Dynaco ST-70: Base Line Testing
- Thread starter dcgillespie
- Start date
Your ignorance is appalling. I said that Kamna pointed out the issue for the ST-70 as a means to explain the high failure rate of the 7199.
RCA systematically worked on the heater-to-cathode leakage issue in the 1950s. It was known before then.
Why are you so hell bent on focusing on trivial issues instead of the bigger picture of heater-to-cathode leakage.
Why do you endlessly stalk me?
I am preparing a monograph on the subject of heater-to-cathode leakage so I'm not yet distributing the full bibliography. It has taken me hundreds of hours of work to assemble the details and analyze them. (This is more for tube failures in general, but HK leakage was a big piece.) I think it explains a lot about tube failure.
I will be happy to send you a draft for review and comments when it's ready if you're interested.
I'm not going to further comment in the forum since nobody else seems to be interested, and I have actual work to do.
Feel free to PM me if you have questions.
s-petersen
Scott
Do the 6GH8A's have as many problems as the 7199's in the Dynaco 70?
I didn't mean to start such a debate over this.
Last edited:
In my experience, the inherent problems of tri/pent tubes are present in all of the 9 pin miniature examples of this type configuration. That is, within any of the tubes identified, it's not hard to find examples that are either microphonic, noisy, hum due to heater/cathode leakage, fail to operate properly in direct coupled AF Amplifier/Cathodyne Inverter applications due to manufacturing tolerances with the screen grid in the pentode section, or any combination of these issues. Therefore, there are a number of things that can make examples of any one of these tube types unacceptable for even general audio use, let alone high quality audio applications. And those examples with screen grid tolerance issues relative to applications involving direct coupling between the sections will invariably not be caught by any tube tester, as the issues causing poor direct coupled performance will typically not show up on a tube tester as having reduced emission or poor Gm (transconductance). In typical Dynaco, Scott, and Heath applications then (as well as a host of others), a number of these tubes don't even make it out of the starting gate due to this concern. As the general supply of any of these tube types are depleted, the inventory levels that remain then are more and more made up of the unacceptable examples that get returned, traded, sold at auction, or are otherwise just passed around, so that more than once I've ended up having to buy well in excess of what I need to find a couple truly good examples of these tube types. Against this backdrop, the cost and availability of the 7199 ends up taking itself out of the game. At many times the cost of the more common types, and getting rarer by the day, it becomes harder and harder to justify not switching to acceptable replacements -- either by the use of adapters where workable, or by the use of boards that have sockets that will directly fit the alternate tube types. At least with the alternate types, supplies are still such that good performing examples can still be reasonably found, and at reasonable prices as well -- without having to buy a whole bushel to find them.
As for the 7199, I've had such problems trying to find truly good NOS NIB examples of this tube -- even from the well recognized best sources out there -- that recently, a client with a Fisher piece I was restoring that used this tube asked me to redesign the circuits involving that tube so that it could use a conventional twin triode tube type in its place. Granted, that is an extreme measure. But the unit meant enough to this person that he didn't want the invariable problems with these tube types in general to become his problem in particular down the road. And it did nail the problem. Think about it: You can buy any of the popular twin triode tube types -- new or old manufacture -- from any of the quality tube venders, and have a good expectation that the tube you receive will work properly as it should. When purchasing any of the popular tri/pent tubes, there's always a gamble factor with audio applications, so make sure the seller accepts returns!
Dave
As for the 7199, I've had such problems trying to find truly good NOS NIB examples of this tube -- even from the well recognized best sources out there -- that recently, a client with a Fisher piece I was restoring that used this tube asked me to redesign the circuits involving that tube so that it could use a conventional twin triode tube type in its place. Granted, that is an extreme measure. But the unit meant enough to this person that he didn't want the invariable problems with these tube types in general to become his problem in particular down the road. And it did nail the problem. Think about it: You can buy any of the popular twin triode tube types -- new or old manufacture -- from any of the quality tube venders, and have a good expectation that the tube you receive will work properly as it should. When purchasing any of the popular tri/pent tubes, there's always a gamble factor with audio applications, so make sure the seller accepts returns!
Dave
s-petersen
Scott
Can the grid of the triode be measured for this defect(leakage) if someone doesn't have a distortion analyzer available?
