Problem Osc. Frequency - IRS2092 and IRFB4227

[nikigada]

Hi I am a long time in the forum but only read from here I learned a lot of things that interest me. Excuse me for bad English but use a translator. I made my first class D amplifier - IRS2092 + IRFB4227, but I have a problem. I can not get to work on oscillatory frequency higher than 190-200Khz. Protection is activated and turn off output. Attached the schematic on which work with an adjustment of the elements. What may be a problem in your opinion? Dead time of transistors is 45 ns and 65ns tried the effect is the same. Output choke is Mikrometal T106-2 - 40 turn 22uH. Gate resistors is - 10 ohms!

Thanks for your help!

[ledsscp7]

Change 4227 for light - IRFB4615 etc.
IRFB4227 big gate charge - it is more than time for opening

[luka]

use max dead time settings, I think it is 105ns
check that your overcurrent settings are "high", since those fets will handle the current just fine.

when does amp turn off, when you want higher then 200k freq.?
do some measurements, tell me what is the voltage between pin 12 and 10

also, does IRS have +/-5.6v at all times on its supply pins?
on my board R18 is not 0R!

nice looking amp, great job!!

[luka]

Originally Posted by ledsscp7

Change 4227 for light - IRFB4615 etc.
IRFB4227 big gate charge - it is more than time for opening

maybe yes, I get 470kHz with no problems with irfp4227, but I have gate buffers, which helps a lot

[nikigada]

when does amp turn off, when you want higher then 200k freq.?

Yes!

do some measurements, tell me what is the voltage between pin 12 and 10

12.1 Volts

also, does IRS have +/-5.6v at all times on its supply pins?

Yes!

IRFB4615 - 150 volts MOSFET, my power suply +-80 Volts

nice looking amp, great job!!

Thanks!

[leadcoma]

I tried the IRS2092 + IRFB4227 combination and added a totem pole gate driver because for my application, the gate drive current from the IRS2092 was too small to quickly turn on the IRFB4227. Gate rise time was long (>500uS). With drivers, rise time could be 100nS with 4.7R gate resistor.

I've read posts indicated some members have used the combination without the gate drivers but for my requirements, they were necessary especially with higher switching frequencies.

[luka]

if you look with scope, then amps shuts off, is it 12.1v right at that moment too?

[nikigada]

if you look with scope, then amps shuts off, is it 12.1v right at that moment too?

Voltage of 12 volts has always there. When stopping output of the amplifier, voltage 12.1 volts does not disappear!

[nigelwright7557]

I have no problems with 4227 and 200KHz.
I did try a faster osc and then the 2092 starts to get warm as if it is struggling a bit so I went back to 200KHz.

My pcb also does drive four 4227 but for that I also have an option for a tc4420 which can be linked in or out depending how many output transistors I use.

[nigelwright7557]

Originally Posted by nikigada

Voltage of 12 volts has always there. When stopping output of the amplifier, voltage 12.1 volts does not disappear!

One problem I had with my pcb was glitches on the 12 volt rail causing clicks on the audio.
I had to decouple the regulator and right next to the 2092 with 10uf.