Need Help. How to fire 2 seperate scr's alternatly

[CrackleZapPow]

I am building an electronic power supply from my own idea. I have 2 seperate thrystors that each have a gate wire and a shunt wire. I want to fire them alternatly. A first, and then B, and back to A and so on at 120 htz.

They will be switching 24 V DC at 80 Amps. The gates can handle 10 volts at around 250 ma.

I know how to build the rest of it, but don't know how to bring it to life. Any ideas on how to fire these up??

 

[itsmoked]

Since they shut off only at zero current, how do you use a thyristor on DC? Or is a thrystor something I've never heard of?

 

[Skogsgurra]

Are you really going to switch DC with thyristors? Or are you building an inverter with forced commutation?

The former will not work since there is nothing to turn the thyristors off, once fired. For the latter, there are loads of application notes and descriptions on the web. Just google what you want to know. For starters, you need triggering transformers to keep the thyristors isolated. I think that you should do some research of your own before you go here for help.

 

[VEBill]

Assuming that you sort out the DC/thyristor issue as mentioned already, then it seems like you should be able to somehow use the alternate polarities of the incoming AC to fire the two devices (each firing at 60 Hz).

If this doesn't work out, then you need a simple divide-by-two flip-flop.

(PS: Whatever you're doing, with only two phases it might run backwards. So be careful where the observers are standing when you fire it up for the fire time. ;-) )

 

[nbucska]

Please read the FAQ.
Can you describe what you are trying to do?

<nbucska@pcperipherals DOT com> subj: eng-tips
read FAQ240-1032

 

[zeusfaber]

Are these simple thyristors, or have you got a pair of GTOs?

A.

 

[CrackleZapPow]

Thank you for the help!

Silly me, I thought a person could use the shunt lead in a curcuit to turn the thrystor off.

What I need is to switch 2 different areas of my design on and off alternatly. A-B-A-B-A-B-A....

(A) must be completly off before (B) switches on.
(B) must be completly off before (A) Switches on.

If I can get the right switches and the right circuit, I can make it come alive.

The switches will be switching DC current like a chopper, but will be doing it alternatly between switch A and B.

If a thrystor is not what I need, then what are some ideas of what to use instead. What would you put there? I need somthing that can handle High amps at low voltage and at around 80 to 120 htz.

My design can be built as big as the componants will allow.

I really appreciate your help
Thanks again

 

[SolderingGunslinger]

A pair of power FETs driven by a flip-flop?

I remain,

The Old Soldering Gunslinger

 

[CrackleZapPow]

The thyristors that I am using right now are IR80RIA40. The old number is IR2N2030. It has 2 small wires coming out of it. A gate wire and a shunt wire.

Gunslinger, you get a medal for a great idea!

If I used a flip-flop, It would have to have 4 seperate FET's. One to turn A on, one to turn B off and then it would have to flop to one to turn B on and A off. That is a great idea. I just don't know how to design the circuit. maybe you can show me one.

Any more ideas? Anyone know a circuit that I can use a Force-Commutated SCR to do the job? The GTO's turn on and off using the same gate but I don't like the fact that I will have to use 1/3 of the amps to turn the gate off. I really haven't looked at the GATT's yet. Maybe a half-bridge circuit would work. The hard part to this is that I am switching DC and that it has to alternate between 2 different switches at above 60 htz.

Thanks again for any input

 

[logbook]

The first thing you have to decide is what guard time is acceptable between the two switching events. You need a non-overlapping drive waveform to ensure that only one circuit is on at a time. If you use a single D-type flip flop and drive one power MOSFET from Q and the other from Q-bar this is guaranteed to allow both circuits to be on at the same time. Typically you could use an edge shaping circuit on say the rising edge of each drive waveform. In other words the turn on is slowed down relative to the flip flop precise timings.

Take the flip flop Q output through a diode/resistor parallel pair, into a capacitor to ground on the output. The diode makes the voltage on the capacitor fall more rapidly than it rises, if you get the diode the right way round. The voltage on the capacitor will necessarily be a poor drive waveform, so square it up again using a logic gate like an inverter. Use this same network on both the Q and Q-bar outputs. Having provided a non-overlapping drive then you can use whatever power MOSFET, GTO, or whatever that you can get your hands on. I think a power MOSFET is the easiest and cheapest way to go.

 

[ScottyUK]

I've gotta ask...

Why are you trying to build a power supply using SCRs, GTOs, and the like? The very reason no-one uses them, except on ultra-high power applications, is because of their unfriendly drive requirements. Small GTOs are pretty rare these days, if you can still get them at all. The voltage and current range you are working with is right in the ballpark for a MOSFET-based design. IGBTs are a bit lossy for this application, and you don't need their high voltage capability.



----------------------------------

If we learn from our mistakes,
I'm getting a great education!

 

[nbucska]

to continue Scotty et al:
Use a comparator to detect zero-xing (polarity) and
use its +/- outputs to switch the MOSFETs. Parallel Rs & diodes in series with the gates can delay the turn-ons if necessary, but at the 0-xing the VTG is low, so a few usec short circuit won't dissipate too much pover considering
the source's --trafo & wiring -- inductance.



<nbucska@pcperipherals DOT com> subj: eng-tips
read FAQ240-1032

 

[CrackleZapPow]

Thanks for the input!

As I look at my design, I see that the 2 scr's that I need to switch are in series with a diode in between them. That would mean that I need 4 seperate mosfet's. I would need to fire them sequentialy.

A off
B on
B off
A on
A off.....

It is starting to look like a half-bridge circuit with a diode in between. The problem again is that I am not switching AC. I will look further at that angle. Any other ideas ? This is just a prototype of a much bigger model, so if I should be using something else to switch DC please guide me.

Thanks again!

 

[ScottyUK]

MOSFETs are ideally suited to switch DC. They are designed for this application. For high voltage, high current look at IGBTs. Unless you are working on megawatt-level designs - which I doubt - you will very probably cover any likely application using either MOSFET or IGBT. The basic design topologies are pretty much interchangeable for these two switching devices.



----------------------------------

If we learn from our mistakes,
I'm getting a great education!

 

[logbook]

I don’t get how two SCRs is replaced by 4 MOSFETs. If it "works" with two SCRs (except they won’t turn off) then it would work with two MOSFETs. Two SCRs in series sounds like a half-bridge, whereas 4 MOSFETs means a full bridge.

It just seems like you are not telling us enough to let us help you :-(

 

[jsolar]

ScottyUK hit it on the head. You only have 24 V to high side switch. The open collector drive from a 555 like timer IC or sw pwr suppy IC will drive a pair of P channel FETS just fine. You did not specify the gurd time needed on your break before make requirement. Or is it just zero time? If so, you can rely on the drive switching speed to take of that.