I'm not sure I understand your question. If you mean the screen grid in the pentode section when used in direct coupled applications, the usual manifestation of screen grid issues is that the pentode section's plate voltage becomes unusually high or low, which then either under or over biases the triode section it directly drives. Due to the nature of the design, the triode section will adjust itself accordingly to any reasonable shifts in the pentode's plate voltage. This is why many manufacturers don't even suggest measuring this voltage due to the wider than normal variations that can occur. But when the shift is unusually great, the resulting performance problems become significant. One of the most accurate ways to check for screen grid issues when a tri/pent tube is operated with direct coupling is to measure the triode's cathode voltage. This will normally be a handful of volts greater than the pentode's true plate voltage due to the bias generated in the triode stage. If the triode's cathode voltage is well off from typical norms, then screen grid issues in the pentode section are a likely possibility. Determining the pentode's plate voltage in this manner prevents the meter -- even high impedance types -- from loading down the plate voltage of the pentode, since it represents a very high impedance itself.
Dave
Dave
s-petersen
Scott
I thought there would be a change in voltage between the cathode and grid of the triode that could be measured. Possibly I misunderstood the issue. I think the way you describe the possible way to test for this when the tube is installed, might work for me.
This has been discussed in another thread.
Basically it has to do with keeping the grid ( pin 9 ) in 7199 on 125 Volt (375 / 3) to have it centered as a
cathodyn phase inverter. This is done through adjusting the screen ( pin 3 7199) to adjust the working point
of the pentode section. If voltage of grid( pin 9 ) is to high then the pentode draws to little current, this
is fixed by decreasing 1.5Mohm resistor to a smaller value. Thus by adjusting the voltage on the
screen one can adjust the anode ( pin 2 ) to be at app. 125Volt.
Note that in SCA35 there has been an invention that stabilized this, the screen is fed from the
cathodyn's cathode, a genious design inprovement !.
s-petersen
Scott
So, what would be the ideal heater Dc voltage above ground for best reliability? is there any downside for the output tubes?
The ideal heater voltage must be a compromise as only one pair ( the triode parts of 6u8/7199 ) is at app 100V,
all other filaments are at a few volts. Having filaments more positive then cathods is more damaging then
the opposite, i would say that at most +20V is a good figure.
I've been following this conversation.
I have a question.
I didn't measure yet so I don't know what they read (well I might have and the end of the rebuild, i don't remember and i doubt now I can find the notes).
Are you saying a handful of volts negative of the triode or pentode section (or would they be so close that it wouldn't much matter?
I have a question.
I didn't measure yet so I don't know what they read (well I might have and the end of the rebuild, i don't remember and i doubt now I can find the notes).
Are you saying a handful of volts negative of the triode or pentode section (or would they be so close that it wouldn't much matter?
As both the pentode and triode and EL34 shares the same filament one has to compromise.
As a general rule, one should minimize the potential difference between filament and catode, but
if possible the cathode should be positive in relation to filament.
The triode sction has cathode at app 100V, the penthode and EL34 a few volts. This reduces the possibilities
to some very low voltage on the filaments. ( say 10-30V )
Thanks
If I understood the explanation
(Which I quite possibly didn't)
You're trying to be close to the cathode (negative) as possible.
So if I'm getting it right, and I get your response, if I were to set up a voltage divider using that triode cathode 100v (obviously I would have to verify this. The filament would be approximately 94-100v (6vdc filament voltage)?
(Which would make you positive of the pentode cathode)
If I understood the explanation
(Which I quite possibly didn't)
You're trying to be close to the cathode (negative) as possible.
So if I'm getting it right, and I get your response, if I were to set up a voltage divider using that triode cathode 100v (obviously I would have to verify this. The filament would be approximately 94-100v (6vdc filament voltage)?
(Which would make you positive of the pentode cathode)
Last edited:
s-petersen
Scott
But then your El34's are at 100v difference, hence the compromise, Ideally if the EL34's were cathode based things could be raised even more
If the heater filament is hotter than the cathode sheath it is heating, wouldn't the normal course be the electron fllow tendency be from hotter to cooler? So the reverse bias to counter would be to make the heater positive relative to the cathode?
s-petersen
Scott
I don't think temperature would change the direction of electron flow, just the electrical potential
If the triode of 6u8 / 7199 cathode was the only tube in town, then raising the filament to 90V would be perfect.
But as other tubes exists, all of them has their cathode only a few Volt above ground one should not
raise the filament more then some tens of Volt.
