Need Some Advice on Using Relays (Preventing Counter-EMFs)

[hayze]

Hello, I've been looking into a few headlight modifications lately that involve SPDT and SPST relays. Now, I pretty much understand everything about the logical functions associated with relays (logic gates). I can hook them up to perform a specific function on my own, and I designed a few modules that I plan to finalize and implement this spring (DRLs and fog light stuff). So I'm good with understanding the switching and true/false logic, but my problem is the general safety of adding relays to car circuits.

Specifically, I am concerned with the counter electromotive force (Counter-EMF). I noticed recently that relays' coils are associated with this phenomenon, and overall, it can damage circuitry components.
http://www.bcae1.com/relays.htm

The general idea I'm focusing on is basically that this is something I need to protect against when using a relay. The solution I read about was using a diode in parallel with the relay coil but oriented so that it does not normally conduct when the coil is energized (facing the + side).

So my question is, do I just simply use a 1- or 2-amp diode in parallel with every relay I add? Or does this solution become more complex as the design becomes more complex? I'm talking about using around 2 to 4 relays, and after installing the module, it will likely then be connected to other OEM relays in the car too (just to note).

If its not such a simple solution, then can I post a few circuit diagrams of modules that I plan to implement? Let me know if you need more details of what I plan to do.

I also came across this concept here:
http://www.candlepowerforums.com/vb/showthread.php?t=256792

Thanks much for any help guys

 

[1 what]

While I can't speak about modern auto electronics I can tell you that a long time ago when I was a working EE we used diodes across all of our relay coils without any problems.

 

[-Virgil-]

Make it easy on yourself; use "suppressed" relays. They contain an internal diode and/or resistor (depending on the specific relay you pick) across the coil to attenuate the spike that can occur on relay switch-off. Make sure to use a good, reliable brand of relay (Omron or Tyco/Bosch I've had good success with; Hella quality is inconsistent, no-name items from low-cost countries are good reminders that you get what you pay for).

Interested to read about what kind of DRL setups you're looking at making -- remember there are fairly specific regs for how various front lighting functions are and aren't allowed to operate in DRL mode.

 

[hayze]

Thanks for the advice guys.

I'll look for suppressed relays. But if I use a non-suppressed relay, then does adding the diode provide the same exact protection?

Just a side note; I'm setting up the DRLs so that they shut off when any other light (besides parklights) turns on. Is this what you mean by 'regs?'

In Ohio, we don't have state inspections nor do we require DRL by law. I'm just doing it for the added safety of day lights and for aesthetics too. In case anyone's wondering about details, I'm switching one of the bulb's ground wire over to a + to run them in series (so they're dimmer).

Again, thanks for your replies. I'm still hoping for some more replies to the original question. It'd be helpful and appreciated.

 

[-Virgil-]

I'll look for suppressed relays. But if I use a non-suppressed relay, then does adding the diode provide the same exact protection?

Yep.

Just a side note; I'm setting up the DRLs so that they shut off when any other light (besides parklights) turns on. Is this what you mean by 'regs?'

Regulations on which front lamps are permitted to serve as DRLs, and at what power levels, and how they may/must/can't be hooked up in relation to other lighting functions.

In Ohio, we don't have state inspections nor do we require DRL by law.

Right, but that doesn't make it OK to just do it however you want. Road safety is at stake here.

Which lights are you planning on running as DRLs, and at what power level?

(Also: you should wire them so they turn off when the parking lamps are switched on.)

 

[greg_in_canada]

Diodes across the coil reduce the back EMF but also slow the opening of the contacts (due to the current shutting off slower). I read a paper from a relay manufacturer that recommended alternate methods. One method was to use a resistor of the right value to reduce the back EMF to a value the driver circuit can handle. Another method was to use a diode plus zener diode to limit the back EMF.

A slowly opening relay doesn't matter much with a resistive load but with an inductive load a slower opening will increase the duration of the arc and wear out the contacts sooner.

Greg

 

[hayze]

Greg, are you saying that my relays will stay closed longer than it takes other relays to switch thus causing malfunctions?

Right, but that doesn't make it OK to just do it however you want. Road safety is at stake here.

The operating conditions I defined for them are pretty strict. They can turn on with the parklights, but the low beams and fog lights serve as circuit breakers. No actual lights can be on with the DRLs (the parklights are small). This should be as safe as if they did shut off with the parklights.

Which lights are you planning on running as DRLs, and at what power level?

It replicates exactly how the stock DRLs work on the Canadian models of the TSX. The high beam bulbs switch to a series (instead of parallel) and get half of the voltage instead of full. I'm not sure what the brightness level is though. How does half voltage map to brightness mathematically? The bulbs are regular 55W incand.

