Hi.
I think about a subwoofer Class D design for large car hifi woofer in halfbridge design without an army of countless single mosfets.
There are industry igbt bricks, but they are slow compared to Mosfets.
I think the lower switching frequency cause larger filter output coils.
max output frequency should be around 200HZ.
I think about following IGBT module (Double IGBT H Bridge) IXYS 400-12E4
Datasheet here http://datasheet.octopart.com/MII400-12E4-IXYS-datasheet-8358599.pdf
tdON: 170ns
tr:60ns
Tdoff: 680ns
Tf: 50ns
Could this IGBT be useful for a design like this ?
I read about 10 times higher switching frequency then Output frequency.
Using 100 times more would be 20khz.
Regards
I think about a subwoofer Class D design for large car hifi woofer in halfbridge design without an army of countless single mosfets.
There are industry igbt bricks, but they are slow compared to Mosfets.
I think the lower switching frequency cause larger filter output coils.
max output frequency should be around 200HZ.
I think about following IGBT module (Double IGBT H Bridge) IXYS 400-12E4
Datasheet here http://datasheet.octopart.com/MII400-12E4-IXYS-datasheet-8358599.pdf
tdON: 170ns
tr:60ns
Tdoff: 680ns
Tf: 50ns
Could this IGBT be useful for a design like this ?
I read about 10 times higher switching frequency then Output frequency.
Using 100 times more would be 20khz.
Regards
I have a defective "Zenon" Ground Zero GZPA 1.0000SPL-HC. A group of output mosfets explode while idle load some seconds after starting the engine.
Its a 8-10kW 2 Channel board.
I have a new irs driver board and new IRFP460LC Fets for repair try.
Its about 70 fets total to remove, to take the whole pcb out. Very bad and boring work.
I now have the idea to make a CAD electric shematic of the SMD driver board.
I have microscope cam, smd desolder tools, tools for measure coils caps resistors and all together. I collected some infos about the Ic's with removed type label values. If all this infos are correct, it must be possible to copy this layout and scale the H bridge parts to the values needed.
Its a 8-10kW 2 Channel board.
I have a new irs driver board and new IRFP460LC Fets for repair try.
Its about 70 fets total to remove, to take the whole pcb out. Very bad and boring work.
I now have the idea to make a CAD electric shematic of the SMD driver board.
I have microscope cam, smd desolder tools, tools for measure coils caps resistors and all together. I collected some infos about the Ic's with removed type label values. If all this infos are correct, it must be possible to copy this layout and scale the H bridge parts to the values needed.
IGBTs aren't efficient until high voltages, as they drop 1.5 to 2.5V all the time, so each amp costs several watts. MOSFETs can be much more efficient particularly below 100V.
With slower devices for the PSU you'd have the issue of needing bigger transformers too.
With slower devices for the PSU you'd have the issue of needing bigger transformers too.
My problem was to find larger mosfet case types. From igbt's i know the different brick. But never discovered same size Mosfets. Meanwhile i see SOT227 package what seem to be ok. But i see there are mosfet modules in igbt brick size like the PHM8001 HTTP 301 This page has been moved this must be enough for a large 9-15V -> 100-150V SMPS. Its not cheap But the pcb May looking much more clear, easier to maintain and repair. But Not sure If the irs 21844s is capable for bigger modules.
IRFP4227 Look pretty good. But i can't believe its a good Idea to run 60A over the the thin copper legs for continious operation. I must take some time to calculate switching losses of different types and frequencys.
The gzpa 1.10000 has 24*IRFP460 (ptot >200W each). Question is how much ptot total is needed.
The gzpa 1.10000 has 24*IRFP460 (ptot >200W each). Question is how much ptot total is needed.
hi nikola , long time. can this fet be used in smps? looks robust and has good current
The thing to watch out for with IGBTs is the 'long tail' current on turn off. Figures for 80% - 20% are much better than how long the last bit off current takes to turn off so at high speed you can get significant power loss through cross conduction. I have had success with IRF WARP2 series in the past up to ~100kHz switching, I'm sure there are even faster parts now available.
Also if you want to do 100kHz switching at high voltage I would go with SiC now (Wolfspeed).
Also if you want to do 100kHz switching at high voltage I would go with SiC now (Wolfspeed).
- Home
- Amplifiers
- Class D
- Using IGBT modules for low freqency Class D ?