Do I need diodes for connecting a relay system for my headlamps? (Daniel Stern)

[niksfish]

Hi.
I want to add a relay system in my motorcycle for the headlamps the way that Daniel Stern explained in his website. I already measured my instalation and I have a 2v voltage drop in both low and high beams, so I need it.

I was researching about relays. I remember that in school we used a diode in anti-parallel with the coil of the relay, to protect the electronic circuit from electrostatic discharge (mainly transistors and chips), like this:

Hotlinked Image

Daniel doesn't say anything about this in his website. I wonder if I might have something damaged if I don't use a diode.

I have searched for the wiring diagram of other motorcycle models. For example the Honda NSR 250 MC21 does have a stock relay but it doesn't seem to use a diode: http://tyga-performance.com/images/00psm/02tpfitman/08fitmantuning/20mc21wired.jpg
The MC28, a newer model, do use diodes. If I understand the diagram correctly, it is only used in the "head light relay" and not the "high beam relay". I can't figure the exact way it is connected:
http://tyga-performance.com/images/00psm/02tpfitman/08fitmantuning/21mc28wired.jpg

So, what do you think? Are diodes needed for relays in this application? Do I have to use a resistor in series with the coil too?

Edit: By the way, what kind of relay should I buy? My bike is 12v. I guess I will use a relay that is commonly used in cars.

 

[Alaric Darconville]

This is a resistive load, not an inductive one. You won't need a diode.

 

[LedTed]

RE: You adding a relay system for your headlamp.
As you pointed out the schematics you included already show a relay.

RE: The relay.
Some relays already have the "anti-kickback" diode internal to their component package.

RE: "This is a resistive load, not an inductive one. You won't need a diode."
The coil of the relay is an inductive load. And as you previously stated in your self-answered question, use of the protection diode is recommended.

Hope this helps,

 

[XeRay]

This is a resistive load, not an inductive one. You won't need a diode.

I believe the coil of the relay would be an inductive load and when suddenly turned off creates a reverse EMF spike which can be bad for other sensitive devices upstream. Some DC contactors (relays) have a diode on or in them internally others do not.
Its not electrostatic discharge as he stated originally, its a reverse pulse of electricity (EMF) that occurs when the magnetic field collapses rapidly from a coil, when the power to it is removed.
It is good practice to prevent this by placing a diode (flyback diode) plenty of info on the internet on this and how to implement it.

Looks like LEDTed answered before I could finish.

 

[Alaric Darconville]

Yet, somehow, Daniel Stern missed out on advising that you'll need a diode?

Add one if you want, it's not a necessity.

 

[-Virgil-]

It's certainly not bad practice to use relays with built-in spike suppression (diode or resistor across the coil). It can't hurt anything and might prevent a problem, albeit one that's very unlikely to happen. As far as I can remember, the kits/harnesses I've bought from Stern have all had suppressed relays.

Why is a spike problem unlikely? Because the systems in question are designed to tolerate such a spike -- consider what happens when a light bulb filament fails while it's powered up. Big blue flash at the moment of failure...!

 

[XeRay]

It's certainly not bad practice to use relays with built-in spike suppression (diode or resistor across the coil). It can't hurt anything and might prevent a problem, albeit one that's very unlikely to happen. As far as I can remember, the kits/harnesses I've bought from Stern have all had suppressed relays.

Why is a spike problem unlikely? Because the systems in question are designed to tolerate such a spike -- consider what happens when a light bulb filament fails while it's powered up. Big blue flash at the moment of failure...!

If a mechanical switch is being used to control the relay (on/off), without the diode the flyback voltage is bad for the switch contacts, arcing and "burning" of the switch contacts can occur.
This will make mechanical switches last longer.
This is besides any other potential bad effects of not having diodes
This is the reason every solenoid (5) and every relay (2) in my Home built (experimental) airplane that I built has a diode added. (some can have a diode built in on DC use solenoids, mine did not)
Solenoids for gear retract (elecro hydraulic) and gear extension (intermittent duty), Main power solenoid (constant duty), Avionics master solenoid (constant duty), engine starter solenoid (intermittent duty).
The other smaller relays are for switching navigation source for autopilot etc.

 

[-Virgil-]

Makes sense, but it doesn't seem to me like there'd be enough of a spike generated by the coil in the mini or micro relays we're talking about for headlamp relay setups to have any bad effect on the headlight switch contacts (which are meant to make/break loads of about 10 to 15 amps for many years). Bigger solenoids and relays for main power, starter motors and other big motor loads, etc would obviously be a different ball of wax.

