Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Sun Mar 06 2011, 04:08AM
OK, here we go...
Purpose: This device will replace my NST and Iggies for corona production, solid and oil insulation chartacterising, and O-scope waveform viewing, and cool pics---possibly burning stuff for fun too.
Implementation: A 555 timer configured for independent high and low driving of a IGBT or MOSFET, switching About 50-60Vdc at 2-3 Amps through the after-production primary.
GOAL: Hopefully 30kV at 200 Watts or there abouts.
Good 555 timer circuit site.
]htg7n60a4d.pdf[/file] IGBT
]stw13nk100z.pdf[/file] MOSFET
555 PWM indepentent high and low. From the above cited web page.
Initial build, aluminum Plate: 7.5" x 7.5" x 0.25"
35kV flyback, crack should not matter as the core is gapped anyway.
I'm considering cutting these pins to half hieght, then epoxing them in. I tried to solder "Ball" that lead such that the field intensity will be reduced due to lack of sharp solder points.
I will start with 10 turns, at 60 volts, im sure it wont take much on-time to cause saturation. Eventually I will enclose this whole flyback in acrylic and an oil tank.
Flyback, Cap bank, LV transformer (60Vdc @ 2A), TO-247, Bridge rect, and 555 PWM--PCB.
Flyback (1 of the 3 I bought from Killa-X)
Re: Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Inducktion, Sun Mar 06 2011, 05:48AM
...why not use a half bridge driver? Or is that what the 555 timer is achieving?
If you use a resonant capacitor the output should be a bit higher, depending on the resonant frequency and the frequency output by the 555 timer. A half bridge version of this should work too. Why not use a GDT or something? o:
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Sun Mar 06 2011, 05:54AM
Inducktion wrote ...
...why not use a half bridge driver? Or is that what the 555 timer is achieving?
If you use a resonant capacitor the output should be a bit higher, depending on the resonant frequency and the frequency output by the 555 timer. A half bridge version of this should work too. Why not use a GDT or something? o:
Trying to keep it simple for first device version, but would a push-pull / half bridge do any good with diodes in series throughout the secondary?
Re: Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Killa-X, Sun Mar 06 2011, 06:53AM
Well, good luck with the project! Let me know if that core starts to cause any issues though. I may plan a 3 series ZVS someday but I will keep a track on your progress and see if I will try it my self for fun!
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
uzzors2k, Sun Mar 06 2011, 10:06AM
Patrick wrote ...
...but would a push-pull / half bridge do any good with diodes in series throughout the secondary?
Not really, you're better off driving these diode split flybacks in flyback mode. Push-pull should be used with full-wave or unrectified flybacks.
Re: Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Sulaiman, Sun Mar 06 2011, 10:26AM
I see two major reliability problems;
200W @ 30 kV = 6.66 mA = overheated secondary winding (3 mA/90W is ok)
driving the lopt primary with a symetrical primary waveform will produce a symetrical secondary waveform, so the stress on the internal diodes is 60 kV total (more with internal 'ringing') = over-volted diodes.
Since you're in the land of 110 Vac (or thereabouts) full-wave rectified mains is perfect for supplying a lopt in conventional flyback mode (as they were designed for)
You will need to isolate (pulse-transformer or opto-isolator) the lopt primary drive circuitry from your low voltage drive circuitry (555 etc.) and connect the 'bottom' end of the secondary to a reliable earth (Protective / mains earth) for safety and ease of experimenting.
With a couple of divide-by-two flip-flops you can drive four lopt with a mark:space ratio of 3:1
(this provides four pulses per 'cycle' for low output voltage ripple)
driven at variable frequency for output power control. This would give 200W RELIABLY.
Don't forget that when there is little load on the output the power that you're driving into the lopts has to go somewhere. I would put a fast >1500V diode from each primary to a 200W voltage clamp, I'd use 20 parallel strings of 3x1.5KE350CP TVS because I've got a lot of them, you would need something equivalent
(e.g. zener diode string to the base of a transistor that can take 200W @ 1.5 kV)
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Sun Mar 06 2011, 02:53PM
Uzzors wrote ...
Patrick wrote ...
...but would a push-pull / half bridge do any good with diodes in series throughout the secondary?
Not really, you're better off driving these diode split flybacks in flyback mode. Push-pull should be used with full-wave or unrectified flybacks.
I will be using the normal single switch flyback mode with TVS's, as I dont want my internal diodes blown out...I see no need to use a symmetric +,- waveform.
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
James, Sun Mar 06 2011, 08:12PM
Don't forget to use a snubber when driving these things hard in flyback mode. There's a good app note here:
Note that quite a lot of power can be dissipated by the snubber. I once used a 60W 120V incandescent lamp in place of the resistor and was able to light it brighter than full brightness from about 40V into the circuit. Measured 76kV out of the SVGA monitor flyback though.
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Inducktion, Sun Mar 06 2011, 10:00PM
Uzzors wrote ...
Patrick wrote ...
...but would a push-pull / half bridge do any good with diodes in series throughout the secondary?
Not really, you're better off driving these diode split flybacks in flyback mode. Push-pull should be used with full-wave or unrectified flybacks.
I meant using a half bridge driver so that you have two mosfets sharing the load equally. It would reduce strain on the fets and allow for more power right?
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Mon Mar 07 2011, 03:13AM
New work
7" tall acrylic oil tank, insulators, and flyback. Also seen, 75 M ohm 20kV resistor.
Does anyone know how well acrylic resists surface carbon tracking? (relative to other plastics: good, fair, poor?)
Re: Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Newton Brawn, Tue Mar 08 2011, 10:43AM
Hi Patric:
Thanks for tjhe 555 site.
Can you shou the complete schematic that you have in mind?
Regards
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Proud Mary, Tue Mar 08 2011, 11:48AM
Patrick wrote ...
Does anyone know how well acrylic resists surface carbon tracking? (relative to other plastics: good, fair, poor?)
Ionic surface contamination (e.g. fingerprints) and ionizable inclusions are the usual suspects in tracking failure, whichever dielectric is used.
But don't take my word for it! Google
comparative tracking index and get your DIY lobotomy kit
absolutely free of charge!
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Tue Mar 08 2011, 08:28PM
Newton Brawn wrote ...
Hi Patric:
Thanks for tjhe 555 site.
Can you shou the complete schematic that you have in mind?
Regards
Yes after my calculus homework is done, i will put up the schematic.
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Thu Mar 10 2011, 04:37AM
Schematic:
OK, everybody check my work, polarity and stuff. I was less than sober while putting this together.
First, on the 555 the pins 4 and 5 are left unconnected. I havent decided what to do with those yet.
Second, the TVS's are hopefully pointed right.
Third, R3 and R4 are pots, the R2 and C3 have not been fully decided yet.
Finnally, the Q1 device would be a common N-channel Mosfet ot IGBT.
So find my errors and make suggestions now before I commit to etch the board
Re: Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Antonio, Thu Mar 10 2011, 01:03PM
D5 cannot be there. It will eliminate the necessary flyback pulse. The usual solution there is a fast diode from the drain to a parallel RC circuit connected to the power supply. A voltage limiter can be put in parallel with the RC circuit.
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Thu Mar 10 2011, 03:15PM
Antonio wrote ...
D5 cannot be there. It will eliminate the necessary flyback pulse. The usual solution there is a fast diode from the drain to a parallel RC circuit connected to the power supply. A voltage limiter can be put in parallel with the RC circuit.
D5 is a TVS invisible until in 600V, are you sure the flyback pulse will exceed 600 V , for a 10 turn primary? or do you mean it's hold off should be bidrictional?
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Antonio, Thu Mar 10 2011, 04:24PM
Patrick wrote ...
Antonio wrote ...
D5 cannot be there. It will eliminate the necessary flyback pulse. The usual solution there is a fast diode from the drain to a parallel RC circuit connected to the power supply. A voltage limiter can be put in parallel with the RC circuit.
D5 is a TVS invisible until in 600V, are you sure the flyback pulse will exceed 600 V , for a 10 turn primary? or do you mean it's hold off should be bidrictional?
The diode in this position will clip the pulse at the drain of the switching transistor at the power supply level, not at 600 V. A bidirectional device will work.
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Thu Mar 10 2011, 05:19PM
OK, yeah i see your point, is this and improvement? 600v held off in either direction, before clipping.
@antonio:
Antonio wrote ...
The usual solution there is a fast diode from the drain to a parallel RC circuit connected to the power supply. A voltage limiter can be put in parallel with the RC circuit.
Can you show me an example? i fail to understand.
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
James, Thu Mar 10 2011, 08:15PM
Patrick wrote ...
OK, yeah i see your point, is this and improvement? 600v held off in either direction, before clipping.
@antonio:
Antonio wrote ...
The usual solution there is a fast diode from the drain to a parallel RC circuit connected to the power supply. A voltage limiter can be put in parallel with the RC circuit.
Can you show me an example? i fail to understand.
Look up RCD snubber. There are lots of application notes out there. As power levels increase though the snubber can dissipate large amounts of energy. I tried once using a 60W incandescent lightbulb as the resistor and could easily light it full brightness with about 40V going into the circuit.
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Antonio, Thu Mar 10 2011, 09:03PM
If the load is not constant and there is no feedback, it's better to use something that recycles the energy in the transformer primary if it is not used in the secondary.
