Oh yeah.
Then there's that 17KV6A HV pulse regulator tube that STILL HAS it's g2 grid!
Can do 275 mA all day. Normal B+, good internal triode, good pentode curves, 21600 gm at 275 mA.
Used to be $2.30 each back in the Antique Electronics Sales Flyers days, but $7 now.
Got a box full of them too.
Yeah, I got boxes full of 35LR6s, 21HB5As, 21JV6s, 6HJ5 and 21LG6A back in the Sales Flyer days too. 38HE7s and 6197s (an improved 12BY7)
were $0.35 each. Can still get a 21EX6 for $3. I haven't seen anything attractive on EBAY for a loooong time.
Then there's that 17KV6A HV pulse regulator tube that STILL HAS it's g2 grid!
Can do 275 mA all day. Normal B+, good internal triode, good pentode curves, 21600 gm at 275 mA.
Used to be $2.30 each back in the Antique Electronics Sales Flyers days, but $7 now.
Got a box full of them too.
Yeah, I got boxes full of 35LR6s, 21HB5As, 21JV6s, 6HJ5 and 21LG6A back in the Sales Flyer days too. 38HE7s and 6197s (an improved 12BY7)
were $0.35 each. Can still get a 21EX6 for $3. I haven't seen anything attractive on EBAY for a loooong time.
Last edited:
It's a good tube, as George has proven a number of times.
For those who have been paying $150 for an EL506, there is the $3 10/6JA5 which has nearly cut and paste curves. Not a bad upgrade for 7868 too, if you can reduce the voltages to fit.
I suspect GE just put a lower V grid2 into EL506 parts on hand to get a linear 19 Watt Vertical Defl. tube for the bigger color TVs. Compactron base now of course.
Then the $1 6JC6A. ($4 now ) The "made in Japan" versions have terrific mu 60 triode curves, the US ones have the best pentode curves. Can throw out those inferior D3As, E280F....
And the $1 6LQ8 pentode in triode mode (mu 50) Super, no competitors except 12HL7 in triode mode. ( both $3 now )
The problem now is the sockets cost more than the tubes did. And the $$$ OTs.
For those who have been paying $150 for an EL506, there is the $3 10/6JA5 which has nearly cut and paste curves. Not a bad upgrade for 7868 too, if you can reduce the voltages to fit.
I suspect GE just put a lower V grid2 into EL506 parts on hand to get a linear 19 Watt Vertical Defl. tube for the bigger color TVs. Compactron base now of course.
Then the $1 6JC6A. ($4 now ) The "made in Japan" versions have terrific mu 60 triode curves, the US ones have the best pentode curves. Can throw out those inferior D3As, E280F....
And the $1 6LQ8 pentode in triode mode (mu 50) Super, no competitors except 12HL7 in triode mode. ( both $3 now )
The problem now is the sockets cost more than the tubes did. And the $$$ OTs.
Last edited:
I forgot to mention the 12/6JN6 and 12/6GE5 Sweep tubes, $1 from yesteryear. The 6JN6s got eaten up by P. Millett's Engineer's Amp, so did the 6HJ5's. But 12/6GE5s are still plentiful. Got a box of them still. Lots of Novar Sweeps around too.
I have some TI parts on order from Newark ( still, waiting over 2 months!! WTF ). OTs should be a $20 PC board by the end of next year.
The other big project here was inspired by some comments in the SS forum that Randy Slone had solved the P-P crossover issue elegantly before he passed away in 2010. A business partner had the notes, but never shared them. After seeing the gm Wing plot for Lateral Mosfets in Cordell's book, it occurred to me what it might be. Should work for tubes too. Work to do. Mysteries are fun.
https://www.diyaudio.com/community/threads/randy-slone-passes.165100/page-2
post # 28
I have some TI parts on order from Newark ( still, waiting over 2 months!! WTF ). OTs should be a $20 PC board by the end of next year.
The other big project here was inspired by some comments in the SS forum that Randy Slone had solved the P-P crossover issue elegantly before he passed away in 2010. A business partner had the notes, but never shared them. After seeing the gm Wing plot for Lateral Mosfets in Cordell's book, it occurred to me what it might be. Should work for tubes too. Work to do. Mysteries are fun.
https://www.diyaudio.com/community/threads/randy-slone-passes.165100/page-2
post # 28
Last edited:
Here are some curves for a 6HZ5/6JD5 using positive grid Volts.
100V/div Horiz., 20mA/div Vert., and 0.5V steps. 4th curve up is 0V on the grid. Top curve is +5.5V
The "knee" side is staying down (LV) pretty well. Should be workable.
I'll look to see if I have any other Beam Triode types around.
