Trigger an SCR to switch a negative voltage

[Riccardo1]

Hi,

Just wondering about how to trigger an SCR for discharging a negatively charged capacitor into a coil.

I assume the gate must be set at some voltage relative to the cathode voltage. If this is the case, the cathode being at up to -1kV relative to ground would need a transformer right?

Is this diagram correct?

View attachment 40870

 

[Alex Tsekenis]

It looks right to me at least the principle of it.

You can treat the circuit as having a ground potential of 1kV and a positive potential of 0V in your schematic.

That SCR of yours looks rather faint hearted...a job for my IGBTs. 

 

[Lyker1]

I would check the holding current of the SCR you use, it needs to be greater than 1A in your circuit. Also, limit the gate pulse current with a resistor. That resistor will also have an affect on the holding current.

 

[Alex Tsekenis]

I would check the holding current of the SCR you use, it needs to be greater than 1A in your circuit.

Why is that? The holding current is the minimum current that must flow from anode to cathode for the SCR to remain in the ON state after the gate trigger pulse dies out.

In this application the current flowing will be much much more than 1A even over short periods of time.

If anything, one would aim for an SCR with the lowest holding current.

The holding current is a specification of the SCR itself. The gate current limiting resistor (if any) will not affect that characteristic like the base resistor in a bipolar transistor switch does not affect the maximum Vc-e rating.

 

[Hero999]

I don't see any need for a zener, the gate should clamp at around 0.7V and the secondary voltage depends onthe primary and the turns ratio.

If it's a sensitive gate triac you can easilly use a pulse transformer which can be made from a ferrite bead and some hookup wire.

 

[Lyker1]

Alex Tsekenis

Depends on how much current is available from the 1kV source, granted it

 

[Alex Tsekenis]

I think you have missed the point of Riccardos circuit. He is trying to dump the energy in the capacitor in the coil. The peak current will be several 10s of amperes after some time after the gate trigger pulse.The current will then decay as the capacitor discharges until it falls below the holding current at which point the scr will switch off with some energy left in the capacitor. R1 will then slowly reacharge the capacitor. Riccardo is not using the 1kV suppky and R1 to supply current to the coil directly. He probably doesnt want to switch it off anyway before the capacitor is discharged.

The holding current for a thyristor of the size of the BT152 is in the 10s of mA. And for a much larger thyristor it will be about a few hundends of mA. In any case much lower than 1A. Is there even a thyroistor with such a holding current?

The two transitor model does not explain the physics in a thyristor, just the latching action. Holding current does not depend on the gate resistor, it is a fundamental specification of the device that depends on temperature.

 

[Lyker1]

I fully appreciate that the idea of the circuit is to dump the energy from the capacitor into the coil, not from R1 directly. My point is this; the current sourced from R1 and the 1kV supply used to recharge the capacitor is enough to keep the SCR latched. For example the holding current for a BT152 may be as low as 16mA. If we take the values in Riccardo

 

[Alex Tsekenis]

Ah, I am with you now.

Maybe the best/easiest thing is to switch off the HV PSU before firing the SCR.

 

[Riccardo1]

Thanks everyone. I'll probably end up switching off the charging source prior to the discharge then.

 

[Riccardo1]

Lyker Says:

A lower resistance from gate to cathode will result in a higher current required to hold the SCR latched.

How would I calculate this?

The short circuit current from the 1kV source will be 270mA, so how can I use a resistor to make the holding current say 300mA?

I'm a bit confused by this idea, I also thought it was not possible to do it like this. Here'a a quote from Wikipedia (yes I know it is not the most reliable source).

It should be noted that once avalanche breakdown has occurred, the thyristor continues to conduct, irrespective of the gate voltage, until both: (a) the potential VG is removed and (b) the current through the device (anode−cathode) is less than the holding current specified by the manufacturer.

SCR is the same as a Thyristor right?

 

[Alex Tsekenis]

So how does the quote from wikipedia relate to gate resistor Vs holding current?

I think SCR is now synonymous to Thyristor, although they used to be slightly different things.

 

[Riccardo1]

Specifically the part saying "...the holding current specified by the manufacturer."

This seems to contradict the idea that a resistor could change the holding current.

 

[Alex Tsekenis]

Last time, holding current is a fundamental and fixed as far as gate resistor is concerned characteristic of the SCR.

 

[Riccardo1]

Yes, that is what I thought, and that is what the wiki says. I was hoping Lyker would explain how he thinks that a resistor can change this as he seems quite adamant that it can be done.

 

[Alex Tsekenis]

He is right, Rgk does affect holding current.

I found these documents:

http://www.littelfuse.com/data/en/Application_Notes/AN1001.pdf
http://www.teccor.com/data/en/Application_Notes/AN1008.pdf
http://us.st.com/stonline/products/literature/an/3563.pdf

Especially the last one on page 7 has a graph that you will find useful.

I was not aware of this. I am not sure why this is so. Anyway, very useful.

 

[sharilewis]

I would check the holding current of the SCR you use, it needs to be greater than 1A in your circuit. Also, limit the gate pulse current with a resistor. That resistor will also have an affect on the holding current.

 

[Lyker1]

An echo or a parrot?  ???