Hello Andy!
It looks like LLC, doesn't it? But why do u need C2, C3 elkos? I d simply split C1 into 2 caps one1 up to supply, one down, each one half value of current C1)... Just an idea...
Did you see that much of efficency / cost gain for this topology using IGBTs over interleaved 2-Switch-forward using CoolMOS to justify the drawback of nearly unregulated output?
What is the intended output power /load impedance for the Amplifier stage supplied by 2 x 120Vdc? I dont think there is any good 300V-MOSFETS for Class -D-stage..
Thats why I am working on a 180V full brigde design (200V devices much better...)
It looks like LLC, doesn't it? But why do u need C2, C3 elkos? I d simply split C1 into 2 caps one1 up to supply, one down, each one half value of current C1)... Just an idea...
Did you see that much of efficency / cost gain for this topology using IGBTs over interleaved 2-Switch-forward using CoolMOS to justify the drawback of nearly unregulated output?
What is the intended output power /load impedance for the Amplifier stage supplied by 2 x 120Vdc? I dont think there is any good 300V-MOSFETS for Class -D-stage..
Thats why I am working on a 180V full brigde design (200V devices much better...)
Just 2 cents
I would not rely on potting for protection. Use micro with memory protection.
I would not use poliester caps in that part of the circuit
I would not use solder to increase traces current capacity - looks and works really low end, or better say N-th tier Chineese.
I would not measure the current through transformer's primary for OVC - control one flowing through IGBT.
What is the primary @ 4kW ~3ph 240/380?
I would not rely on potting for protection. Use micro with memory protection.
I would not use poliester caps in that part of the circuit
I would not use solder to increase traces current capacity - looks and works really low end, or better say N-th tier Chineese.
I would not measure the current through transformer's primary for OVC - control one flowing through IGBT.
What is the primary @ 4kW ~3ph 240/380?
Hi:
C2 and C3 on the example are the main caps from the PFC output, I have a very good efficiency with CCM PFC + quasi-resonant output stage, I am using IGBT+SIC diode combination on the actual third revision, the amp feeded is 2 x 2200w rms at 4 ohm class D full bridge 180v supply, +/-120V was only for initial supply testing with dummy load. OverCurrent protection is now on IGBT legs for fast single cycle bi-directional detection and primary is single phase 180 to 265 volts a.c. can be used for 90 to 120v mains also resizing some values on the PFC stage. Regarding resonant cap I use polypropilene 630v one.
Nevertheless if I would of said that it's better to control the current the way it's originally posted you would of agreed as well?
Same goes for cap type...
BTW no cap with single wire lead can't handle 20 A RMS, particularly of the size shown.
Your latest response is that the cap is the resonant one. Previously your circuit was quasiresonant.
Quasi resonance with series tank require series diodes with the switches in order to limit conduction angle to half cycle.
That's in turn require much higher voltage for the driver and a different driver.
This confusion leads me to believe that you are trying not to disclose your design - why have you published it at the first place?
This power level is easily achievable and readily available.
I do agree though that unregulated resonant circuits gives the ultimate efficiency.
Same goes for cap type...
BTW no cap with single wire lead can't handle 20 A RMS, particularly of the size shown.
Your latest response is that the cap is the resonant one. Previously your circuit was quasiresonant.
Quasi resonance with series tank require series diodes with the switches in order to limit conduction angle to half cycle.
That's in turn require much higher voltage for the driver and a different driver.
This confusion leads me to believe that you are trying not to disclose your design - why have you published it at the first place?
This power level is easily achievable and readily available.
I do agree though that unregulated resonant circuits gives the ultimate efficiency.
Excuse my English. I am Romanian. I have some questions for you:
Why is using fixed frequency synchronization using NE556? As I can see from the picture.
QSC using same technique, into their amplifiers.
Any chance of showing us real voltage drop at 3500W constant load? Because you say your supply drops only 10V per rail, I don’t think this is true. It’s impossible in my opinion.
What about temperature in *C* for the resonant coil in case running this SMPS for 5 minutes @ 4000W? Because I think it will fire itself with that weak clumsy *litz wire used*. + That low current capacitor used in series with the resonant coil. Either I can believe this SMPS will deliver that power, using only 4x3300Uf at the primary side. Or using that clumsy small bridge rectifier you are using........
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.