SCR precharge

[alexhanyuan]

Hey all! Have some more free time now that I'm finished up with school. Wanted to share a little SCR precharge circuit that has been working quite well for me.

Originally for my DRSSTCs I had been using a precharge resistor and relays, but it's difficult to find cheap relays that are rated for more than 30A. Didn't want to waste the room on large contactors, so Steve recommended I try out SCRs. Big thanks to him for getting me started with this.

I found a bunch of cheap Semikron dual SCR modules that were the same Semipack 1 package as the diode modules I had been using so it was an easy swap. Plus, I got to get rid of two relays and the precharge resistor. I get to charge as slow or fast as I'd like (down to 1 second haha). But the main benefit of this is that it is much more robust. I have yet to blow up any SCRs despite some of the abuse I have subjected them to.

My github has some more info and all of the design files:
https://github.com/alexhanyuan/scr_precharge

[davekni]

Nice simple circuit.  Even though you have it working well, two suggestions:

1) Either remove C7 and C9 or make R6 and R8 larger than 0 ohms.  If you increase R6 and R8, it would be even better if R9 and R10 (10k pull-down resistors) were moved to the other sides of R6 and R8, directly on logic outputs.  That way FET gate drive voltage isn't reduced by the series resistance.
As is, C7 and C9 are directly on ATTINY output pins, causing large glitches of switching current.  That induces ground bounce inside the chip package.  Although it is working now, the next batch of chips might be more sensitive.  Or the next layout have a bit more ground or +5V trace inductance.

2) If this circuit is used on a single phase (230V in Europe), neutral needs to be X2-1, not X2-2.  (I'm presuming that neutral is roughly the same voltage as ground as would be expected here in USA.)  A label for such might help anyone else copying this.  Or, change to a fully differential comparitor input circuit.

[alexhanyuan]

Dave, thanks for the input! Now that I look at it, I'm not even sure why I included those 1n caps in the first place. Also, you are correct that it should be X2-1, typo on my part. Fixed!

[davekni]

Also, you are correct that it should be X2-1, typo on my part. Fixed!

I didn't realize that either notation specified which was neutral.  BTW, if using this circuit on a line-neutral circuit, you could leave off the neutral connection.  Almost no current flows there presuming neutral and ground are at roughly same potential.  Operation will not change if the neutral connection is left open.
If using in a hot-to-hot situation where ground potential is between the two, then the X2-1 connection to - comparitor input is of some value, keeping comparitor input common-mode voltage closer to center supply voltage.