I was actually thinking about getting an additional resistor to reduce the brightness even more. So the bulbs would be in series with each other and with a resistor. What kind of resistors can I use for such a high load?

Thanks for all your advice guys.

 

[-Virgil-]

The operating conditions I defined for them are pretty strict. They can turn on with the parklights

That's legal but not the best choice from a functional perspective -- it makes it too easy for the driver to mistakenly to drive with the parking lamps + DRLs on at night (instead of full headlamps) which is unsafe. Better to have the DRLs go off when the parking lamps come on.

It replicates exactly how the stock DRLs work on the Canadian models of the TSX. The high beam bulbs switch to a series (instead of parallel) and get half of the voltage instead of full.

That'll work OK. High beam DRLs like this are legal. Not very good, but legal.

How does half voltage map to brightness mathematically?

(operating voltage ÷ rated voltage)^3.4 gives you your multiplier, which you multiply by the rated output of the bulb to get the output at the operating voltage.

I was actually thinking about getting an additional resistor to reduce the brightness even more.

I'm sort of curious what the overall goal is here...you'd be better off from just about every persepctive leaving the headlamps completely out of the DRL circuit and implementing turn signal DRLs instead. Better DRL safety performance, lower power consumption, much longer bulb life, no glare, no chance of mistaken DRL use after dark, equally legal, etc.

 

[Kuryakin]

It's a non issue. The time difference is immaterial in this application. Also, the contacts, once the armature breaks free of the coil's pole piece, it snaps open without any increased contact wear vs coils without a snubber diode. The contacts don't start to open until the bulk of the stored energy is dissipated anyway.

So, use a diode across each coil! The 1N400x series works just fine, as will pretty much any diode with at least a 100 mA rating, 50 PIV for these apps.

Diodes across the coil reduce the back EMF but also slow the opening of the contacts (due to the current shutting off slower). I read a paper from a relay manufacturer that recommended alternate methods. One method was to use a resistor of the right value to reduce the back EMF to a value the driver circuit can handle. Another method was to use a diode plus zener diode to limit the back EMF.

A slowly opening relay doesn't matter much with a resistive load but with an inductive load a slower opening will increase the duration of the arc and wear out the contacts sooner.

Greg

 

[hayze]

Thanks for the replies. So let me ask, would you feel safe using a diode like this in all the relays in a DRL module or a fog light circuit modification?

And, what does "50 PIV " mean (last post)?


Just a few more questions about EMFs:

What specifically becomes damaged during a back-EMF from a relay coil? Is it just that relay's coil, or are other components in the circuit (including OEM components) at risk too?

The diode conducts during the EMF because its oriented toward the + side. So I'm guessing it redirects the EMF somewhere? Which pathway does the EMF take?

 

[Kuryakin]

PIV = Peak Inverse Voltage. Usually, there's no issue should you not use a snubber diode or resistor. However, there is a spike, and if you're worried, a snubber will stop this. Adding diodes doesn't have a downside, technically. Just another thing you need to either spec or add.

As far as DRL modules go, well, can't really have a DRL module active AND have relays, as the relays won't open when they're supposed to.

Thanks for the replies. So let me ask, would you feel safe using a diode like this in all the relays in a DRL module or a fog light circuit modification?

And, what does "50 PIV " mean (last post)?


Just a few more questions about EMFs:

What specifically becomes damaged during a back-EMF from a relay coil? Is it just that relay's coil, or are other components in the circuit (including OEM components) at risk too?

The diode conducts during the EMF because its oriented toward the + side. So I'm guessing it redirects the EMF somewhere? Which pathway does the EMF take?

 

[-Virgil-]

As far as DRL modules go, well, can't really have a DRL module active AND have relays

Sure ya can...as long as the DRL module operates a different set of lights than the relays do (such as the turn signals).

 

[Disaster]

I'm sort of curious what the overall goal is here...you'd be better off from just about every persepctive leaving the headlamps completely out of the DRL circuit and implementing turn signal DRLs instead. Better DRL safety performance, lower power consumption, much longer bulb life, no glare, no chance of mistaken DRL use after dark, equally legal, etc.

I agree from a lighting/awareness standpoint. Seems to me turn signals on stick out more than just dim white lights. However, I didn't think they were legal for DRLs....like driving with the park lights on. The argument I've heard against them is people might see the turn signal light momentarily and assume the person is turning.

It seems if one uses the headlights for DRL's should cause less legal issues because you are allowed to run with your headlights on all the time and some jurisdictions require it for dusk and dawn driving.

Sure ya can...as long as the DRL module operates a different set of lights than the relays do (such as the turn signals).

Wouldn't that interfere with the turn signal function?...or at least make it more complicated? You'd have to wire it so that the turn signals would turn off the DRL.

