Of course one could always built a cage/enclosure to keep the heat and voltage out of reach.
Didn't J/J and Svetlana make octal versions of the EL509 or 6P45S, respectively, without plate caps? Drawback is the EL34 pinout with the plate pin #3 next to the heater pin #2...
Best regards!
pin 3 is simply too close to pin 2 for comfort, even with the el34 i got arcovers at plate voltage of 440 volts...
that is why i much prefer the 12 pin compactrons...
5-6 years ago, I tested the load capacity of the 17KV6A tube by permanently connecting it to Ua ~ 300V and Ug2 ~ 150V, and with -Ug1 I set the plate dissipation to Pa ~ 28W (Ia ~ 93mA).
Even in the darkest darkness, there were no traces of any overheating of the plate and g2, even after 24 hours of the test!
I only noticed that the drain resistor g1 must have a relatively small value (at a constant value of such a large anode dissipation!) which must not exceed Rg1 ~ 47kOhms.
I tested 2 tubes (RCA and Syjvania), and after testing and cooling, the tubes on the static emission measurement did not show any signs of falling emissions and damage.
Even in the darkest darkness, there were no traces of any overheating of the plate and g2, even after 24 hours of the test!
I only noticed that the drain resistor g1 must have a relatively small value (at a constant value of such a large anode dissipation!) which must not exceed Rg1 ~ 47kOhms.
I tested 2 tubes (RCA and Syjvania), and after testing and cooling, the tubes on the static emission measurement did not show any signs of falling emissions and damage.
I'm not too sure that this might help reliably. In my experience, most pin 3 to 2 arcing occurs at the tube holder's top side or even at the tube socket where no one would put shrink tube on. Ceramic tube holders instead of bakelite provide some security, though.
Best regards!
I only use ceramic octal sockets, and never had a problem up to 600 volts. But I started putting on heat shrink when I noticed that the electrical charge would cause the plate lug to collect dust that seemed to “grow” over time. Like the “tin whiskers” on old germanium transistors, which would eventually lead to a short. Shrink wrapping them seemed to prevent this - probably by getting rid of the exposed sharp edges that develop locally high E field. I do often have resistor or capacitor leads “flying” over socket lugs when going point to point, and if high voltage differential is present I’ll put some insulation on the leads.
The tv tube data sheets often state the Rg1 needs to be at least a megohm, but that only applies to horizontal sweep service. As amplifiers, Rg1 does need to be less. Even common audio tubes like the 6550 and especially 7591 don’t like high Rg1 and can run away. I think the dissipation ratings on sweeps have more to do with runaway than on actual average plate temperatures anyway. Maximum bulb temps for most sweeps are rated 220C, rather than the 250 for 6L6GC.
You won’t get excessive ig2 when biased at 300V, 93 mA regardless of tube type. Ig2 won’t spike till the plate voltage falls.
I have also seen the arc happen inside the base of the tube. This was not uncommon on 6L6GC's used outdoors in the heat and humidity of Miami Florida, but I have also seen it on amps not used outside. Heat + humidity + dust, and maybe a little sweat or beer and a fireball can erupt inside that brown stuff that glues the glass into the black Bakelite base.
Once an arc happens either inside the base of the tube, across the bottom of its base, or in a Bakelite socket, the plastic will carbon track virtually guaranteeing a repeat performance. This will usually kill the rectifier tube or those black plastic diodes in early Fender amps with solid state rectifiers. This should blow the fuse, but sometimes the user stuck a 20 amp car fuse in there.......
A ceramic tube socket will not carbon track so it doesn't become permanently damaged if an arc should occur.
Remember that the plate voltage can go to 2 X the B+ voltage in a HiFi amp played without clipping. I have seen spikes to about 2500 volts in a cranked guitar amp played into speakers. The resonant impedance peak of a guitar speaker is often inside the guitar's frequency range.
...because it was in common use for tube holders and bases .
I'm scratching my head about what RCA was thinking when they put the 6L6 plate pin adjacently to the heater's, and why Philips/Mullard did the same with the EL34 and, on top of that, allowed it's plate voltage to be up to 800 V.
Best regards!
.I'm scratching my head about what RCA was thinking when they put the 6L6 plate pin adjacently to the heater's, and why Philips/Mullard did the same with the EL34 and, on top of that, allowed it's plate voltage to be up to 800 V.
Best regards!
I now have a bunch of 6JC5 to play with, but yeah, rp is shown on the data sheet as 50k ohms. I'll probably strap those triode and run them at about Vp = 300V and Ip = 60mA. I see nothing wrong with a 10W per channel PP triode amp. I figure an 8kp-p primary OPT should work for that. I have some of those...
6GE5 are $6 each... ?
Back when they first made 6L6’s (no bloody A,B,C or D) no one ever expected plate voltages to go to 800. I suppose it might have made sense to move pins around on the EL34 and the KT stuff, but I suppose they wanted keep them 7AC-ish so you could put in “bigger tubes” in an older design.
Socket 7AC had two spare pins, this should not have happened.
Compatibility has a big downside later on. Look at the 8088 follow on chips we still use, along with all the bus changes and Windows software changes.
Non-compatible compatibility.
We are all still "driving" Chevy Chevette engines with 10 layers of super-charging added on.
-------------------
12GE5 are $3 at Vacuumtubes.net. Used to be on the $1 list.
So were 21HB5A. $6 now.
What has happened to all those ESRC tubes?
Someone needs to tell some Estate Settlement lawyers and clients about tubes going gassy with time. And hobby demand dying off with the Pandemic.
Compatibility has a big downside later on. Look at the 8088 follow on chips we still use, along with all the bus changes and Windows software changes.
Non-compatible compatibility.
We are all still "driving" Chevy Chevette engines with 10 layers of super-charging added on.
-------------------
12GE5 are $3 at Vacuumtubes.net. Used to be on the $1 list.
So were 21HB5A. $6 now.
What has happened to all those ESRC tubes?
Someone needs to tell some Estate Settlement lawyers and clients about tubes going gassy with time. And hobby demand dying off with the Pandemic.
Last edited:
I know that as of Christmas week 2019 there were about 300 26HU5's, because I personally climbed the ladder to get them, and put them back on an upper shelf because Angelo, Stan's warehouse guy had the week off.
I was in Orlando with my daughter's extended family including 4 grandkids dropping a few kilodollars at the Kingdom of the Mouse.
I only had enough leftover cash and van space for a few, but I did pick out the shiniest tubes with good boxes. Unfortunately that would be the last time I ever saw Stan.
"F.ex. the EC/PC86 triode will outperform ther popular ECC88/6922 types with lower noise and higher gain & linearity."
-----
I have to respectfully differ with this assessment in a few ways.
1. Although the Gm is slightly higher for the ec/pc86, the noise differs by only 230 vs 300 ohms.
One will obtain more noise difference between resistor types and values than tube differences.
2. I am not sure where the claim of the ec/pc86 being more linear than a 6dj8/ECC88/E88cc/6922 tube types came from.
Was the source using the printed family of curves?
For general public's info, the JJ E88cc harmonic distortion (HD) measures -79db down, preamplifier 2 volts rms output. Even
the touted 6SN7 have some 19db higher HD (approximately 10 times higher under their best conditions). Other 6DJ8/E88cc
brands as well.
The one disadvantage of the JJ is its transparency is so good that associated poor quality parts will be exposed. As such,
the JJ brand requires a higher quality design.
3. With a much higher plate resistance, the ec/pc86 has about half the bandwidth (BW). The rise (attack) and fall times are half
as fast. In an audio system, the rise and fall times of each stage add, thus altering the attack times, perceived dynamics etc.
I hope this adds to one's knowledge base, and no, I am not associated with JJ in any way.
cheers
pos
-----
I have to respectfully differ with this assessment in a few ways.
1. Although the Gm is slightly higher for the ec/pc86, the noise differs by only 230 vs 300 ohms.
One will obtain more noise difference between resistor types and values than tube differences.
2. I am not sure where the claim of the ec/pc86 being more linear than a 6dj8/ECC88/E88cc/6922 tube types came from.
Was the source using the printed family of curves?
For general public's info, the JJ E88cc harmonic distortion (HD) measures -79db down, preamplifier 2 volts rms output. Even
the touted 6SN7 have some 19db higher HD (approximately 10 times higher under their best conditions). Other 6DJ8/E88cc
brands as well.
The one disadvantage of the JJ is its transparency is so good that associated poor quality parts will be exposed. As such,
the JJ brand requires a higher quality design.
