Is there a way to test a SPD without spending thousands ?

[Avo Mk8]

This simply injects a surge and check when the SPD trips and shows you the trip V. Then you can check if the trip V is the same as the one that the SPD manufacturer gave you in the datasheet.

The test does not involve a 'surge'. The trip mechanism is generally a one-shot disconnect, as you'll see from post #7, so triggering the device is not helpful!
The test involves slowly ramping up the applied voltage (from a high impedance source), and measures the voltage at which the MOV begins to conduct, as you describe. That doesn't harm the SPD as the current involved is small.

 

[SirLouen]

The test does not involve a 'surge'. The trip mechanism is generally a one-shot disconnect, as you'll see from post #7, so triggering the device is not helpful!
The test involves slowly ramping up the applied voltage (from a high impedance source), and measures the voltage at which the MOV begins to conduct, as you describe. That doesn't harm the SPD as the current involved is small.

AFAIK, from what I've read, transients don't trip, only permanents. The only way to "test" transients is by breaking the cartridge on purpose. In fact, any transient SPD has a switch in the device, only the cartridge and a red/green indicator like @brianmoooore commented

How are you going to test this without breaking it?

 

[Avo Mk8]

How are you going to test this without breaking it?

Well it depends on what test you want to conduct. I was just referring to the test that MFT's typically carry out, which do not trip the SPD, but check that its MOV is alive and well.

 

[pc1966]

Exactly as @Avo Mk8 says, you can test the SPD is functional by using a MFT or IR tester, basically it should sustain 370-430V when tested at a higher voltage (i.e. 500V or 1kV) and the power involved in that is trivial so no risk of damage to the SPD.

If you wanted to type-test the SPD to see if it can absorb X Joules of surge energy, etc, as per the data sheet then it could very well be destructive, but that is far different from verifying it is present and connected as expected.

 

[pc1966]

I had a feeling that is was trademarked in the USA by one of the manufacturers, but a quick look at Wikipedia suggests otherwise.

I found a data book from around 1987 that shows I mis-remembered it, the trademark was for GE-MOV and not the generic Varistor or MOV notation:

 

[pc1966]

Those GE-MOV parts were eventually absorbed in to the Littlefuse brand. They had various factories around the world, including one in Ireland. But most were shut down bar China along with a weasel-worded statement about "shareholder value".