IR2113 blowing ! help needed...

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Jump to Latest
kASD

kASD

Member
Joined 2004
Charles, I think I have already said them as 2 resistors ,one series with input, second one from the output filter to the inverting input of comparator and one integrating cap parallel with that resistor.....from the filter output to -input of comparator

Also I would also use inverse parallel schottky diodes from -Input of Comparator to the ground to prevent overdrive during clipping...;)
 
A

alexclaire

Member
Joined 2005
hi Fredos,

Still experimenting on my class d amp based on your D1200 front end design...
Changed the fets for IRFP240 (IRFP 4227 before) : no change
placed ground plane under output inductor: no change

at idle output signal is 250kHz sinus 3Vpp with ringing at 12Mhz (20Vpp) every half period, how could I get rid of theses oscillations ?
My filter is 54uH T106-2 core and 470nF MKP
otherwise sound is great but LF output sinus signal is polluted by these ringings...not audible ;-)

still testing the unit at 580WRMS@1kHz (4 ohms load)

thanks.
 

Attachments

  • img_0310.jpg
    img_0310.jpg
    57.4 KB · Views: 1,598
F

fredos

Member
Joined 2005
d-amp.com
If you play too much with gate resistor, you will slow down the rise and fall time of you waveform. If you cant reduce deadtime trought the gate resistor, go at the input of the 2113. A good first attempt for dead time is to measure the current at idle with a load, no signal. Optimum point will be the lowest current. Booth gate may not be the same value due to difference in the 2113 delay!

Fredos

www.d-amp.com
 
kASD

kASD

Member
Joined 2004
fredos said:
Hickup means that power supply work for fews cycle, then shut off, then work for fews cycle, etc. In this way, effiency is greatly improved, and the way I have implement it, keep voltage stable anyways, unless more power is needed. In this case power supply work in normal way.

Fredos
Click to expand...

For how much time its shut off and for how much time its in working mode...
 
F

fredos

Member
Joined 2005
d-amp.com
Alexclaire

Not enought and too much can do the same effect...

Workhorse

On and off time depend of the load...At low level and idle, this could be every 2-3 second, always on at full power!

The trick is to do a regulated PWM power supply without rectifier inductance, with 10% more turn ratio on TR than DC buss needed!
Feedback should be as ''dry'' as possible!

Fredos
 
A

alexclaire

Member
Joined 2005
Amplifier prototype is ready for final module ;-)

just need 2 modules clock synchronised.

Sound is very great ! no more problem at clipping !

144Vpp @1kHz (4 ohms load) !
Mosfet temp rise very low !!

Thanks for all Fredos and long life to D-AMP

don't worry for the critics just go ahead...

Alexis.
 
A

alexclaire

Member
Joined 2005
Fredos,
I can imagine the way you work ;-) , the way great improvements have been done in earlier physics (I'm teacher ;-) ).
Don't pay attention to these people always trying to find out problems, the results are here to proove the efficiency of your design...
Have a nice day !
Alexis
 
A

alexclaire

Member
Joined 2005
Hi luka,
clock generator and pwm modulator are from d amp 1200 (Fredos) and output stage is from Crest LT series schematics available on the net...

output ripple 1,6vpp was normal (+-80V square wave @250kHz filtered with LC with f0=25kHz means -40dB so near (1/100)*80 sinus wave ripple)

30Vpp ringing were generated at each switching, I lowered them by increasing gate resistors to 65 ohms same resistor for each side.this setting was made with adjustable resistor till ringing felt to a minimum that is 3Vpp.
I couldn't get much better I think...

I don't have drawn schematics yet so have a look at fredos's and Crest's ones ;-)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom