I have one interesting problem with all my prototypes of classD, they donot clip cleanely, like classAB, instead they clip like this (one photo finded on this forum).
I have tryied many feedback schematics, including self-osc or clocked designs, comparators LM311-LM319-AD8561, dead time with CD4070-74HC02-74HC14, drivers IR2110-6N137 and TC429, the same results. Even a Behringer PMH5000 distorts like this, unmoddified.
I have developed one clipper schematic, sensing supply voltages and preventing this clipping, but i think this is a band aid.
I want to know if this clipping is a normal thing in a full range classD, without any tricks (clipper, limiter).
Even my sims (Philips Ucd, discrete design) indicates problems with recovery from clipping.
I have tryied many feedback schematics, including self-osc or clocked designs, comparators LM311-LM319-AD8561, dead time with CD4070-74HC02-74HC14, drivers IR2110-6N137 and TC429, the same results. Even a Behringer PMH5000 distorts like this, unmoddified.
I have developed one clipper schematic, sensing supply voltages and preventing this clipping, but i think this is a band aid.
I want to know if this clipping is a normal thing in a full range classD, without any tricks (clipper, limiter).
Even my sims (Philips Ucd, discrete design) indicates problems with recovery from clipping.
Attachments
Last edited:
They clip on both positive and negative, meaninig this is not a problem regarding bootstrap circuit. Offcourse, i have tryiet with floating small smps, the same result.
Photos take it from Ucd based feedback, 370khz at idle, AD8561, 74HC02, IR2110, IRFB4332, +/-100v, T130-2, 40uH and 2uF.
A self-oscillating design will always stop oscillation when driven into clipping, and the reduction in switching frequency just before (and after) clipping is normal. The reasons for this are fairly well documented on the web. A well laid-out UCD design will exhibit a somewhat cleaner entry and exit of the clipping region than is shown in the photos, but what you show is fairly normal.
Don't forget that the bootstrap capacitor will be slowly discharging when the circuit is not switching, so make sure it has eenough capacitance.
Don't forget that the bootstrap capacitor will be slowly discharging when the circuit is not switching, so make sure it has eenough capacitance.
I have had problems with clipping on the IRS2092.
Driving it hard is fine but if I hard clip it then the chip will sometimes gets upset and resets. IR just said dont overdrive it into hard clipping, useful reply lol They said if I apply more than 10 v pk to pk then it will upset the input and cause the chip to reset.
Driving it hard is fine but if I hard clip it then the chip will sometimes gets upset and resets. IR just said dont overdrive it into hard clipping, useful reply lol They said if I apply more than 10 v pk to pk then it will upset the input and cause the chip to reset.
- Status
- Not open for further replies.