 

[-Virgil-]

I agree from a lighting/awareness standpoint. Seems to me turn signals on stick out more than just dim white lights. However, I didn't think they were legal for DRLs....like driving with the park lights on.

Incorrect. Turn signal DRLs are legal and common in the USA and Canada. Corvettes since the C5, many Cadillacs, GMC/Chevrolet full-size vans for many years now, several Toyota and Lexus products, '96-'00 Chrysler minivans, certain recent Lincolns, the latest-generation Chrysler Pacifica, and other vehicles have turn signal DRLs as original equipment. This is the front turn signals we're talking about -- the bright amber ones, not the dim white or amber parking lights.

Wouldn't that interfere with the turn signal function?...or at least make it more complicated? You'd have to wire it so that the turn signals would turn off the DRL.

Just toss in a turn signal DRL module like this one and be done with it.

 

[hayze]

Just so you guys know how it works...

There are 2 relays. The first relay switches the high beams to a series at this intersection: (ACC/ON) AND (Out-Of-Park). It does this by switching the ground of one bulb to a +. The bulbs are in parallel stock, so it becomes a series with this one change.

The second relay breaks the first one (by switching it's coil) when the passing lights OR low beams are on. For this, I will tap into the combination switch output lines in the steering column.

So the whole gate is:

(ACC/ON) AND (out-of-park) AND NOT(Passing-Lights OR Low-Beams)

As far as DRL modules go, well, can't really have a DRL module active AND have relays, as the relays won't open when they're supposed to.

What do you mean by this? Are you saying the module won't work?

 

[Disaster]

Just toss in a turn signal DRL module like this one and be done with it.

Thanks for the link. I'm going to pick one up for our car which doesn't have DRL. May also pick one up for the Avalon because I'm not a fan of it's immediate lights. With turn signal indicator lights you can also replace the incandescents with LED's and get even better efficiency.

 

[-Virgil-]

With turn signal indicator lights you can also replace the incandescents with LED's and get even better efficiency.

No, don't. "LED bulbs" don't work safely or effectively in lamps intended to use filament bulbs. This applies to all "LED bulbs"; it's a concept problem, not an implementation problem (i.e., it doesn't matter who makes the square wheel or what color it's painted or what it's made out of, it's still a square wheel).

 

[mrb]

No, don't. "LED bulbs" don't work safely or effectively in lamps intended to use filament bulbs. This applies to all "LED bulbs"; it's a concept problem, not an implementation problem (i.e., it doesn't matter who makes the square wheel or what color it's painted or what it's made out of, it's still a square wheel).

Totally agree. The "efficiency" is so great they'll flash at double-speed but in daylight nobody will see them anyway.
LED bulbs only really "OK" for interior use at a push.

 

[Disaster]

No, don't. "LED bulbs" don't work safely or effectively in lamps intended to use filament bulbs. This applies to all "LED bulbs"; it's a concept problem, not an implementation problem (i.e., it doesn't matter who makes the square wheel or what color it's painted or what it's made out of, it's still a square wheel).

I understand the concern about bulb substitution with lights that work on reflectors and filaments to illuminate, but how big an issue is it with a reflector/marker style light that just needs to light up? Shouldn't a 180 degree bath of light...or two or three...or an array of LEDs, light the reflector up so that it visibly works?...ie...can be seen well by others. Perhaps you are suggesting bright enough LED drop ins, aren't available yet to meet the requirements.
***this is assuming that the circuit is corrected for electrical load or a LED compatible flashing unit is used.***

Totally agree. The "efficiency" is so great they'll flash at double-speed but in daylight nobody will see them anyway.
LED bulbs only really "OK" for interior use at a push.

The efficiency isn't about when they are flashing. It is about when they are illuminated all the time, as DRLs. What do you mean "at a push" and why do you feel LED's are only good for interior use?

 

[Alaric Darconville]

I understand the concern about bulb substitution with lights that work on reflectors and filaments to illuminate, but how big an issue is it with a reflector/marker style light that just needs to light up? Shouldn't a 180 degree bath of light...or two or three...or an array of LEDs, light the reflector up so that it visibly works?...ie...can be seen well by others. Perhaps you are suggesting bright enough LED drop ins, aren't available yet to meet the requirements.

Turn signals designed for a particular filament bulb have reflectors that are designed to work with that filament's size and placement. LED "bulbs" just have a bunch of LEDs plastered on them and light is beamed out willy-nilly (or worse, just a single emitter pointed in one direction). Turn signals, side marker lamps, license plate lamps, reverse lamps, and brake lamps, just like headlights, have certain photometric requirements that must be met. They can't be met unless they are using the bulb they were designed around.

I could be abysmally wrong with this statement, but it would seem that the bulb is chosen first, and then the fixture is designed.