Weekend DRSSTC

Reaching, Sat Aug 19 2006, 01:21PM

okay, this project grew up yesterday, cause i have sooo many parts of my old drsstcs and experimental primaries etc liying around that i decided to use them in a quick but not dirty drsstc build and tested in a weekend. yesterday i searched for parts to use and came up with some really nice parts i found, such as

fullbridge of 40n60 igbts (almost finished)
primary with 6mm copper tubing and 7turns, inner diameter of 21cm, height of 10cm, strikerail etc (finished)
a nice hdpe case i build a few weeks ago already with a 80mm fan installed (finished)
10x wima fkp1 caps 100nf 1,6kV for a mmc of 300nF 3,2kV DC (finished)
some nice high frequency cases
yesterday i finished the driver (from my new drsstc6) and tested it ok
finished the mmc and the 1/1000 current transformer today.

still have to build the 1/140 feedback transformer and put it in a shielded case , have to wind the 2 gdts for the bridge and add some decoupling caps, tranzorbs etc to make the bridge work. here are a few pics.

Link2 (here is what i started with yesterday)
Link2 (finished driver in the small package,hehe)
Link2 (finished mmc, double sided pcb full tinned)
Link2 (the 1to1000 current transformer in a hf box

more to come very soon, still have to finish the bridge, the secondary and the toroid
secondary will be a 11x40 or 11x35cm wound with 0,17mm wire to give a fres of approx 90khz with a 10x30 toroid, but i doubt i can finish them this weekend cause i want a very nice toroid and that needs some time
Re: Weekend DRSSTC
Wilson, Sat Aug 19 2006, 02:55PM

Man! Thats an incredibly tiny driver!! suprised
There are components on the other side of the PCB ... right?
Re: Weekend DRSSTC
Reaching, Sun Aug 20 2006, 12:22PM

small update, started to wire the things up as professional as i can, that needs some time and i have to build up a comparator charging circuit for the electrolytic cap too. dont want to use this old charging circuit i have on hand, its too stoneage like rolleyes
but things seems to do great and this will be a very powerful small portable drsstc when im finished. heres a pic of the current status

Link2

heres the front

Link2

okay, not finished this weekend, but the next weekend is only 5 days away, hrhr shades

more updates

working on a comparator charging circuit for the 2200yf electrolytic cap. status leds for charging and ready, seems to work fine on my breadboard, now im on building it on a pcb. when it works then i can finish the complete electronics today and can make first decisions on how to make a nice toroid and secondary shades

update

finished the charging circuit, it charges the cap up to 250volts dc via 2 charging resistors 190ohm 8 watt and when the cap reaches 250volts a relay bridges the charging resistors, so i can pull up the bridge relatively smooth without tripping the breaker while loading up the big bridge cap. therefore there are 2 leds one shows the charging status the second led goes on when the cap is full (~310volt dc)
there are 2 switches for this device, one to switch it on and the second to switch the bridge rectifier on and start the charging , when the switches are off the cap discharges automatically through a another resistor
Link2
Re: Weekend DRSSTC
Reaching, Thu Aug 24 2006, 05:33PM

new update

had a failure in my charging circuit due to a bad voltage divider but now it works fine and everything does like it should. finished the electronics today, and tommorow i can start winding the secondary. then i can do the first low voltage testruns and when everything works i can have first full power tests on saturday or sunday. cant wait to see this beast in action.

here are two pics of the current status

Link2 (okay not industrial standard but better than nothing)

Link2 (inside full of electronics)
more to come soon
Re: Weekend DRSSTC
Hazmatt_(The Underdog), Thu Aug 24 2006, 08:54PM

very very nice.

