I am looking at the implementation of hum pot for preamp tubes as I plan to apply AC to their heaters. I will elevate the heater to around 20V above the cathode as well as use hum pot tech to reduce/ eliminate hum from AC filament. Is that necessary to use 1 hum pot for each filament/tube or I just need to use 1 hum pot to take care a series of filament from tubes wired in parallel to a single 6.3V winding? Besides, is it necessary to place the hum pot as close to the filament/tube as possible ? Please share your views on that. Thanks.
Last edited:
Basically, unless you have a separate filament winding for each tube, you can only have one hum pot. The hum pot sets the ground of the filament winding at a ratio of the AC filament voltage.
Do you have filaments or heaters in your preamp tubes?
If you have heaters, the hum pot is unlikely to be necessary.
If you have filaments, the hum pot is unlikely to be sufficient for quiet operation.
If you have heaters, the hum pot is unlikely to be necessary.
If you have filaments, the hum pot is unlikely to be sufficient for quiet operation.
Even if you have Indirectly Heated tubes, if one of the tubes has a cathode that does not have a bypass capacitor to ground, you may have a hum problem.
Depends on what stage in the amp chain (the input tube is the most sensitive to hum that is caused by the filament to cathode interface).
And it depends on the quality of the tube, some are better than others of the same tube type.
A preamp that is made for a moving magnet or variable reluctance phono cartridge is quite sensitive, although with a good quality tube, is probably OK with AC filaments.
An input tube for a moving coil cartridge is super sensitive, and should almost certainly have DC on its filament.
Depends on what stage in the amp chain (the input tube is the most sensitive to hum that is caused by the filament to cathode interface).
And it depends on the quality of the tube, some are better than others of the same tube type.
A preamp that is made for a moving magnet or variable reluctance phono cartridge is quite sensitive, although with a good quality tube, is probably OK with AC filaments.
An input tube for a moving coil cartridge is super sensitive, and should almost certainly have DC on its filament.
Think of a filament and a cathode.
They are extremely close together.
Even with insulation, there may be some resistive leakage from the filament to the cathode.
And, because of the close spacing and the large area, there is capacitance from filament to cathode.
Either one of the above is a path for the AC filament voltage to transfer to the cathode.
No bypass cap, means that the interference signal is not shunted to ground.
I had a commercial kit integrated amplifier that drove me crazy.
The filament to cathode transfer of the filament voltage was not shunted to ground, because the circuit did not have a bypass cap to ground, and that circuit would have to be highly modified to be able to use a bypass cap there.
It was not just the 60Hz and 120Hz that came through. Some mid and high frequency noise that came through was much more offensive.
Just my experience.
They are extremely close together.
Even with insulation, there may be some resistive leakage from the filament to the cathode.
And, because of the close spacing and the large area, there is capacitance from filament to cathode.
Either one of the above is a path for the AC filament voltage to transfer to the cathode.
No bypass cap, means that the interference signal is not shunted to ground.
I had a commercial kit integrated amplifier that drove me crazy.
The filament to cathode transfer of the filament voltage was not shunted to ground, because the circuit did not have a bypass cap to ground, and that circuit would have to be highly modified to be able to use a bypass cap there.
It was not just the 60Hz and 120Hz that came through. Some mid and high frequency noise that came through was much more offensive.
Just my experience.
We learn as we go.
Sometimes we learn to late, and sell the offending product off, hopefully with
a statement of the problem that it has, to the purchaser.
Sometimes we learn to late, and sell the offending product off, hopefully with
a statement of the problem that it has, to the purchaser.
Dear All,
same "wish" or issue here.
I was living so far with a tiny bit of AC hum with my headphone amplifier - 6N1P-EV input, 6N6P output.
I have a 6.3V secondary, sadly when I had my trafos made couple of years back I didn't think about DC heaters and 9V.
Anyway.. I'm willing to sweat and learn.
How do I best implement a humdinger pot ? Can a multi-turn precision trimpot be used here ?
I'm just looking at one's datasheet: 100 Ohms, 22 turns, 750mW.
There are also 500R, 1k, 5k and 10k variants of the very same type (all 750mW).
I'm already using a virtual center tap with 2x 100Ohms resistors (if I remember well) towards ground the usual way. Not sure what causes that little remaining hum: the tiny differences between the two resistors (which is normal) or can I take the DC balance "done" and I still need to address AC balance ? What do you think ?
The trigger of the whole story: I'm listening quite a lot with this amp, the sound is just ahh, amazing, although a very simple design.
I needed to swap the 6N1P-EV-s and the 6N6P-s too.
By changing the 6N6P-s to new ones, nothing happened, works great..
