In the process of re-biasing. Set P1 for 190mV across source resistor R15 on both channels in Amp 2. Oddly, the offending channel is now behaving like the good channel.
One thing I noticed was the "+" speaker binding post on the offending channel was loose. I couldn't determine if it was electrically contacting the chassis so I just tighten up the backing nut. Could the binding post have shorted to chassis and caused the biasing issue and overheating heatsink?
As I'm waiting for the amp to cook, figured I'd ask. Will be back in an hour with measurements.
One thing I noticed was the "+" speaker binding post on the offending channel was loose. I couldn't determine if it was electrically contacting the chassis so I just tighten up the backing nut. Could the binding post have shorted to chassis and caused the biasing issue and overheating heatsink?
As I'm waiting for the amp to cook, figured I'd ask. Will be back in an hour with measurements.
Not sure what the sign convention is. I put the red probe on the top lead of the resistor, black probe on the bottom. For Iq, did V/0.47.
R Ch (prior offender)
R16 -271mV; Q3 Iq 0.577A
R17 -303mV; Q4 Iq 0.645A
R18 -331mV; Q5 Iq 0.704A
R19 294mV; Q6 Iq 0.626A
R20 313mV; Q7 Iq 0.666A
R21 300mV; Q10 Iq 0.638A
L Ch
R16 -358mV; Q3 Iq 0.762A
R17 -335mV; Q4 Iq 0.713A
R18 -245mV; Q5 Iq 0.521A
R19 333mV; Q6 Iq 0.709A
R20 308mV; Q7 Iq 0.655A
R21 300mV; Q10 Iq 0.638A
Temperature of the retaining bolt for all MOSFETs is around 60C. Behaving very differently than last night. Next step is to try it in the system but that may wait until tomorrow night. Sleep time.
R Ch (prior offender)
R16 -271mV; Q3 Iq 0.577A
R17 -303mV; Q4 Iq 0.645A
R18 -331mV; Q5 Iq 0.704A
R19 294mV; Q6 Iq 0.626A
R20 313mV; Q7 Iq 0.666A
R21 300mV; Q10 Iq 0.638A
L Ch
R16 -358mV; Q3 Iq 0.762A
R17 -335mV; Q4 Iq 0.713A
R18 -245mV; Q5 Iq 0.521A
R19 333mV; Q6 Iq 0.709A
R20 308mV; Q7 Iq 0.655A
R21 300mV; Q10 Iq 0.638A
Temperature of the retaining bolt for all MOSFETs is around 60C. Behaving very differently than last night. Next step is to try it in the system but that may wait until tomorrow night. Sleep time.
Ok, I didn't wait. Sounds like the other F4, sweet yet detailed. Currently listening to Brahms' String Sextets performed by the Belcea Quartet on Alpha Classics (96kHz, 24bit). Guessing it was the binding post at this point. Tomorrow, may back the biasing voltage down to get a little margin in the heat dissipation. Will definitely try the two as balanced amps and listen for awhile.
)The offending L ch is still offending. Put it into the system as a balanced monoblock and it is clearly lower volume than R ch. When I turn up the volume on the Benchmark DAC, it starts to distort. Using the monoblock as a stereo amp with SE inputs, sounds fine. Next steps are to figure out a troubleshooting plan.
In regards to the MOSFET Iq's, can you point me towards a How-To on matching MOSFETs?
Hi 6sX7: for basic matching, start here: https://firstwatt.com/pdf/art_matching.pdf
Please have a look at the testing article as well: https://firstwatt.com/pdf/art_mos_test.pdf
Please have a look at the testing article as well: https://firstwatt.com/pdf/art_mos_test.pdf
me: "ok, ok, ok"even Sparrows around are familiar with every word written on firstwatt.com pages
I would swap
- R18 of R-ch and L-ch
- R20 of R-ch and R21 of L- ch
Don't forget to re-bias after the swap.
Cheers,
-fred-