Five Equals Four with a Diode OR: Something Didn't Add Up Here!

How this Gadget Freak Began

U-Haul rents trailers with 4 wires, called a 4Flat. My van has a combination tail light that needed a 'diode OR' circuit in order to properly run their lights. Would this be a good idea for another Gadget Freak article?

Five Equals Four with a Diode OR Original Gadget Freak Case File

Original Submitted Van Trailer Build Instructions

Trailer rental companies insist that tow vehicles provide operating lights – turn, tail and brake lights for every towed vehicle rented. Easy to do if the tow vehicle (our van) was wired that way. It was not. The standard connector on rental trailers is called a 4-Way Flat – one female and three male pins. The female is a ground return and the 3 male pins bring 12 VDC power to the tail lights and the right and left turn and stop lights. The combination turn and stop lights are where our troubles lie – our van, a 1996 Dodge Grand Caravan, has separate turn and stop lamp circuits. The solution was a diode OR circuit for each side’s lights, then either the van stop signal lit both trailer lamps or that side’s turn signal would flash only that side’s lamp. I used Schottky diodes for low voltage drop, rated 3 amps, which would handle 36 watts of trailer lights. Throw in a few quick splice connects for the rear tail light wires and the mate to the trailer’s 4-Way Flat and you’re set quicker than you can say “You haul!” Oh, yeah – be sure the vehicle and tow hitch can handle the trailer, cargo and tongue weight.

Bill of Materials

The Gadget Freak Debate

Randy L Wrote: Did anyone such as the author test the circuit given in this latest design idea?

It is very simple (too simple), and it looks to me like only the turn signal function will work. If the brake pedal is pushed, it will cause not only the trailer stop/turn lights to turn on, but also all four turn signals on the tow vehicle to turn on. This would be front/back and left/right.If the brake pedal is pushed with the turn signal activate, the blinking turn signal will stop and again, all four tow vehicle lights will turn on solid, along with the trailer lights.This would seem like a dangerous situation.Adding two more diodes so that the brake light power would not backfeed into the two vehicle turn signals would improve the situation, but it still does not work properly since there will be no turn signal blinking on the trailer when the brakes are applied. I don't believe proper operation can be done with diodes alone.
Tony G: This Euro to US tail light problem has been coming up for years and the solution given is probably the first thing that comes to an engineers mind since it seems at first like an obvious solution. You will not find a commercial version of this solution because, I suspect, it is illegal in all US states (probably outside the US as well).

The problem with the solution is that while it allows brake lights light the trailer lights and it allows turn signals to operate on the trailer, it does not allow turn signals to function on the trailer when the brake lights are on. Let us assume you are getting ready to make a right turn at the next corner. You apply brakes to slow down and the trailer brake lights turn on. Next you turn on the right turn signal. The trailer right brake light will not flash because it is forced on by the brake light feed. There will be no indication of turning until you take your foot off the brake after turning the corner. More logic is required than two diodes to make the lights flash in a legal manor.

This solution provides a possibility for accident, injury and subsequent lawsuits. At a minimum the driver can get a citation. I did not see any disclaimers so legal action might occur against Mr. Lindstrom, Allied or Reed Business should any serious accident occur based on someone else implementing this gadget.
You might want to think about this and pass it by someone on your staff that is familiar with vehicle codes in your state.

Patrick D: The diode or solution for trailer lights will not work in the case where you are stopped at a stop sign with your blinkers on. The stoplight signal will override the on/off action of the turn signal through the diodes. In most cars, stopping will also turn on all your sidemarker indicator lights (not a desired function). This is a poor design and should not be relied upon.

Dick N: Looks to me like with the diodes wired in as shown, stepping on the brake will turn on all the tow vehicle’s Turn lights, fore & aft! He needs 2 more diodes, one in series with each Turn signal lead from the tow vehicle, to isolate the vehicle's Turn lights from the OR connection. These would be added in series with the Brn/Red and DGrn/Red wires, cathodes to the cathodes of the existing diodes.

Attached is the newer, better design. **SHEESH**

I feel terrible as an engineer- like my spaceship just crashed.