When you're drilling your PCB's don't push down so hard, its deforming the fiberglass. Just let the drill do its job.
Re: Weekend DRSSTC
Reaching, Sat Aug 26 2006, 07:52AM

started doing some first low voltage resonant tests and got dissapointed cause the thing wont start to oscillate, but no problem, i found the failure on my driver pcb, i have 2 wrong turned diodes which leads the feedback signal to ground. now i have to change the diodes and hopefully the thing will do its job. more info to come soon, and a few scopeshots if everything works after fixing the failure

started doing first testruns and got first light today with around 50volts input some 10cm discharges. after fixing some problems with my driver i can now do first tests on mains voltage, hope that doesnt end in a disaster tongue

update

yeah, and it ended in a disaster and i cant find the failure, my fullbridge is just popping away when i test it on mains voltage. i thought it was the driver powersupply which was not enough smoothed , after fixing this i tested again, but the same failure, it is like the bridge is just turned on and ends like a bakon dead , . my thought is the feedback ct which has only 115turns , but why does this thing work on 50volts with all signals ok but not on mains voltage?
now im on building a new driver version with (almost the same like steve ward describes in his drsstc guide) hope the 1/1000 ct´s will fix the problem, hopefully,
Re: Weekend DRSSTC
Steve Ward, Mon Aug 28 2006, 10:35PM

Seems unlikely that the CTs would be a problem. Ive used 1:100 CTs for feedback in the past, they seemed to work just the same as the 1:33 to 1:33 CT arrangement that ive been using more recently.

You might want to make sure you dont have some voltage transients killing the IGBTs.
Re: Weekend DRSSTC
williamn, Tue Aug 29 2006, 02:58AM

I have used 1:100cts for the feedback circuits on my past two coils and they work fine. I still use cascaded CTs for my current sense transformers becasue using a low value loading resistor casued all kinds of problems with faulty triggering. Your Cts should be fine. Could your clamping schema for feedback Ct be faulted causing your 5V logic to fry? Have you checked the operation of your logic?
Re: Weekend DRSSTC
Reaching, Tue Aug 29 2006, 10:27AM

yeah, logic and everything is fine. the only thing i wondered about was the feedback which sometimes stops at 20volt input on the bridge when i turn the on time higher or the off time lower, this problem is gone when i use 50volts on the bridge. and i dont think i kill the igbts in cause of voltage transients, okay i have not tvs on the bridge, but i never used them in my 6 drsstcs i build before, so the problem is a secret for me, .im very sure that it has to do with my high side cap charging circuit or with my driver (its a revised version of my drsstc6 controller which is a revised version of steve wards drsstc3 driver)

now a question, where should i put the ct and feedback transformer in and how does it behave , after the mmc or before the mmc ?? okay the current transformer should be mounted after the mmc to catch the real pulsed amperage but where does the feedback ct "work" better?. its like a noobish question for one who already build 6 drsstcs but 4 of them are secondary feedback which worked better for me than any primary feedback i tried so im in the thought to switch back to secondary base feedback with overcurrent protection.
i never popped a fullbridge by switching mains on it and never fried any igbts during a testrun when i worked with secondary feedback which is so "bad for the igbts" cause of no zero current switching etc, i dont share this opinion after 4 of my good igbts died so fast
Re: Weekend DRSSTC
williamn, Tue Aug 29 2006, 12:12PM

I always put the feedback CT before the MMC. If your logic is ok, I would suspect it has something to do with your gate drive, did you ever get a chance to use a scope and check for shoot through?
Re: Weekend DRSSTC
Steve Conner, Tue Aug 29 2006, 12:58PM

It's a series circuit, so it doesn't matter if you put the current sensor "before" or "after" the tank capacitor, the current flowing through it is the same.
Re: Weekend DRSSTC
Reaching, Tue Aug 29 2006, 03:49PM

i dont know, its the first drsstc i spend so much effort into to make it look good and now, nothing works, dissapointing. some parts arrived today, jeah and some 400volt tvs,, i made a pcb wich is a 1:1clone of steve wards circuit he describes on his drsstc guide, maybe this will do the trick, but i dont know, i used my version of this driver with success in my drsstc6, the only thing i really changed is to use a lm393 instead of an lm311 and i changed a part of the overcurrent logic with a 7400 to get away from this funny rc combination, and it works perfect, on the first german teslathon this year i had an interrupter failure which made my interrupter go like mad with on times of about 2mS ,and i ran the coil for minutes with this settings without blowing parts, the sparks made a rattling loud noise and were brighter and longer i ever seen before, hehe okay blew some 10A fuses but no blown igbts, so really my logic works.
im not sure but i think its the high side cap charging circuit.
but i dont know, its very simple, i use an lm393 with an 1/100 voltage divider to measure the cap voltage. when power switch is on the cap charges via 190ohms resistors until i get 250volts dc on the cap, then the comparator switches a relais on via some bc238 transistors .the relais bridges the charging resistors, .i have 2 leds to show the status, charging and ready, nothing too special. okay its a high side charging circuit cause the negatives of the cap and the comparator are the same, but the circuit is an stand alone with own bridge rectifier (but same transformer)so it should have nothing to do with the driver causing the bridge to go boom.
Re: Weekend DRSSTC
Bennem, Tue Aug 29 2006, 05:07PM

