UCD 25 watts to 1200 watts using 2 mosfets

Yep! Thank you.

I do see also that the sic mosfets who needs negative bias voltage even as this nias is not needed in a D amp there is the problem that the lower mosfet gets 2 x negative voltage less output voltage who do make dissipation higher, this is 6 to 8 volts, because the driver can not give negative voltage above rail, then we need to set it above it, maybe I do see it not right but I think it is. There are sic Jfets who do not need negative voltage but has speed, so we need this then, I go for normal mosfet now, and in time look at the newer ones, but first I go do the circlotron.

for low distortions in a d amplifier the speed is very important, and there are coming mosfets who can switch in as low as 22 nS, a high speed comparator is also needed, or a fast discrete power comparator as these.

The input opamps low pass filters has poles, these will be difficult to get right for very low distortion as I did read, maybe adjustable capacitors do help.

Ohh yes, do you now why the lower mosfet is nog complete switched on with the IR 2011 I think something to do with model from @analogman? I have not yet answer for that.

Also the square signal is not nice, you see to switch to rail in plus side but not on the lwer side, means mosfet is not complete switched on.

thanks


regards
 

Attachments

  • ScreenHunter_314 Jun. 30 17.01.jpg
    ScreenHunter_314 Jun. 30 17.01.jpg
    128.7 KB · Views: 1,668
Last edited:
hi all still continuing with the 2step and 3step h pcb ab/h

the class d 200wt using n and p fets is very stable reliable and silent the only downfall is the voltage has to be up to +/-42vlts max which is abt 200wts in 4ohms .:( i wanted to add it to class h like manoj style with tlo72 2step and 3step please advice


1. i have mounted multiple d amps of 200wd in one casing without issues of noise and cross conduction.
2. less heat than ab amp
3. can handle lower impedance stably
4. d amp is price friendly

diysmps

.also i have the below smd package opamps instead of tl072.
will they work in the step stage with this opamp ic 4550, cct8002 , 4558 :confused::confused:
 

Attachments

  • My class D amp - Page 63 - diyAudio 1661_n.jpg
    My class D amp - Page 63 - diyAudio 1661_n.jpg
    102.8 KB · Views: 1,247
  • My class D amp - Page 63 - diyAudio Class D with Class H_001.jpg
    My class D amp - Page 63 - diyAudio Class D with Class H_001.jpg
    853.2 KB · Views: 1,325
  • My class D amp - Page 63 - diyAudio Class D with Class H.pdf
    30.4 KB · Views: 466
Disabled Account
Joined 2005
Hi Stewin,
Just asking, why do you need stepper for class D.
Stepper is usually used to avoid linear drop losses because of Vce of Vds voltage. In class D, Vds is almost zero.

From my point of view, I don't see any benefit. You can find easily 150V, 200V, 250V, 300V mosfet. If not enough power still you can go with easier way: fullbridge or BTL.
 
Hello. I wanted your help to understand. I did the test of my class amp d. I measure the temperature of the heatsink. Psu + -70v output amplifier loaded by resistive dummy load. 8ohm 56 ° after one hour. 4 ohm 85 ° after half an hour and protection cuts. The test is done with hard rock music at clipping. is it normal. Now I do not have the speaker that support this power to test in real load. Do the speaker require less power. Help me please to understand if this temperature is normal. With 6 ohm after one hour 65 °
 
Did you guys tri sic mosfets? these are very fast, like 14-24 49 nS for 90 amps version, but needs a special driver as I did read, needs -2 to -5 bias voltage and 20 volts gate drive, these are as such complicated?

But in sim I get pretty low distortions, even on 8 amps 4 ohms.

with 1.4 amps output it is very very low, for simplicity there is not yet a proper dead time.

The feed forward network on output is quite nice to recover fase behavior for stable oscillation and frequenty setup together with the low pass opamp in input, I did read this stuff and have now sim it, I go try also with the IR 2011S chip, that way I can make better pcb with lower parasitics..

regards

What kind of "UCD" type feedback is this. Can you post some link related? Resembles one old Oersted univerity file. Still have you every tried such feedback in real life. What about stability, etc ? (you put square input, what you get at output, etc). Even author of original file from Oersted did not try this in real life.. As we see 12 pF caps, etc.

Still this may be superior comapred to "UCD+" from russian forums, if THD is constant, and not dependent on frequency. Still I wonder how this behaves when we change speaker impedance 2 ohm 4 ohm 8 ohm, can it work without output filter change. Classic UCD and UCD+ according to spice, can.

but I am really afraid if we put here square input that output may not be stable. (will not resemble as square also). you can see my previous posts...(similar feedback) but not that advanced.

Output stage like output stage, but am an interesting in feedback,. good feedback can make crappy output stage very low hd... (Lety say big dead time)
 
Last edited:
What kind of "UCD" type feedback is this. Can you post some link related? Resembles one old Oersted univerity file. Still have you every tried such feedback in real life. What about stability, etc ? (you put square input, what you get at output, etc). Even author of original file from Oersted did not try this in real life.. As we see 12 pF caps, etc.

Still this may be superior comapred to "UCD+" from russian forums, if THD is constant, and not dependent on frequency. Still I wonder how this behaves when we change speaker impedance 2 ohm 4 ohm 8 ohm, can it work without output filter change. Classic UCD and UCD+ according to spice, can.

but I am really afraid if we put here square input that output may not be stable. (will not resemble as square also). you can see my previous posts...(similar feedback) but not that advanced.

Output stage like output stage, but am an interesting in feedback,. good feedback can make crappy output stage very low hd... (Lety say big dead time)

I just tried this feedback in general D class setup (generic) ideal switches and fixed deadtime..

Seems not stable,

square wave input, graph shows pwm section and output
 

Attachments

  • Untitled.png
    Untitled.png
    69.3 KB · Views: 568