The circuit needed is an 'XOR' switch, more parts but totally correct operation.
No beg-offs with dim bulbs, sorta phantom operation...

Click here for John L's better trailer hitch design.

Dick N on the revised schematic: I think this one has a problem too. When B is active and T is not, T is connected to B through RLY1's upper contact and RLY2's NC contact. This applies power to T, which lights up the tow vehicle’s Turn lamps, and also turns off RLY1, making a buzzer out of it!

It's not clear where "To other Turn Ckt" goes, but I would assume it's to the corresponding point in that circuit. This only connects the two B's together whenever (B & ~TL & ~TR) where TL and TR are the car's L & R Turn signals.

Aren't they tied together anyway in the car's wiring?

It would suffice to use an XOR function, i.e. L = B^TL, R = B^TR. This would, of course, invert the Turn flashing when the brake is on, but it's not that unusual to see front and rear turn signals flashing in complimentary fashion, so why not likewise the trailer and tow vehicle? To accomplish in-phase flashing of the trailer's and tow vehicle's turn signals while the brake is ON would require a signal indicating the turn function is active, even during the OFF part of the flashing cycle. This signal is unavailable at the tail-light.

It could be synthesized with a 1-shot if in-phase flashing is really important.

Since RLY1= B&~T, and RLY2= T&~B, the XOR could be implemented by combining the two FET outputs. RLY1|RLY2 is equivalent to B^T. It's tempting to consider using 1 relay per trailer lamp, but then we'd need a +12v that's always ON - also not available in the harness. If it is available and each XOR is implemented as 2 FET's and 1 relay, remember to use series diodes to isolate the Drain terminals. Otherwise T and B will backfeed into each other via the Drain-Source diodes.
It might suffice to connect the trailer lamps in place of the relay coils in this modification, but beware the 6 amp max of the IRF7241. This connection would not require the always-on +12v, but would still require the isolation diodes.

Randy L on the revised schematic: I took a look and unfortunately I don't think this design will work correctly either.

It would be clearer to the designer if the entire circuit was drawn, then he/she can be seen what interactions will take place.The FET and relay circuit block itself seems fine. Use of the surface mount package FET might give the implementer some challenges in assembling the unit.I think the turn signal relay should have the connection to the NO terminal instead of the NC terminal. With this fix, the brake light function and the turn signal function would each work separately. It is not clear what the "T" input to the brake light control relay is connected to. The right turn or left turn?There still is a problem with correct operation when both brake and turn signals are activated at the same time. Unless I don't understand how the designer intended the "other" turn signal circuit to be connected.

Tony G on the revised schematic:I see a couple of drawing errors. RLY1 points to DPDT (double pole double throw) contacts and the 821-4200 relay called out is SPST (single pole single throw). The problem with the DPDT relay needs to be fixed. If he meant a DPST then the 841-2400 part number would be correct. I don’t think the logic works when I analyze the circuits unless errors were introduced going from an original working drawing to the PDF version I received.

The author may need to draw out the actual three subcircuits with interconnect to show how he intended this to actually work and where the Bs and Ts get hooked together. The schematic pdf with some minor changes seems to work for one turn signal on the trailer. I put one brake and two turn signal circuits into my Multisim simulator last night and tried to get it to work by moving the wiring around. Brakes seemed to work but one turn signal made both trailer lights flash. The brake circuit in the upper part of the diagram has a T input. There are two turn signals coming from the car, left and right, but only one input for the circuit. I suspect the two turn signals are diode ORed before they reach the T in the upper left of the schematic. Before I spend too much more time looking at this, please get the author to give the entire schematic needed to make this actually work. I may play around with the wiring some more tonight to see if I can get it to work.

Just for reference, the circuit must meet all, or depending on the state nearly all, the following criteria:

Additionally, the circuit load on each light circuit must be low in order to keep older cars with thermal flashers from flashing faster when the trailer is attached. The GF author meets this requirement with the FET input. The relay chosen could also be used without the FET as it is very low power. Also, as indicated by the author, the trailer conversion circuit cannot affect the tow cars light operation. The author does this with diodes on the circuit inputs.