By changing the 6N1P-EV-s to new 6N1P-s, hum has doubled at least or even tripled, it's annoying now. (EV is the strengthened military version with 5000hrs).
Trying to get rid of the increased hum and due to lacking 9V secondary on my toroid i thought a humpot is the way to maybe go below hearing treshold.
(Just for reference, circuit schematics attached for all who're interested).
Any help appreciated.
same "wish" or issue here.
I was living so far with a tiny bit of AC hum with my headphone amplifier - 6N1P-EV input, 6N6P output.
I have a 6.3V secondary, sadly when I had my trafos made couple of years back I didn't think about DC heaters and 9V.
Anyway.. I'm willing to sweat and learn.
How do I best implement a humdinger pot ? Can a multi-turn precision trimpot be used here ?
I'm just looking at one's datasheet: 100 Ohms, 22 turns, 750mW.
There are also 500R, 1k, 5k and 10k variants of the very same type (all 750mW).
I'm already using a virtual center tap with 2x 100Ohms resistors (if I remember well) towards ground the usual way. Not sure what causes that little remaining hum: the tiny differences between the two resistors (which is normal) or can I take the DC balance "done" and I still need to address AC balance ? What do you think ?
The trigger of the whole story: I'm listening quite a lot with this amp, the sound is just ahh, amazing, although a very simple design.
I needed to swap the 6N1P-EV-s and the 6N6P-s too.
By changing the 6N6P-s to new ones, nothing happened, works great..
By changing the 6N1P-EV-s to new 6N1P-s, hum has doubled at least or even tripled, it's annoying now. (EV is the strengthened military version with 5000hrs).
Trying to get rid of the increased hum and due to lacking 9V secondary on my toroid i thought a humpot is the way to maybe go below hearing treshold.
(Just for reference, circuit schematics attached for all who're interested).
Any help appreciated.
Attachments
The new tube is the problem, it is not quite up to spec if the previous one hummed much less. Try another of the same if you can. A hum adjust might not help enough. Get a 100 ohm pot, put pin one to one side of the heater and pin three to the other end of the heater. Pin two goes to ground. Remove the virtual ground resistors you already have, this replaces them. If you have both too much current is being grounded which might cause hum and wastes power which you transformer might not have to spare. A single turn pot of decent quality is all you need. Half watt or more, more is better. Set the pot to the mid point to start out!
Ok thanks, well the 6N1P and the 6N1P-EV-s are 99.99% 🙂 identical, a simple switch to the other type shouldn't be the problem. I'll try to get another set of 6N1P-EV-s.
What if I apply the trick you mentioned with the 500R pot ? What does some more resistance mean here in this application ? Less precise ? Or less offset-range to correct (if I need to) ?
What if I apply the trick you mentioned with the 500R pot ? What does some more resistance mean here in this application ? Less precise ? Or less offset-range to correct (if I need to) ?
Another option is to add a higher value, maybe 1K Ohm or 2K Ohm, pot in parallel to the existing 100R + 100R. The larger value makes it a finer adjustment, so multiturns aren't really needed. This is a good plan generally, reducing DC current in the pot's wiper..
All good fortune,
Chris
All good fortune,
Chris
It’s the individual tube not the type that migh be causing the hum, just a badly made tube. Another tube made on another day might be perfect. A 500ohm pot will work but 100 ohm is better because of the current going through it to ground. It is harder to adjust the 500 because a tiny turn is five times the same turn on a 100 ohm unit. Just more finicry but doable.
Thanks Chris, I'll try that, I just have a 1k trimpot laying around (aforementioned 750mW), I can try it tomorrow morning. Dangerous to solder late in the night (01:10 now here in Europe).
mdpaudio: I see, alright, I won't try 500 Ohms for now then. I understand. Thanks for the detailed explanations, it helped a lot.
To the tubes: both 6N1P-s behaving the same. Some more hum with them, compared to the more rugged EV variants. Strange. Anyway, I'll give this humpot a chance, let's see what I can do here.
mdpaudio: I see, alright, I won't try 500 Ohms for now then. I understand. Thanks for the detailed explanations, it helped a lot.
To the tubes: both 6N1P-s behaving the same. Some more hum with them, compared to the more rugged EV variants. Strange. Anyway, I'll give this humpot a chance, let's see what I can do here.
A humpot allows the symmetry of the AC heater voltage around 0V (or an elevated voltage) to be adjusted for more asymmetry to one end of a heater compared to the other end. That can help neutralise the net hum signal reaching a high impedance input grid of a valve - especially an input stage grid. If one side of a heater circuit has wiring that is physically closer to the input grid than the other side of the heater winding, then the input grid receives more capacitance coupling of hum signal from one side of the heater than the other. Most circuits and layout have quite low asymmetry, so a humpot provides no discernible improvement, but every layout and build is different, so a humpot may be able to noticeably reduce that form of hum ingress. That form of hum does not relate directly to how the cathode is bypassed (that is a different effect due to current flow through the heater-cathode insulation).
