Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Relay/Diode/Back EMF with a new twist

Status
Not open for further replies.

VtMtnMan

New Member
I have run into a problem wiring a set of driving lights on my 2009 Subaru Outback. Google led me to this forum, and I can see that there are more than enough knowledgeable people here to set me straight. I closely read two threads about putting a diode in parallel with a relay in order to keep back EMF from causing mischief with the car's circuitry, but neither one addressed my particular issue.

I have installed driving lights in all but one of my vehicles over the last 31 years, always successfully. What has changed is car electronics.

I always use a Bosch 30 amp relay. There is no problem with the power circuit. Terminal #30 of the relay is connected to the "+" terminal of the battery with a fused 12 ga. wire. Terminal #87 of the relay is connected to the driving lights, which ground through the bumper.

The issue is with the trigger circuit, terminals #85 and #86 of the relay, which connect to the relay coil. On every installation until now, I simply connected terminal #86 to the wire supplying current to the driver's side high beam and terminal #85 to ground through a toggle switch in the passenger compartment. When the toggle was up, the driving lights came on and off with the high beams. I put a Radio Shack diode parallel to terminals #85 and #86 to protect the car's circuitry. Perfect!

It's more complicated this time around. The connector for the high beam has two wires going to it, one green and the other white. The bulb is well insulated from the headlight shell, and the path to ground has to be through one of the wires. I disconnected the connector from the high beam bulb and tested for the presence of voltage to ground with a simple circuit tester. I clamped the lead of the tester to ground and probed the connector and here is what I found:

With the low beams selected, *both* the green wire and the white wire have voltage to ground; and

With the high beams selected, *only* the green wire has voltage to ground.

That is, the green wire is always "hot" even though it's connected to the high beam. What I am assuming is going on is that this is a variation of a switched ground. Instead of cutting off the path to ground, the car puts a voltage on the white wire equal to the voltage on the green wire when the low beams are selected and, as a result, the high beams remain off.

If this is the case, I could make my driving lights come on and off with the high beams by connecting terminal #86 of the relay to the green wire and terminal #85 to the white wire (through my toggle switch). We come now to my questions:

Assume I tie the green (always hot) wire to terminal #86 and the white wire (sometimes ground) to terminal #85 and place a diode in parallel with the two terminals, oriented so as to protect the green wire and whatever is connected to it.

1. When the coil of the relay is energized and then de-energized, will there be a voltage spike (from back EMF) in one direction or two? That is, will a voltage spike appear on terminal #86, terminal #85, or both?

2. Will the diode act to shunt the voltage spike from the back EMF down the white wire?

This was never an issue before because one side of the circuit, the ground side, needed no protection since it consisted only of a wire straight to ground through a heavy duty toggle switch. But in my Outback, both the green and white wires are apparently tied to microprocessor controlled switches, which a voltage spike from back EMF might fry.

To rephrase my questions, if I place a properly oriented diode in parallel with terminals #86 and #85 of my relay, wired as described above, will the white wire be affected by back EMF when the relay coil is de-energized?

I would be grateful for any help you can give me.

VMM
 
I see I was too long-winded; query condensed

Lots of views of my query and no replies, I think because I included way too much unnecessary information. I have read, here and elsewhere, that it is a good idea to install a diode in parallel with the coil of a relay to prevent back EMF from harming sensitive components connected electrically to the coil. My questions, simply stated:

1. Does placing the diode in the circuit prevent the voltage spike from back EMF from even happening or simply direct it to someplace it won't do harm? and

2. If I place a diode in the circuit, will it protect the wire providing voltage to the relay (and the circuitry connected to that wire), the wire to ground (and the circuitry connected to that wire), or both?

Many thanks for any light you can shed.

VMM
 
A diode across the relay coil simply provides an alternate path for the inductive coil current to keep flowing when the switch powering the relay coil is opened. Thus the EMF spike is limited to the forward drop of the diode (typically 0.7V). It will protect anything else in the coil circuit both on the voltage side and the ground side.
 
Connection point?