When you drive your car around town you should be able to notice that your car lights work like described above.
I had not expected this to turn into such a big project. It is just a “simple” trailer light converter. I would like to see this project get finished though.

Randy L on the revised schematic: Here is my version of a simple circuit to provide stop/turn signals for a trailer when the tow vehicle has separate turn and brake signals.

I believe this circuit will have the fewest parts and be the simplest to build. It requires two relays (plus one optional one for the parking lights) and one in-line fuse holder with fuse. And wires of course.

The parts to use are
qty 2 Relays - Allied part number 886-9981
qty 1 fuse holder - Allied part number 846-0100

I used the Allied part numbers since this seems to be the preferred source from the original article. Equivalent relays and fuse holders can also be bought at any automotive supply store and are very common.

The circuit operation is very non-intuitive and anyone looking at it would initially think that it does not work. Its operation depends on some assumptions on the tow vehicle and may not work with all vehicles. But this allows the use of only two relays.

The circuit works because the vehicle brake bulbs and turn signal bulbs provide the ground return path for the relay coil, allowing the relay to act as an XOR gate for the brake and turn signals.

Looking at the Left Turn/Brake Relay. If the Brake is applied, (no turn signal active) 12V will be applied to terminal 86 of the relay. Current will flow out terminal 85 and return to ground through the filaments of the Left Turn bulbs. This works because the relay coil resistance is in the order of 70-90 ohms, and the resistance of two turn signal bulbs in parallel will be 1-3 ohms. The vehicle turn signal bulbs will see a &&1V drop and will not light. The Relay will see >11V, will turn on, and provide 12V power to the trailer Left Turn/Brake light bulbs.

If the Brakes are off, but the Left Turn signal is active, then 12V will be applied to terminal 85 of the relay and the current will exit terminal 86 and return to ground through the filaments of the two brake light bulbs. Again the relay coil will see > 11V and activate, and the brake light bulbs will see &&1V and not turn on. With the relay activated, 12V power will be provided to the trailer Left Turn/Brake light bulbs.

If both the brake and Left Turn signal are activated, then both terminal 85 and 86 of the relay will see 12V and the relay will be off. The trailer light will blink in opposite phase as the vehicle turn signal. I.E. when the vehicle turn signal light is ON, the trailer turn/brake light will be OFF. This should not be a problem. Some GM vehicles do this opposite phase blink with their side marker lights.

I would be interested to hear feedback on this circuit. It is almost as simple as the original diode only solution.

Patrick D on the revised schematic: Agreed on the debate thing - there is a possibility of backdriving each light through the 4.7k resistors. I had thought of this, but the resistance of an incandescent bulb is rather low, so this shouldn't be a big issue.
Looking forward to hearing the outcome.

I had given this some thought, but haven't picked parts. This is what I came up with - it's a bit simpler, but relies on mosfets with good Vgs resistance. But whatever, all that is really needed is an XOR function.

Tony G on the revised schematic: Attached is the schematic I have used. Took a while to find a pdf writer that worked. It is very similar to the three relay version except that it is self powering and requires low power relays. There are probably a dozen ways to create a working circuit, each with different features. I think the relay based circuits are the closest to the “grab some parts and throw it together” feel that the original article had. There may not be power in the trunk so the powered ones could be a problem. The circuit attached gets by this problem but introduces a loading problem for pre 70s era cars that might have thermal flashers. Thermal flashers will need the total light wattage to remain about the same with the trailer attached. The means you might use slightly lower wattage bulbs or switch to a heavy duty flasher. This was the way we handled the flash rate problem in these cases when I was younger. Today’s solid state flashers will not have this problem.

The diversity of the designs shows there are many ways to solve the 5 to 4 problem. I have to admit Randy’s use of the low resistance in the car’s lights was pretty clever. Thinking outside the box trailer

There are plenty of other designs using small CPUs (8-pin PICs?) or CMOS IC logic with FETs, diodes plus relays, etc. but we need to keep this very simple.

