Green and yellow are for "Stop and turn" signals, white is ground, and the brown is the running lights.

If you do a diode from green and from yellow, that should keep things happy because the single light will flash for either turn and stop. I don't know about the running lights, they are usually separate, and getting one bulb hooked up to be everything is a bit more complicated.

You probably can do it by having a resistor from the brown to the light, and hooking the diodes directly to the light, bypassing the resistor. That way the running light would be dimmer, and the turn signal would be brighter. I do not know what is a good resistance, that depends on the bulb you have.

Might be best to put a diode in the brown wire too, so as not to back feed it with the turn signals That might not be a problem, but I do not know what the computer would do... it might think there is a fault in the running light controls.

What is the reason for it all? I do not know if blinking with the turn signal is a good idea.

If you just wanted it to be on, no turn signals, the brown wire and white is your connection.

JST, thanks for your reply. With the ford 4-wire connection, 1 is ground/white, 2 is tail lights/brown, 3 is left turn & brake/yellow, 4 is right turn & brake/green. I want the hitch light to be on whenever the brake lights are on or the left or right turn signals are on. I made an error above, I do not want the light on when the tail lights are on.

-Dave
 
JST, thanks for your reply. With the ford 4-wire connection, 1 is ground/white, 2 is tail lights/brown, 3 is left turn & brake/yellow, 4 is right turn & brake/green. I want the hitch light to be on whenever the brake lights are on or the left or right turn signals are on. I made an error above, I do not want the light on when the tail lights are on.

-Dave

Solved a similar challenge on a '68 FIAT I had traded by '72 for a BMW, so "a while ago".

The trick was to use transistors - cheap & common TO-3-cased 2N3055's for the lamps, 2 other small, common TO-220 cased types I no longer recall - instead of just Diodes.

That allowed putting enough resistance in the control input for high impedance "sensing", rather than direct switching, that the rest of the electricals were blissfully unaware of what I was switching, why, or when.

It wasn't all that costly, nor particularly complex.

"Somebody" in the Marine/RV/work-truck/towing accessory industry probably sells a general-purpose circuit already packaged to do that very sort of thing for brake, TS, running lighting, specifically to get around that inbuilt auto-fault-sensing you are fighting. Research should find one.
 
Bill and Doug, thanks for your suggestion. You both are correct, there are other solutions available. Call me stubborn, but I am looking for a simple solution that adds a few cheap components to the $6 HF light, rather than a solution that either picks up the wiring at a point where the computer is not so picky (e.g. at the tail light wiring), or involves using a black box that makes the computer happy but doesn't solve the problem the way I want it solved. The HF light comes with a 4-pin plug that only has pins 1 & 4 on it, so rewiring it with a new 4-pin plug plus a couple of diodes would be simple, if that solution works. I think someone makes 4-pin adapters similar to a null modem rs232 connector, I.e. they provide a simple breakout box where you can do some simple wiring, so that would work too. What I really am looking for is some specifics about how to pick up both turn signals, feed them into a single wire without allowing any current traveling the wrong way (back into the vehicle on the unenergized wire).

i called etrailer, and their suggestion was to install a load resistor to pin 3, Putco LED Light Bulb Load-Resistor Kit - Ceramic Putco Accessories and Parts P234C

that is a solution I was aware of, but it would mean the hitch light would only blink when the RTS is on.

Dave
 
Bill and Doug, thanks for your suggestion. You both are correct, there are other solutions available. Call me stubborn, but I am looking for a simple solution that adds a few cheap components to the $6 HF light, rather than a solution that either picks up the wiring at a point where the computer is not so picky (e.g. at the tail light wiring), or involves using a black box that makes the computer happy but doesn't solve the problem the way I want it solved. The HF light comes with a 4-pin plug that only has pins 1 & 4 on it, so rewiring it with a new 4-pin plug plus a couple of diodes would be simple, if that solution works. I think someone makes 4-pin adapters similar to a null modem rs232 connector, I.e. they provide a simple breakout box where you can do some simple wiring, so that would work too. What I really am looking for is some specifics about how to pick up both turn signals, feed them into a single wire without allowing any current traveling the wrong way (back into the vehicle on the unenergized wire).