It's kind of an academic argument, though, because of (again) relays with built-in spike suppression being supplied in reputably-sourced headlamp relay kits or harnesses.

 

[niksfish]

I believe the coil of the relay would be an inductive load and when suddenly turned off creates a reverse EMF spike which can be bad for other sensitive devices upstream. Some DC contactors (relays) have a diode on or in them internally others do not.
Its not electrostatic discharge as he stated originally, its a reverse pulse of electricity (EMF) that occurs when the magnetic field collapses rapidly from a coil, when the power to it is removed.
It is good practice to prevent this by placing a diode (flyback diode) plenty of info on the internet on this and how to implement it.

Yeah, I'm sorry, bad choice of words. You are absolutely right, what you said is that happens. What I meant to say is that you will get a high voltage suddently when you disconect the coil.

Yet, somehow, Daniel Stern missed out on advising that you'll need a diode?

Add one if you want, it's not a necessity.

Yeah, maybe he didn't mention anything because it isn't needed.

If a mechanical switch is being used to control the relay (on/off), without the diode the flyback voltage is bad for the switch contacts, arcing and "burning" of the switch contacts can occur.
This will make mechanical switches last longer.

Hmm, then I will use a diode.

This is besides any other potential bad effects of not having diodes

My main concern was the CDI / computer of the bike. Maybe nothing happens because it is in parallel, not series with the coil (like the transistor in the pic I posted at the beginning, it was in series with the coil and it would be affected directly). Here is a drawing (beautiful, right?)

Makes sense, but it doesn't seem to me like there'd be enough of a spike generated by the coil in the mini or micro relays we're talking about for headlamp relay setups to have any bad effect on the headlight switch contacts (which are meant to make/break loads of about 10 to 15 amps for many years). Bigger solenoids and relays for main power, starter motors and other big motor loads, etc would obviously be a different ball of wax.

Well, I don't really know about that, but the spike can have a high voltage even with small relays, enough to ruin transistors.

It's kind of an academic argument, though, because of (again) relays with built-in spike suppression being supplied in reputably-sourced headlamp relay kits or harnesses.

Yeah, the solution is easy and cheap. I will use a diode. But anyway, I'm interested in the academic argument to learn a little more.



So... This is what I'm going to do. I'm buying a relay, and I'll test it to see if it already has the diode or not. If it doesnt, then I'll add one.
Also, I found that there are mini relays that occupy little space. For bikes this means an easier instalation.

 

[-Virgil-]

I'm buying a relay, and I'll test it to see if it already has the diode or not.

How do you plan to do the test? The diode is in parallel with the coil.

 

[Alaric Darconville]

So... This is what I'm going to do. I'm buying a relay, and I'll test it to see if it already has the diode or not. If it doesnt, then I'll add one.

The test should be easily enough done if you have some 1/W resistors, a 12V power source, and a VOM and know what you're doing.
But instead of adding a diode, get one that already has one if you're so set on having a diode.

Also, I found that there are mini relays that occupy little space. For bikes this means an easier instalation.

And if they're not made by a reliable company (for example, Tyco-Bosch) they may not be any good (even if they happen to have a diode present).

 

[XeRay]

I believe it would be easy and pretty obvious to test for a diode by metering across the 2 coil pins (not connected to any circuit). If the same resistance reading is found metering in both directions (ie switch the leads from the meter to the contacts).
If the resistance is identical both "directions" of metering DC flow, there is no diode. If there is a notable resistance difference with swapped leads polarity, then there is a diode, Also the markings on the relay should really show a diode symbol with its pin schematic printed on it or a spec sheet of it at minimum.
Obviously, adding the diode also creates a polarity to the DC coil connections needs to be observed, which would not be the case without the diode.

 

[niksfish]

How do you plan to do the test? The diode is in parallel with the coil.

I didn't think of that

And if they're not made by a reliable company (for example, Tyco-Bosch) they may not be any good (even if they happen to have a diode present).

I'll search for a decent brand. Do you know which ones are good? Besides the one you mention.

I believe it would be easy and pretty obvious to test for a diode by metering across the 2 coil pins (not connected to any circuit). If the same resistance reading is found metering in both directions (ie switch the leads from the meter to the contacts).
If the resistance is identical both "directions" of metering DC flow, there is no diode. If there is a notable resistance difference with swapped leads polarity, then there is a diode.

Thanks!

Also the markings on the relay should really show a diode symbol with its pin schematic printed on it or a spec sheet of it at minimum.
Obviously, adding the diode also creates a polarity to the DC coil connections needs to be observed, which would not be the case without the diode.

Oh yeah you are right, if it does have the diode there must be an indication of polarity. Also, I'll google the spec sheet before buying anything.