John, it is your article so you should pick a design or modify one of these to your liking if you want and send it back to Karen. I would not get offended if you used mine and I hope the others on this list feel the same about theirs. Each one has strengths and all have minor flaws (possible lack of 12V power, clicking sound of relays, loading of the circuits on the car, parts too small, would not want to try to build in 20 minutes before taking trash to the dump, etc.). Check that it meets the basic requirements pointed out below plus the one I missed about “brake lights must always work when pressed”, pick some part numbers out of the Allied catalog and you are done.

Randy Lon the revised schematic: Here is another version that takes 3 relays instead of 2, but will work with all vehicle types and does not depend and tricky ground paths through light bulbs.  It is simply a relay implementation of the 3 way switches uses in house to have two switches control one light. Nothing really groundbreaking.

Feel free to pass on any or all of my diagrams for comment.

Dick N on Turn/Break Signal i/f: I think we finally have a variety of solutions that all work, and their limitations are on the table (e.g., old flashers that require a fixed load, cars with LED tail lights). I have some comments to add re: LED tail lights - I think this is a very different problem from cars with incandescent lamps, and it requires a different design. I don't think the same design will be compatible with both incandescent and LED tails. Problem is the car's use of PWM to change the intensity of the LEDs. Some cars have LEDs only for the high-mount brake light and these are not a special case. Maybe somebody has one with all LEDs that he uses for towing and can submit his solution.

Hope the above is not too confusing. One of these days I'll have to spring for a PDF maker.

Dick N on Turn/Break Signal i/f: While incandescent lamp filaments will snub relay coils just fine, if the Turn or Brake light is an LED it would be a good idea to add a diode to ground from that terminal. This will prevent a reverse-bias spike from a relay coil damaging the LED's.

2: I haven't seen schematics of any vehicles that use them, but observing the lights reveals that at least some cars with LED tail lights use the same LED for Tail and Brake. PWM is used to dim the LED for Tail-light. Not true for cars with incandescent taillights and LEDs only in the high-mount, e.g. a string of them in the spoiler. There's also a movement afoot in the automotive industry to minimize copper. FAIK, they might be running a serial bus out to the taillights and sending them commands! None of these designs would work properly with PWM, and there might not even be a place to connect to. I think an Editor's Note pointing this out would be justified in the final version, hopefully inspiring somebody who's solved that problem to submit his design.

That's my 2 cents worth. This is JL's article and he should get to process all this input.

PS - Thanx, Randy 4 tip re: "Cute PDF" utility. I've been looking for one too.

Dick N on Turn Brake Schematic: I can't take credit for the 3-relay "bull's eye", that was Randy's. Guess it's hard to tell the hornets apart =:o

I think all the designs with "unswitched" +12v input could get it from 3 diodes
(1 from each of L, R & B), since none of them have any relays active when no light is on in the tow vehicle. The +12 is needed only when a light is on.
(comments, anyone?)

I believe you've swatted all the bugs this time. BTW, I have a Toyota T-100 pickup which came with a trailer wiring harness. The truck has sep. brake & turn lamps but the L & R trailer light terminals behave per the XOR rule. If you're on the brake and then hit the L turn signal, the immediate response is the truck's L turn lamp comes ON and the trailer's L lamp goes OFF. I don't know who hooked up the trailer harness, it comes from the nether regions so I assume it's an option. Bottom line is: there are systems out there that work just like that.

re my concern about LEDs:

The LED taillights I've seen use the same LED both for tail/parking and for brake. The brightness change is done by switching the LED on and off rapidly in tail/parking light mode to make it appear dimmer. For brake, the duty cycle is 100%, yielding full brightness. You couldn't use this PWM signal, but you probably could use the high-mount brake light for the brake signal. You've protected the LEDs against reverse voltage already with the diodes across the relay coils, so that's covered. My only remaining concerns about LEDs are the fusing and possible electronic control, since LEDs draw less current. Check out the AD8240 LED driver/monitor from Analog Devices for an example of the trend.

Anyway, a system that has problems only with LEDs would still be compatible with most of the cars & trucks currently on the road. I don't know about the
2007 models (my truck is a '96 and my car is a '95).

Randy L on Turn Brake Schematic: I think Tony's design is good and he has summarized all of the designs well.

