2

I am designing a circuit that has a nominal 24VAC input. This circuit is protected by a 4A Fast-Acting fuse. I would also like to insert a transient voltage surge protection. For achieving that, I am willing to add a 26VAC varistor between the input terminals. I am well aware that a long duration over-voltage may lead to a catastrophic failure in the varistor. However, for this voltage there is no TMOV device available, not even THT. My question is: is the only solution to insert an extremely ugly, huge and Thermal Cutoff near the varistor terminals?

Edit 1:

As suggested by Dwayne Reid, I think it would be better to use a TVS Diode instead of a varistor. Besides, he pointed that the 26VAC rating is underestimated. I agree, and think that a 11% variation in the mains 230V as well as a 25% voltage regulation from the transformer should be considered. A 35VAC rating makes much more sense.

Edit 2:

Well, I have finally decided to use 2x 1812 35VAC varistors in parallel. The TVS diodes are a good option, but can only withstand much lower surges. As far as I have read, the fast acting fuse at the input should protect the varistor against catastrophic failure. Another worry with the TVS Diode is that most manufacturers do not specify failure condition. It could only damage the P-N junction in a way it appears to be an open circuit. In that way, the user would never notice that he is currently unprotected against surges.

Thanks.

gstorto
  • 669
  • 1
  • 7
  • 18
  • 1
    By THT thermal cutoff, do you mean a PTC polyfuse? – George Herold Feb 04 '15 at 16:56
  • No, I meant a thermal cutoff: http://en.wikipedia.org/wiki/Thermal_cutoff. Not sure if it is the right solution, but in an application note of a TMOV, they tested it against a thermal cutoff + MOV. – gstorto Feb 04 '15 at 17:23

2 Answers2

2

One of the main differences between a TVS and a Varistor is that the voltage tolerance of a TVS is much tighter than a Varistor. They also don't appear to have the "wear" factor that Varistors have.

Varistors that are repeatedly subjected to transients appear to have their clamping voltage lowered as a result of the stress. Eventually, they become conductive at the operating voltage and then fail catastrophically. That is: they get very hot, then turn black, then eventually catch fire.

Many of the better AC power bars that have Varistor clamps within also have a thermal cutoff device mounted between two Varistors. The thermal cutoff device disables the power bar when a Varistor fails.

In terms of what happens to a transient suppression device when it is exposed to voltages in excess of their rated clamp voltage: they get hot. The will, in fact, get hot enough to fail.

When a Varistor fails because of over-temperature, it generally becomes restively-conductive. That is, it turns black, then catches fire. Your over-current device may or may not operate, depending on how low the resistance of the Varistor has dropped.

When a Tranzorb fails because of over-temperature, it generally becomes a short-circuit. This causes sufficient current to flow to cause your over-current device to open.

Dwayne Reid
  • 23,390
  • 2
  • 35
  • 67
  • Ok, so you mean that the TVS diode would not go catastrophic due to extended wear from surges. But it does not mean that a long duration over-voltage may not cause catastrophic failure in the TVS diode as well, does it? So the final question is: do I have to protect the TVS diode against extended over-voltage, if so, how? – gstorto Feb 04 '15 at 18:40
1

There is an alternative to a Varistor called a "Tranzorb". These are available in many more voltage ratings (E12 steps, if I recall correctly) and have many other advantages over Varistors. However, they tend to be significantly more expensive than Varistors.

Be aware that Tranzorbs can be specified as uni-polar or bi-polar, depending upon your requirements.

Internally, a uni-polar Tranzorb acts like a zener diode. A bi-polar Tranzorb looks like TWO zener diodes in series in opposite polarity. This allows them to work with bi-polar or AC signals.

Wikipedia has a good description of Tranzorbs: Transient-voltage-suppression diode

All that said: I think that you are making a mistake in trying to suppress transients so close to your desired input voltage. Far better to design your circuit to withstand normal variation in supply voltage and choose a transient suppression device that clips transients significantly above that voltage.

For example, we build a LOT of HVAC equipment that runs from 24Vac. All of our boards have a relatively small Varistor on board rated at 35V. Specific part number is S07K35 from Epcos and readily available from many suppliers including Digikey S07K35.

The only times that a Varistor in any of our products has failed is when the installer tapped the input transformer for 120V but connected to a 230V supply. This damaged the Varistor and tripped the Polyswitch over-current protection device before the input transformer was destroyed by the over-voltage. We use certified Class 2 transformers with internal non-resetable thermal / over-current devices. No smoke or flames but the installer wrecked several systems before discovering his mistake.

Dwayne Reid
  • 23,390
  • 2
  • 35
  • 67
  • Thanks. I was already searching for TVS Diodes thinking it may be the solution. However, I have never seen any mention they are safer than a MOV, just that they are quicker. So ok, for 0.3EUR for a TVS Diode I think I will add a bidirectional one. But the question remains: with a constant over-voltage will they catch fire or something like that? Or may I consider that the fuse won't let it happen? Second remark, I did a simple calculation (11% tolerance for input 230V), 25% voltage regulation from the transformer, and I arrived to a maximum input of 33.3V, close to the 35V you recommended. – gstorto Feb 04 '15 at 17:13