 

[John_Galt]

Timely thread. I buikt a relay harness for my 4runner approximately 2 years ago. It recently started exhibiting a new behavior- eith the headlights off, I can flash high beam circuit as normal. With the low beams on, if I flash highbeam or if I click the stalk switch forward to activate highbeams, it will kick the lowbeam off and activate highbeam, however when the switch is pulled back to the lowbeam position, the high beam will remain energized aling with the lowbeam. Turning the headlight switch to the off position clears the issue.

After inspecting the harness and swapping relays around, no change in behavior, no damage wire insulation.

Am I right in assuming that I 1- need to replace my headlight switch (contacts are arcing in the stalk, leaking enough voltage across the contacts to keep the high beam relay closed/triggered), and 2- I need to replace my relays with relays that have flyback diodes in them to prevent this from occurring with the replacement switch?

Can anyone recommend 4 terminal relays that have a diode in them?

 

[XeRay]

Google is your friend ( "relay with diode" ) find one that meets your needs, number of contacts and amp rating.

 

[-Virgil-]

Timely thread. I buikt a relay harness for my 4runner approximately 2 years ago. It recently started exhibiting a new behavior- eith the headlights off, I can flash high beam circuit as normal. With the low beams on, if I flash highbeam or if I click the stalk switch forward to activate highbeams, it will kick the lowbeam off and activate highbeam, however when the switch is pulled back to the lowbeam position, the high beam will remain energized aling with the lowbeam. Turning the headlight switch to the off position clears the issue.

I don't think it's your headlight switch. And I don't think your un-dioded relays (if they are) suddenly started causing problems after two years of not.

Any chance you've got fog lamps? If so, try pulling the fog lamp relay and see if the problem disappears. Typically fog lamp relays are wired with their coil ground tied into high beam power, thus automatically disabling the fog lamps when the high beams are on. When the high beams are off, the fog lamp relay coil grounds through the high beam filaments. But if the high beam power is now tripping relays instead of powering filaments, it interferes with that circuit logic and you can get exactly the behavior you describe.

 

[John_Galt]

-Virgil- My 4runner does have the wiring from the fuse box down to each side under the bumpernfor the fog lights, but the oem fog relay is not in the fuse box, and the headlight stalk switch mine is equipped with does not have the option. There are two headlight switch arrangements, one for the trim package with fogs, and one without. I have not pulled the trim for the steering column to verify, but I suspect given the exterior wiring, that I could addthe fog light stalk switch and it would plug into the existing harness, and them I would just need the relay and I would be ablento activate the circuit.

It's the original switch, which is over 20 years old at this point ('99 model year), so given the change in behavior, I think it is the likely culprit.

My relay harness has 3 relays, one for lowbeam, one for high beam, and there is a leg off the highbeam which I ran back to a switch on my dash which runs back to the 3rd relay. Which is for my auxiliary driving lights, so I can turn those on/off with the highbeam circuit, or choose to leave them off. My auxiliary light don't pull very much, I am running a set of LED Cibies, but I wanted to to have them on a relay in case I ever replaced them with something with a higher draw.

After checking the wiring for chafed insulation or signs of any arcing or shorting and finding no issues there, I pulled each relay and rotated them through each of the other two legs of my harness, which resulted in no change in behavior. Plugging the oem harness back into the bulbs does not result (as far as I can tell) with the same behavior, which led to my conclusion that the headlight switch is leaking enough voltage between the low and highbeam switch contacts to keep the highbeam relay closed, but not enough to noticably energize the highbeam filament when the relay hsrness is removed from the equation.

 

[-Virgil-]

Well, I could easily be wrong (it's happened a few thousand times in my life). Please let us know how your repair efforts go.

 

[Alaric Darconville]

My 4runner... '99 model year

Seems like you can get pretty much any combination switch from a mid '90s, early 2000's Toyota or Lexus (and probably from a broader range of years than that) from a Pull-A-Part or similar and be done with it.

 

[John_Galt]

Seems like you can get pretty much any combination switch from a mid '90s, early 2000's Toyota or Lexus (and probably from a broader range of years than that) from a Pull-A-Part or similar and be done with it.

Haha, every local scrap yard is closed. There are some people a few hours from me parting out a 3rd gen, but the drives not worth it, got an aftermarket on order. It should be here tomorrow, I'll be able to swap it and see if that's the solution, if not, back to the drawing board.

As far as relays go, these aren't Bosch, but what are your thoughts on the brand?
https://www.waytekwire.com/item/75879/Picker-PC792B-1A-C-12S-RN-X-Mini-ISO-Relay-/