transient voltage in dc circuit

What happens with a zener diode if it receives a transient? Is current at a voltage below the zener's rating that continues to flow, and then the zener conducts current to ground, to reduce the voltage, or will it be a loss of power situation whilst it's conducting?

When it's conducting, is it a 0-ohm short, or is there some resistance such that it will get warm, but recover after the transient is over?

I'm using an automotive application and so if I receive a transient greater then 16V, it'll fry some ICs.
I'm thinking i'll put a 15V zener on input power to ground, but just want to make sure what behaviour will occur during breif transients.
I imagine there wouldn't be total loss of power during the transient ?

I'm also not looking to have limited current available in the circuit as a side effect of adding the zener - what I think will happen: I can go along happily consuming current, and after the voltage exceeds 15V, the zener will take care of it.

What I don't want to happen: I draw 4A at 15V, and the zener limited by it's 5W rating breaks.

A Zener diode breaks down to Gnd but maintains the rated voltage on the cathode. It must be able to handle the excess current from the source. You may think,of it as a variable R that loads the source to maintain the voltage.

As it is only rated at a certain power, you must limit the power that can be applied. A Zener regulator usually has a power R input to dump most of the excess voltage. The voltage drop is calculated using the load current and Zener current. The zener is used to regulate the voltage that is in effect on a voltage divider.

If you simply put a Zener on the input with high current capacity, it will burn out. It needs a R to limit the current to a safe level.

If you are only expecting short pulses, select a R that will limit the Zener current to less than its peak pulse current. As it is not carrying current continuously, it can handle higher currents on pulses.

Weedpharma

A zener is not a protection device on its own. As you have worked out, if it needs to conduct 4A to keep the input voltage within tolerance, then it will burst.

It's best used 'behind' a protective resistor. This is useful to protect individual inputs on the IC (or Arduino.) It's not useful as a power supply protection. Then you need some sort of voltage regulator.

In the image attached is what I was first thinking… since VIn will routinely be below the zener voltage, wouldn’t it therefore conduct 0mA over the zener, until the transient occured, where it would probably have blown the zener?

With a resistor inline with it, limiting the zener to it’s 5W rating, would that still work? Load could be > 100W (greater than the zener rating) and the zener will be fine since it only conducts when the fault is present?

Excuse my messy fritzing mock up.

A car battery and a normal zener diode is like a mouse protecting an elephant.
Spikes are better handled by TVS diodes. Spike devouring super zeners...
The second diagram won't work.
Best to have an LC filter between battery and sensitive load, and a 16volt TVS diode.
Leo..

The second diagram is wrong as you need to take the output from the Zener, not the resistor.

As drawn the R protects the Zener but not the equipment.

Fully describe what you are wanting to do so we can advise with the knowledge rather than guessing.

Weedpharma

I want to protect the ICs from intermittent spikes above 16V (the lowest maximum voltage of the ICs on the board).

TVS diode is probably the answer, the zener diode came up as a possible quick fix (TVS diodes aren't on the counters around here).
I hoped to simply put the diode in line with my existing boards power input.

Sigh! XY problem yet again!

You really need to tell us what these ICs are and what the circuit in question is.

If you wanted surge suppression you could use SADs (Silicon Avalanche Diodes). They have a very flat response to surge pulses. I know because I tested countless units with a 5x20 uS High Voltage Pulse Generator. Here is an example of a similar product:
SAD Surge Suppressors

The Silicon Avalanche Diode (SAD) devices include such devices as TransZorbs, Zeners, Sidactors, etc. They are typically characterised by a predictable low let-through voltage, a fast response time, a very low surge rating, and a very high cost.

Silicon devices typically have a lower clamping voltage and better clamping ratio for the same MCOV of a MOV device. For a 220-240V rated piece of equipment a let-through of 600V-1000V is desired.

There's two things that require protection:

  • The power supply coming into your box.
  • Every other wire which leaves the boundary of your box.

I think this question is asking about #1 and is keeping the details on #2 secret. Some of the suggestions above seem to be relevant to lightning protection, not a wild automotive alternator. Lightning is a very difficult problem: any spark which has jumped hundreds of feet through the air won't have any problem going around any so-called protection you might have.

I like the good old 7805 voltage regulator. They're hard to kill and, with any basic attempt at reverse-voltage protection, don't seem to conduct too many problems through to the downstream equipment.

Paul__B:
Sigh! XY problem yet again!

You really need to tell us what these ICs are and what the circuit in question is.

I was trying to minimise pointless detail.

For example, telling you that the IC I am protecting in my circuit is a UCC27424 from TI will only tell you that it's datasheet maximum rating of supply voltage is 16V, and it's a mosfet driver.

The mosfets on the other side of the UCC27424 are rated at 18V, so pointless detail because if I'm protecting against transients for 16V, I'm also getting the same for 18V.

The voltage regulator to the arduino isn't seen as much of a threat, but it'll get protected as well if I simply protect the input power to the board from transients.. which I can do as I have terminals to interface the input power to.

I suppose there's merit, I have decoupling capacitors on the supply, but these I don't see holding back a transient of over voltage, as they are rated at 25V..

So I do need to protect against transients, I thought a zener would do the job, but it was never going to work, as I'd have to put a resistor in, which limits the power needed to drive the mosfets.. not good enough.

Remaining solutions I can see:

  • Use a voltage regulator to deliver regulated 12V - but that's not doable as supply won't regularly be above 14.5 (the minimum of a 7812). Low drop out? Can't get them over the counter, so I have to shop online. Shopping online introduces delay, so other solutions are worth considering.

  • Use a TVS diode (avoided in the first place as can't get over the counter, so introduces delay). Better option then the voltage regulator.

  • Swap out the UCC27424 for another component, yet have the same problem: need to protect the mosfets from transients.

Given the above, I'm probably going with TVS diode, which has it's own considerations (I want one that has a reverse breakdown voltage greater than 14.5 (charging voltage), and won't minimum breakdown voltage of 16V (the lowest of the max rating of the components being protected), and I want it fast.

There is not a lot of room between ~14 and 16volt.
As said, keep spikes OUT with a proper LC filter.
Use a 15v TVS for extra security.
Leo..

That's basically what I was getting at. You need a LDO regulator rated for a higher input voltage. In fact, you should be able to synthesise such a function with discrete transistors, since close regulation is clearly not as important as LDO and surge isolation.

(Not everything comes in ICs.)

You should be aware that, in addition to load dump transients and high voltages radiated from ignition circuits, some breakdown services use 24 V batteries for jump starting and there is a possibility of a battery being connected in reverse by accident.

This application note might be worth reading: http://www.st.com/st-web-ui/static/active/en/resource/technical/document/application_note/CD00181783.pdf

Russell.

Remember that a regular LDO is a slow device, and is not very good at stopping spikes.
Inductors and capacitors are much better at that.
Leo..

http://www.maximintegrated.com/en/app-notes/index.mvp/id/760
I like figure 1 on the link, and the LDO regulator combination…

If I apply Figure 1 to the input of the low drop out reg, it should give me a fair bit of protection from transients down stream (at the mosfets) and protect from 24Vdc.

If the drop out reg fries, it’ll trip the fuse in a short condition, and provide no cooling in the open circuit condition.

That’s a nice simple solution.

Russell.