This is a flyback circuit that I made some time ago with an RCD snubber:
The resistor R3 was eating a lot of power. This is a modified version that worked much better, with C3 and D3 returning unused energy to the power supply:
Don't look at the switching transistor driving circuit, that is very inefficient and has insufficient base current capability for that transistor.
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Thu Mar 10 2011, 11:11PM
OK antonio which one?
or this one?
im still thinking about pins four and five? i dont like them left floating.
Re: Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Proud Mary, Fri Mar 11 2011, 12:33AM
Don't you think a buffer - such as a simple NPN/PNP totem pole - between the 555 and the gate, would make switching a bit more snappy? The capacitance of both the gate and your TVS diode have to be charged up afresh on each cycle, so the gate waveform could start to look as mole hills do after a shower of rain, and where would your dv/dt be then?
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Fri Mar 11 2011, 12:50AM
Proud Mary wrote ...
Don't you think a buffer - such as a simple NPN/PNP totem pole - between the 555 and the gate, would make switching a bit more snappy? The capacitance of both the gate and your TVS diode have to be charged up afresh on each cycle, so the gate waveform could start to look as mole hills do after a shower of rain, and where would your dv/dt be then?
Alright I suppose I could tolerate more components in exchange for more V. I assume you say totem pole as that would be your first choice?
should i use To-220's or the To-92's ?
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Antonio, Fri Mar 11 2011, 02:21AM
Pin 5 shall be connected to ground with a 0.1 uF capacitor. Pin 4 goes to Vcc. You can use pin 5 to implement a protection against excessive current. Something like this:
Your second circuit is better. The capacitor shall be chosen so the voltage at the drain of the transistor doesn't get too high when it switches off.
Re: Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Fri Mar 11 2011, 02:28AM
let me run the simulator on a totem pole configuration, ill post more soon.
Re: Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Sat Mar 12 2011, 08:06AM
Im going to switch over to a 556 timer, one 556-1/2 will be setting the on time (monostable), while the other 556-2/2 will be triggering the first 1/2 astable (frequency).
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Proud Mary, Sat Mar 12 2011, 11:58AM
Patrick wrote ...
Im going to switch over to a 556 timer, one 556-1/2 will be setting the on time (monostable), while the other 556-2/2 will be triggering the first 1/2 astable (frequency).
Something like this?
Re: Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Thu Mar 17 2011, 08:18AM
That was exactly the circuit I wanted Proud Mary
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Proud Mary, Thu Mar 17 2011, 05:46PM
Patrick wrote ...
That was exactly the circuit I wanted Proud Mary
Don't forget that CMOS varieties like the TLC556 series can motor along at up to 2 MHz.
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Thu Mar 17 2011, 06:05PM
Proud Mary wrote ...
Don't forget that CMOS varieties like the TLC556 series can motor along at up to 2 MHz.
This particular circuit will top out at 25kHz, but I guess near vertical sides will be helped out at 2Mhz. Then again the transistors will have more to do with rise/fall time then the timer driving them.
When I used the double diode trick for independent timing resistors, I forgot that changing one on/off time also changes the other off/on time. So 556 it is.
I beleive I will be able to adjust the on-time for best voltage/spark length, while the frequency can be increased or decreased as needed for average power.
Re: Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Proud Mary, Thu Mar 17 2011, 11:18PM
Lá Fhéile Pádraig!
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Thu Mar 17 2011, 11:54PM
Proud Mary wrote ...
Lá Fhéile Pádraig!
Dont know what this means.
Re: Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Proud Mary, Fri Mar 18 2011, 12:13AM
St Patrick's Day
- today!
Re: Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Fri Mar 18 2011, 12:20AM
In which languange?
Re: Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Proud Mary, Fri Mar 18 2011, 12:42AM
Gaeilge - i.e. Irish Gaelic
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Fri Mar 18 2011, 06:57AM
new work.
Oil tank, nylon screws on top, and thermistor for ferrite, thermistor for oil.
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Fri Mar 18 2011, 07:18AM
I guess the rest of you wEll be drinking more thn usuall?
(resisting the urge to drink and HV)
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Proud Mary, Fri Mar 18 2011, 12:27PM
Patrick wrote ...
new work.
Oil tank, nylon screws on top, and thermistor for ferrite, thermistor for oil.
Will you be taking steps to reduce the chance of oil leaks or bursting under thermal expansion?
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Fri Mar 18 2011, 06:56PM
Proud Mary wrote ...
Will you be taking steps to reduce the chance of oil leaks or bursting under thermal expansion?
Yes. but havent committed to a design for that yet. Any ideas, anyone ?
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Proud Mary, Fri Mar 18 2011, 07:12PM
Patrick wrote ...
