Hello everybody.
I’m new on this forum as a writer (up to now I read it only).
First of all I apologize about my English, is not well but I hope you will forgive me
.
I try to build +/- 1000 Vdc, +/- 10 mA PSU. In my short adventure with a switched PSU I’ve build several AC/DC converters (flyback), but output voltage finished at 48 Vdc. All PSU I build based on VIPer series from ST.
So first, I took my standard 230Vac / 24Vdc PSU (flyback) and I recalculated secondary transformer windings with control loop. Normally transformer had 100t/18t (primary/secondary) after modification 100t/400t (duty cycle has changed).
Power on … and don’t works as I would like to. Primary over-current protection acts all time. I’ve found secondary has too large leakage inductance. So I tried bigger core with 40t/160t.
Better but I forgot about reflected voltage… is too high (150/45)*1000 ~ 330V. Adding to mains voltage 310V + 330V = 640V on the drain of the VIPer. The snubber unsolders themselves … (transil 1,5KE200).
The question is, am I going the right way?
I try to build +/- 1kV (10mA) PSU in the flyback topology.
Maybe easiest way is to power it from 24V (no over voltage on transistors – no snubber).
Is it right way?
Maybe push-pull will be better to such task?
What is your opinion?
Have you any experiences with such voltages?
Best regards,
CodiJack
I’m new on this forum as a writer (up to now I read it only).
First of all I apologize about my English, is not well but I hope you will forgive me
.I try to build +/- 1000 Vdc, +/- 10 mA PSU. In my short adventure with a switched PSU I’ve build several AC/DC converters (flyback), but output voltage finished at 48 Vdc. All PSU I build based on VIPer series from ST.
So first, I took my standard 230Vac / 24Vdc PSU (flyback) and I recalculated secondary transformer windings with control loop. Normally transformer had 100t/18t (primary/secondary) after modification 100t/400t (duty cycle has changed).
Power on … and don’t works as I would like to. Primary over-current protection acts all time. I’ve found secondary has too large leakage inductance. So I tried bigger core with 40t/160t.
Better but I forgot about reflected voltage… is too high (150/45)*1000 ~ 330V. Adding to mains voltage 310V + 330V = 640V on the drain of the VIPer. The snubber unsolders themselves … (transil 1,5KE200).
The question is, am I going the right way?
I try to build +/- 1kV (10mA) PSU in the flyback topology.
Maybe easiest way is to power it from 24V (no over voltage on transistors – no snubber).
Is it right way?
Maybe push-pull will be better to such task?
What is your opinion?
Have you any experiences with such voltages?
Best regards,
CodiJack
The issue with flyback will be always too high reflected voltages, both on the main switch during flyback phase and on the output diodes during charging phase.
You should either use a 1000V primary switch with proper snubber and >2KV output diodes or try another topology like the two-switch forward converter or a half-bridge.
Also, as an alternative, I would consider a bipolar transistor as a switch in a self-oscillating topology, because some bipolars may whitstand even 1500V during flyback period provided a optimum snubber is employed. On the other hand, high voltage MOSFETs are clumsy devices but 1200V IGBTs may be an alternative.
You should either use a 1000V primary switch with proper snubber and >2KV output diodes or try another topology like the two-switch forward converter or a half-bridge.
Also, as an alternative, I would consider a bipolar transistor as a switch in a self-oscillating topology, because some bipolars may whitstand even 1500V during flyback period provided a optimum snubber is employed. On the other hand, high voltage MOSFETs are clumsy devices but 1200V IGBTs may be an alternative.
Hi,
flyback powered from 12V or 24V is very suitable technology for your requirements. You can get away by using 100 or 200V mosfet. The most difficult part to obtain is secondary HV diode. I once used BYV26G from Philips but is now discontinued. You might look at http://www.voltagemultipliers.com/.
But you can also split each secondary in half and use one diode between both halfs of the secondary and one diode at the end of the winding. This technique reduces secondary winding capacitance. I used it sucessfully in 24V to 25kV capacitor charging converter (about 20 secondary windings and diodes in series). Also be sure to use HV specified resistors for feedback divider. Ordinary metal oxide resistors tend to open in the presence of high E field.
Best regards,
Jaka Racman
flyback powered from 12V or 24V is very suitable technology for your requirements. You can get away by using 100 or 200V mosfet. The most difficult part to obtain is secondary HV diode. I once used BYV26G from Philips but is now discontinued. You might look at http://www.voltagemultipliers.com/.
But you can also split each secondary in half and use one diode between both halfs of the secondary and one diode at the end of the winding. This technique reduces secondary winding capacitance. I used it sucessfully in 24V to 25kV capacitor charging converter (about 20 secondary windings and diodes in series). Also be sure to use HV specified resistors for feedback divider. Ordinary metal oxide resistors tend to open in the presence of high E field.
Best regards,
Jaka Racman
Sch3mat1c said:
Vacuum tubes… nice thing, but what I know about it is produce warm light (specially in the darkness) and about 10 exist in the old grandpa radio.
Yes, I thought about a half-bridge, it will be the "next victim" if I give up with flyback converter….Eva said:
My 10mA of working current sounds like leakage current of 600A modules ...mzzj said:
To Jaka:
Jaka Racman said:
Hi,
I thought the same - to decrease primary voltage to eliminate high reflected voltage (wire cross section will increase – something for something).
To split secondary winding is a good idea (thanks, i'll test it) – also should solve reverse voltage of the diodes (which will be total output voltage divided by number of secondary sections – am I right?)
Would you tell me more details about your transformer (prim./sec. turns, frequency)?
Best regards,
CodiJack.
low voltage supply are much safer for 1KV supply than AC mains.
battery+ black&white TV horizontal output transformer + almost any power transistor + feedback control , this will give +1000V.
for -1000V, you can use another black&white TV horizontal output transformer, wind another isolated primary around the core.
or wind a Xformer your self. diode...FR107 are 1000V fast recovery diode, use several in series, maybe with parallel resistor to equalize.
battery+ black&white TV horizontal output transformer + almost any power transistor + feedback control , this will give +1000V.
for -1000V, you can use another black&white TV horizontal output transformer, wind another isolated primary around the core.
or wind a Xformer your self. diode...FR107 are 1000V fast recovery diode, use several in series, maybe with parallel resistor to equalize.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.