SCR/ Thyristor Gate driver

HI,

SCR manfacturers provide detailed application nodes.
I´ve designed drivers according this information and they work perfectly.

Klaus

KlausST said:

HI,

SCR manfacturers provide detailed application nodes.
I´ve designed drivers according this information and they work perfectly.

Klaus

Click to expand...

Do you happen to have any electronic copies of those application notes? can you pls point to them? I have spent quite a bit of time and could not find any. Most of the documents I found show some transistorized circuits. I want to use simple IC, just like gate driver IC for IGBT and Mosfet. Do you know any IC that can do this job?

As I experimented with a triac I planned to make it control speed of a small fan on our wood stove. The usual drive circuit was printed on the Radio Shack package. Possibly the alternate drive circuit containing a neon bulb in place of a diac.

I read that thyristors turn on when a sufficient volt level is applied to the gate. I figured I'd try varying that voltage with a 1M potentiometer. It seemed to work without any complicated circuitry. I made my speed control for the low wattage fan and it worked. It resembles a simulation below. The important thing is to find minimal current which is sufficient to turn on the thyristor. (Falstad's model has default 10mA and the 15K resistor dissipates 2 W.)

In real life I was able to get by with a 1M pot, adding a safety resistor or two considering the unit was hooked up to 120VAC. I'm quite certain the fan had a squirrel-cage motor because my more powerful ordinary window fan blew the whole unit instantly when I tried it.

High current SCR controllers are mostly using gate driver transformers for isolation. Because SCR are pulse triggered, it's the most robust and technically perfect solution. Transformer driver circuit is often build discrete, I append a datasheet of obsolete Infineon TCA785 to give an idea how an IC solution could look like.

Controller design depends of course on the intended control method which wasn't yet mentioned, e.g. full wave switching or phase angle control.

Hi,

an simple internet search for "scr gate drive application note" gives a lot good application notes from various manufacturers.

Before you can do a search for useful driver ICs you need to decide what volatge and current they need to provide.
Electronics design works with volts and currents. No informations given so far.

Depending on your values you may even find MOSFET/IGBT drivers suitabe for your SCRs.

Klaus

FvM said:

High current SCR controllers are mostly using gate driver transformers for isolation. Because SCR are pulse triggered, it's the most robust and technically perfect solution. Transformer driver circuit is often build discrete, I append a datasheet of obsolete Infineon TCA785 to give an idea how an IC solution could look like.

Controller design depends of course on the intended control method which wasn't yet mentioned, e.g. full wave switching or phase angle control.

Click to expand...

Thanks. Actually I wanted to avoid the transformer, modern gate driver have capacitive or optical or some form of isolation and plenty are available almost 10A peak. I think current is not problem, upto 20V I think they are available. I am surprised no one is giving any information. It seems SCR is a lost act in history. My application is DC but I need bi-directional SCR ( not triac) so back to back. 1000V battery and 350A, SCR part is MCC312-16io1 has 2 parts, I am seeing all the info on internet is very old and no modern ICs are shown.
For example, I see this part https://www.onsemi.com/download/data-sheet/pdf/moc3043m-d.pdf
it is through hole, has limited isolation and limited drive capability but clearly shows how to drive SCR back-to-back and has isolation too.





My SCR needs 220mA and I would like to avoid to design or get into pulse transformer, not my strength, looking for simple gate driver which can turn ON.

On second thought Looking at SCR datasheet it seems it needs whooping 2.5W- 20W power ( I assume for short time but more than IGBT on time), Thyristor world seems to be more complex than I thought..

the gate drive voltage and current depends upon time but the graph for Ig and vg is so complex


Do you know how to read figure 5?

Also is this device MCC312-16io1 a GTO or not ? how do we know that?

thanks

PS: I am very new to SCR design..

Surely no GTO, just standard high current SCR. In case you want an on/off DC switch, SCR doesn't work.

With isolated gate driver you safe the driver transformer but need isolated power supply instead. Typically more effort.

Simple opto-triac are an option for SCR control, but not suited for > 600 V.

Hi,

chaitanyab said:

I am surprised no one is giving any information.

Click to expand...

With "no one" you mean forum members?

For me speaking:
* General informations are given by the semiconductor manufacturers. All informations you need to know.
* now I´m waiting for your decisions, your ideas, your requirements, your values.

Klaus

KlausST said:

Hi,

With "no one" you mean forum members?

For me speaking:
* General informations are given by the semiconductor manufacturers. All informations you need to know.
* now I´m waiting for your decisions, your ideas, your requirements, your values.

