I am using AC for heater on my tube phono amp. How should the heater be referenced to B+? Do I have to do anything? Thanks.
B+ may be too high, and could exceed heater-cathode voltage limits. The heater must be referenced to something. Try ground. If this works, fine. If not, try somewhere around +40V (use a potential divider).
I see this used quite often (referencing the heater to a +ve voltage with respect to the cathode). Typical values I've seen are +30 to +40 Volts. I used +50V in both my latest HiFi Amp and Guitar Amp.
Last night I looked up the topic in RDH and I note that they recommended +70V.
Has anyone used this sort of "stand-off" voltage. Does that seem a bit high?
Cheers,
Ian
Last night I looked up the topic in RDH and I note that they recommended +70V.
Has anyone used this sort of "stand-off" voltage. Does that seem a bit high?
Cheers,
Ian
One can use any voltage which is safe for the lowest heater-cathode rating. I am not aware of anything less than 100V. But I am not sure of a high voltage advantage; as far as I have read, some +30V to +40V is enough to cope with any heater-cathode leakage. Only if there is e.g. a cathodyne phase splitter or Schmitt (the LTP) where a signal carrying cathode is positive, a little consideration is necessary. The 'bias' voltage can then go up to a safe 80V, so as not to leave the signal cathode (at perhaps some +100V to +120V plus peak signal voltage) too positive w.r.t. the heater.
Heater-cathode leakage resistance increases with voltage away from 0V for most tubes it seems, with most change occuring by 10-15V difference. However the resistance even at 0V has been meassured for a few common valves of different ages to be pretty bloody high, and effectively negligible in terms of noticing any improvement, as the coupling is effectively swamped by heater/cathode capacitance (which is not changed by DC elevation). And if you don't bypass the DC elevation correctly then the hum/noise may get worse from ripple injection via the HV supply. It seems best initial aim is to do all standard hum reduction techniques wrt cabling, and use a humdinger pot to tune out gate/heater residual hum.
In "Getting the most out of vacuum tubes" the author (I forget his name - you can Google) recommends +40V as giving the best life IIRC. This may or may not coincide with lowest noise/hum.
Tomer.
It's important to note that you would want that 40V (or whatever you choose) to be with respect to the cathode voltage, not ground!
It's important to note that you would want that 40V (or whatever you choose) to be with respect to the cathode voltage, not ground!
SY!! Don't do that .... 😉
As SY said, but that is not always possible e.g. when max. heater-cathode voltage is 100V and a cathode sits at +120V. One then does as well as one can.
This brings me to the point that the problem is not with heater-cathode (leakage) resistance as such, but the fact that sometimes a positive cathode can pick up heater "hum" by acting like the anode of a heater-cathode diode (depending on circuit circumstances obviously). I have had a few occasions where the heater-positive state has cured hum-pickup. This is possibly a rare occurrence, but still it is easy to get the heater chain positive, just in case a rogue tube comes along acting up sometime.
As SY said, but that is not always possible e.g. when max. heater-cathode voltage is 100V and a cathode sits at +120V. One then does as well as one can.
This brings me to the point that the problem is not with heater-cathode (leakage) resistance as such, but the fact that sometimes a positive cathode can pick up heater "hum" by acting like the anode of a heater-cathode diode (depending on circuit circumstances obviously). I have had a few occasions where the heater-positive state has cured hum-pickup. This is possibly a rare occurrence, but still it is easy to get the heater chain positive, just in case a rogue tube comes along acting up sometime.
The behaviour around 0V is not a classic 'diode' characteristic from measurements I've seen, with heater-cathode leakage resistance increasing with voltage away (positive or negative polarity) from 0V. This appears to not typically be a symmetric characteristic about 0V, but there is not much in the way of measurements to look at.
When trimming hum to low levels with 12A?7 tubes, the datasheet indicates there may be some benefit in choosing the 2nd half for the lower signal level use - although there doesn't seem to be any difference in the capacitive coupling specs between the halves.
When trimming hum to low levels with 12A?7 tubes, the datasheet indicates there may be some benefit in choosing the 2nd half for the lower signal level use - although there doesn't seem to be any difference in the capacitive coupling specs between the halves.
Most interesting - only we are talking about two different things?