A diode across the relay coil simply provides an alternate path for the inductive coil current to keep flowing when the switch powering the relay coil is opened. Thus the EMF spike is limited to the forward drop of the diode (typically 0.7V). It will protect anything else in the coil circuit both on the voltage side and the ground side.

Thank you! That was the answer I was hoping for.

Am I correct that I don't have to connect the diode to the relay terminals themselves but can splice it in to the wires to and from the relay a couple of inches below the the terminals?

VMM
 
Am I correct that I don't have to connect the diode to the relay terminals themselves but can splice it in to the wires to and from the relay a couple of inches below the the terminals?
Yes, connecting the diode a few inches from the coil connections will work just fine.
 
Yes, connecting the diode a few inches from the coil connections will work just fine.

Yes, thanks. What little I used to know about circuits is starting to come back to me, and it was a dumb question. Thanks for not pointing that out.
 
What diode to buy?

So long as it is in parallel with the relay you will be fine, close or far :O....

Many thanks. I now know why I'm going to put in a diode and where I'm going to put it (and how to orient it). In a 12 volt auto circuit using a Bosch 30 amp relay, is there a certain diode I should look for? One article I found says I can use pretty much any type of switching diode, such as 1N4001, 1N4002, etc. or Radio Shack part numbers 276-1101, 276-1102, etc. What would work best in this application or doesn't it matter?

Thanks again.

VMM
 
There will not be so much current on the relay on the On/Switch side (< 1A?) as on the Active side (30A). The basic diodes should work, 1N4001 or 1N4002. Make sure it is rated for up to 1 amp just in case.
 
There will not be so much current on the relay on the On/Switch side (< 1A?) as on the Active side (30A). The basic diodes should work, 1N4001 or 1N4002. Make sure it is rated for up to 1 amp just in case.

Perfect. I'll make sure it is rated for 1 amp, which is the value of the fuse for the relay trigger circuit. Any current large enough to knock out the diode will then blow the fuse, so I'll know right away.

Thanks!

VMM
 
Zener diode in series?

I have been doing some more research, and it seems that inserting a simple switching diode parallel to the relay coil solves one problem and creates others. See:

https://www.electro-tech-online.com/custompdfs/2009/09/13c3264.pdf

https://www.electro-tech-online.com/custompdfs/2009/09/13c3311.pdf

The preferred solution seems to be to include a zener diode in series with the switching diode. I am warned:

"the only design precaution is to select a zener with an appropriate breakdown voltage and impulse power specifications adequate for the relay in its application."

This is in a car, so the supply voltage is 12 volts nominal, probably 13.5 +/- volts max. What do I ask for at Radio Shack tomorrow?

Thanks as always.

VMM
 
Firstly, if you have HID (High Intensity Discharge) bulbs you shouldn't go anywhere near the bulbs. The voltages on them are nowhere near 12 V.

It sounds like the ground is switched. You need a wire from one side of the bulb, through the switch, through the relay and to the other side of the bulb.

The diode will mainly protect your switch, which probably doesn't need protecting. The relay current will carry on flowing for microseconds when you turn off the switch and cause an arc across the switch. Most relays have a resistor inside them, in parallel with the coil, and the current can flow through that after the switch opens. If the resistor normally takes 1/4 of the coil current, when the switch opens it has to take all the coils current, so the voltage is about 4 times as much and not hundreds of volts.

The best thing to add is a bidirectional transient voltage suppressor in parallel with the coil or the switch. It must be rated well above 12 V. 18V or so is a good start. That will limit the voltage to 18 V. As they are bidirectional, you don't need to worry which way round to fit it.
 
No HID

Firstly, if you have HID (High Intensity Discharge) bulbs you shouldn't go anywhere near the bulbs. The voltages on them are nowhere near 12 V.

....

The diode will mainly protect your switch, which probably doesn't need protecting....

The best thing to add is a bidirectional transient voltage suppressor in parallel with the coil or the switch. It must be rated well above 12 V. 18V or so is a good start. That will limit the voltage to 18 V. As they are bidirectional, you don't need to worry which way round to fit it.