I understand it's already a challenge to have dead silent AC heating for an MM stage and for MC, almost-always DC heating is the preferred. (Let's leave that here for now).
In a normal tube amp (power amp), how likely is AC heating noticable on the output tubes ? (Given that all other small signal tubes are on DC). I see many designs where heating for the output stages is done with AC, I assume there is already so big difference between outgoing signal vs. that little hum that we can't always hear that or it's simply diminishing and goes below noise floor.
What about driver stage tubes ? Are they also sensible to hum and better to use DC, or due to already amplified voltage levels by input stage, is it less of an issue staying on AC ?
In a normal tube amp (power amp), how likely is AC heating noticable on the output tubes ? (Given that all other small signal tubes are on DC). I see many designs where heating for the output stages is done with AC, I assume there is already so big difference between outgoing signal vs. that little hum that we can't always hear that or it's simply diminishing and goes below noise floor.
What about driver stage tubes ? Are they also sensible to hum and better to use DC, or due to already amplified voltage levels by input stage, is it less of an issue staying on AC ?
Because hum can enter the signal stream in so many ways, it is a little simplistic to point to certain spots in a circuit other than where the signal level is lowest (ie. input stages) or where Rhk can often be a concern (such as in a concertina stage where the lower leg is not bypassed). Imho, every amp likely needs it own testing, as well as tube rolling.
trobbins you're right.. the suggestion above didn't help so I opted for DC at 6.3V secondary. Just to try it, 6V heating is still acceptable and if it falls below that, that's too weak then (although specs say 5.7-7V).
Anyway, using 4x good-old FR607 diodes (1000V, 6A, Vdrop=1.3V, 500ns) and 3 caps with altogether 18000 uF after it, I got ~8.5V DC unloaded and 6.45V on full load by the heaters. Wow. So apparently I could use my overspec'd 6.3V secondary for DC heating these 4 little tubes, however, without stabilizer.
So far everything is fine but I still need to search because as you told me, hum enters in many ways and it's still there. Insane.
Now I got an advice from a friend: I shall maybe try and change the 1N4007 diodes on the anode voltage side to my pack of Hexfreds from Vishay (600V, 15A, 23ns). Since the charging time for capacitors are short with Graetz bridges, it's important not to use slow recovery time diodes and now the 1N4007-s are really slow with their 1.5 us rating, that's 1500 ns compared to the Hexfreds' 25ns.
The heaters have the FR607 diodes now, providing DC.. 500ns.. maybe I'll swap them too.
Anyway, I learned a lot about the hum pot, many thanks for you guys..
Soldering iron is still hot, trying to grab the cooler side and use it now. 😉 Cheers.
Anyway, using 4x good-old FR607 diodes (1000V, 6A, Vdrop=1.3V, 500ns) and 3 caps with altogether 18000 uF after it, I got ~8.5V DC unloaded and 6.45V on full load by the heaters. Wow. So apparently I could use my overspec'd 6.3V secondary for DC heating these 4 little tubes, however, without stabilizer.
So far everything is fine but I still need to search because as you told me, hum enters in many ways and it's still there. Insane.
Now I got an advice from a friend: I shall maybe try and change the 1N4007 diodes on the anode voltage side to my pack of Hexfreds from Vishay (600V, 15A, 23ns). Since the charging time for capacitors are short with Graetz bridges, it's important not to use slow recovery time diodes and now the 1N4007-s are really slow with their 1.5 us rating, that's 1500 ns compared to the Hexfreds' 25ns.
The heaters have the FR607 diodes now, providing DC.. 500ns.. maybe I'll swap them too.
Anyway, I learned a lot about the hum pot, many thanks for you guys..
Soldering iron is still hot, trying to grab the cooler side and use it now. 😉 Cheers.
Perhaps the simplest first action in your case may have been to temporarily power the heaters from a separate 6V battery, as that removes most of the heater related hum causes (depending on where you disconnect the 6.3Vac supply). That is one of the initial diagnostics described in the link ( https://www.dalmura.com.au/static/Hum article.pdf ).
There is some description on generating dc from ac for heaters, including what voltages can be achieved for different diodes, in section 8 of this link: https://www.dalmura.com.au/static/Power supply issues for tube amps.pdf .
There is some description on generating dc from ac for heaters, including what voltages can be achieved for different diodes, in section 8 of this link: https://www.dalmura.com.au/static/Power supply issues for tube amps.pdf .