I find triode-pentodes useful for many applications. Types ECF80/6BL8 and ECF82/6U8 are fairly obtainable here and not too expensive.
Getting some hum in a recent design, I hooked up a circuit to test for cathode-heater leakage for either triode or pentode, and got the shock that few tubes are 'kosher' in this department. Of about 10 tested, only 2 had negligible heater ac transfer to the signal output. (The test was done with the alternative of switching over to 6V clean dc on heaters. Heater potential was at some +70V wrt common.)
The 7199 may be suggested as a more reliable alternative, but the gm is a little low for my purpose and they come expensive here. Most of the triode-pentodes with higher gm were designed for rf use, where no particular care was taken to keep heater leads away from signal leads. I could also not notice significant differences in internal layout with a magnifier, so much of the undesired transfer must take place inside cathodes.
The alternative would be to use separates e.g. EF86 + triode(s) or a dc heater supply, but I am still puzzled at the inconsistancy of the test results.
Does that mean that one had better steer away from the typical tv application tubes? I do see them in audio circuits. I would appreciate anyone's experiences here.
Getting some hum in a recent design, I hooked up a circuit to test for cathode-heater leakage for either triode or pentode, and got the shock that few tubes are 'kosher' in this department. Of about 10 tested, only 2 had negligible heater ac transfer to the signal output. (The test was done with the alternative of switching over to 6V clean dc on heaters. Heater potential was at some +70V wrt common.)
The 7199 may be suggested as a more reliable alternative, but the gm is a little low for my purpose and they come expensive here. Most of the triode-pentodes with higher gm were designed for rf use, where no particular care was taken to keep heater leads away from signal leads. I could also not notice significant differences in internal layout with a magnifier, so much of the undesired transfer must take place inside cathodes.
The alternative would be to use separates e.g. EF86 + triode(s) or a dc heater supply, but I am still puzzled at the inconsistancy of the test results.
Does that mean that one had better steer away from the typical tv application tubes? I do see them in audio circuits. I would appreciate anyone's experiences here.
A trick is temporarily up the heater volts to 7V; clout the tube and see if noise drops.That would indicate some impurity between htr & cathode.
As you mention, the other solution left is to DC all heater volts and lift 50V+ off ground or supply heater from separate source.
Going to 7199's; I abandoned this move some yrs ago as these are getting rare and pricy. What RCA and GE stock I have remaining have similiar same htr & cathode leakage problems and far worse, screen voltage variations. I find Sovtek's 7199 vary sonically.
In price comparison I find ECL series, from Aero, Ei, and 7643 group quite consistent.
richy
As you mention, the other solution left is to DC all heater volts and lift 50V+ off ground or supply heater from separate source.
Going to 7199's; I abandoned this move some yrs ago as these are getting rare and pricy. What RCA and GE stock I have remaining have similiar same htr & cathode leakage problems and far worse, screen voltage variations. I find Sovtek's 7199 vary sonically.
In price comparison I find ECL series, from Aero, Ei, and 7643 group quite consistent.
richy
richwalters said:
Hi Rich,
What do you mean by "clout the tube," even as a native English speaker I am not sure I have a clue as to what you mean. I suspect non native speakers might be even more confused. Do you mean tap the tube or something else?
I also found that Sovtek 7199s hum uncontrollably in a lot of applications in addition to their generally very mediocre sound quality.
I switched from 7199 to 6GH8A for my Dynaco amp due to availability and price. I had to use adapter plugs but it works well if you find the right 6GH8A. I had hum problem with supposedly reputable RCA (NOS) but since it was so cheap, it didn't bother me much to toss them. Then I went with Amperex brand and it works fine. Again, it was very cheap.
richwalters said:
My observation would have to be the same on that one. I don't toss them as otherwise they are fine, I just save them for an application where that is not an issue or heat them with dc.
The Sovtek 7199 are a major disappointment, I've had ones that had internal shorts in the connections that were visible to the naked eye. That is when I could still see with the "naked" eye.. Unfortunately no longer..
Thanks to all for input.