100V/div Horiz., 20mA/div Vert., and 0.5V steps. 4th curve up is 0V on the grid. Top curve is +5.5V
The "knee" side is staying down (LV) pretty well. Should be workable.
I'll look to see if I have any other Beam Triode types around.
I don't have any schematics, but Google can probably find one. Look for a tube type color TV made after about 1971 or so. I worked in a Philco shop until 1971 and all of them still had the 6BK4 secondary side regulator.
Color TV's had the typical 6BK4 HV triode used as a constant duty (not pulsed) shunt regulator on the secondary side of the flyback, after the typical 3A3 rectifier. They were built this way from the mid 50's until X-rays were discovered (made public) around 1970. The CRT was not the main X-ray emitter, the 6BK4 regulator was. For a few years many variants of the 6BK4 were made, some with leaded glass for shielding, some with modified guts, and some with some really funny looking contraptions inside. shielding was added to the HV box too.
Then some brainiac invented the primary side pulse regulator. Its plate connects to a tap on the primary side of the flyback, the focus rectifier circuit, or even wired directly in parallel with the horizontal (line) output tube. The tube used was the HV beam triode discussed here. it was held in cutoff for most of the time, then pulsed at the appropriate time to chop the top off the flyback pulse keeping the high voltage in check. I have seen several circuits, but it was 40+ years ago.
Those are likely real measurements made on a "bogey" tube. They use a fancy curve tracer that pulses the tube very quickly so that the average dissipation is quite low. Someone we both know has gone to 600+ watts on a 36LW6 using a push button with about a half second ON time to make curves on a couple tubes. In operation the tube is OFF for most of the time, then pulsed ON for a small percentage of the overall line trace time. In a real late 60's / early 70's TV pulses were not so precise, so some excess dissipation was generated. I assume these curves and the additional numerical data provided to TV manufacturers allowed the engineers to figure out how to run these tubes without melting them.
My experiments with these tubes often did a great job of making power oscillators and TV jammers. Clip lead and flying wire hookups don't go well with high GM, high Mu tubes driven by a mosfet. I did make possibly the simplest amp ever made though. It used two transformers and a tube. I used a 1:1 driver transformer from my Sony Discman (yeah it was long ago) with the secondary wired from cathode to grid. I had a power supply and an OPT wired between the cathode and plate. The beam plates were connected to the plate. I turned up the power supply until I got some useful current then adjusted for best sound. I was somewhere north of 700 volts. The sound was not great, and the bass sounded like the kid's car across the street, without the license plate rattle.
I saved a box full of several kinds for future experiments but haven't got to them yet.
Smoking Amp -
Thanks for the effort. It will come in handy when I actually try to do this P-P amp. I'm interested to see if the mu and transconductance for the 6HV5A hold up at more modest values of plate current (probably not nearly as high as the data sheet values, but the tube has a closely spaced frame grid, after all). I plan to use feedback to make the output device even more triode-like, as the plate resistance is noting to write home about by itself.. I suspect that some compensation will be needed on the inner loop based on results I've already obtained using partial feedback with 6AH6 and 6L6GC. The P-P drive will likely be done using a couple of triodes and folded cascode, with feeback capacitively coupled to the cascode load resistors (big HV film caps...). For output transformers, maybe a pair of big Hammonds, or a pair of guitar transformers I got off of E-pay a few years ago that utilized selected output tubes at high plate voltage. The transformers, if I remember correctly, were 8K primary impedance with 8 ohms load, requiring high plate voltage to get the Watts I want, but also making better use of available output bias current. I have a nice selection of 6HV5As that I've quietly acquired on Ebay over the years. The most promising are from GE and Westinghouse, with larger plates than some of the other vendors. I'd have to dig some out to take a peek at this late date, but I suspect the Westinghouse tubes are likely re-marked GE types.
Thanks for the effort. It will come in handy when I actually try to do this P-P amp. I'm interested to see if the mu and transconductance for the 6HV5A hold up at more modest values of plate current (probably not nearly as high as the data sheet values, but the tube has a closely spaced frame grid, after all). I plan to use feedback to make the output device even more triode-like, as the plate resistance is noting to write home about by itself.. I suspect that some compensation will be needed on the inner loop based on results I've already obtained using partial feedback with 6AH6 and 6L6GC. The P-P drive will likely be done using a couple of triodes and folded cascode, with feeback capacitively coupled to the cascode load resistors (big HV film caps...). For output transformers, maybe a pair of big Hammonds, or a pair of guitar transformers I got off of E-pay a few years ago that utilized selected output tubes at high plate voltage. The transformers, if I remember correctly, were 8K primary impedance with 8 ohms load, requiring high plate voltage to get the Watts I want, but also making better use of available output bias current. I have a nice selection of 6HV5As that I've quietly acquired on Ebay over the years. The most promising are from GE and Westinghouse, with larger plates than some of the other vendors. I'd have to dig some out to take a peek at this late date, but I suspect the Westinghouse tubes are likely re-marked GE types.
