SCR Recovery Speed

I have an SCR.

I put voltage on the SCR.

I pulse gate current to the SCR.

The SCR conducts, and the voltage diminishes.

The voltage goes negative (there's a cap and a coil involved).

What next? Yes, the SCR turns off -- but how quickly? The SCR data sheet indicates a "circuit commutated turn-off time" that's considerably longer than the resonant frequency of my cap and coil. Does this mean that the SCR will conduct, even in the negative direction? Or does it just mean that if I don't wait long enough and put a positive voltage on the thing that it'll conduct?

I need to know more about SCR behavior...

--
My liberal friends think I'm a conservative kook.
My conservative friends think I'm a liberal kook.
 Click to see the full signature
Reply to
Tim Wescott
Loading thread data ...

There's probably a spec on slew rate and turn-off characteristics. Got a data sheet? I haven't used an SCR for 30-40 years :-) ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
 Click to see the full signature
Reply to
Jim Thompson

Yes.

SCRs act like regular old diodes when reverse recovery occurs. The difference is, because the layers are stuffed full of charge carriers from forward operation, and there's no way to extract the charge, they only dissipate through recombination. The anode junction (which carries the voltage in reverse bias) may recover relatively quickly (since there's a path for recovery current), but when the voltage swings around again, the rest of the device may still be conducting. Even if it doesn't turn on immediately, it may turn on due to breakdown voltage or rate of rise, both of which are more sensitive during the recovery tail.

At work, we have some old induction heaters that use stacks of small puck SCRs and operate as high as 85kHz. The stacks are necessary because SCRs over 200V are really slow. Trick is, frequencies that high only work at high Q, so that it keeps ringing for a while, delivering energy over a relatively long time period, until the SCRs can turn off and repeat the cycle. The inverter itself might run at a few kHz.

Tim

--
Deep Friar: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms
 Click to see the full signature
Reply to
Tim Williams

"Tim Wescott"

** Yes to the latter.

** SCRs are slow, switching speeds may go to a few kHz at best.

You know what the alternatives are.

.... Phil

Reply to
Phil Allison

it's the same problem slow switching diodes have.

Normally, they are triggered on with the shorted pulse width time possible on the gate/cathode.

Hence the reason you don't see Triacs used in bridged circuits. The aftermath is worse.

Jamie

Reply to
Jamie

"Jamie" wrote in message news:G4OMq.6908$ snipped-for-privacy@newsfe01.iad...

Actually, they are most often driven by pulse trains. This is nice because you can get away with a much smaller transformer than the operating frequency would suggest, and allows for variations in conduction angle.

The drivers they have in the power lab at school produce a couple initial cycles of overdrive and a variable number of normal cycles thereafter. I forget who made them, but they looked like some ridiculously expensive commercial model. Something of a waste to use with the junkbox TO-220 SCRs students get from the lab supply desk.

Tim

--
Deep Friar: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms
Reply to
Tim Williams

I believe it varies with the current thru the device. Once the current reaches a certain level (pinch off?), the device will stop conducting. Many designs incorporate a inductor to help the SCR stop conducting. You might want to read up on GTO's, another weird device, an SCR (or SCS 'Silicon Controled Switch') designed to switch DC.

Cheers

Reply to
Martin Riddle

Possibly. Closest I got were the mains commutated converters for battery chargers up to 600A, 36 years ago.

Special circumstance there was the battery load, and each SCR dipping into reverse early, we used half controlled bridges, three SCRs high side, and three diodes for the low side.

There's a big reverse current spike like a slow diode, but I've not read that stuff since the 1975 GE SCR Handbook, which I don't have any more. SCR only conducts a bit in reverse until it turns off like oirdinary diode, I guess you're really asking how long to wwait before applying positive voltage?

I suggest that's when the SCR runs out of them electron whatsits inside, you'd visualise the physics of it better than my memory of them.

Simple, slow, noisy, high forward drop, some of the bigger ones neeed a fairr kick on the gate to get going, and too much dI/dt will kill one due to hotspots.

That's my vague input, IGBTs and MOSFETs took over from them for high power?

Solves the commutation problem, unless you're running nice slow mains commutated converters, and why would anyone build the transformer for that these days? And, an output inductor about half the size of the transformer, airgapped until it no longer saturated at full output current.

Grant.

Reply to
omg

It'll conduct in reverse for the turnoff time, then block. However, if the dV/dt of your resonant circuit is high enough, it'll turn right back on again on the next available positive half-cycle without being triggered.

Engineer to colleague: "what's the turnoff time of that SCR?"

Colleague: " Don't know, I've been trying to get it off since last night"

--
"For a successful technology, reality must take precedence 
over public relations, for nature cannot be fooled."
 Click to see the full signature
Reply to
Fred Abse

Previous post canceled when I realized what I'd said. This is what I meant.

You may get a reverse spike due to device capacitance, but the SCR will be off. You have to wait the duration of the turnoff time before it will

*stay* off.

One thing to look out for is dV/dt turn on. Your resonant load could turn the SCR back on without any applied triggering.

--
"For a successful technology, reality must take precedence 
over public relations, for nature cannot be fooled."
 Click to see the full signature
Reply to
Fred Abse

Thanks everyone.

Tim I'm responding to your post for the general thanks, because, your explanation about the carriers in the various layers makes sense to me, and jibes with what I think will happen.

Since this is exactly the behavior that I can live with (turn-off when the resonant circuit goes negative, then wait for a long time compared to the carrier lifetime), I'm reasonably happy. Now, as long as I don't go blowing up SCRs...

--
My liberal friends think I'm a conservative kook.
My conservative friends think I'm a liberal kook.
 Click to see the full signature
Reply to
Tim Wescott

ies

d
.

igh

o

ttdesign.com- Hide quoted text -

The thyristors turn-off when the terminal current goes to zero. This is not necessarily in phase with the voltage. Another thing is there are roughly two categories of thyristor, those intended for line frequency apps and others for higher speed switching inverter apps. Sounds like you want the switcher variety.

Reply to
Fred Bloggs

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.