Darn it, @#%$&%! I will have to go to our local supplier and investigate every such topology (tube kind ... I must now and then use a fancy word to make others think I am Engresh) for internal construction to see if anything better exists. Not to many of them; I will start with 6AN8 - thanks Carlos.
Some further detail: I was going to use this in an RIAA tube pre with a hi-Gm triode at high current for low noise, as input. Testing as stated earlier, I found an up to 20x variation in hum induction via the triode cathode with 1K bias resistor to which RIAA feedback also went. The few good ones introduced nothing.
This is not a one-off; the circuit will get published, so tapping etc. no go. One might be forced to use a separate triode; hope ECC88 has no similar shenanigans.
Well, this is opposite to what I was told! That is that hum can be the result of unwanted emission from heater to somewhat positive cathode. Making the heater positive with regard to all cathodes (say +40V) would obviate that. I have experienced that in the past. But by 'pushing the HK rating' you might literally mean running at maximum spec or higher - then agreed, but that is not what I meant (and which one should not do).
But then the Scottish blood in me says involving an extra tube will be more expensive than making a dc heater supply, with other advantages of the latter .... decisions, decisions...
(why don't I just go to SS) Oops, sorry!
Darn it, @#%$&%! I will have to go to our local supplier and investigate every such topology (tube kind ... I must now and then use a fancy word to make others think I am Engresh) for internal construction to see if anything better exists. Not to many of them; I will start with 6AN8 - thanks Carlos.
Some further detail: I was going to use this in an RIAA tube pre with a hi-Gm triode at high current for low noise, as input. Testing as stated earlier, I found an up to 20x variation in hum induction via the triode cathode with 1K bias resistor to which RIAA feedback also went. The few good ones introduced nothing.
This is not a one-off; the circuit will get published, so tapping etc. no go. One might be forced to use a separate triode; hope ECC88 has no similar shenanigans.
Tweeker said:
Well, this is opposite to what I was told! That is that hum can be the result of unwanted emission from heater to somewhat positive cathode. Making the heater positive with regard to all cathodes (say +40V) would obviate that. I have experienced that in the past. But by 'pushing the HK rating' you might literally mean running at maximum spec or higher - then agreed, but that is not what I meant (and which one should not do).
But then the Scottish blood in me says involving an extra tube will be more expensive than making a dc heater supply, with other advantages of the latter .... decisions, decisions...
(why don't I just go to SS) Oops, sorry!
Any experiences with the 9DX base T-P tubes?
These were all video tubes, and 60 Hz hum feedthru would not have been tolerated on the TV screen. Surprising these haven't been picked up for DIY, with so many variants that fit the same socket. Tube rolling heaven.
9DX:
6AU8 6AW8 6BA8 6BH8 6CX8 6EB8 6GN8 6HF8 6HZ8 6JE8 6JL8 6JT8 6JV8 6KR8 6KS8 6KV8 6LB8 6LF8 6LQ8 6LY8 6MV8 6JA8 10LW8 7716
and higher filament Volt versions
Don
These were all video tubes, and 60 Hz hum feedthru would not have been tolerated on the TV screen. Surprising these haven't been picked up for DIY, with so many variants that fit the same socket. Tube rolling heaven.
9DX:
6AU8 6AW8 6BA8 6BH8 6CX8 6EB8 6GN8 6HF8 6HZ8 6JE8 6JL8 6JT8 6JV8 6KR8 6KS8 6KV8 6LB8 6LF8 6LQ8 6LY8 6MV8 6JA8 10LW8 7716
and higher filament Volt versions
Don
Smoking-amp,
Mmmmm - that IS serious. My RCA Tube Manual does not give that; in fact it describes many of the other tubes given above in the same manner.
Then I remembered I had a Philips Special Tubes manual somewhere. Complete data given for E80CF and E82CF.
DON!
Did you have to do this to me just before I went to bed? The linearity of both triodes are awful. Others listed by you earlier are not readily available locally. Oh well, back to further research, or separates.
Thanks for pointing this out to me. I will take a sleeping tablet (perhaps two).
Mmmmm - that IS serious. My RCA Tube Manual does not give that; in fact it describes many of the other tubes given above in the same manner.
Then I remembered I had a Philips Special Tubes manual somewhere. Complete data given for E80CF and E82CF.
