Yet another version of IR audio amp ??

[Gorilla]

Would you possibly want to add C46 to your board to improve the turn on of your high side MOSFET?

Trying switch on your MOSFET with charge from the 10uF/C39 (I guess electrolytic or tantalum with a larger than wanted impedance at higher frequencies) I wouldn't think would be very good.


Edit:
Also I always though it to be good practice to run your power traces via your decoupling caps (C62, C64 in this case) to help isolate your DC buses from the high freqency current loops created.

[fredos]

Hello Zeljko Kaiser



As per information on IR web site, they suppose to work with TTL to, but in recommended operation condition, they claim that Vin lo is near 6 volt in worst case and Vin hi is 9.5V in worst case. If you dont trim your 2110, I think you could have some bizard result as you have. But anyways, the notch you have at every 2 cycle seem to come from bad shape of your triangle wave, maybe some **** over it. Try to add some 0.1uf across the power supply of your opamp and speaker ground, directly at the terminal of your amp, or best, directly to the GND of the power outlet. It seem more a probleme of decoupling, but the fact that you use your 2110 near is limit could not help you too. Another place to add capacitor, is between the 15V you feed to the 2110 and chasis ground too. I built class-d amplifier for a while and this is some point I have discover that will kill noise (and distortion!). I drive my IR2110 with a CMOS 4001 in toggle with dead time, and I dont have any probleme. By the way, just to let you know that I hate auto-osciating amplifier...Think about crystal controled, you should be able to have more control over your parameter! Sorry for my english!

Have a nice day!

Fredos

www-d-amp.com

[phase_accurate]

If you want to drive a 2110 from low voltage logic then you'd have to reduce the supply voltage for the input logic accordingly ! Input logic levels are proportional to the supply voltage.

Regards

Charles

[zkaiser]

I did some improvement with 0.1 caps and some pieces of thick wire like you might see in this pic.
Now I am able to achieve 400kHz but I have around 40 deg. celsius on fets and 45 on ir2110 so I don't know what to do or think but one thing is for sure.

I would definetely gain some signal quality if I make some metal cage (or case) for oscilator and ir2110 to make them isolated from all outer side effects, since there are 80% chances that this all is due to pcb design.

And I added some clampig diodes but they do not improve anything.

[zkaiser]

these are voltages and current drawns of 1 channel of amp since I've disabled one

[zkaiser]

and this is what It has to done due to too much dissipation(s)
!!!!

[zkaiser]

is it ok to put some deadtime passive control like in this pic, cos. in this way turn on off time on fets can be short since I am inserting deadtime before ir2110 and he has some schmit trigger on all inputs, so this must not be the problem ?
I also supply logic side of 2110 with same 5V as I supply 7404 (little smd) and power side is supplied with 15V.

So my thinking is, if I get managed to insert dead time befor level/power shifting of 2110 then I may use lower vaule for gate resistor (like 10ohm or so) and make fets open/close faster to get lower dissipation.
But one thing confuse me, why I have heating on ir2110, do I perhaps have some power interferencing that makes him heat?

regards
Zeljko Kaiser

[fredos]

Hello zkaiser

Your approach seem good, but I suggest you higger value for your dead time cap and lower value of resistor to avoid parasistic pick up at the input of your 2110....Dissipation of the 2110 is directly proportional the the frequency and current feed to the mosfet...So try to reduce frequency to 250-300 Khz, that was enought by the way for audio if you take feedback after the output coil, and irf640 should be drived with 15-22 ohms gate resistor....By experience at 300Khz!

Have a nice day!

Fred

www.d-amp.com

[fredos]

I forgot....Supply 5V to the VDD and dont forget 15V to the VCC for the mosfet...5V will not completly switch on mosfet!

Bye!

Fred

www.d-amp.com

[fredos]

Hello Zeljko Kaiser

I just finish my final amp with IR2110. It's post coil feedback, Class BD at +/-145V, 250Khz, with APT5010LLVR 500V 47A mosfet! That kick! THD is 0.5% at full power, 1500W 4 ohms, 2000W 2 ohms (protection trigger some time at 2 ohms...Have to work a bit more on this!), 35 to 22 000Hz. SMPS with 40A 600V IGBT at 62.5Khz, resonant mode with PLL on control side of SG3525, regulation by syncronous rectifier (phase modulation). 96% effiency from line to output, no heat sink on IGBT, just for syncronous mosfet for the rectifier. Look at this!

By the way, if anyone close to Montreal Canada have some time, I'm looking to hire someone to drive the certification process for this new serie, 6-8 week job, good pay as usual! I'm really busy and anyone in house have time to make this....

Thank's

Fred

www.d-amp.com