Thank you very much for this advice. Possibly I am worrying needlessly about potential harm from back EMF, but there is obviously a lot of microprocessor controlled switching in the lighting circuit of my 09 Subaru, and I sure don't want to fry it.

No HID lights in my car. If I had 'em, I probably would not be adding driving lights.

VMM
 
Mail order source for zener diode or suppressor?

I went to town today looking for a bidirectional transient voltage suppressor as recommended or an 18 V or 24V 1 amp rated zener diode to put in series with a 1N4001 switching diode. No luck at the auto parts place or, surprisingly, Radio Shack (they had only 2 zener diodes, 12V and 5.somethingV.

Can anyone recommend an online source for these parts?

Many thanks.

VMM
 
You don't need the transient diode in addition to the regular diode, you only need one or the other. The only reason to use a transient diode across the coil instead of a regulator diode is to reduce the relay turn-off time (it speeds up the current collapse). And I don't think relay turn-off time is a factor in your application, so just stick with the regular diode.
 
Last edited:
You don't need the transient diode in addition to the regular diode, you only need one or the other. The only reason to use a transient diode across the coil instead of a regulator diode is to reduce the relay turn-off time (it speeds up the current collapse). And I don't think relay turn-off time is a factor in your application, so just stick with the regular diode.

I have read that the best suppression scheme is to use either (1) a transient suppressor diode, or (2) a regular switching diode in series with a zener diode. These articles suggest that my relay will operate longer and more reliably (avoiding sticking problems from microwelds due to slowing down the relay release time) using either of these schemes than if I use a switching diode alone:

https://www.electro-tech-online.com/custompdfs/2009/09/13c3311-1.pdf
https://www.electro-tech-online.com/custompdfs/2009/09/13c3264-1.pdf

I bought a couple each of IN4001 and IN 4003, so I already have the switching diodes. Zener diodes seem inexpensive and it would be no trouble to add one if i could find a suitable one. My understanding is that I should exceed the car's voltage, so 18V or 24V.

The IN400X are each rated at 1 amp. The zener diodes I have found so far online seem to be rated in watts and in the range of only a fraction of a watt to about three.

Please set me straight: Am I looking at the wrong specification? How do I determine what is a suitable zener diode (I think I've got the voltage covered; I mean the power rating) and where can I buy one?

Thanks for any help.

VMM
 
Buy 4 off 12 V zenners.

Put all 4 in series, with two one way round, two the other way round. Fit the string in parallel with the relay coil.

That will do what the bidirectional TVS does.

The only advantage of the bidirectional TVS, and the only reason for zenners over a diode, is that polarity doesn't matter. Either the bidirectional TVS or the zenners will work either way round.
 
The Zeners you are looking at are rated in terms of continuous power. Look and see if they have a surge current specification, or a pulsed power specification. They will only be seeing a very brief high current surge when the relay shuts off, so they can handle more than what their continuous power rating is. The little TVS diodes you can buy can handle 5kW for about a microsecond, but only a few watts continuous. Zeners are similar, but they don't clamp quite as fast as a true TVS diode (at least I don't think they do) and zeners aren't as good at pulsed power handling as true TVS diodes (at least I don't think they are). I could be wrong.

In any case, the continuous power rating doesn't mean a lot, since it will only be seeing occasional current pulses when the relay is switched off.
 
Here's a good example. A little 3W standard zener in a surface mount package:

https://www.electro-tech-online.com/custompdfs/2009/09/1SMB5913BT3-DPDF.pdf

Look at Page 4, Figure 2. It will handle 50W for a 10ms rectangular pulse if the starting temp was 25 degrees C. That's the sort of info you need to pick a good diode. You also need to have an idea of how long the pulse is going to last from the relay spike, kind of hard to know without a scope though. I would think 10ms would be a totally fine guess.

Also, I would personally use a TVS zener. Hardly any price difference when you're buying single unit quantities, and better at pulse handling (and they will ALWAYS have a chart for pulse handling in the data sheet).
 
Last edited:
Status
Not open for further replies.

New Articles From Microcontroller Tips

Back
Top