Dave

Dave, I started out with the same goal. "Can't get there from here".

One might... put a high-enough value resistor in series with a diode to not load the 'puter's sensors to trigger-piont, but then.. there'd not be be enough current flow permitted to energize the LED's.

There's the part that's hard to fight if the LED's are any more power-needful than tiny "in cockpit" brake-light "health" indicators.

Which is all mine WERE, BTW. No on-board 'puters back in that era. Even so, I ended up having to use transistors as sensors/amplifiers.

It simply takes more energy to light up your LED'S than the Fordputer's "failure" sensor can be easily blinded to.

Catch 22.
 
I use two diodes to do basically what you want to do, to simplify my motorbike turn signal wiring. If you buy the stock 50 dollar flasher it has an indicator
terminal to light the single turn indicator lamp on the instrument cluster.

50 dollar turn signal relay? Hell no. I put a pair of 1N2007 diodes in. Diodes point from each turn signal bulb, to the single indicator lamp.

If you want to light the hitch light whenever each turn/brake line goes hot, this will do it for you. No other components needed. The diode listed above
is good for about one amp of current. Often it's cheaper to buy a bridge rectifier and only use two of the diodes if you want something good for 5 amps.
 
Not sure why there is a problem about the computer. If it can light the light, and nothing goes back down a line to trigger a fault, it should be happy
 
Not sure why there is a problem about the computer. If it can light the light, and nothing goes back down a line to trigger a fault, it should be happy

Without the diodes, 'something' will go back down the line.

" What I really am looking for is some specifics about how to pick up both turn signals, feed them into a single wire without allowing any current traveling the wrong way (back into the vehicle on the unenergized wire)."
 
Without the diodes, 'something' will go back down the line.

" What I really am looking for is some specifics about how to pick up both turn signals, feed them into a single wire without allowing any current traveling the wrong way (back into the vehicle on the unenergized wire)."



Which is WHY I support the diode idea, obviously.

It seems likely that a signal sent when there should not be one might trigger a fault code that means the brake light Mosfet or relay has failed "on".

The question is really why someone would think the diode solution would NOT work. So long as it is drawing current at a reasonable level, it appears all will be well. It would not surprise me if the computer were to know if the current is low enough that a bulb must be out, or high enough to trigger shutting off the drive.
 
Some bulb control modules are fussy. But I suspect the two diode fix would not be a problem.

Fussy: new (to me) motorbike with a fussy bulb control module. Attempted to fit an LED 1157 lamp to the tail light, it caused a fault
light on the dash. Attempted a fix by putting a shunt resistor in parallel with the brake portion, no good, still faults. Attempted a
fix by putting a small incandescent lamp in parallel, fault at startup gone, but fault light goes on whenever brakes are applied. Grrr.

Final fix is back to the regular incandescent 1157 lamp with an LED brake bar in parallel. sigh.
 
Attempted a
fix by putting a small incandescent lamp in parallel, fault at startup gone, but fault light goes on whenever brakes are applied.

So long as it is transient / self-resets, that could be seen as a "feature" rather than a "bug".

Lets you know everything is working as per (revised) plan, no?

That's what my "pre OBDC" DIY circuit on that '68 FIAT 124 Coupe did.

Pair of green lamps were alight, filaments were intact, providing the sense transistor with bias path to ground.
Brakes applied, bias was pulled toward supply rail, red lamps lighted-up.

Red lamp, either side, when brakes NOT applied, filament was open or socket was corroded.

Why the extra care?

With good tires, brakes, alert driving, I could stop safely. That not always being true of the person(s) behind me, I had been tail-ended thirteen times in the first 25 years of Metro-DC commuting. Once every two years, on average, for a quarter of a century, straight.

:(

It always helped with the inch-hoorance to be CERTAIN I had working brake lights at all times.
 
"So long as it is transient / self-resets, that could be seen as a "feature" rather than a "bug".

I'm a firm believer in the "no red lights unless something is wrong" approach. Vechicles should not
have flashing hoo haa lights if there's nothing amiss. Distracting, and tends to make the operator
iignore lights that mean there really ARE problems.

See also the pilots who flew their plane into the ground because they were all worried about a burned out
light bulb. And nobody was driving.
 
Some bulb control modules are fussy. But I suspect the two diode fix would not be a problem.

Fussy: new (to me) motorbike with a fussy bulb control module. Attempted to fit an LED 1157 lamp to the tail light, it caused a fault
light on the dash. Attempted a fix by putting a shunt resistor in parallel with the brake portion, no good, still faults. Attempted a
fix by putting a small incandescent lamp in parallel, fault at startup gone, but fault light goes on whenever brakes are applied. Grrr.

Final fix is back to the regular incandescent 1157 lamp with an LED brake bar in parallel. sigh.



It "knew" about the current the thing should pull, maybe even knew about incandescent bulb inrush..... I wondered about that, since the whole point of LED lights is to pull less current/power, and presumably (maybe not in actuality) demand less power from the alternator and engine. If it is sufficiently fussy, it may urp at the LED, as in your case.

Just remember, every one of these safety oriented "lamp provers" is another thing to go wrong. Each one, if it goes wrong, is a potential need to replace $1200 worth of vehicle computer (that cost $75 to build).
 
It "knew" about the current the thing should pull, maybe even knew about incandescent bulb inrush..... I wondered about that, since the whole point of LED lights is to pull less current/power, and presumably (maybe not in actuality) demand less power from the alternator and engine. If it is sufficiently fussy, it may urp at the LED, as in your case.

Just remember, every one of these safety oriented "lamp provers" is another thing to go wrong. Each one, if it goes wrong, is a potential need to replace $1200 worth of vehicle computer (that cost $75 to build).

And don't think it doesn't bug the heck out of me. Battery, switch, light bulb. Can't be simpler, right? No. There has to be
a lamp control module to see which one is there or not. Grrrr.

I checked to see how much current was going through the lamp when off, to see how to fool it. Far as I can tell it looks for zero
volts across the lamp in the off state (via test current) and a certain minimum current through the lamp in the on state. Fail either
one or both and it lights a large red light on the instrument cluster. I know guys who simply black tape that red light.
 
I use two diodes to do basically what you want to do, to simplify my motorbike turn signal wiring.

I put a pair of 1N2007 diodes in. Diodes point from each turn signal bulb, to the single indicator lamp.

If you want to light the hitch light whenever each turn/brake line goes hot, this will do it for you. No other components needed. The diode listed above
is good for about one amp of current. Often it's cheaper to buy a bridge rectifier and only use two of the diodes if you want something good for 5 amps.

Jim, I have a bunch of diodes left from repairing Monarch 10EE modules. They are 1N5408. I think they are rated for 3A, so I will give them a try. They look pretty substantial.

Dave
 
The hitch light is now working properly. A pair of 1N5408 diodes was installed, one each from pin 3 & 4 on the 4-pin connector to the + connection on the hitch light circuit board.
Now the hitch light blinks when either the lts or rts is blinking, it is on steady when the brake lights are on, and the computer no longer thinks the left trailer light is burned out so both sides blink at the same rate. The only hard part for me was getting the diodes soldered to the circuit board traces, since I had to drill through the circuit board and the traces were narrow.
 
I don't think you really, really want the center light to blink with the turn signals.


You may be cited by a cop for this operation you describe.


Turning right ? You want only lights on the right side to blink.
 
Glad this worked out the way you wanted.

As far as police concern, heck around here folks drive at light with NO lights on at all and they never get stopped.
Blue headlights, green headlights, weedburner headlights, anything goes.
 








 
Back
Top