'Been over this ground already (#37).
The Zobel network is supposed to be first so that it does its intended job of snubbing HF noise and oscillation at the source and limiting supersonic bandwidth for stability. That has nothing to do with reducing specified audio bandwidth, such as you suspect is occurring here.
The coil is ideally placed as far as possible from switching currents of the output stage and PSU to avoid passing that interference back into the amplifier.
Plenty of amplifiers show poor design practice by having components placed wherever it suits the designer, according to how much they know or care about the issues. As far as the distances involved on the OP PCB are of concern, it probably doesn't make much difference in their present close proximity. The point of interest for your concerns is that it wont affect any simple audio low pass filter effect here, if that exists.
The Zobel network is supposed to be first so that it does its intended job of snubbing HF noise and oscillation at the source and limiting supersonic bandwidth for stability. That has nothing to do with reducing specified audio bandwidth, such as you suspect is occurring here.
The coil is ideally placed as far as possible from switching currents of the output stage and PSU to avoid passing that interference back into the amplifier.
Plenty of amplifiers show poor design practice by having components placed wherever it suits the designer, according to how much they know or care about the issues. As far as the distances involved on the OP PCB are of concern, it probably doesn't make much difference in their present close proximity. The point of interest for your concerns is that it wont affect any simple audio low pass filter effect here, if that exists.
Too big millercaps could make such problems too.
But this is not anything consistent to this error.
You should thace the amp with a scope, to see where the rolloff is found.
Yes it takes time, but You should be good if choosing 1Khz as reference and an as high as possible frequency to act as Your marker. Two measurements at each step of testing.
But this is not anything consistent to this error.
You should thace the amp with a scope, to see where the rolloff is found.
Yes it takes time, but You should be good if choosing 1Khz as reference and an as high as possible frequency to act as Your marker. Two measurements at each step of testing.
on the simulation its showing very good rolloff after 100khz forthe same above schematic...
it seems something is wrong with my tweeter crossover...I have a question today I tried to push the bias into very little amount of class A like 4 watts of class A
at the Emitter resistor of .22ohms 3w the measured voltage is 80mv and it made up the current of 400ma. I have checked out the bias was increasing after 80mv but it was relatively stable at 50mv..
after sometime the voltage across that resistor increased and in about 4 mins it reached to 180mv and at the time I felt that its too high and turn off the psu some thing catch fire i just pulled off the plug from the socket.
i have checked out the output transistor didnt burn but the resistors of R7 R12 and R10 have burnt. I just thought to replace them with higher power but how many watts they have to be if im using the amp for 30w in class A I do agree the voltage is 53v which is high but never expected the resistors would burn..
the emitter resistor is .22ohm 3w so according to the current into it which is 1amp lets take at max so its .200mv (max ) x 1 amp which is still .2W so the emitter resistor is safe..
but what about those other resistors..
the current what I measured across the resistor is 1A but is it peak?
at the Emitter resistor of .22ohms 3w the measured voltage is 80mv and it made up the current of 400ma. I have checked out the bias was increasing after 80mv but it was relatively stable at 50mv..
after sometime the voltage across that resistor increased and in about 4 mins it reached to 180mv and at the time I felt that its too high and turn off the psu some thing catch fire i just pulled off the plug from the socket.
i have checked out the output transistor didnt burn but the resistors of R7 R12 and R10 have burnt. I just thought to replace them with higher power but how many watts they have to be if im using the amp for 30w in class A I do agree the voltage is 53v which is high but never expected the resistors would burn..
the emitter resistor is .22ohm 3w so according to the current into it which is 1amp lets take at max so its .200mv (max ) x 1 amp which is still .2W so the emitter resistor is safe..
but what about those other resistors..
the current what I measured across the resistor is 1A but is it peak?
If those resistors are fried you must have some shorted transistors. You CANNOT turn a Class AB amp into a Class A amp by just increasing the idle current, you saw the results. This is one reason you don't see any cheap Class A amplifiers with huge output, if it was that easy everyone would be doing it.
Craig
Craig
In an AB amp the bias is "usually" set just enough to eliminate crossover spike as seen on an o'scope. But you usually never see that in the service manuals, usually a number is given, so many mv across a resistor or ma across an open fuse holder, whatever. These numbers will get you past the crossover spike but if you were to use the proper test equipment and set the idle current to just to eliminate the spike you will find the number is usually lower than specified.
Craig
Craig
i did bias it at 500ma and found the heatsink hot the heatsink is an igbt one which measures 150mm length x 126mm width and 126mm height and weighs about 5kg...
but still it gets hot... I thought about 50w class A I can just imagine just with 500ma bias 4w how much hot it becomes...
so I measured across the emitter resistor .22 ohms so is this peak current or continuous?
but still it gets hot... I thought about 50w class A I can just imagine just with 500ma bias 4w how much hot it becomes...
so I measured across the emitter resistor .22 ohms so is this peak current or continuous?
but altogether i enjoying the 4w class A at that time now just biased the other channel to 250ma and still able to get 2 watts of classA at lower volume the treble is very transparent and mids are very involving... i dont know how much precisely the wattage at the output but the heatsink is almost hot...
i am working on LME49830 with about 4 pairs of IRFP240 and irfp9240 which might give better current sharing altogether but I have underestimated the heat problems and thermal runaway with higher bias..
i am working on LME49830 with about 4 pairs of IRFP240 and irfp9240 which might give better current sharing altogether but I have underestimated the heat problems and thermal runaway with higher bias..
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.