Any Theories on 6U8A Hum?

I have found it necessary to elevate the heater on 7199’s as well . I usually elevate them by 25v or so . I also find a 0.1 uf film cap from the voltage divider junction to ground eliminates the last bit of buzz .
 
The lack of elevated heaters on the 7199 explains why the Dynaco ST-70 eats them like candy.

25 V, however, is insufficient as the alumina insulation is really only rated for 90 V at that thickness. Yes, the datasheet specifies more, but at those ratings the RCA papers report dramatically reduced lifespan. That higher number is, I believe, a fiction created to allow the 7199 to be used as disposable items in a tube TV, its intended application, which was itself a disposable item. But what we know is that the continual stress eventually causes insulation failure, hum, noise, and eventually heater-to-cathode shorts.

In addition to the stressing of the insulation is metal migration which occurs at any potential difference above about 5 V.
 
The trick, and I've mentioned this previously, is that the driver tubes generally share a heater string with the rest of the tubes. Get the driver H-K within 5 volts and you're probably some 50-90 volts different on the rest of the tubes. Short of adding a separate heater supply for those tubes, about the best you can do is find a happy medium that keeps everything reasonably happy. Not perfect, but better than nothing.
 
Heater transformers are inexpensive and small, and that way all the tubes are within limits.

Costs a whole lot less than unobtainium tubes. Just sayin'.
 
So I now have 39VDC to the hum pot wipers on the ASR-433 due to the circuit modification.
I am going to see how this works out on tube longevity.
Or, would it be worth the trouble to install a separate filament transformer for the 6U8A's, remove the current circuit modification, and increase the DC voltage to the 6U8A heaters only?
 
So I now have 39VDC to the hum pot wipers on the ASR-433 due to the circuit modification.
I am going to see how this works out on tube longevity.
Or, would it be worth the trouble to install a separate filament transformer for the 6U8A's, remove the current circuit modification, and increase the DC voltage to the 6U8A heaters only?
It's not possible to optimize the filament offset in this amp as the penthode section has filament in common
with the triode that is used as cathodyne. The penthode has it's cathode only a few volts above ground, the
triode has ( according to the schematic i have ) 70V above ground. Thus 39V seems to be a very reasonable
compromise in this case.
 
Ok, Peter- thank you for the explanation.
I am going to leave well enough alone and just enjoy listening.
 
Heater transformers are inexpensive and small, and that way all the tubes are within limits.

yes but ....

It's not possible to optimize the filament offset in this amp as the penthode section has filament in common

so the best you can really do is compromise. I believe all of the commonly used triode/pentode tubes have the same situation with a single heater connection for both elements.
 
yes but ....



so the best you can really do is compromise. I believe all of the commonly used triode/pentode tubes have the same situation with a single heater connection for both elements.
Sort of like anti-skate, it's an approximate compromise, and some is better than nothing.
 
It's not possible to optimize the filament offset in this amp as the penthode section has filament in common with the triode that is used as cathodyne. The penthode has it's cathode only a few volts above ground, the triode has ( according to the schematic i have ) 70V above ground. Thus 39V seems to be a very reasonable compromise in this case.

Yes, you are absolutely correct.

I'd forgotten about the shared cathode in this instance. As a general rule one should optimize for individual tubes but with cathode sharing one's hands are tied.

So 39 V is the best one can do under the circumstances.
 
I checked the schematic for my Silvertone/ Warwick console amp, they tie the heater chain to the output tube cathode bias.
 
I checked the schematic for my Silvertone/ Warwick console amp, they tie the heater chain to the output tube cathode bias.

Since I don't have the schematic in front of me, some minor speculation which needs validation.

If the elevation voltage is pulled from the cathode bias resistor the voltage slightly fluctuates, even with a capacitor. Removing the degeneration is the purpose of the capacitor, but it cannot be fully removed.

Any fluctuation in heater voltage modulates the heater winding and thus the cathode. True, this may be minor, but it can still inject noise. Any foreign signal in this context is noise.
 
Since I don't have the schematic in front of me, some minor speculation which needs validation.

If the elevation voltage is pulled from the cathode bias resistor the voltage slightly fluctuates, even with a capacitor. Removing the degeneration is the purpose of the capacitor, but it cannot be fully removed.

Any fluctuation in heater voltage modulates the heater winding and thus the cathode. True, this may be minor, but it can still inject noise. Any foreign signal in this context is noise.
There was some bypass I replaced substantially more low esr capacitance with film augmentation to each channel during the upgrade.
 
Many of them do tie the cathodes to the heater string. With 6bq5 thats typically somewhere about +13v. Not all phase inverters run the cathode at a lot of voltage either, some are basically at ground potential so there isn't much need to elevate it. Split load / cathodyne / concertina where it has output from both plate and cathode to the output tubes are where you have issues. Also designs with direct coupled stages tend to have at least one cathode that is a bunch of volts above ground.

Deriving it from B+ is probably the better way, but the cathode voltage was "good enough" and it cost zero extra parts to use.
 
One of the advantages of the paraphase front end (Magnavox) is the cathodes being nearly 0 volt bias. Elevating the heater chain via the output cathode bias is free and more than adequate.
Mine is a cathodyne phase splitter.
 
Since I don't have the schematic in front of me, some minor speculation which needs validation.

If the elevation voltage is pulled from the cathode bias resistor the voltage slightly fluctuates, even with a capacitor. Removing the degeneration is the purpose of the capacitor, but it cannot be fully removed.

Any fluctuation in heater voltage modulates the heater winding and thus the cathode. True, this may be minor, but it can still inject noise. Any foreign signal in this context is noise.
silvertone-sears 528.69160.jpg
 
It must be amazing and impossible then to have a @100% quiet (common inobtrusive PS ripple/ear to the speaker hum excluded) amp with AC heaters lifted by only 7v of DC from output tube cathodes with minimal bypassing.

Instruments will reveal the modulation.

Besides, the issue is not merely hum, but degradation of the insulation leading to failure.
 
One of the advantages of the paraphase front end (Magnavox) is the cathodes being nearly 0 volt bias.

I suppose there has to be something nice to say about it :)

I really need to do this on my amp with the LTP inverter. It runs some 90 volts on the cathode, and the heater string runs right to ground. 12AX7's, maybe not as hum-prone but at the same time why abuse them? No room for a filament trafo so I'll probably just elevate it 30-40 volts and call it better than it was.
 
Back
Top Bottom