DON!
Did you have to do this to me just before I went to bed? The linearity of both triodes are awful. Others listed by you earlier are not readily available locally. Oh well, back to further research, or separates.
Thanks for pointing this out to me. I will take a sleeping tablet (perhaps two).
Johan,
As you are working on a phono preamp, use the 6GK5 in the 1st gain block position. The type offers a rare combination of high mu and high gm. Also, excellent stock of Japanese manufacture is available at sane prices.
I will not say that AC heating in a phono preamp is impossible. With VERY great care, it can be done employing 7025s and 6EU7s. Don't go there without a spiral wound, hum bucking, heater.
It's EASY to build regulated 12 VDC heater supplies, in an economical fashion. Greinacher voltage double 6.3 VAC via Schottky diodes and HUGE capacitance in the stack. Follow with a 7812 3 terminal regulator. Wire "6" V. heater types in series.
As you are working on a phono preamp, use the 6GK5 in the 1st gain block position. The type offers a rare combination of high mu and high gm. Also, excellent stock of Japanese manufacture is available at sane prices.
I will not say that AC heating in a phono preamp is impossible. With VERY great care, it can be done employing 7025s and 6EU7s. Don't go there without a spiral wound, hum bucking, heater.
It's EASY to build regulated 12 VDC heater supplies, in an economical fashion. Greinacher voltage double 6.3 VAC via Schottky diodes and HUGE capacitance in the stack. Follow with a 7812 3 terminal regulator. Wire "6" V. heater types in series.
You discovered my secret weapon. AC on the filament has not caused me any problems, though I have not done a back-to-back comparison to DC. Does your 6.3V winding have a center tap? Make sure the center tap is at AC ground potential, but it can be floated DC-wise. Perhaps it may be worth to try a heater pot with the wiper connected to AC ground as with a DHT to null hum.
Eli,
Thanks again for info. I presume you have used the 6GK5 in practice. The dc heater topology you described I have used in the past without the regulator.
As you say, one can get by with 6Vac on heaters. I have hardly ever used dc in the past, but with tubes such as EF86 and 6AU6 as inputs (RIAA circuit). Just careful construction.
The problem with valves here is that it is convenient to stay with what our local agent (there is mainly one) stocks. That means that she is able to import in bulk, while DIYers must do so as best they can one/two at a time. In the USA and presumably in Europe there is no such problem.
Jeb-D,
I do use a 100 ohm pot to get best balance with heaters at ac. My previous experiments were like that, compared to smooth dc. With most of the unsuitable tubes the situation was best with the pot slider at one end - the end where the heater of the triode was closest to triode g1.
Smoking-amp,
Taking another look at the ECF80/82 triode non-linearity, the signal input is at most 10mV, i.e. output at some 400mV. Though graphic analysis is approximate because of size, it seems that at such low output, amplitude non-linearity might still be only some 0,1 - 0,2%, before global NFB. (I will need to measure this.) Local availability might then yet force the use of one of those.
But advice is gratefully noted.
Thanks again for info. I presume you have used the 6GK5 in practice. The dc heater topology you described I have used in the past without the regulator.
As you say, one can get by with 6Vac on heaters. I have hardly ever used dc in the past, but with tubes such as EF86 and 6AU6 as inputs (RIAA circuit). Just careful construction.
The problem with valves here is that it is convenient to stay with what our local agent (there is mainly one) stocks. That means that she is able to import in bulk, while DIYers must do so as best they can one/two at a time. In the USA and presumably in Europe there is no such problem.
Jeb-D,
I do use a 100 ohm pot to get best balance with heaters at ac. My previous experiments were like that, compared to smooth dc. With most of the unsuitable tubes the situation was best with the pot slider at one end - the end where the heater of the triode was closest to triode g1.
Smoking-amp,
Taking another look at the ECF80/82 triode non-linearity, the signal input is at most 10mV, i.e. output at some 400mV. Though graphic analysis is approximate because of size, it seems that at such low output, amplitude non-linearity might still be only some 0,1 - 0,2%, before global NFB. (I will need to measure this.) Local availability might then yet force the use of one of those.
But advice is gratefully noted.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.