I got it off eBay. Don't remember the seller because the original build was almost 4 years ago.
If I put my ear right up against the speaker, I can hear a very faint hum, so all good. The main power ground is tied to the chassis at the back left mounting bolt of the power transformer. The ground for the bias supply goes to the back right bolt. The signal grounds are just the terminal strips. The ground from the IEC is tied to it's own dedicated bolt. The chassis is aluminum.
drt
As to your chassis wiring... that 330K resistor connecting the blue wire on the IO valve holder? Where is this on the schematic ? I am a little perplexed by the lanky layout that is creating spurious oscillations. Can you elaborate !
The 100K grid stopper on the output stage is a complete overkill.
rJ
As to your chassis wiring... that 330K resistor connecting the blue wire on the IO valve holder? Where is this on the schematic ? I am a little perplexed by the lanky layout that is creating spurious oscillations. Can you elaborate !
The 100K grid stopper on the output stage is a complete overkill.
rJ
The 330k resistor was for plate to plate feedback. I have since removed it and converter the feedback scheme to GNFB. The THD+N at 1 W, 1kHz was 2.15% with the plate to plate feedback. With the GNFB, it's down to 0.3%. The grid stoppers on the 6AV5GA are 1k and not 100k. Absolutely no issues with oscillation on sweeps into the hundred-kilohertz domain. I had a temporary problem with the right channel acting stupid, but I traced it down to the large coupling capacitor in the wrong orientation with the outer foil interacting with the chassis.
Interesting that you brought up this amp. I just had this sucker on the bench this morning. When I designed the cascode front end, the plate load was 30k. As the B+ and the plate voltage of the output tubes were about the same, with a 330k plate to plate feedback resistor paralleled with a 33k load resistor, the plates saw 30k. I left the 33k plate resistors in there after I removed the 330k resistors and didn't change it to 30k. This morning, I temporarily soldered a 330k resistor in parallel with the 33k plate resistor. Made absolutely no difference on the scope or the distortion measurements, so I left the 33k load resistor in there.
Interesting that you brought up this amp. I just had this sucker on the bench this morning. When I designed the cascode front end, the plate load was 30k. As the B+ and the plate voltage of the output tubes were about the same, with a 330k plate to plate feedback resistor paralleled with a 33k load resistor, the plates saw 30k. I left the 33k plate resistors in there after I removed the 330k resistors and didn't change it to 30k. This morning, I temporarily soldered a 330k resistor in parallel with the 33k plate resistor. Made absolutely no difference on the scope or the distortion measurements, so I left the 33k load resistor in there.