Hello,
I recently obtained a pioneer sx-110 (1965 manufacture date). It has 7868 finals. I’m studying the schematics before replacing all the original (bad) paper-oil caps, and they suggest a rather unusual operating point compared to all other 7868 receivers and amplifiers I’ve looked at. Unless I’m reading something wrong (or the schematic is wrong), it looks like it has a really high B+ for the tube, yet a low quiescent current.
I see 495v and 94mA going to all four plates; or 23.5 mA for each tube.
Usually, for other amplifiers using the 7868, I see B+ around 425-475ish and 35-40mA-ish going into each tube.
Other pioneer receivers or amplifiers from the same era with 7868 finals seem in line with just about every other I’ve seen:
SX-2000: 460v, 35.75 mA
SA-810: 425v, 36 mA
Whereas:
SX-110: 495v, 23.5 mA
Here is the relevant part of the schematic. Could I be reading a number wrong? Bad math? Could it be a misprint? If it’s correct, what do you think the designers were going for with high voltage but biasing the tubes rather cold?
Thanks for the help!
I recently obtained a pioneer sx-110 (1965 manufacture date). It has 7868 finals. I’m studying the schematics before replacing all the original (bad) paper-oil caps, and they suggest a rather unusual operating point compared to all other 7868 receivers and amplifiers I’ve looked at. Unless I’m reading something wrong (or the schematic is wrong), it looks like it has a really high B+ for the tube, yet a low quiescent current.
I see 495v and 94mA going to all four plates; or 23.5 mA for each tube.
Usually, for other amplifiers using the 7868, I see B+ around 425-475ish and 35-40mA-ish going into each tube.
Other pioneer receivers or amplifiers from the same era with 7868 finals seem in line with just about every other I’ve seen:
SX-2000: 460v, 35.75 mA
SA-810: 425v, 36 mA
Whereas:
SX-110: 495v, 23.5 mA
Here is the relevant part of the schematic. Could I be reading a number wrong? Bad math? Could it be a misprint? If it’s correct, what do you think the designers were going for with high voltage but biasing the tubes rather cold?
Thanks for the help!
I think it is the data sheet suggestion. Datasheet's 450V is for full power, which in fix bias would likely un-sag to 490 or 500V. Which is well within the 550V limit. Ah, yeah. The box is rated 40Wpc so this is the 44W condition through transformer loss. Not exact same-as datasheet but a practical implementation of the tubemaker's test-bench suggestion.
Interesting, I can buy that! So the other two pioneers also claim “40W”, despite having quiescent B+ 35-70v less. Does this mean the sx-110’s spec might be more in line with reality? (The spec sheet also claims a frequency response of 15Hz-100kHz at rated power. I’ll believe it when I see it though!)
How about the low bias current, then? As it stands, it would only be dissipating 11.6W, compared to 16.5 and 15.3 of the other two. Is there an obvious engineering reason that stands out?
(When I get the caps changed, I’d measure voltages and distortion anyway. Just curious coming into this, because it still seems the values they chose are noticeably different than other 7868-based amps I’ve read about)
How about the low bias current, then? As it stands, it would only be dissipating 11.6W, compared to 16.5 and 15.3 of the other two. Is there an obvious engineering reason that stands out?
(When I get the caps changed, I’d measure voltages and distortion anyway. Just curious coming into this, because it still seems the values they chose are noticeably different than other 7868-based amps I’ve read about)
When replacing coupling caps, don't feel temptated to increase them insanely. Too high cap value may cause too low frequencies appearing in the transformer primary, saturating the core and possibly arcing itself, tubes and sockets.
1. The first rule with coupling caps is to make sure they have a high enough voltage rating.
What happens when the B+ increases a lot, because the output tubes are bad or are not warm, draw no current, the input tubes are not warm, draw no current, and so the very large B+ voltage goes to the cap to the next stage grid leak resistor to ground. Ouch!
A solid state B+ rectifier, or a direct heated tube rectifier both warm up quicker than the (Indirect Heated Cathode) input and output tubes.
2. Why do engineers pick different B+, different plate current, different output transformer primary impedances, etc.?
The reasons are many.
That is just like the differences between 3 Tacoma Narrows bridges (the first one went down into the Puget Sound). Ouch!
What happens when the B+ increases a lot, because the output tubes are bad or are not warm, draw no current, the input tubes are not warm, draw no current, and so the very large B+ voltage goes to the cap to the next stage grid leak resistor to ground. Ouch!
A solid state B+ rectifier, or a direct heated tube rectifier both warm up quicker than the (Indirect Heated Cathode) input and output tubes.
2. Why do engineers pick different B+, different plate current, different output transformer primary impedances, etc.?
The reasons are many.
That is just like the differences between 3 Tacoma Narrows bridges (the first one went down into the Puget Sound). Ouch!
Yeah, I'm not going to mess with the values of the coupling caps - just going to replace them with same capacitance, but polypropylene. I replaced similar decayed Suzuki paper/oil coupling caps in an sx-304B with orange drops, and that worked really well.
As far as the high B+, looking at a load-line calculator suggests higher power, but also higher distortion at the operating point they chose, but I don't think I'll get a real picture until I actually replace the caps and can safely start measuring response to input signals. Was hoping a glance at the schematics (or general experience or wisdom) might result in an obvious conclusion to somebody
As far as the high B+, looking at a load-line calculator suggests higher power, but also higher distortion at the operating point they chose, but I don't think I'll get a real picture until I actually replace the caps and can safely start measuring response to input signals. Was hoping a glance at the schematics (or general experience or wisdom) might result in an obvious conclusion to somebody