DRSSTC - cannot get oscillating
Marko, Mon May 29 2006, 07:00PMWell I dragged my mess out this morning for first, low-voltage run.
It's a halfbridge one with steve ward's primary feedback circuit.
Voltage to halfbridge is filtered by one big 400V 1000uF cap and divided by two smaller 470uF 200V caps (there are also transzorbs, decoupling caps and mess).
I ran it on 30V (so I get 15V per igbt) but it just seems to be ''working'' on interrupter pulses, and secondary rings to some point where it lights neon bulb 5cm away.
Primary feedback CT cascade gives no signal at all, zero.
I really have no clue what could be happening, I tested the board bunch of times before and it seemsed to work perfectly with resistor-limited signal generator input.
Only thing I can guess is that primary current is simply way to low for such a cascade to feel it.
Output of CT is at zero voltage all the time, and I need some 3V to trigger 74HC14 input
But If I add more input I thing I'l just blow something up because it will continue to work on interrupter frequency.
And from what I figured out it really should get running at 30V in...
Re: DRSSTC - cannot get oscillating
Dr. Shark, Mon May 29 2006, 08:20PM
What about using a higher input voltage but current limiting it (e.g. with a lightbulb), so you cannot blow up your IGBTs if it does not work? I am not a expert in current transformer, but I could well imagine that a transformer designed for 500A or so wont give much of a signal at only 5A.
Dr. Shark, Mon May 29 2006, 08:20PM
What about using a higher input voltage but current limiting it (e.g. with a lightbulb), so you cannot blow up your IGBTs if it does not work? I am not a expert in current transformer, but I could well imagine that a transformer designed for 500A or so wont give much of a signal at only 5A.
Re: DRSSTC - cannot get oscillating
Steve Ward, Mon May 29 2006, 10:19PM
The coil should oscillate just fine at 15V or better applied to the bridge (mine usually oscillates with only 5V or so). Its designed so that even a few A on the primary of the CT will give plenty of signal (its a very high gain system).
Did you use an oscilloscope to verify that the feedback is staying at 0V? Do you detect a short on the CT input when the CT is disconnected? Also, you may need to *reverse* the phasing the the CT, by either passing it through the wire the other way around, or reversing the output leads.
Also, each IGBT sees the full supply voltage (its not divided because the transistors are in "series").
Steve Ward, Mon May 29 2006, 10:19PM
The coil should oscillate just fine at 15V or better applied to the bridge (mine usually oscillates with only 5V or so). Its designed so that even a few A on the primary of the CT will give plenty of signal (its a very high gain system).
Did you use an oscilloscope to verify that the feedback is staying at 0V? Do you detect a short on the CT input when the CT is disconnected? Also, you may need to *reverse* the phasing the the CT, by either passing it through the wire the other way around, or reversing the output leads.
Also, each IGBT sees the full supply voltage (its not divided because the transistors are in "series").
Re: DRSSTC - cannot get oscillating
HV Enthusiast, Mon May 29 2006, 10:21PM
What are the specifications of your current transformer? Assuming you have a 1000:1 transformer, you should have no problem getting a signal with even 5A of current. Remember, the transformer you have takes the sampled current and gives an secondary current equal to 1/1000th of the primary current. However, voltage is the inverse, and since you have no burden resistor in this circuit, the voltage can be quite high, hence the need for the clamping network. So, with even a 5A primary current signal, you should have ample voltage to drive your circuit.
In fact, you should get oscillation without even having power on the bridge circuit. This is what that initial buzzing sound you may get when you first plug in your circuit.
I suspect your polarity of your transformer may be reversed. Try doing that, and see if it makes a difference.
HV Enthusiast, Mon May 29 2006, 10:21PM
What are the specifications of your current transformer? Assuming you have a 1000:1 transformer, you should have no problem getting a signal with even 5A of current. Remember, the transformer you have takes the sampled current and gives an secondary current equal to 1/1000th of the primary current. However, voltage is the inverse, and since you have no burden resistor in this circuit, the voltage can be quite high, hence the need for the clamping network. So, with even a 5A primary current signal, you should have ample voltage to drive your circuit.
In fact, you should get oscillation without even having power on the bridge circuit. This is what that initial buzzing sound you may get when you first plug in your circuit.
I suspect your polarity of your transformer may be reversed. Try doing that, and see if it makes a difference.
Re: DRSSTC - cannot get oscillating
Marko, Mon May 29 2006, 10:31PM
Of course I tried to reverse CT.
