3

I understand that bidirectional TVS diodes have no polarity and can be used to limit positive and negative swings in an AC application.

Is there a down side to using bidirectional TVS diodes in 12 V DC applications, such as across the coil of the relay or solenoid, or to protect other sensitive equipment?

"Why?" you may well ask. Because I've got a bunch of them to the spec I need.

ocrdu
  • 8,705
  • 21
  • 30
  • 42
Max
  • 31
  • 2
  • They may conduct when the 12 volt relay is activated from a 12 volt supply. Do you want that? – Andy aka Dec 21 '22 at 10:16
  • Similarly to @Andyaka's comment, in the other direction they may also _not_ conduct, if the reverse EMF doesn't have a voltage beyond the TVS breakdown voltage. Meaning that it won't act as freewheel and the reverse current will then go somewhere else. – Lundin Dec 21 '22 at 12:45

3 Answers3

3

Usually the reason why not is reliability.

As its name suggests, a Transient Voltage Suppressor (TVS) is designed for suppressing the effects of infrequent and unusual transients into a board connector or cable. It's connected directly across that input, to dissipate ESD, lightning strikes etc.

It's not designed for dissipation of normal operating power in a circuit. Its body package typically cannot conduct power well compared to other components because it doesn't need to. It's datasheets reflect this, with discharges through it rated at a very low pulse width and repetition rate indeed to keep the peak power and average power dissipation right down.

So using a TVS outside of its rated behaviour will decrease its reliability and it will fail more quickly and unpredictably than its specified life.

A TVS is just an electronic component, like any other. And like any component, it has:

  • characteristics that define all of its behaviour
  • specifications that define some of its behaviour as performance and reliability

The manufacturer will list only some of its characteristics as specifications, in datasheets and reliability data etc. The manufacturer will qualify the part's behaviour and reliability against the specifications and guarantee against them, but not against all of its characteristics.

Other characteristics aren't specified. And if you go outside the specs, and rely on unspecified component characteristics, you're on your own in uncharted territory with a good chance of sinking. For one or a few circuits, you may choose to take your chances and it may work out OK. But it may not and for many/most mass production situations, that's unacceptable because the cost of failures in the field swamps the cost of properly spec'd parts.

So, with a TVS as an operating current clamp, there simply are better alternative components and circuits to use that do give specified reliability.

And, as said above, usually lower business costs in the long and short term.

TonyM
  • 21,742
  • 4
  • 39
  • 62
  • It's not clear to me that TVS have any reliability problems as long as power or temperature ratings are respected. Further reading: https://electronics.stackexchange.com/questions/647527/technical-explanation-desired-are-tvs-safe-for-continuous-dissipation-why-why – Tim Williams Jan 18 '23 at 21:51
  • @TimWilliams, your linked question refers to my answer yet misquotes it and seems to interpret it wrong. It does appear like you've not read what's written here but seen what you want to see. – TonyM Jan 18 '23 at 22:10
  • 1
    Could you clarify (in both places) to clear up any confusion? Since the question here is repetitive peak service (across a coil etc.) it seems like you're saying a TVS would be perfectly suited here, but you then say "there simply are better alternatives". Or am I also misreading "operating current clamp"? (It's a voltage clamping device, I assume this is just a typo, or implying to clamp *excess* current?) – Tim Williams Jan 18 '23 at 22:18
  • @TimWilliams, I really don't understand your confusion so I have nothing to clarify, I've written it clearly. Nor do I understand in starting a separate question that refers to this answer without (a) notifying the poster (myself in this case) and/or (b) posting comments here which I would have taken to chat so we could have got somewhere productively. – TonyM Jan 18 '23 at 22:29
  • 1
    Well, it's a standalone question in that it's something I've heard (or at least have interpreted by implication) several times here on EESE, and haven't read anywhere else; this answer is just a particular example. Perhaps I should've asked; I apologize. But there are other instances, which asking just here wouldn't have covered. – Tim Williams Jan 18 '23 at 22:39
  • 1
    As to my reading, this answer seems to give a generally negative response: "usually the reason why *not*"; "It's *not* designed for dissipation of *normal operating power* in a circuit" <-- you don't mention if this case applies to the OP, but mentioning it seems to imply so; "So using a TVS outside of..." <-- similar implication; "So, with a TVS as an operating current clamp..." <-- (read: "to clamp [excess] current" e.g. inductive flyback) saying OP's suggested application is ill suited; have I grossly misread these, and if so, in what way? – Tim Williams Jan 18 '23 at 22:41
  • 1
    Negative, as in, "don't do that"; a generally negative or cautionary sentiment or tone; more cautionary against, than encouraging for, the action asked. I'm not sure that anyone else has [mis]read it in this way, but someone appears to have just upvoted these comments so it might not be just me? – Tim Williams Jan 18 '23 at 23:19
  • Re-read the question, which asks "*Is there a down side...*" and "*to using bidirectional TVS diodes in 12 V DC applications*". That's asking for reasons 'why not' and the answer starts "*Usually the reason why not is...*". That should clear your confusion. ("*it might not be just me*" Question's been viewed 110 times so far, I think it's hardly anybody.) – TonyM Jan 19 '23 at 08:26
  • Right, a negative question ("why not..") begets a negative answer ("don't because..."). So you're saying it's in general (the "12 V DC applications" part), not in reply (at all?) to the "such as..." part? – Tim Williams Jan 19 '23 at 16:36
  • That then contradicts your earlier point. – TonyM Jan 19 '23 at 19:07
  • I... you're telling my points are wrong. So, finding a contradiction should be a good thing, right??! You aren't giving up anything, are you. – Tim Williams Jan 19 '23 at 22:02
  • You must see the contradiction, though. – TonyM Jan 19 '23 at 22:53
  • What do you consider "normal operating power"? TVSs are widely suggested (including *by manufacturers*) to deal with EMR coil surge; are you disagreeing with that recommendation? – Matthew Mar 03 '23 at 20:10
1

Bidirectional 12 V TVS limit swings to +12 V in one direction and -12 V in the other direction.

Unidirectional 12 V TVS limit swings to +12 V in one direction and about 1 forward diode drop in the other direction.

If your application will not encounter or will not be damaged by negative transient voltages, then I guess bidirectional will be OK.

I am not sure I would put a TVS across a relay coil. I mean, it could be OK but you need to make sure the repetitive cycling is acceptable. I think that usually people let the coil energy freewheel through an ordinary diode.

But if you need to turn off the coil fast, then I guess you could use a TVS instead of a freewheel diode.

user57037
  • 28,915
  • 1
  • 28
  • 81
  • The advantage of a (unidirectional) TVS is that you'll also get ESD/spike protection at the same time. Some ordinary freewheel diode like 1N4007 won't help against such. Related topic: [What should be considered when picking a flyback diode?](https://electrical.codidact.com/posts/280393) – Lundin Dec 21 '22 at 12:50
  • A regular diode also delays the contact opening time. They're great for protecting stuff *besides* the relay, but not so good for the relay itself. (*Omitting* surge protection is actually best for the relay itself, but as is often the case in engineering, you're looking at a trade-off.) – Matthew Mar 03 '23 at 20:13
0

I am reading your questions somewhat like "Can I use a bidirectional 12V TVS across a solenoid coil as free-running diode?".

In this case: The point of a free running diode is to clamp the negative swing on the voltage across a coil to as close as possible too -0V. So by introducing the -12V bidirectional behavior you can harm your circuit.

In other cases like limiting the amplitude of a +/- analog signal they are just the easiest way to go. Feed the signal through a current limiting resistor and "clamp" it with the diode relative to GND on the full +/- amplitude band. If your resistance is big enough there will be no issues with power dissipation.

CAUTION: Using this circuit will introduce a RC-lowpass as TVS diodes (besides special types) can have significant capacitance!

This is the circuit I commonly use to limit external +/-5V analog signals and also provide ESD-protection. Even when used with two unipolar diodes and an extra cap I get f-3dBs at around ~130kHz which is good enough for my usual 100kHz uppper band-limit requirements.

Input

JYelton
  • 32,302
  • 33
  • 134
  • 249
ElectronicsStudent
  • 2,746
  • 5
  • 18