Large DRSSTC [update] / Projects / Forums

Forums

4hv.org :: Forums :: Projects

« Previous topic | Next topic »

Large DRSSTC [update]

first previous 2 3 4 5 next last

Move Thread

LAN_403

Linas

Mon Apr 13 2009, 02:06PM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

Finn Hammer wrote ...

I wish I could understand what those 2 BS170'es are doing......

Tdon delay (but not tdoff delay)
two BS170 doing same as if i use on 3K resistor ultrafast diode, like then you driving IGBT, put on 10R 1N5819 or similar

i have problems with feedback. i madded small 100W induction heater from my last GDT driver, signal from CT was very strong, and burn out all my resistors (2W 1K)
what to do ?

Steve Ward

Mon Apr 27 2009, 05:52AM

Registered Member #146 Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055

Excellent work so far... just a couple thoughts:

The RCD snubbers may in fact not help provided you already use the de-coupling capacitors at the bus bar connections (it depends on inductances, but your RCD will almost always have more L than the single cap). But, the snubbers do look very well laid out, good job there

. It cant hurt really

.

Second, i suspect you will need more like 2500-3000 Apk for the power level you are talking about at only 311VDC.

How does your de-sat protection work? Is it not detailed in your gate driver schematic? I have wondered whether or not desat protection would help much for DRSSTCs. I think your best bet is to overdrive the gates so much that you are well clear of desaturation risks during normal operation. If you somehow get an inverter fault to ground or something, then your de-sat circuit would be able to save the day, which i cannot say about anything ive designed, so in that respect its good. You have to softly turn off the IGBT for proper protection, though, otherwise the resulting voltage spike might contain more energy than your snubbers can absorb and you blow the IGBTs... at least thats what most power engineers claim.

Linas

Mon Apr 27 2009, 02:39PM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

Hi Steve
Over current protection based on MC33153, i connected just signal input, desat input, desat output, and have some 10K load on MC33153 output just in case.
every thing is simple. MC will know then IGBT are desaturating to much, then it 1 in fault out, that signal i will take with isolator (TI, 5ns delay ) and connect other isolator side to LM311 output, 555 input. It will work same as CT. So desaturation protection turn off igbt, but it don't let continuous operation some time. ( that is problem, i can start resonate...)
For Ipeak, i still can't get capacitors for large peaks currents. I want to use 0,5uF or so.
Now have just 150nF and 6KV

BSVi

Mon Apr 27 2009, 05:15PM

Registered Member #1637 Joined: Sat Aug 16 2008, 04:47AM
Location: Kiev, Ukraine
Posts: 83

i have problems with feedback. i madded small 100W induction heater from my last GDT driver, signal from CT was very strong, and burn out all my resistors (2W 1K)
what to do ?

The more turns of wire you have on your CT, the less power you have to dissapate with your resistors. I think, 1000turns would be enough for your.

Linas

Mon Apr 27 2009, 05:28PM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

BSVi wrote ...

The more turns of wire you have on your CT, the less power you have to dissapate with your resistors. I think, 1000turns would be enough for your.

yep, if you are using 1:33 >>1:33, but i don't want to use two cores, it will give some delay. But if i use 1000t on single toroid, it will have enormous L :(
I found bug in circuit. Feedback was same as SSTC, so i replaced 1K to 2R2, it forks just fine :)
Also I want to say thanks to Finn, he give me brass bolts, thanks a lot

BSVi

Tue Apr 28 2009, 08:18PM

Registered Member #1637 Joined: Sat Aug 16 2008, 04:47AM
Location: Kiev, Ukraine
Posts: 83

Steve wrote (
I should also mention a bit about current transformers. I read a paper that basically said, for RF use, cascaded (1 CT feeds the next CT) CTs are better than a single CT with lots of turns on it. This makes life much easier! For my DRSSTC-1 using primary feedback i use 2 cascaded CTs, each a 1:33 ratio. I use an identical set of CTs for the over current detection. It seems to work great. One other bit of crucial information. Terminate the CTs (either it be the diodes on the feedback CT or the resistor on the OC CT) as close to the board as possible (i mount the resistors/diodes right on the PCB). This greatly helps reject noise on both circuits.

Dont worry about delays. You are building big coil, and big coil resonates at low frequency, so delays are going to be negligible.

Dr. Dark Current

Tue Apr 28 2009, 08:35PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Linas wrote ...

BSVi wrote ...

The more turns of wire you have on your CT, the less power you have to dissapate with your resistors. I think, 1000turns would be enough for your.

Delay? The capacitances of the windings themselves are quite negligible and do not cause any delay in my opinion.

Linas

Wed Apr 29 2009, 05:14AM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

5-30ns it's lot

i guess that my circuit (with IGBT) can give more than 800ns delay

also i think to wind on single toroid 200-300T for feedback, i know who have winding machine, so isn't problem for me :)

Linas

Mon May 18 2009, 04:03PM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

Here is interrupter. two 555 for bps and on time control
Principal:

for signal transmitting i think to use fiber optic. also thinking soft bridge turn on trigger on interrupter panel. I don't think use variac, because it is lot of weigh

LithiumLord

Mon May 18 2009, 07:01PM

Registered Member #1739 Joined: Fri Oct 03 2008, 10:05AM
Location: Moscow, Russia
Posts: 261

Err,why no 556? ;) Nice board anyway :p

first previous 2 3 4 5 next last

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob