I know that these are identical topologies, but the ST-70 has a .02uf cap in series with the 6.3VAC center tap to ground, and the Mark IV schematic shows the same center tap connected to ground, reason I ask is I've put together my own version of the ST-70 using the same filament scheme as the ST-70 and I'm getting heater buzz, left channel is almost dead silent, right channel buzzes like crazy!! So it's not the power supply for sure! All tubes are super good, I'm not sure about the cap in series, with the ground! I've built two perfectly sounding Mark IV's without that .02 cap, what's the deal!
Swap driver tubes and see if the hum follows. If so, you can try find a quieter tube. Depending on the operating point of the driver stage, the tube may be exceeding its heater-cathode voltage rating. And a "noisy" tube may just be one that doesn't bias properly in the DC coupled driver circuit. Ideally 25 to 30% of the supply voltage on the triode splitter cathode. Distortion also suffers if it's not in this range.
"Floating" the heater winding allows it to go to a positive voltage from heater-cathode leakage. This results in less hum, since the cap charges to the peak value rather than having conduction across every AC cycle. A DC voltage divider is the preferred way to do this, but Hafler saved a few cents by just using the cap. And he might have chosen a larger cap value, if it didn't cost more...
"Floating" the heater winding allows it to go to a positive voltage from heater-cathode leakage. This results in less hum, since the cap charges to the peak value rather than having conduction across every AC cycle. A DC voltage divider is the preferred way to do this, but Hafler saved a few cents by just using the cap. And he might have chosen a larger cap value, if it didn't cost more...
Last edited:
I will probably just go ahead and elevate the heater CT probably 100VDC
Would be sufficient, nearly every schematic I've seen that has uses a cathodyne
Phase splitter,uses a voltage divider off the B+ to elevate the heaters.
I will probably just go ahead and elevate the heater CT probably 100VDC
Would be sufficient, nearly every schematic I've seen that has uses a cathodyne
Phase splitter,uses a voltage divider off the B+ to elevate the heaters.
Heater supplies should never be floated; they should always have a DC reference voltage (which may be zero, but could be higher). No electrode within a valve envelope should be floated, as then you have no control over what potential it will get.
In post 1 he said that the 6.3V CT was connected to ground via a cap, or directly (in two different circuits). No mention of a DC bias for the cap version. I was commenting on this:
Tom Bavis said:
I believe the caps are there to be DC blockers in case someone used a preamp with DC biased heaters. Since the 70 was an everyman's kit it could be used by someone who had other equipment and not familiar with the potential problems of a CT ground.
But this is one place where a DC blocker is the opposite of what is required!
The correct thing to do, if flexibility of connection to other equipment is needed, is to put a highish value resistor to ground (say, 100k). This ensures a DC reference but can easily be over-ridden by some other DC reference. A cap in parallel with it may do no harm, but probably will do no good either unless it has a sufficiently high value; 0.02uF might be useful in a radio set (to stop RF getting along the heater wiring) but not an audio amplifier.
The correct thing to do, if flexibility of connection to other equipment is needed, is to put a highish value resistor to ground (say, 100k). This ensures a DC reference but can easily be over-ridden by some other DC reference. A cap in parallel with it may do no harm, but probably will do no good either unless it has a sufficiently high value; 0.02uF might be useful in a radio set (to stop RF getting along the heater wiring) but not an audio amplifier.
Nothing is "required" there. DC biased heaters don't get grounded. DC "powered" heaters need the ground. The 70 would provide the AC to the preamp if someone wanted to use the front connectors. The caps are there to stop any DC from getting a ground from the heater winding.
Last edited:
Great discussion guys!
This is something that I have always wondered about.
A friend of mine brought over his all tube stereo power amp for some repairs and I had noticed that the heater CT had that same cap serial to ground. The amp was a push pull KT88 amp, relatively simple. After the repair I didn't like the fact the power supply didn't have bleed down resistors and there was 425v sitting under the hood for someone to get hit by if they were ever so curious to go poking around so I pulled the heater CT caps and referenced it to a voltage divider off the B+, the resistors do double duty as a potential divider and they safely bleed the power supply. The amp was quieter too, the left channel had a slight hum before this which sounds like the same problem you are having.
A short while later I had a conversation with the builder of the amp, Craig from "Nosvalves" and he was irate about what I did to his amp. I said well it wasn't your amp at the time it was my buddies, this slight technicality didn't stop him from telling me I had depreciated the value of his amp by doing this "mod". He also said it was detrimental to the performance of the amp. When I tried to question him about why what I did was so bad and why his way is superior he told me that he wasn't going to educate me for free, whatever that means. So when I saw this thread I got excited because it may just shed some light on whether I did a good thing or a bad thing to that poor amp. I do know it couldn't be any worse than the cold solder joint to one of the KT88's control grid causing a runaway bias condition that I found it in 🙂
Good luck to anyone that has to have customer support for any of his products.
This is something that I have always wondered about.
A friend of mine brought over his all tube stereo power amp for some repairs and I had noticed that the heater CT had that same cap serial to ground. The amp was a push pull KT88 amp, relatively simple. After the repair I didn't like the fact the power supply didn't have bleed down resistors and there was 425v sitting under the hood for someone to get hit by if they were ever so curious to go poking around so I pulled the heater CT caps and referenced it to a voltage divider off the B+, the resistors do double duty as a potential divider and they safely bleed the power supply. The amp was quieter too, the left channel had a slight hum before this which sounds like the same problem you are having.
A short while later I had a conversation with the builder of the amp, Craig from "Nosvalves" and he was irate about what I did to his amp. I said well it wasn't your amp at the time it was my buddies, this slight technicality didn't stop him from telling me I had depreciated the value of his amp by doing this "mod". He also said it was detrimental to the performance of the amp. When I tried to question him about why what I did was so bad and why his way is superior he told me that he wasn't going to educate me for free, whatever that means. So when I saw this thread I got excited because it may just shed some light on whether I did a good thing or a bad thing to that poor amp. I do know it couldn't be any worse than the cold solder joint to one of the KT88's control grid causing a runaway bias condition that I found it in 🙂
Good luck to anyone that has to have customer support for any of his products.
On the contrary, a DC reference is required - which the resistor provides. DC biased heaters also provide a DC reference. DC powered heaters may or may not provide a reference (depends on whether the DC supply is floating or not) - but they should provide a reference. A resistor to ground in the amp (if non-zero DC reference is not used) provodes a DC reference in the amp, but does not preclude providing an alternative DC reference if the heater supply is then shared with another unit such as a preamp.20to20 said:
I don't know why people make a fuss about this: heaters need a reference voltage because all electrodes within a valve need a reference voltage. Simple, easily provided: a resistor does it, a capacitor does not. It baffles me why people seem to think that using just a cap does anything useful.
Heed the words of DF96; heater supplies should never be left floating.
The single capacitor allows the heater to float; some value of purposeful resistance is needed. Whether this resistance ties to common (ground) or some other DC reference is another issue. It really is that simple.
The single capacitor allows the heater to float; some value of purposeful resistance is needed. Whether this resistance ties to common (ground) or some other DC reference is another issue. It really is that simple.
The resistance value of the cap on the heater winding CT to the AC @ 60hz is .13R. Is that enough ground reference?
I find it difficult to attach meaning to this sentence. Possible meanings are:20to20 said:
1. The cap is extraordinarily leaky and has a DC resistance of 0.13ohms - needs replacing.
2. The cap has a value of 20000uF and so has a reactance at 60Hz of 0.13ohms - presumably it is an electrolytic and so has sufficient leakage current to act as a DC leak.
I'm going to guess that neither of these meanings were intended?
Capacitors store charge. Very leaky capacitors may be capable of dissipating charge. The idea is that any element inside the tube envelope may accumulate charge, which is undesirable as the voltage of that element increases to unknown/uncontrolled positive or negative values. A reasonably intact capacitor does not dissipate this accumulated charge.
Or perhaps we could set the bias conditions of tubes with only capacitors, no longer needing any resistance. How's that grid leak working out for you?
Or perhaps we could set the bias conditions of tubes with only capacitors, no longer needing any resistance. How's that grid leak working out for you?
- Status
- Not open for further replies.