And output is scoped and completely dead, with just some fluctuations at few hundred milivolts.
I don't see why should I get anything, even close to 5A trougt the primary when it is running just on interrupter pulses. I would probably blow it up if I increased the voltage more.
(Interrupter 'enables' , inverting UCC goes low and uninverting high, and this pulses the pri mary without feedback starting to oscillate.)
IGBT will see 300V; but it will switch only half of it because of voltage divider on primary.
Marko, Mon May 29 2006, 10:31PM
Of course I tried to reverse CT.
And output is scoped and completely dead, with just some fluctuations at few hundred milivolts.
I don't see why should I get anything, even close to 5A trougt the primary when it is running just on interrupter pulses. I would probably blow it up if I increased the voltage more.
(Interrupter 'enables' , inverting UCC goes low and uninverting high, and this pulses the pri mary without feedback starting to oscillate.)
Also, each IGBT sees the full supply voltage (its not divided because the transistors are in "series").
IGBT will see 300V; but it will switch only half of it because of voltage divider on primary.
Re: DRSSTC - cannot get oscillating
HV Enthusiast, Mon May 29 2006, 11:15PM
Using a scope trace your signals. Use a signal generator to input a 5V square wave at the primary current feedback input and trace this through your circuit. I think you'll probably find out why its not working.
HV Enthusiast, Mon May 29 2006, 11:15PM
Using a scope trace your signals. Use a signal generator to input a 5V square wave at the primary current feedback input and trace this through your circuit. I think you'll probably find out why its not working.
Re: DRSSTC - cannot get oscillating
Steve Ward, Tue May 30 2006, 05:23AM
Huh? When one IGBT is ON, the other is OFF, and has the entire supply voltage across it... end of story. I dont care about a "voltage divider" for the return of the primary circuit, its passing RF, so it all looks like a loop anyway (capacitors are low Z to RF).
There should be some noise at least on the output of the CT when the primary is pinged from the interrupter going HI.
Figure out the surge impedance of the primary, and use V=IZ, and that will give you exactly what the initial primary current would be from the very first pulse applied. My DRSSTCs have a surge Z around 6 ohms, so at 30V, thats 5A already.
What type of diodes are you using to clamp the CT? And what arrangement?
Steve Ward, Tue May 30 2006, 05:23AM
IGBT will see 300V; but it will switch only half of it because of voltage divider on primary.
Huh? When one IGBT is ON, the other is OFF, and has the entire supply voltage across it... end of story. I dont care about a "voltage divider" for the return of the primary circuit, its passing RF, so it all looks like a loop anyway (capacitors are low Z to RF).
There should be some noise at least on the output of the CT when the primary is pinged from the interrupter going HI.
I don't see why should I get anything, even close to 5A trougt the primary when it is running just on interrupter pulses. I would probably blow it up if I increased the voltage more.
Figure out the surge impedance of the primary, and use V=IZ, and that will give you exactly what the initial primary current would be from the very first pulse applied. My DRSSTCs have a surge Z around 6 ohms, so at 30V, thats 5A already.
What type of diodes are you using to clamp the CT? And what arrangement?
Re: DRSSTC - cannot get oscillating
Marko, Tue May 30 2006, 05:46PM
I guess my last post was somewhat confusing :P
Never mind.
I tried to use antenna on feedback input, and got first light (or at least first oscillations).
With 30V input I got a 1cm streamer that could be drawn to a total of 2cm, when primary was around best tune. I also tried a signal generator (welcome back to Jimmy Hynes age
) but it was quite worse than antenna, and I needed double tuning...
I dug the whole bridge trying to debug the feedback, but it simply seems to be too weak.
CT's are completely fine,
All connections are good, and DRSSTC actually works (sort of 'works').
Your arrangement, of course. two 5V zeners series with 4148's in reverse.
ANt that seemed to clamp the antenna signal very well.
I get a tiny streamer from the antena but it doesn't bother the circut at all.
Marko, Tue May 30 2006, 05:46PM
Huh? When one IGBT is ON, the other is OFF, and has the entire supply voltage across it... end of story. I dont care about a "voltage divider" for the return of the primary circuit, its passing RF, so it all looks like a loop anyway (capacitors are low Z to RF).
I guess my last post was somewhat confusing :P
Never mind.
I tried to use antenna on feedback input, and got first light (or at least first oscillations).
With 30V input I got a 1cm streamer that could be drawn to a total of 2cm, when primary was around best tune. I also tried a signal generator (welcome back to Jimmy Hynes age
) but it was quite worse than antenna, and I needed double tuning...I dug the whole bridge trying to debug the feedback, but it simply seems to be too weak.
CT's are completely fine,
All connections are good, and DRSSTC actually works (sort of 'works').
What type of diodes are you using to clamp the CT? And what arrangement?
Your arrangement, of course. two 5V zeners series with 4148's in reverse.
ANt that seemed to clamp the antenna signal very well.
I get a tiny streamer from the antena but it doesn't bother the circut at all.
Re: DRSSTC - cannot get oscillating
HV Enthusiast, Tue May 30 2006, 06:13PM
Sounds like you might have something out of whack. Check your primary / secondary resonant calculations again or measure the resonance and see if they are at least close together.
HV Enthusiast, Tue May 30 2006, 06:13PM
Sounds like you might have something out of whack. Check your primary / secondary resonant calculations again or measure the resonance and see if they are at least close together.
Re: DRSSTC - cannot get oscillating
Reaching, Tue May 30 2006, 06:36PM
do you have clamping diodes on the secondary side of the CT, (i guess)? . if so, check the polarity of them, maybe they are wrong and let your signal going to gnd or so.
Reaching, Tue May 30 2006, 06:36PM
do you have clamping diodes on the secondary side of the CT, (i guess)? . if so, check the polarity of them, maybe they are wrong and let your signal going to gnd or so.
Re: DRSSTC - cannot get oscillating
vasil, Tue May 30 2006, 09:01PM
Bad solders on the PCB? It was one of my problems....
Did you conect right the pins of the 74 HC 14? All the chips are in right position? sometime in a hurry it is possible to make it wrong.
The input for the drive has to be very well filtered. Thats why i bought a dedicated 12 V for it.
Where did you get from the GDT cores?
vasil, Tue May 30 2006, 09:01PM
Bad solders on the PCB? It was one of my problems....
Did you conect right the pins of the 74 HC 14? All the chips are in right position? sometime in a hurry it is possible to make it wrong.
The input for the drive has to be very well filtered. Thats why i bought a dedicated 12 V for it.
Where did you get from the GDT cores?
Re: DRSSTC - cannot get oscillating
Marko, Tue May 30 2006, 09:31PM
Guys, give me a break a bit..
The board works, it was tested lots of times before I used it for this coil.
I put my signal generator to very good use there.
I made gate drive signals to perfection far before I actually hooked the board to IGBT gates.
And I surely didn't do things like turning diodes wrongly or something.
The Coil oscillates nicely with antenna, giving some 2,5cm arc (at 30V in).
Since I get nice sparks form external 1:100 osciloscope CT (without burden) I guess one in cascade on the bridge probably got shorted (I receive only a small amount of noise on output). Bridge will get an autopsy next weekend
(Im leaving town tomorrow and I can't continue much work now)
Marko, Tue May 30 2006, 09:31PM
Bad solders on the PCB? It was one of my problems....
Did you conect right the pins of the 74 HC 14? All the chips are in right position? sometime in a hurry it is possible to make it wrong.
The input for the drive has to be very well filtered. Thats why i bought a dedicated 12 V for it.
Where did you get from the GDT cores?
Guys, give me a break a bit..
The board works, it was tested lots of times before I used it for this coil.
I put my signal generator to very good use there.
I made gate drive signals to perfection far before I actually hooked the board to IGBT gates.
And I surely didn't do things like turning diodes wrongly or something.
The Coil oscillates nicely with antenna, giving some 2,5cm arc (at 30V in).
Since I get nice sparks form external 1:100 osciloscope CT (without burden) I guess one in cascade on the bridge probably got shorted (I receive only a small amount of noise on output). Bridge will get an autopsy next weekend
(Im leaving town tomorrow and I can't continue much work now)
Re: DRSSTC - cannot get oscillating
HV Enthusiast, Wed May 31 2006, 01:02AM
2.5cm output arc with only 30V in sounds reasonable.
HV Enthusiast, Wed May 31 2006, 01:02AM
2.5cm output arc with only 30V in sounds reasonable.
Re: DRSSTC - cannot get oscillating
Marko, Fri Jun 02 2006, 08:51PM
I removed and disassembled both CT's, and put hem to test separately.
One works with oconsional arcing to core, while the second seems to be totally shorted.
I can get shocks just from single CT when coil is operating just at 30V, and I didn't take some great care for insulation! (just a bit of tape).
Probably II also managed to damage thick enameled wire I used for second CT, since output was complete dead-short...