I took some measurements of grid current for the 6HZ5/6JD5 tube above. These are avg. grid current while doing a plate V sweep 0V out to 750V
0 mA at 0V
10 mA at +1.7V
20 mA at +2.6V
30 mA at +3.4V
40 mA at +4.1V
50 mA at +4.8V
60 mA at +5.4V
70 mA at +6V
Rummaging thru the "junk" box here I found a 6JK5 (a long plate version) which is somewhat similar to the 6HS5 discussed earlier. I also found a 6JH5 (long plate version). So I'll get some curve traces for these too.
The TEK 576 here can do 80 uSec and 300 uSec pulses. I don't normally use that, I think it gives you a series of dots for curves. The tracer I have here is modified for tubes ( 0 to +/- 170V grid step range) and is also modified to do just 750V peak on the max plate V scale, instead of the standard 1500V.
I didn't want to have Electric Chair voltages pop up accidentally if I turned the selector knob too far. And there is no plastic safety shield here (bypass switch added), with lead wires going to the dangling tube sockets. After a couple of shocks over the years, vigilant caution is the norm. I mostly use the 350V max setting.
As it is, I'm tracing the Beam Triode tubes on the 220 Watt scale to get the 750V range, but that's really 110 Watt max with the reduced 750 V Vmax. And with the sweep of plate V and grid V steps (hence current), is more like 28 Watts avg. The 35 Watt tube doesn't red plate or break a sweat while I'm adjusting the camera, their tuff.
0 mA at 0V
10 mA at +1.7V
20 mA at +2.6V
30 mA at +3.4V
40 mA at +4.1V
50 mA at +4.8V
60 mA at +5.4V
70 mA at +6V
Rummaging thru the "junk" box here I found a 6JK5 (a long plate version) which is somewhat similar to the 6HS5 discussed earlier. I also found a 6JH5 (long plate version). So I'll get some curve traces for these too.
The TEK 576 here can do 80 uSec and 300 uSec pulses. I don't normally use that, I think it gives you a series of dots for curves. The tracer I have here is modified for tubes ( 0 to +/- 170V grid step range) and is also modified to do just 750V peak on the max plate V scale, instead of the standard 1500V.
I didn't want to have Electric Chair voltages pop up accidentally if I turned the selector knob too far. And there is no plastic safety shield here (bypass switch added), with lead wires going to the dangling tube sockets. After a couple of shocks over the years, vigilant caution is the norm. I mostly use the 350V max setting.
As it is, I'm tracing the Beam Triode tubes on the 220 Watt scale to get the 750V range, but that's really 110 Watt max with the reduced 750 V Vmax. And with the sweep of plate V and grid V steps (hence current), is more like 28 Watts avg. The 35 Watt tube doesn't red plate or break a sweat while I'm adjusting the camera, their tuff.
Last edited:
So - it looks like some solid state augmentation will be necessary to support the grid current you are observing. I thought the plate curves looked rather similar to a typical screen grid-driven sweep tube. It's a large, long grid, so it maybe can support big grid current pulses for peak power.
The TEK 576 will do up to 2 Amps for "grid current", made for low Beta transistors I think. I could test a Beam Triode up to grid or plate destruction to find out, but I only have a couple of those tubes, and the plate current limiting is just a slow circuit breaker.
The tiny wires visible on those frame grids will get hot a some point.
For normal Beam power tubes (with a grid 2), the screen V power supplies here have fast current limiting, so I don't generally damage tubes. But a 12HL7 I tested once at 60 mA steady plate current didn't seem to have as much emission after that, although it improved some after a cool down. So I'm a little leery of frame grid tubes near max ratings.
The tiny wires visible on those frame grids will get hot a some point.
For normal Beam power tubes (with a grid 2), the screen V power supplies here have fast current limiting, so I don't generally damage tubes. But a 12HL7 I tested once at 60 mA steady plate current didn't seem to have as much emission after that, although it improved some after a cool down. So I'm a little leery of frame grid tubes near max ratings.
Last edited:
The X-ray issue was known to the trade well before 1970. I recall in the early 1960s being told not to sit too close, and not because it would make us blind, but "X-rays".
Radio TV News Nov 1958:
https://worldradiohistory.com/Archive-Radio-News/50s/Radio-News-1958-11-R.pdf
My father didn't really fix the TV but he knew this.
Some years later a specific tube-type was recalled: take certain offending tubes out of your customer's sets, get a upgraded replacement and a few buck bounty.
