Removing the 6072 will cause the phase splitter to turn on, and possibly reduce its anode impedance. So the hum could be entering at either stage.
Does the hum disappear if you AC ground the 6072 anode via a suitable capacitor? If so, then it is getting in via the 6072. If not, then a later stage is to blame.
Does the hum disappear if you AC ground the 6072 anode via a suitable capacitor? If so, then it is getting in via the 6072. If not, then a later stage is to blame.
User68,
Some photos might help. Maybe it's being radiated and received over-the-air. i.e. It could be a wiring/layout problem that creates loop areas between conductor pairs, which would allow any AC fields in the air to induce currents in them, which would then induce voltages across any impedances in the loops. See Faraday's Law.
The main one to worry about, in that respect, is probably your input signal/ground pair. If they are not running close together (preferably twisted together and also shielded), everywhere, then they could have an AC current induced in them which might induce a hum voltage across your grid input resistor, and your input source impedance, and even across the wire impedances themselves to some extent.
Do you get the same hum with and without an input connected? If it's there with nothing connected, then this is probably not your problem, since there would be no loop then (unless maybe it's grounded wrong). But if it's not there with no input connected and is there with a shorted or source-connected input, then this probably IS your problem.
Also, signal input ground should only be grounded at the ground end of the grid input resistor. Otherwise, you will have created a larger loop between the signal input conductor and its ground return path.
Similarly, if there is any appreciable length to them, your AC input pair should be twisted together, and secondary pairs, DC power/ground pairs, output pairs, etc.
And if ground returns don't follow next to the entire path where their current comes from, as much as possible, there's loop area created (i.e. yet another antenna for time-varying magnetic fields).
All small-signal conductors should be kept away from all AC and large-signal or power conductors (and devices). And they should never run in parallel with any of those. And if they must cross, or even come near each other, they should be at a right angle to each other.
I prefer to have all small-signal input pairs inside a shielded cable, with the shield grounded where the input enters the chassis, but not on the other end, where they usually connect across the grid input resistor. Note that the shield must not carry the signal ground, in that case.
Cheers,
Tom
Some photos might help. Maybe it's being radiated and received over-the-air. i.e. It could be a wiring/layout problem that creates loop areas between conductor pairs, which would allow any AC fields in the air to induce currents in them, which would then induce voltages across any impedances in the loops. See Faraday's Law.
The main one to worry about, in that respect, is probably your input signal/ground pair. If they are not running close together (preferably twisted together and also shielded), everywhere, then they could have an AC current induced in them which might induce a hum voltage across your grid input resistor, and your input source impedance, and even across the wire impedances themselves to some extent.
Do you get the same hum with and without an input connected? If it's there with nothing connected, then this is probably not your problem, since there would be no loop then (unless maybe it's grounded wrong). But if it's not there with no input connected and is there with a shorted or source-connected input, then this probably IS your problem.
Also, signal input ground should only be grounded at the ground end of the grid input resistor. Otherwise, you will have created a larger loop between the signal input conductor and its ground return path.
Similarly, if there is any appreciable length to them, your AC input pair should be twisted together, and secondary pairs, DC power/ground pairs, output pairs, etc.
And if ground returns don't follow next to the entire path where their current comes from, as much as possible, there's loop area created (i.e. yet another antenna for time-varying magnetic fields).
All small-signal conductors should be kept away from all AC and large-signal or power conductors (and devices). And they should never run in parallel with any of those. And if they must cross, or even come near each other, they should be at a right angle to each other.
I prefer to have all small-signal input pairs inside a shielded cable, with the shield grounded where the input enters the chassis, but not on the other end, where they usually connect across the grid input resistor. Note that the shield must not carry the signal ground, in that case.
Cheers,
Tom
Last edited:
Tom,
Thanks for your reply, it could well be a problem created by the layout. The signal wires are not twisted and unshielded and although they are short, are running very close to the twisted power wires. I'll post a photo tomorrow.
The hum is the same with or without an input connected. Before I consider changing the layout and perhaps covering the ground wires with tubular braiding I'll try grounding the 6072 anode to locate if that's the problem, as per DF96's suggestion.
DF96, can you advise what a suitable capacitor size would be or how I'd go about calculating it?
Thanks
Thanks for your reply, it could well be a problem created by the layout. The signal wires are not twisted and unshielded and although they are short, are running very close to the twisted power wires. I'll post a photo tomorrow.
The hum is the same with or without an input connected. Before I consider changing the layout and perhaps covering the ground wires with tubular braiding I'll try grounding the 6072 anode to locate if that's the problem, as per DF96's suggestion.
DF96, can you advise what a suitable capacitor size would be or how I'd go about calculating it?
Thanks
More tests have narrowed the hum down to occurring only when there's a tube inserted in the left 5687 socket. Taking the 6072 does stop all hum but then I replaced it and then tested with and without the 5687s. I found:
5687s in both sockets = hum
A 5687 in the right only = no hum
Same 5687 in the left socket only = hum
I'll try to trace the problem from the left tube socket
5687s in both sockets = hum
A 5687 in the right only = no hum
Same 5687 in the left socket only = hum
I'll try to trace the problem from the left tube socket
Well, I like puzzling.
User, have you tried a grid stopper (try 470 ohms - 1k) in the entrance tube (6072) and the grid (2) of the phase splitter. Note that it is probably RF entering, and Tom's suggestions are valid, while this is a remedy too.
Furthermore, the RF could sometimes be induced by the heater, so in the past some people have user a HF stopper ring (what's the proper name for it I can't recall right now) on the heater wires for the sensitive tube.
albert
User, have you tried a grid stopper (try 470 ohms - 1k) in the entrance tube (6072) and the grid (2) of the phase splitter. Note that it is probably RF entering, and Tom's suggestions are valid, while this is a remedy too.
Furthermore, the RF could sometimes be induced by the heater, so in the past some people have user a HF stopper ring (what's the proper name for it I can't recall right now) on the heater wires for the sensitive tube.
albert
Yes, it could be RF causing hum modulation. Grid stoppers would cure this, and if small enough won't do any harm anyway. Otherwise look for differences in lead layout between the left and right sockets. Anything carrying AC, whether a signal or PSU, should be screened or twisted to reduce induction and kept well away from other such leads. Where possible, route them close to the chassis too so you get some electrostatic screening.
.- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.