Not sure I understand correctly, Trobbins. As said I see the heater-grid insulation problem acting more from an emission point of view than an ohmic one. In that sense it does not matter when the heater is positive (heater acting as anode) to the cathode. It is when the cathode is positive to the heater (cathode acting as anode) that faults in heater insulation can cause electronic emission to the cathode, thus getting 50/60 Hz into the signal line. (After all, one does have 8000 to 16000 mVp ac in the proximity of a few mV signal.)
Regarding tube construction, I visually inspected. For the 12A*7 construction strapping 'grid to pin 2' lies close to the heater center tap, and strapping 'grid to pin 7' close to the heater strapping going to pin 5. In certain models these come quite close, a small fraction of a mm. But in one 12AU7 I have, the grid posts lie on a radius from tube center. That then created different mechanical proximities.
It might be an interesting experiment to hook up both triodes similarly and investigate these effects - have never thought about that myself.
Not sure I understand correctly, Trobbins. As said I see the heater-grid insulation problem acting more from an emission point of view than an ohmic one. In that sense it does not matter when the heater is positive (heater acting as anode) to the cathode. It is when the cathode is positive to the heater (cathode acting as anode) that faults in heater insulation can cause electronic emission to the cathode, thus getting 50/60 Hz into the signal line. (After all, one does have 8000 to 16000 mVp ac in the proximity of a few mV signal.)
Regarding tube construction, I visually inspected. For the 12A*7 construction strapping 'grid to pin 2' lies close to the heater center tap, and strapping 'grid to pin 7' close to the heater strapping going to pin 5. In certain models these come quite close, a small fraction of a mm. But in one 12AU7 I have, the grid posts lie on a radius from tube center. That then created different mechanical proximities.
It might be an interesting experiment to hook up both triodes similarly and investigate these effects - have never thought about that myself.
The heater to grid interaction is effectively stray capacitive coupling via the wiring path, which includes the wire geometry within the valve socket and at the lead-ins to the heater/cathode tube (and also under-chassis wiring). That stray coupling can be a different magnitude for one 'end' of the heater, versus the other 'end', which I believe is why a humdinger pot is often 'off-centre' to achieve a hum null from the contributions of the two heater 'ends'.
The other hum mechanisms involving the heater-cathode interaction are a leakage resistance coupling (which is affected by voltage difference between heater and cathode), and a capacitive coupling due to the proximity of heater within cathode.
Measurement of the leakage resistance level does not typically show a 'diode' character - eg. where resistance is hugely different for one polarity compared to the other polarity, or where a 'junction potential' is observed.
Ciao, Tim
The other hum mechanisms involving the heater-cathode interaction are a leakage resistance coupling (which is affected by voltage difference between heater and cathode), and a capacitive coupling due to the proximity of heater within cathode.
Measurement of the leakage resistance level does not typically show a 'diode' character - eg. where resistance is hugely different for one polarity compared to the other polarity, or where a 'junction potential' is observed.
Ciao, Tim
Jeepers Tim - things are a disappearin' round here... 😱
The post where I apologised for the above mistake plus a few more comments 'appears to have disappeared' from the thread, even after I checked for its presence after having clicked the 'submit' button.
Thus to repeat; SORRY! The above should have read "the heater-cathode insulation of course. Hope that did not confuse you.
I respect your examination; I never dit that myself. We did occasionally find, with certain rogue tubes, that hum suddenly (i.e. not gradually) disappeared when the heater-cathode voltage was raised above a certain level. That seemed to confirm some sort of diode effect of a positive cathode being able to collect emission from a (faulty) emitter. This possibility was generally known at the time and we gave it no further thought. (There were not enough examples around to test anyway - think I found this only two - three times.)
Now let me try again ....
No probs Johan. The only measurements I have come across are:
http://dalmura.com.au/projects/Heater%20cathode%20insulation%20performance.pdf
http://dalmura.com.au/projects/Heater-cathode%20conduction%20plots.pdf
I haven't got any new 12AX7 to check for comparison - all those valves are used.
http://dalmura.com.au/projects/Heater%20cathode%20insulation%20performance.pdf
http://dalmura.com.au/projects/Heater-cathode%20conduction%20plots.pdf
I haven't got any new 12AX7 to check for comparison - all those valves are used.
- Status
- Not open for further replies.