It will work with 4227 as that is what I am using.
The VCC supply also charges up the VB supply when vs is at zero volts.
This takes a chunk out of the vcc supply and resets the 2092.
Adding a 47uf to the vcc supply fixed this problem.
I don't understand, first it didn't work with 4227's, and now it does? And you just added 47uf?
Yes. I can only guess there was something wrong with my 12 volt supply.
Adding 47uf instead of the 10uf was enough to get rid of the 12 volt glitches.
What was happening was the the first HO pulse reset the 2092.
This happened because the vb took power from the 12 volts through the diode and 22uf and this caused the 12 volts to drop out a volt or so and so resetting the 2092.
Might just have been my 12 volt supply was poor and wasnt decoupled enough.
Last edited:
I have used the circuit in irs2092 datasheet.
Only difference is I have used maximum deadtime and changed the vcc 10uf to 47uf.
Rgate are 10R.
I use +/- 35 volts supply.
Download the datasheet for the irs2092.
Not the application note or iraudamp5.
I use same freq elements.
I havent tried a higher voltage yet.
Due to the TO220 package I am having trouble getting more than 5 amps out of the pcb tracks, so currently I cant get more than 100 watts without using 2 oz copper. WIth 2 oz copper i can get 400 watts.
you can click icon down below for my email, but I have pdf of 2092 from IR, so I see datasheet
Ok now I got is working at about 324k
elements from 1138 are way of for me and my setup
What was wrong is too low voltage for fet drives, but when I lower the resistance
(from gnd goes via R to zener for 12v supply(lower side)), I got clean sound
Thanks nigel for pushing me to test it again, i'm very happy now
What I might do is use floating supply for lower side, no more resistors like now
elements from 1138 are way of for me and my setup
What was wrong is too low voltage for fet drives, but when I lower the resistance
(from gnd goes via R to zener for 12v supply(lower side)), I got clean sound
Thanks nigel for pushing me to test it again, i'm very happy now
What I might do is use floating supply for lower side, no more resistors like now
Last edited:
Great to hear it is working.
I found mine took a bit of work to get going.
It sounds very good, well worth the extra effort.
Just a shame the datasheet wasnt quite right for it to work first time.
Nice to hear about it.
Its best to use separate winding from trafo of 0-15V and get a regulated 12-15V for the driver supply. Peaks upto 4A are normal in occurrence through boot diode when the output is clipping and high side boot cap needs sufficient current pulse to get charged rapidly. A low esr bulk cap is must for driver supply close to boot diode.
And that is what I will do, don't really know why I didn't do it first time around. Having 470u for low and 47u for hi side, its really all about even getting that voltage up to 12 or more volts
Basic one is almost useless, I mean, have seen better, more detailed from them
I found i could get away with a dropper resistor from zero volts and a 12v regulator. The 2092 takes about 125mA when running.
Sure but then again its about the same work for me to make another secondary and have floating voltage all the time
What if you have high voltage rails near +/-85VDC and then the series resistor will have to drop as much as 75V on each side which results in big heat dissipation through resistor and its not a good thing as per my view.
Still there are many things which can be improved with this chip such as its current limit, converting feedback to post filter and so on.....
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.