Thanks for help anyway, I'm now glad I didn't try to push the coil more in this state..
Marko, Fri Jun 02 2006, 08:51PM
I removed and disassembled both CT's, and put hem to test separately.
One works with oconsional arcing to core, while the second seems to be totally shorted.
I can get shocks just from single CT when coil is operating just at 30V, and I didn't take some great care for insulation! (just a bit of tape).
Probably II also managed to damage thick enameled wire I used for second CT, since output was complete dead-short...
Thanks for help anyway, I'm now glad I didn't try to push the coil more in this state..
Re: DRSSTC - cannot get oscillating
Steve Ward, Sat Jun 03 2006, 01:22AM
This is another reason i switched to using 2 cascaded CTs so that each would have only a few 10's of turns. This allows you to use plastic insulated wire, greatly reducing the risk of shorting out the winding! Usually, just winding the wire up against the ferrite (even if it has the slightest edge) will damage the insulation enough to make it useless. You have to coat the entire core with something plastic, or just use plastic insulated wire.
Hopefully you will get it working and not be led to think that my circuit is a phoney
.
Steve Ward, Sat Jun 03 2006, 01:22AM
This is another reason i switched to using 2 cascaded CTs so that each would have only a few 10's of turns. This allows you to use plastic insulated wire, greatly reducing the risk of shorting out the winding! Usually, just winding the wire up against the ferrite (even if it has the slightest edge) will damage the insulation enough to make it useless. You have to coat the entire core with something plastic, or just use plastic insulated wire.
Hopefully you will get it working and not be led to think that my circuit is a phoney
.Re: DRSSTC - cannot get oscillating
Marko, Sat Jun 03 2006, 11:11PM
Yep, it works now.
I rewound CT's with PVC wire and hooked them up, but now I was able to get only about 5mm spark, giving my best with tuning.
Frequency match is almost perfect (says osciloscope) but for some reason the coil doesn't wnat to draw much power and secondary rings down almost imediately after beginning of the cycle.
Maybe it is time to put some more volts to the bridge and see if it blows up.
Marko, Sat Jun 03 2006, 11:11PM
Yep, it works now.
I rewound CT's with PVC wire and hooked them up, but now I was able to get only about 5mm spark, giving my best with tuning.
Frequency match is almost perfect (says osciloscope) but for some reason the coil doesn't wnat to draw much power and secondary rings down almost imediately after beginning of the cycle.
Maybe it is time to put some more volts to the bridge and see if it blows up.
Re: DRSSTC - cannot get oscillating
Dr. Drone, Sun Jun 04 2006, 01:32AM
Dr. Drone, Sun Jun 04 2006, 01:32AM
Re: DRSSTC - cannot get oscillating
Steve Ward, Sun Jun 04 2006, 07:32AM
Only 5mm sparks at 30VDC input? Can we have the full specs of the setup? Operating frequency, number of cycles per burst, tank L and C values (approximate is fine, just to get an idea). IF you are running only say 5 cycles of RF, then you will only get puny sparks at low input voltages. If you can, scope the output voltage of the half-bridge and compare it with the current output, make sure its reasonably close to ZCS. If that looks ok, then you are probably alright at higher voltages. A variac is still really nice to have... 30V to 320V is quite a jump!
Steve Ward, Sun Jun 04 2006, 07:32AM
Only 5mm sparks at 30VDC input? Can we have the full specs of the setup? Operating frequency, number of cycles per burst, tank L and C values (approximate is fine, just to get an idea). IF you are running only say 5 cycles of RF, then you will only get puny sparks at low input voltages. If you can, scope the output voltage of the half-bridge and compare it with the current output, make sure its reasonably close to ZCS. If that looks ok, then you are probably alright at higher voltages. A variac is still really nice to have... 30V to 320V is quite a jump!
Re: DRSSTC - cannot get oscillating
Marko, Sun Jun 04 2006, 02:25PM
It seems to repair a bit at 100V input to bridge.
Now I have some 7-8cm spark, either using primary feedback or antenna.
Specs of coil are:
Primary cap: 125nF, 3,2kV
Primary inductance: about 10uH
Secondary: 110*450mm, PVC tube, o.2mm wire for inductance of 80mH
Topload is 10pF, self capacitance 8pF, and resonant frequency about 120kHz.
After some 100us of ON time I get almost no more spark length.
Coil draws only ten watts at 100bps, seems quite low!