Hi Reaching,

In the picture of your lay out, you use a lot of wire (inductance)
from your buss cap to your bridge,
could induce spikes at higher buss voltage and kill your IGBT's?
Plus you said that you don't use any tvs on your bridge.

Just a thought.
Re: Weekend DRSSTC
Reaching, Tue Aug 29 2006, 05:46PM

but that wouldnt pop a bridge just in time when you switch on the interrupter, btw never had any problems with voltage spikes etc, and i use 12mm² wire, thats a low inductance compared to my other designs.
Re: Weekend DRSSTC
Bennem, Tue Aug 29 2006, 06:25PM

Do you use a variac to slowly increase buss voltage?
its a good idea to use one whilst de-bugging your system.
If fuses blow at lower buss voltage, problem could be GDT phasing problem,
if things work ok at lower buss voltage but start blowing fuses/IGBT's at higher voltage,
problem could be spike problem.

I had problems with my UCC driver ic's causing all weird and wonderfull problems
when i built my last small DRSSTC.
Re: Weekend DRSSTC
Part Scavenger, Tue Aug 29 2006, 10:59PM

Reaching, if you don't have a variac, you can make one out of an MOT. It's not "linear" it's really more of a tapped transformer, but I rewound one and made a switch bank that canceled out each lower switch (not hard if you think about it). I was going to put it on my site because it worked great, but I acidentally shorted it one day and killed it. Soon after I got a real variac so I didn't pursue it any longer. I'll see if I can dig up a picture if you want.

The advantage of this is you can have whatever-power variac you want for really cheap, however, it's not continuous. (IE. current is shut off briefly during switching.)
Re: Weekend DRSSTC
Reaching, Wed Aug 30 2006, 10:25AM

i have a variac rated 2,2kW but thats not the question here. why should i put in a variac to slowly bring up the voltage until the igbts blow upt the third time,.. i know there is something wrong in my circuit or whatever and i want to find the failure until i blow more igbts. btw i have to build a new bridge cause i have only 2 40n60 miniblocks left, and the others are already build in in my other drsstcs. but i have 10 of the to 247 pack 40n60 but without diodes , i think i can use them when i can find some usable fast freewheeling diodes such as mur1560 or whatever. seems that nothing would work i build a new driver (exact 1:1 clone of steve wards drsstc guide driver) but it wont work properly, i know its not a failure in steves circuit, more a failure in my pcb but i spend 2 hours of debugging and didnt found the failure. maybe someone of you had the same problem. i tested the driver with a signal generator at the feedback input and with an interrupter on the interrupter input. the only thing the driver makes is to switch the 100khz feedback signal through the driver ics, the interrupter signal is lost on pin 7 of 74hc109,,,, rolleyes
Re: Weekend DRSSTC
Bennem, Wed Aug 30 2006, 05:51PM

Hi Reaching,

so you have no interrupter signal at pin 7?

have you chosen the correct values of R9 and C14?
this delay RC should be chosen acording to your coils running frequency.
I think Steve uses a 1nf and a 22k for his coils frequency.

Just a thought. smile

On second thoughts.....ignor the above, even if you had the wrong values, an interrupter signal should still be present at pin 7 of your flip flop ic. ill
Re: Weekend DRSSTC
Reaching, Wed Aug 30 2006, 06:31PM

yeah i know, i build several types of flip flop sync circuits so im somewhat confirm with them, but i cloned steves circuit 1:1 and it doesnt work, i checked everything 5 times ,looked for shorts on the pcb etc but found nothing, the circuit is exactly that in the shematic and it doesnt work, i tried out 5 different 74hc109 different 74hc14 different 555s ucc chips etc but nothing, i dont know the reason. on pin 1 of the 74hc109 is the inverted interrupter signal, pin 2 and 3 are connected to +5volt pin 4 is the feedback signal, pin 5 is the noninverted interrupter signal and on pin 6 and 7 is nothing,sometimes the right signal but only 0,5volts amplitude, seems like a pin7 short on the pcb without a short on the pcb, lol, already build the 74hc109 on a breadboard and connected pin for pin by showing the current draw of the circuit which is nearly 200ma without a load and the 74hc109 gets warm thats not normal i know. ahh i dont know cry
Re: Weekend DRSSTC
Steve Ward, Wed Aug 30 2006, 06:52PM