Klaus

Click to expand...

Apologizes with "no one", I meant the semiconductor manufacturers and app notes from them. Not the forum members sorry for the confusion.

Requirements

1. Application Voltage - 1000VDC
2. Normal Application continuous current - 400A (for 20ms only)
3. Short Circuit/ Surge/ Inrush current capacity - 4kA
4. General information about - Basically a breaker kind of application. So controlled turn off is necessary.
5. Bi-directional application - In system charger in one direction and consuming load in another, both side operation

I think the short circuit current requirement drove me to the SCR route as they have very high surge current, apparently all SCRs are not GTOs. I don't know GTOs can be configured in dual Bi-directionally or not and I am thinking must be expensive.

FvM said:

Surely no GTO, just standard high current SCR. In case you want an on/off DC switch, SCR doesn't work.

With isolated gate driver you safe the driver transformer but need isolated power supply instead. Typically more effort.

Simple opto-triac are an option for SCR control, but not suited for > 600 V.

Click to expand...

In your experience does the 250mA current requirement typically I am seeing for lot of HV SCRs for around 100usec seems to be more than the ns requirements for IGBT/Mosfet peak currents, does the driver power consumption a concern here? Just for my knowledge I am asking.

I am wondering normal IGBT Isolated gate driver is probably not going to meet, even though they are available up to 1200V.

Hi,

chaitanyab said:

Application Voltage - 1000VDC

Click to expand...

chaitanyab said:

So controlled turn off is necessary.

Click to expand...

Quite confusing requirements.
1000V DC .... while an SCR is an AC device

controlled turn OFF ... While an SCR is a controlled turn ON and latching device ... that only turns OFF on (AC) zero cross of current.

Please explain.

Klaus

KlausST said:

Hi,


Quite confusing requirements.
1000V DC .... while an SCR is an AC device

controlled turn OFF ... While an SCR is a controlled turn ON and latching device ... that only turns OFF on (AC) zero cross of current.

Please explain.

Klaus

Click to expand...

I think SCR/Thyristor can be AC/DC application. Voltage source does not matter, just similar to the diode. See https://www.electronicshub.org/scr-applications/
go to middle section of the document.

The controlled turn off is called GTO (Gate Turn OFF) this is a special class of thyristors which can be switched off by Gate, control similar to an IGBT can have, just different gate drive requirements.

Hi,

I don´t know what your application/schematic looks like.
Thus I recommend:

Try it.
You may simply use a 12V battery and an incandescent light bulb.

Klaus

The above might help.

Shows turnoff techniques in DC applications.

Regards, Dana.

Exc

danadakk said:

SCR MANUAL : General Electric, Auburn, NY : Free Download, Borrow, and Streaming : Internet Archive

Established In 1987, we are an INDEPENDENT e-commerce SHOP with the flexibility of focusing on providing a quality product with personal service. With over...

archive.org

The above might help.

Shows turnoff techniques in DC applications.

Regards, Dana.

Click to expand...

excellent book, I think the class E commutation shown on page 160 is the best ! Simple transistor and isolated voltage or pulse transformer to switch off, thanks a lot !
Now I need to see how the timing performance is. Has anyone switched off or commutated a thyristor?

chaitanyab said:

Has anyone switched off or commutated a thyristor?

Click to expand...

I saw one article telling how it can be and is done with a single brief burst of opposing voltage at the thyristor terminal (current-carrying terminal). This stops current flow through the device by the identical manner which textbooks describe a thyristor shutting off. It's hard to wrap one's mind around this method since to ordinary thinking it sounds like an error which risks destroying components. I forget just what is the optimum circuit to build to deliver the turn-off pulse. Of course it needs to be carefully designed.

BradtheRad said:

I saw one article telling how it can be and is done with a single brief burst of opposing voltage at the thyristor terminal

Click to expand...

Below is the LTspice sim of an example commutation circuit that uses a pulse from a second SCR (U2) to turn off U1:

That circuit is very nice - an inductor in series with C1 would form a resonant reset and limit the peak current and di/dt - which can be very high indeed with the given circuit - SCR's have a rating for di/dt at turn on - which should not be exceeded.

No doubt that an SCR turn-off circuit of this type can work, it's e.g. frequently used for exposure control with Xenon flashlights. I think the discussion has nevertheless departed from original problem, most clearly specified in post #10 under requirements. An open point, shall the breaker be able to disconnect 4 kA or only 400 A?