Proud Mary wrote ...
Will you be taking steps to reduce the chance of oil leaks or bursting under thermal expansion?
Yes. but havent committed to a design for that yet. Any ideas, anyone ?
Funny you should ask that!
If you take a short section of 20mm polypropylene corrugated flexible duct, and block off one end of it, you can use it like an aneroid bellows. Fit a 20mm standard duct termination into the top of the case - but upside down going inward - and fit your short terminated section onto this, sealing it all with RTV, or what have you.
As the liquid volume in the case increases, the bellows will be compressed and shorten, and
vice versa, to be equalised with atmospheric pressure.
This system is used in some oil-filled X-ray units, but using one or two metal bellows.
Re: Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Fri Mar 18 2011, 08:39PM
I will try this, but I was going to use gasket sealer, which would squeeze out, like the cheeze whiz in a can stuff.
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Proud Mary, Fri Mar 18 2011, 10:55PM
Patrick wrote ...
I will try this, but I was going to use gasket sealer, which would squeeze out, like the cheeze whiz in a can stuff.
Here in England we have no end of cheap and nasty aneroid barometers - the kind of thing with printed gothic lettering and antiqued wood for which there was once a vogue - so you could take the bellows out of one of those if you wanted a metal concertina. There is also a type of corrugated flexible aluminium trunking which you could experiment with, though I have no experience with it myself.
A different approach might involve a compressible mass of closed cell foam like neoprene , or simply a free space above the oil filled with a stable, incombustible gas.
Re: Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Fri Mar 18 2011, 10:58PM
I will definately leave a dry air bubble.
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Sun Mar 20 2011, 05:22AM
Proud Mary, can you advise me as to what kind of totem pole i should use? maybe one like this.
From all about circuits.com
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Proud Mary, Sun Mar 20 2011, 08:22AM
Patrick wrote ...
Proud Mary, can you advise me as to what kind of totem pole i should use?
I was thinking of something very simple like this, which you could adapt to your needs:
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Sun Mar 20 2011, 06:51PM
Proud Mary wrote ...
Patrick wrote ...
Im going to switch over to a 556 timer, one 556-1/2 will be setting the on time (monostable), while the other 556-2/2 will be triggering the first 1/2 astable (frequency).
Something like this?
Proud Mary the output from this circuit is connected to the 2nd timers reset (pin10), instead of 2nd timers output (pin9). Is this right?, where did you get this pic?
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Proud Mary, Sun Mar 20 2011, 07:12PM
Patrick wrote ...
Proud Mary wrote ...
Patrick wrote ...
Im going to switch over to a 556 timer, one 556-1/2 will be setting the on time (monostable), while the other 556-2/2 will be triggering the first 1/2 astable (frequency).
Something like this?
Proud Mary the output from this circuit is connected to the 2nd timers reset (pin10), instead of 2nd timers output (pin9). Is this right?, where did you get this pic?
Oh dear, what a calamity! I've had the circuit on file for ages, and cut and pasted it from somewhere, but I don't remember where it came from now. I should have verified the circuit before handing it on to you, and I'm sorry for any trouble my error may have caused.
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Sun Mar 20 2011, 09:20PM
Proud Mary wrote ...
Oh dear, what a calamity! I've had the circuit on file for ages, and cut and pasted it from somewhere, but I don't remember where it came from now. I should have verified the circuit before handing it on to you, and I'm sorry for any trouble my error may have caused.
Dont worry about it, you've saved me to many times for me to be mad at. Also I havent been so foolish as to etch an unproven circuit. Besides thats what sims are for! (though I did notice it before putting it into MultiSIM 10)
EDIT: the sim just puked 1's and 0's all over, im trying to figure the circuit out.
Re: Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Mon Mar 21 2011, 06:11AM
Im going to use the ICL8038 function generator IC, the 555/556 are to unstable and difficult.
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Proud Mary, Mon Mar 21 2011, 09:37AM
Patrick wrote ...
Im going to use the ICL8038 function generator IC, the 555/556 are to unstable and difficult.
I've always thought that 8038 looked good fun, so I'll be interested to see how you get on with it. Intersil have it marked as "Obsolete Product. No recommended replacement," so I guess what stocks there are will be the last of them.
Re:
Flyback (collaboration) for Corona, Insulation, Waveform purposes.
Patrick, Mon Mar 21 2011, 06:10PM
Proud Mary wrote ...
Patrick wrote ...
Im going to use the ICL8038 function generator IC, the 555/556 are to unstable and difficult.
I've always thought that 8038 looked good fun, so I'll be interested to see how you get on with it. Intersil have it marked as "Obsolete Product. No recommended replacement," so I guess what stocks there are will be the last of them.
I only have 6 left !!!