I tought it will overload my tiny 40VA isolation transformers but for some reason it is not fery power-hungry (even in perfect tune)
I need to mock up an additional CT to measure primary current precisely.
I guessed my pri mary was overcoupled, so I raised secondary about 5cm more.
Interestingly it didn't affect performance or power draw by any mean.
Marko, Sun Jun 04 2006, 02:25PM
It seems to repair a bit at 100V input to bridge.
Now I have some 7-8cm spark, either using primary feedback or antenna.
Specs of coil are:
Primary cap: 125nF, 3,2kV
Primary inductance: about 10uH
Secondary: 110*450mm, PVC tube, o.2mm wire for inductance of 80mH
Topload is 10pF, self capacitance 8pF, and resonant frequency about 120kHz.
After some 100us of ON time I get almost no more spark length.
Coil draws only ten watts at 100bps, seems quite low!
I tought it will overload my tiny 40VA isolation transformers but for some reason it is not fery power-hungry (even in perfect tune)
I need to mock up an additional CT to measure primary current precisely.
I guessed my pri mary was overcoupled, so I raised secondary about 5cm more.
Interestingly it didn't affect performance or power draw by any mean.
Re: DRSSTC - cannot get oscillating
HV Enthusiast, Sun Jun 04 2006, 03:31PM
Are you running it with a 40VA isolation transformer???
If so, thats probably your problem. The wire in these is quite thin at that power level, and you are in essence "choking" your dc bus cap and not allowing it to charge up enough when operating. Get rid of the isolation transformer and try again.
HV Enthusiast, Sun Jun 04 2006, 03:31PM
Are you running it with a 40VA isolation transformer???
If so, thats probably your problem. The wire in these is quite thin at that power level, and you are in essence "choking" your dc bus cap and not allowing it to charge up enough when operating. Get rid of the isolation transformer and try again.
Re: DRSSTC - cannot get oscillating
Dr. Shark, Sun Jun 04 2006, 03:35PM
I fact I am using a one of those 40VA xfmrs on my DRSSTC as well, and I don't see it hurting performance. Of course, if I turn the bps up too high, it gets terribly hot (I think it is supplying more than 100W that way), and at some point the voltage drops, but for low power runs, I think it is perfect. In fact, it has the added benefit of the "series lightbuilb trick", limiting the current in case of a fault.
Dr. Shark, Sun Jun 04 2006, 03:35PM
I fact I am using a one of those 40VA xfmrs on my DRSSTC as well, and I don't see it hurting performance. Of course, if I turn the bps up too high, it gets terribly hot (I think it is supplying more than 100W that way), and at some point the voltage drops, but for low power runs, I think it is perfect. In fact, it has the added benefit of the "series lightbuilb trick", limiting the current in case of a fault.
Re: DRSSTC - cannot get oscillating
Marko, Sun Jun 04 2006, 03:49PM
They are actually two small transformers, one of them wired backwards. Interestingly I get very little voltage drop on it, and coil draws only few watts at low breakrates with 100us ontime.
Annd coil simply doesn't want to draw any more power!
I can't get 'rid' of it, since I wanted lower voltage on bridge for low power test.
EDIT: I measured primary current.
With current 100V input it hits about 40-50A.
Marko, Sun Jun 04 2006, 03:49PM
They are actually two small transformers, one of them wired backwards. Interestingly I get very little voltage drop on it, and coil draws only few watts at low breakrates with 100us ontime.
Annd coil simply doesn't want to draw any more power!
I can't get 'rid' of it, since I wanted lower voltage on bridge for low power test.
EDIT: I measured primary current.
With current 100V input it hits about 40-50A.
Re: DRSSTC - cannot get oscillating
HV Enthusiast, Sun Jun 04 2006, 05:25PM
Primary current seems much too low for 100V on the bridge. You should be on the order of a few hundred amps based on my experience, even more if your primary is tuned lower than natural frequency of secondary.
HV Enthusiast, Sun Jun 04 2006, 05:25PM
Primary current seems much too low for 100V on the bridge. You should be on the order of a few hundred amps based on my experience, even more if your primary is tuned lower than natural frequency of secondary.
Re: DRSSTC - cannot get oscillating
Marko, Sun Jun 04 2006, 07:13PM
Yep, I easily hit over 100 amps when I detune primary to lower frequency.
I'l also try to find some more powerful source of about 100 volts...
Marko, Sun Jun 04 2006, 07:13PM
Yep, I easily hit over 100 amps when I detune primary to lower frequency.
I'l also try to find some more powerful source of about 100 volts...
Print this page