Do you have rail-rail oscillation when there is no excessive current draw? = NO
But if I remove the output mosfets and play some sinewave thru the RCA I can see nice square waves at the low side gates as they should be.
Is the circuit before the IR2010 like the one in the diagram I posted? = in general mostly yes., but there is no lm311, lm555 and u16 (74AHC1G04 Inverter SOT-235)
U9 + U16 are actually the same as 74hc02d in my case.
But if I remove the output mosfets and play some sinewave thru the RCA I can see nice square waves at the low side gates as they should be.
Is the circuit before the IR2010 like the one in the diagram I posted? = in general mostly yes., but there is no lm311, lm555 and u16 (74AHC1G04 Inverter SOT-235)
U9 + U16 are actually the same as 74hc02d in my case.
Last edited:
The problem could be deadtime or a damaged/shorted output filter inductor. If you twist the output filter inductor when it's drawing excessive current,, does it make a difference?
To check the deadtime, you can check the input to the 2010. You can drive a high-frequency signal (100k) into the circuit beyond the crossover. Pin 2 of U2 may be a good point. This will be better to show a deadtime problem than a low frequency signal. If you supply high-side B+ to the IC, you can check the output of the IC for deadtime (should be the same as the input but it's something you can easily check.
To check the deadtime, you can check the input to the 2010. You can drive a high-frequency signal (100k) into the circuit beyond the crossover. Pin 2 of U2 may be a good point. This will be better to show a deadtime problem than a low frequency signal. If you supply high-side B+ to the IC, you can check the output of the IC for deadtime (should be the same as the input but it's something you can easily check.
The problem could be deadtime or a damaged or shorted output filter inductor. If you twist the output filter inductor when it's drawing excessive current, does it make a difference? = No.
I don't have anything capable of producing a 100 kHz sinewave. U2—this is my U2 or the U2 in the infinity amp schematic?
Okay, I've just noticed that if I rotate even slightly the LPF filter knob, it goes excessively overcurrent immediately, with no load and no RCA plugged.
My theory is that maybe during the rail filter caps explosion some of the small SMD ceramic capacitors have been damaged (I've already changed 2 of these as the original ones have fallen apart, and since I do not know their capacitance, I just used the 100 nF highest voltage rating I had) which are being used in the preamp and the class D section due to their proximity to the rail caps. But without the service manual, there is no way I could know their voltage ratings and capacitance.
I don't have anything capable of producing a 100 kHz sinewave. U2—this is my U2 or the U2 in the infinity amp schematic?
Okay, I've just noticed that if I rotate even slightly the LPF filter knob, it goes excessively overcurrent immediately, with no load and no RCA plugged.
My theory is that maybe during the rail filter caps explosion some of the small SMD ceramic capacitors have been damaged (I've already changed 2 of these as the original ones have fallen apart, and since I do not know their capacitance, I just used the 100 nF highest voltage rating I had) which are being used in the preamp and the class D section due to their proximity to the rail caps. But without the service manual, there is no way I could know their voltage ratings and capacitance.
As a side note, you can make a square wave generator from a TL594. Make it a single output (not inverting like that used in amp power supplies) by grounding pin 13.
I've changed a couple of smd caps, taking as a reference the infinity schematic and looking at their values. Pulled the mosfets out. At the low side gate I can see square waves at 68Khz. I suppose it's self-escalating design. But once I fit the mosfets in...it's not stable at idle anymore. Draws a lot of current and the mosfets are heating up really fast. Is there any information in Your guide Perry how is the deadtime set or it is more like design/driver ic related ?
If you feed the high-side a B+ supply and look at both the high and low sides do you see any overlap at the output gate pads (no FETs installed) at 68k?
I may sound dumb but how should I do that ? I know from the guide the HIP4080 bridging the two pads....but how is it done with this class D design ?