Each design has tradeoffs and the choice design could depend on the situation and the type of trailer. If the trailer is a simple utility trailer with only a pair of stop/turn/tail lights, I would use Tony's design because it does not require a separate 12V feed.If the 5W to 4W conversion is required for a large travel trailer with many side marker lights. Then this trailer also will possibly have a 7 Wire plug with a 12V feed and I would chose the 12V powered relay versions since getting a 12V feed will not be a problem and the trailer will be more isolated from the two vehicleThe list of tradeoffs between the designs is long, and I'll stop here.I'll also leave it to John to decide which design to present to his audience.

Tony G on Turn Brake Schematic: I am in agreement with Randy. I believe Randy and I also agree that the relay coil load does not require a low current relay as most regular automotive relays pull a lot less current than a single light bulb. This will make selecting the relays from the Allied part list much easier.

Tony G on Turn Brake Schematic: It is now up to you John.

Don’t get too excited about this. I would bet that nearly everyone that initially tries to solve the 5 to 4 problem starts with the original diode solution. I know that was the first thing I thought of when I got my first trailer (Volvo sedan and the back half of a Ford pickup truck). I suspect many use it and never think twice about it. You are probably right that most police will let you slide on the light operation. Problem is some will not let you slide, or if a serious accident happens because of the circuit, the plaintiff will come looking for you or reedbusiness for restitution. The latter problem is really why I commented on it in the first place.

We can debate reasons and designs for the next month but I would just like this closed and your GF article back on line.

Though it looks like you kicked the wrong hornets nest, it just shows that people are interested in these articles and take the time to think about them. If you have another GF idea, don’t let this adventure keep you from submitting it.

I hope to see this GF article back on line in August.

John L on Turn Brake Schematic: Here’s my answer and some circuits. Thanks for all the help; at first it seemed like
piling on, but I learned a lot…
07/24
07/20 Turn Brake Schematic

Dick N: I can't take credit for the 3-relay "bull's eye", that was Randy's. Guess it's hard to tell the hornets apart =:o

I think all the designs with "unswitched" +12v input could get it from 3 diodes
(1 from each of L, R & B), since none of them have any relays active when no light is on in the tow vehicle. The +12 is needed only when a light is on.
(comments, anyone?)

I believe you've swatted all the bugs this time. BTW, I have a Toyota T-100 pickup which came with a trailer wiring harness. The truck has sep. brake & turn lamps but the L & R trailer light terminals behave per the XOR rule.
If
you're on the brake and then hit the L turn signal, the immediate response is the truck's L turn lamp comes ON and the trailer's L lamp goes OFF. I don't know who hooked up the trailer harness, it comes from the nether regions so I assume it's an option. Bottom line is: there are systems out there that work just like that.

re my concern about LEDs:

The LED taillights I've seen use the same LED both for tail/parking and for brake. The brightness change is done by switching the LED on and off rapidly in tail/parking light mode to make it appear dimmer. For brake, the duty cycle is 100%, yielding full brightness. You couldn't use this PWM signal, but you probably could use the high-mount brake light for the brake signal. You've protected the LEDs against reverse voltage already with the diodes across the relay coils, so that's covered. My only remaining concerns about LEDs are the fusing and possible electronic control, since LEDs draw less current. Check out the AD8240 LED driver/monitor from Analog Devices for an example of the trend.

Anyway, a system that has problems only with LEDs would still be compatible with most of the cars & trucks currently on the road. I don't know about the
2007 models (my truck is a '96 and my car is a '95).

See Figures Below:

Figure 1
Figure 2
Figure 2A
Figure 3

Figure 3A Figure 4
Figure 4A

John L : Good idea with the diodes to develop +12VDC. I don't look forward to stringing a fused 12VDC in a safe, kid-proof manner.
Also, I started this GF with the idea that it could be used on any 5 wire harness to light up a 4 wire rental trailer. A gadget freak may have as little as a 12V light bulb to verify wires; I wanted a Gadget idea for them, forgoing the shop manual, etc. Find the 12V going out to the intended bulb and use it without 'too much'
loading or interaction.

