I finished up the last of the actual electronic work/mods I wanted to do on the 340B prior to aligning the receiver, a final cleaning and bringing it up for a longer-term listening test.
I removed the common 3.3 ohm cathode resistors from the 7591A output tubes and installed matched 10 ohm 1/4-watt carbon film resistors on each cathode. This gives the ability to use the DC balance and bias controls now to tune each tube for proper cathode current. Unfortunately, it also reduces the utility of the nice, above-chassis test points for setting cathode current. You have to separate the cathodes for each pair in order to install individual resistors, so you lose the ability to use the test points to set the bias a pair at a time. I left them hooked up for a "quick check" feature, but they only are connected to V5 (L channel) and V105 (R channel) now.
When I put the new resistors in, I was still getting >10 ohms from each cathode to chassis, which was puzzling. Recall above that I said the original 3.3 ohm resistors had gone up in value--not so! The problem turned out to be a bad riveted chassis ground point. I traced the ground for the cathode resistors all around the chassis and finally found the chassis ground point off the common can of the C401-C404 multi. Everything was good up to the point where the can was grounded to a tie-point riveted connector, with the connection showing several ohms of resistance to the chassis! I drilled the rivet out and put a screw, nut and star washer on it and now the ground was good and all the resistors read 10 ohms to ground now. Before correcting, I would have thought I had 10 ohms for setting bias, but would have had 12-15 ohms and the bias setting would have been artificially low (under-biased) because of the higher (and changing?) resistance.
I could have just terminated the cathode resistors to a point nearer the output tubes, but I figured the Scott engineers had done it that way for a reason and I didn't want to introduce any weird ground loops. I wanted to straighten out the ground problem as well. I had read that sometimes these riveted chassis points can go bad, but this was the first I have found myself.
Then, I installed 100 ohm, 1/2-watt screen protection resistors on each output tube off a 14 gauge wire buss (picture). I put a tie strip on each end of the buss--one end off a screw holding the 8K 20W resistor to the chassis and the other fastened to a strip attached under one of the PT nuts. Took great care to make sure this buss didn't touch anything else, as it carries 430VDC, and left the insulation on the wire except at the solder points. I replaced the 330 ohm 2W carbon comp R404 with a 3 watt Vishay/Dale 500V metal film unit to bridge to the buss.
While I was in, I replaced the chassis isolation "death cap" with an XY rated unit and a new 820K resistor. Kind of tucked in back there, but got it in OK.
Powered up without the output tubes after triple checking everything and all was good voltage-wise. Put the tubes in and biased each to 350 mV to give 35 mA quiescent cathode current per tube, per the manual.
Sound check was good, but I can tell the tuner needs alignment. Even with the pointer aligned with the stops, stations are about 0.8 mHz higher than the pointer in frequency and I'm getting a couple of stations at more than one spot at the bottom of the dial, so alignment is next!
Pictures are : 1 and 2, cathode and screen resistors; 3 bad chassis ground point repaired; 4 new chassis safety cap/resistor.
Dave
I removed the common 3.3 ohm cathode resistors from the 7591A output tubes and installed matched 10 ohm 1/4-watt carbon film resistors on each cathode. This gives the ability to use the DC balance and bias controls now to tune each tube for proper cathode current. Unfortunately, it also reduces the utility of the nice, above-chassis test points for setting cathode current. You have to separate the cathodes for each pair in order to install individual resistors, so you lose the ability to use the test points to set the bias a pair at a time. I left them hooked up for a "quick check" feature, but they only are connected to V5 (L channel) and V105 (R channel) now.
When I put the new resistors in, I was still getting >10 ohms from each cathode to chassis, which was puzzling. Recall above that I said the original 3.3 ohm resistors had gone up in value--not so! The problem turned out to be a bad riveted chassis ground point. I traced the ground for the cathode resistors all around the chassis and finally found the chassis ground point off the common can of the C401-C404 multi. Everything was good up to the point where the can was grounded to a tie-point riveted connector, with the connection showing several ohms of resistance to the chassis! I drilled the rivet out and put a screw, nut and star washer on it and now the ground was good and all the resistors read 10 ohms to ground now. Before correcting, I would have thought I had 10 ohms for setting bias, but would have had 12-15 ohms and the bias setting would have been artificially low (under-biased) because of the higher (and changing?) resistance.
I could have just terminated the cathode resistors to a point nearer the output tubes, but I figured the Scott engineers had done it that way for a reason and I didn't want to introduce any weird ground loops. I wanted to straighten out the ground problem as well. I had read that sometimes these riveted chassis points can go bad, but this was the first I have found myself.
Then, I installed 100 ohm, 1/2-watt screen protection resistors on each output tube off a 14 gauge wire buss (picture). I put a tie strip on each end of the buss--one end off a screw holding the 8K 20W resistor to the chassis and the other fastened to a strip attached under one of the PT nuts. Took great care to make sure this buss didn't touch anything else, as it carries 430VDC, and left the insulation on the wire except at the solder points. I replaced the 330 ohm 2W carbon comp R404 with a 3 watt Vishay/Dale 500V metal film unit to bridge to the buss.
While I was in, I replaced the chassis isolation "death cap" with an XY rated unit and a new 820K resistor. Kind of tucked in back there, but got it in OK.
Powered up without the output tubes after triple checking everything and all was good voltage-wise. Put the tubes in and biased each to 350 mV to give 35 mA quiescent cathode current per tube, per the manual.
Sound check was good, but I can tell the tuner needs alignment. Even with the pointer aligned with the stops, stations are about 0.8 mHz higher than the pointer in frequency and I'm getting a couple of stations at more than one spot at the bottom of the dial, so alignment is next!
Pictures are : 1 and 2, cathode and screen resistors; 3 bad chassis ground point repaired; 4 new chassis safety cap/resistor.
Dave