That particular article discussed X-rays emitted from the CRT itself. These were usually weak and absorbed or scattered by the shadow mask in the CRT. The real issue was the HV regulator triode, usually a 6BK4 which shot a fairly strong beam of X-rays downward from the long cylindrical plate structure. Some were scattered, but most came out the bottom of the tube. The old style 6BK4's were collected by amateur X-ray machine builders, and often sold on Ebay for that purpose. The newer 6BK4B and 6BK4C were reduced X-ray versions that are shunned for DIY X-ray systems. Oddly most DIY X-rayers do not apply heater voltage and old worn tubes are said to work better than NOS.
https://www.ebay.com/itm/393884650223
https://www.ebay.com/itm/393884650223
I saw a schematic for an 811A SE amp with a 6V6 buffer dumping right into the 811A grid without any resistor/current source to gnd. Maybe that's how a beam triode P-P will end up looking. If I get squeamish about driving the grid real hard on the beam triode, I can always revert to a mondo sweep tube,, like I previously mentioned. I don't have to run the mosfet driver drain off a huge voltage. Maybe I'll incorporate a limiting resistor on the mosfet driver drain It'll be interesting to see what difference the plate feedback makes to the drive.
If you read the article, it is clear that hours on the CRT face is low exposure and seconds inside the HV cage is significant exposure.
Here are some curves for a Sylvania 6JK5 long plate Beam triode. Same scaling as before. (the last one was an RCA 6HZ5)
100V/div Horiz. and 20 mA/div Vert.. 0.5 V grid steps, 4th curve up is 0V, Top curve is +5.5V
A lot more pull away from the 0V plate axis than before. A lower HTR power tube. 6.3V 1.8 Amp
Grid currents:
0.12 mA 0V
10 mA 2V
20 mA 3.1V
30 mA 4V
40 mA 4.8V
50 mA 5.5V
60 mA 6.2V
70 mA 6.7V
100V/div Horiz. and 20 mA/div Vert.. 0.5 V grid steps, 4th curve up is 0V, Top curve is +5.5V
A lot more pull away from the 0V plate axis than before. A lower HTR power tube. 6.3V 1.8 Amp
Grid currents:
0.12 mA 0V
10 mA 2V
20 mA 3.1V
30 mA 4V
40 mA 4.8V
50 mA 5.5V
60 mA 6.2V
70 mA 6.7V
Last edited:
And here is the 6JH5 Beam Triode, Sylvania, long plate. Same scaling: 100V/div Horiz., 20 mA/div Vert., 0.5V steps, 4th curve up 0V, top curve +5.5V
Similar curves to the 1st 6HZ5. Both tubes take higher HTR power. 6.3V 2.4 Amp
Grid currents:
0.07 mA 0V
10 mA 2.2V
20 mA 3.2V
30 mA 4.1V
40 mA 4.9V
50 mA 5.5V
60 mA 6.2V
70 mA 6.8V
Similar curves to the 1st 6HZ5. Both tubes take higher HTR power. 6.3V 2.4 Amp
Grid currents:
0.07 mA 0V
10 mA 2.2V
20 mA 3.2V
30 mA 4.1V
40 mA 4.9V
50 mA 5.5V
60 mA 6.2V
70 mA 6.8V
Last edited:
I read through most of it. The average TV user may spend hours watching TV, but zero time inside the HV cage. If he goes inside with the power on, he will only do that once! I learned that at an early age. I had my own TV made from three identical Admirals fetched from trash, by age 10 and the only color TV in the neighborhood when the Apolo moon landing that was televised in color happened. I spent considerable time at the trash dump as a kid takinf parts out of discarded TV's. It was my parts store, one with unbeatable prices. Some B&W TV's only had a cage over the flyback and HV rectifier. Some had no cage at all. I didn't see the guts of a color TV till I got a 1957 vintage Emerson in 1967. I rebuilt it and used it until 1973.
GE experimented with liberal amounts of their newly created Polycarbonate plastic called Lexan in the HV cage of some of their 25 inch ( big for 1970) color TV's. The experiment might have done well elsewhere but was a colossal failure in South Florida, especially in homes without AC. Dust + HV + humidity makes for carbon tracking and lots of fireworks. I even saw a few meltdowns.
We did mostly Philco TV's which were sold and financed through Firestone tire stores. I believe that the Goodyear stores sold RCA TV's. GE's weren't too common in the area where I lived and worked. Sylvania made a color TV in a metal box. They rarely died beyond the usual sweep tube failure. They just had a nasty habit of shocking the crap out of the user due to the "death cap" often found in hot chassis radios. Most of the older houses still had two wire ungrounded electrical outlets.