3. With a much higher plate resistance, the ec/pc86 has about half the bandwidth (BW). The rise (attack) and fall times are half
as fast. In an audio system, the rise and fall times of each stage add, thus altering the attack times, perceived dynamics etc.
I hope this adds to one's knowledge base, and no, I am not associated with JJ in any way.
cheers
pos
Last edited:
My goodness, that is a significant difference!
If that's all true, I really need to try a couple of those JJ E88CC tubes.
I'm currently using some NOS (original boxes and all) PhilipsECG 6DJ8 tubes I bought back in the 1990s. Maybe it's time to leave them behind?
I have tried them and about every other, including 16 different Bugleboys, PQs, Sovteks, Telefunkens, Philips etc,
but the JJs are more accurate. But more needs to be done, or the JJs may not pan out. I know, a trap of continuous
testing, adjusting, tweaking. Something that could also immediately help.
I can only recommend some general ideas.
1. I eliminated all the electrolytic decoupling capacitors in favor
of poly caps, in my entire system, including the dac. One will probably not be able to replace
them all, but every one replaced seems to help. Mundorfs EVO and EVO in oil are excellent
caps to use. Might even check out Solens or small Solens in parallel vs electrolytics.
I found point 2 to be important for full potential.
2. If you have a coupling capacitor of 0,1uf to 0,68uf, there are problems in the design.
Worse yet if more than one stage uses such a small value. the -db frequency response adds.
The small size is to prevent low frequency oscillations, often called motor boating. Feedback, through
the power supply causes this problem. Can one add an additional power supply with filters stages
(and even a transformer) to isolate the stages? If not, then you are limited.
3. Bypass inductors next to decoupling capacitors if possible by using a resistor of value of inductor
winding resistance. The inductors cause non linear isolation between the decoupling cap and previous cap,
distorting the musical quality. If one must be used, try to keep it as close as possible to the rectifier, and
as far as possible from active devices.
Improvement can be obtained. it is just a matter of how much improvement one wishes, the cost, and
time involved in testing.
All the best.
pos
Hi, I recently scored a few dozen TV tubes with 150v screen grids… (6/12/17DQ6, 6/17GE5, 6BQ6, 6JN6, 6CD6, etc…) And I would like to use them as my everyday listening tubes and save the nicer KT88s and 6L6GCs for when I have guests haha.
I have a SE amp originally designed to run 6L6 with 300v B+ (and switchable Pentode/UL/Triode modes) And i would like to modify the G2 connection to run the lower 150v~175v voltages.. What is the best way to achieve this?
I’ve tried the following:
- OD3 to supply a constant 150v to G2 is successful but i do not want to add an additional tube. This would also only enable pentode mode.
- built a VVR which is more compact but it oscillates between 50v and 200v?? I may have to revisit this with some sort of bypassing scheme..
Things I have not tried yet:
Would it be possible to maintain the UL and Triode modes ?
Are there “minimum” screen voltages that need to be respected? If so imagine any sort of modulating voltage feeding into the screen needs some sort of scaling and positive offset to ensure operation within an acceptable window…
Sorry if this has been asked a bunch of times before, I am unable to find a “satisfying” answer. Yes I did see the “UNSET” thread but it really has me scratching my head…
Am I naively asking for the holy grail?
I have a SE amp originally designed to run 6L6 with 300v B+ (and switchable Pentode/UL/Triode modes) And i would like to modify the G2 connection to run the lower 150v~175v voltages.. What is the best way to achieve this?
I’ve tried the following:
- OD3 to supply a constant 150v to G2 is successful but i do not want to add an additional tube. This would also only enable pentode mode.
- built a VVR which is more compact but it oscillates between 50v and 200v?? I may have to revisit this with some sort of bypassing scheme..
Things I have not tried yet:
- Zener dropper of 130v…
- Just resistors…
- combination of zeners and resistors?
Would it be possible to maintain the UL and Triode modes ?
Are there “minimum” screen voltages that need to be respected? If so imagine any sort of modulating voltage feeding into the screen needs some sort of scaling and positive offset to ensure operation within an acceptable window…
Sorry if this has been asked a bunch of times before, I am unable to find a “satisfying” answer. Yes I did see the “UNSET” thread but it really has me scratching my head…
Am I naively asking for the holy grail?