Boy, did I kick the wrong hornet’s nest! Original GF circuit as intended was Fig 1.
I have no excuses, but I’ve been closing out an old job, moving and getting up to speed at a new one. After the valid objections about feedback, and brake signal override I submitted Fig 2. I admit that circuit drawing left out a bunch and assumed the rest, but it too was flawed. Fully drawn, in all it’s error as Fig 2A.
The circuit I needed was basically an XOR. Or so I thought…

In these FET switches, the FET is a P Channel with Vgs of -1 to -3V (ON with gate slightly lower than source) and a Vgs, max of 20V, thus the 2k for a divider, not just a pull-up. In the four signal cases, both brake and signal OFF is easy.
Either signal ON means that signal’s FET biases itself ON and applies it’s voltage thru the now made NO contact to that light.
When both turn and brake signals are ON, they block each other’s gate voltage and when the turn signal goes OFF, the brake light flashes ON during the turn signal’s OFF time.

I used the blocking diodes because if someone uses a lower R than 4.7K, no interaction will occur and because I didn’t want to depend on the turn signal’s impedance. I wanted the impedance to be either high volts or floating.

John L: I looked at Fig 2 and found Big Oops #1 – both brake lights flash a turn signal! It’s a coupled XOR and try as I might, I couldn’t split the 2 brake sides without adding another relay. And after looking at Dick Neubert’s elegant 3 relay bulls-eye, all I can say is, “No fair, he used a 12V line!” I tossed around a few diode override ideas but nothing felt solid.  The result is Fig 3. The relays are all SPST, NO so now both Left and Right act independently. Each side’s brake/turn signal is decoupled since the relays break before they make - so no Schottkys to drop the 12VDC. Trailer lights can now echo the car’s flasher, etc. in this fashion: they do what the rear brake/turn signal/running lights do, except that they flash 180’ out w/respect to the rear turn lights and they steal this power from the rear lights’ current. The recommended FETs are about a buck, the relays are about 3, so I’m up to $16 without pcb or passives.

************************************************************
Now that I’ve split the sides, most objections/ questions become informative discussions (I hope). I’ll try to answer comments in a time line order, starting with Randy Law’s –

Surface mount packages are a pain to prototype, yes, so I found a TO220, #273-0114,
An IR IRF5305 with Rds, ON = 06 ohms.

Randy’s ckt. is fine and a novel use of lamp filament resistance. I hadn’t thought of LED resistances, but I do know some polarized relays mean just that and won’t activate otherwise. Hard lesson there for me. (** See below )I also didn’t want to find a spare or run a new +12V line. This is meant to be a gadget, not a chore and I’m assuming 1-3 spare amps of current; sorry if I’m being optimistic. Randy’s design answers the separate power hog concern. My circuit works if and when the brake and turn signal lights work. Also, as mentioned, I don’t know or want to assume an OFF impedance from any signal – how about if turn signals were tri-stated? Other than those objections, a fine solution, hits all marks, and his 3 relay design is even better, for at least the lamp impedance objection.

I’m not clear about Dick Neubert’s concern. A signal turns on its own FET; a signal applied to a FET’s gate turns it OFF. Again, I apologize for my hasty schematic. I agree about the asynchrony of the flashes; I don’t care as long as they both flash. Should I be? Wouldn’t an asynchronous flash be more attention grabbing? Emergency vehicles do this.

Tony G: Yes, 821-4200 is a SPST Form A and 821-2400 is the DPST Form A .
My bad.
I object to some of the list; the immediate flash spec first. Maybe it’s a law but who’s to know? I’m not trying to be glib or smart-ass, I accept the fact my earlier ckt. had many faults but lean a bit toward the comment the “cops are glad there are ANY lights showing…” If a light flashed a little late OK, bad in principal, good on the road – especially since I’m not selling these mass-market.
I flat don’t design/use one-shots unless there is no alternative. YMMV.
Item #2 – I agree, this should happen.
3. Yes, also.
4. Yes, also.
5. My circuit Fig 3 will mimic the rear light signals (I hope).
 What flashes when 4 ways flash – the turn signals, the brake lights or both?
6.  My circuit should do this.