Home brew mosfet power amplifier
Here are three photo's of a mosfet power amplifier I built a few years ago.
The amplifier is of dual monaural design with each channel having it's own power supply. The toroidal mains transformers are 625VA. I incorporated dc offset loudspeaker protection and a power output monitor. Continuous sinusoidal output power is 270W per channel into 8 ohms. The power mosfets are Exicon laterals and the supply rails are +80-0-80v.
Here are three photo's of a mosfet power amplifier I built a few years ago.
The amplifier is of dual monaural design with each channel having it's own power supply. The toroidal mains transformers are 625VA. I incorporated dc offset loudspeaker protection and a power output monitor. Continuous sinusoidal output power is 270W per channel into 8 ohms. The power mosfets are Exicon laterals and the supply rails are +80-0-80v.
Attachments
The caps on the mains input filter panel are class x.
I remember I salvaged the mains filter from a scrap Farnell SMPS.
The driver circuit is based upon the old Maplin 1KW mosfet power amp design. However, I have changed components to improve deep bass response and lower the gain of the amplifier to 22.
After many years of reliable service I recently had a problem with the left channel. Although the sound quality was fine I noticed that the heat sink was cooler than the the heat sink on the right channel.
After investigation I found one of the ECF10N20 n-channel mosfets had failed and wasn't conducting at all. I removed the mosfet and found it showed 160 ohms resistance between gate and source. I have absolutely no idea why the mosfet failed.
I remember I salvaged the mains filter from a scrap Farnell SMPS.
The driver circuit is based upon the old Maplin 1KW mosfet power amp design. However, I have changed components to improve deep bass response and lower the gain of the amplifier to 22.
After many years of reliable service I recently had a problem with the left channel. Although the sound quality was fine I noticed that the heat sink was cooler than the the heat sink on the right channel.
After investigation I found one of the ECF10N20 n-channel mosfets had failed and wasn't conducting at all. I removed the mosfet and found it showed 160 ohms resistance between gate and source. I have absolutely no idea why the mosfet failed.
Hi Mooly and thanks for your reply.
Inspecting the mosfet that had failed I notice that it has a dark area on it's cap which I cannot remove using alcohol. Seems like the metal has changed colour due to over heating.
When I removed the faulty mosfet I noticed that the nuts and bolts that hold it down to the heat sink bracket weren't as tight as they should be.
I think that's because the lock washers under the nuts have compressed over the years.
Also the nuts holding down the other three mosfets needed tightening.
After replacing the failed mosfet I checked the voltages across each source ballast resistor and found the mosfets weren't sharing the quiescent current very equally. Normally each mosfet should pass about 100mA quiescent current but I found one passing 45mA and another passing 64mA. On the other channel all the mosfets pass the same current plus or minus a few mA.
I have therefore decided to replace all the mosfets on the left channel which will hopefully restore quiescent current balance. I thought that lateral power mosfets were immune to overheating. I've ordered the mosfets from Profusion and I will install them at the end of next week. I will report back my findings.