All ICs have proper 5V measured at the pins? I really cant imagine what the problem might be unless its simply a bad connection somewhere. Ive had bad sockets before, where some pins didnt make contact, and id spend hours trying to figure out what went wrong. Maybe use a continuity checker on each pin of the IC to the trace on the PCB. Many of the problems i encounter with my PCBs is either forgetting to supply power or ground to the chip (its the most basic thing, so i tend to overlook it of course).
Re: Weekend DRSSTC
Reaching, Wed Aug 30 2006, 07:56PM

found a failure but dont know the reason, when i measure the voltages on every ic socket while all ics are out i measure the exact 5 volts on the logic and 15 volts on the driver ics and lm311 and ne555, but when i put in the 74hc14 or 74hc 109 the logic voltage increases to 11 volts amazed and i dont know why, i checked the whole pcb and there are no shorts or something else on the back and its only a 1 sided pcb . please, if someone has time, check out my pcb and compare it with steves drsstc guide driver, i dont know what to do,

heres the board file in eagle format

Link2
Re: Weekend DRSSTC
ragnar, Wed Aug 30 2006, 09:12PM

I had issues with schottky diodes clamping rails doing that....

15V on UCCs, 5V on 4046...

Except as soon as you added the diodes, I don't know if they sent the regulator crazy, but everything went wrong. =P No, they weren't incorrectly biased. ^^
Re: Weekend DRSSTC
Marko, Fri Sept 01 2006, 05:24PM

Doh, I don't see any obivous bugs in board.

Only difference I used is a resistive divider to assure HC14 input will fall to 0V if interrupter is accidently disconnected.


Can't you just go for a soog old signal generator and scope method..? from the beggining to the end of circuit and you'l surely find what's wrong..

If board is OK only GDT can be in bad phase or shorted...
Re: Weekend DRSSTC
Reaching, Fri Sept 01 2006, 11:23PM

gdt was not connected, its only the driver which draws around 200mA without any load, in cause of this obvious failure that the 5 volt logic goes up to 12 volts when a 74hc14 or 74hc109 is in the socket confused
Re: Weekend DRSSTC
Steve Ward, Sat Sept 02 2006, 09:37PM

Is your interrupter a 12V signal? I bet that somewhere, you are pumping too much voltage into the 5V chips, and the clamping diodes inside the chip are sending that energy to the supply, charging it up to 12V.
Re: Weekend DRSSTC
Reaching, Sat Sept 02 2006, 09:41PM

interrupter is 9volt signal with an 1k damping resistor ,this is what i used in all my drsstcs so it cant be the problem, . and the problem still exist without any interrupter connected
Re: Weekend DRSSTC
Marko, Sun Sept 03 2006, 10:40AM

Is your interrupter a 12V signal? I bet that somewhere, you are pumping too much voltage into the 5V chips, and the clamping diodes inside the chip are sending that energy to the supply, charging it up to 12V.

I considered that but I erased the sentence from the post cheesey

With series resistor that couldn't happen, and chip's diode will just clamp off small current to supply rail.

If there was no resistor NE555 output or 74HC14 chip itself would die much before they manage to mess up the regulator and increase logic supply voltage like such.

It just seems to be a dead regulator or etc. (78xxx sometimes do very wierd things when they die)
Re: Weekend DRSSTC
Reaching, Sun Sept 03 2006, 07:35PM

yeah cheesey found the failure on my pcb, its within the board file, a wrong jumper was set through pin3 of the ucc driver ics to +12 volt causing pin 4 from 74hc14 and pin 7 from 74hc109 floating at +12volt confused , now everything should work, still on testing