Thyristor problem

[Kim Taylor]

I use a thyristor to discharge a small 1kV capcitor through a transformer coil to generate a 20kV spark which is used to trigger a 3.5kV spark gap. The 1kV supply used to charge the capacitor is intermittent (0.5 seconds ON and 10 seconds OFF). The capacitor therefore has to hold its charge for at least 10 seconds. Unfortunately all the available high voltage thyristors I have tried leak to such an extent that the capacitor discharges in less than 1 second. The now discontinued TN22-1600D works perfectly in this application. My question is: can anyone suggest an equivalent to TN22 that DOES NOT LEAK?

Most grateful for your help

Kim Taylor

[magpie]

The Thyristors you are using are still available in all packages from RS Components.

[Kim Taylor]

Hi Magpie
Thank you for your reply to my query about a thyristor. I get most of my components from RS and indeed my few remaining TN22s came from there. But RS now states that this component has been discontinued. I have emailed RS asking for them to suggest an equivalent but they have not replied. I don't know where to go next. Any suggestion would be most welcome.
Kim

[David]

Are you sure that the leak is not in the capacitor or other parts of the circuit? RS has a good selection of 2kV thyristors in its online catalogue at various prices. It may be appropriate to use a higher rated one than the TN22-1600D, which is a 2A one, intended for fluorescent tube starting.

[Kim Taylor]

Hi David

Thanks for your suggestions. I am pretty sure that it is not the capacitor which is leaking since there is no problem with the TN22 thyristor. The 1kV capacitor is only .02u so it is unlikely that the thyristor is being over-loaded by the discharge. However, I will try other thyristors. The latest one from RS that failed the test was TYN1012RG. It is annoying that I can find no way of testing these components without soldering them into the circuit board. I can detect no leakage at low voltage.

Thanks again for your help.

Kim

[magpie]

You don't have to solder them in to test. Take out the thyristor, solder in three wires and bring them out with small crocodile clip attached to the ends. You can test various ones to see how they perform. If you need to hold them in place use a bit of Blue-tack and keep the contacts separated with a bits of paper to make sure they will not touch.

[David]

It is always difficult to give precise advice at long range. I am wondering about your first post that "the cap discharges in less than 1 sec". Is this a complete discharge to near enough zero, or to (say) 50%? I am trying to get an idea of the time constant of the cap and the equivalent leak's conductance here. What is the voltage rating of the cap? what is its construction (paper, mica, polycarbonate ...)? Do you put exactly 1KV on the cap? Is the transformer a car ignition coil?

A circuit diagram with voltages & values of components would be extremely useful here. It may be that we will need to put in some clamping diodes or other surge protection.

In the meantime, keep experimenting with magpie's excellent suggestion. Try to use thyristors with a higher voltage rating than that applied to the cap.