Remove Hum from AA5

[maxhifi]

As per title... the ever popular AA5 power supply has a half wave rectifier, and a small input filter cap. The result is audible hum, under all circumstances.

Can't increase value of the filter cap without exceeding the peak current capability of the 35W4, and shortening its life. Good thing about a half wave rectifier though, is the ripple frequency is 60Hz, which is well below what the output transformer and speaker of a typical AA5 are capable of reproducing. During the day, this is no problem, but for night time listening at very low levels, there's residual hum audible in all of my radios.

A long time ago, I made a massive DC power supply, using four 6AX4 TV damper diodes in parallel, into a 1500uf capacitor. Plug the AA5 into DC, and the hum is gone - it's very cool. But now the 50C5 isn't getting the voltage it needs for full output power. Probably not a big deal.

In a few old designs I've seen ripple voltage added in reverse phase to the input of an audio amp, at just the right amplitude to achieve hum cancellation. I'm thinking this may be the easiest strategy to try. This how my RCA theatre amplifiers get away with having only a 14uF filter capacitor and inaudible him, but it requires periodic adjustment.

Aside from the two ideas above, or using a hum bucking winding on the output transformer, is anyone aware of any clever hum mitigating circuits from days gone by? I have a little silvertone clock radio by my bed which I listen to ever night, and due to the clock it can't be used on DC.

[bob91343]

Just increase the filter capacitor to about 30 microfarads. If you are worried about surge current, add a small resistor to limit it. And take a close look at the present capacitor to see if it's still got capacitance and a low esr.

[Chad Hauris]

I personally have not noticed any audible hum in a properly working 5-tube radio with new filter capacitors...I usually use 47 mfd for the first capacitor after the rectifier cathode and 100 mfd for the second one. Have had no trouble with that.
You may be able to just hear hum if you put your ear right up to the speaker but not otherwise.

You certainly could just leave the 35W4 filament in the circuit and shunt the AC input to a solid-state power supply maybe using a 1N4007 diode with a surge resistor ahead of it if you wanted to go for more capacitance. You probably could build it on a terminal strip inside the radio.

[bob91343]

I see no reason to modify the radio. I am sure that with good components the hum will be negligible. Getting fancy with hum bucking will only cause trouble, since it's hard to duplicate the wave with its harmonics and feed it into the amplifier. And while adding a solid state rectifier might be easy enough, it will produce higher voltages than the tube; combine that with the very common extra voltage we have these days and it invites early failure.

[peverett]

One thing to check is the heater to cathode leakage of the 50L6 or 50C5. If the leakage is great enough, this can introuduce hum. If you cannot check this(my tube tester does a good job of this, some do not). Try NOS tubes or tubes from other radios that do not have hum.

[Reece]

All good points to check listed above. One more: sometimes a first audio tube (12SQ7 or =) will be sensitive to hum coming right through the glass to it. I've found radios that will quiet down using a metal XSQ7. For this to work, octal pin 1 needs to be grounded. To test on any first audio tube, wrap foil around the tube and test clip it to chassis to see if a "real" shield would help.

[bob91343]

Perhaps it would be pertinent to decide what kind of hum it is. There are basically three types. The open grid hum, sort of high pitched raspy sound coming from poor audio shielding. The poor filtering hum, a sort of low pitched growl. And the 60 Hz leakage hum, a smooth deep bass tone.

[maxhifi]

Hey guys, thanks for all the contributions.

What I am dealing with is a little of all three types of hum, but more so the poor filtering hum. Modeling the power supply of an AA5 using PSUD from duncanamps, it is quite amazing how much ripple is present as it's designed - it's huge, and a single ended output stage with no negative feedback has no built in hum rejection at all.

That said, I tried swapping tubes in the Silvertone with no real improvement. Keep in mind I am talking about reducing the hum in a properly operating AA5, so that I can use it for very quiet night time listening beside my bed - not repairing something defective. I have about 15 AA5 radios, and all of them exhibit the same problem, more or less. The ones with the smallest speakers have the least hum, simply because they are unable to reproduce it. If I use it during the day at a normal volume range, the hum level isn't an issue at all.

[bob91343]

Ah that changes the picture! I thought you were trying to repair it.

So you want to redesign it to eliminate the hum. One thing you can do is cut off the low end of frequency response. Is the hum there with volume set at zero? You can add a bit of feedback (degeneration, actually) by removing the cathode bypass capacitor on the output tube.

You can also attenuate the audio with a small network at the voice coil; this will reduce maximum output level but you want it for quiet listening so that shouldn't matter. About 10 dB ought to do it, maybe a bit more. So the hum will drop that much, and the maximum output power will also, down to perhaps about 100 mW, plenty for your needs.

[old_coot88]

As long as the heater string is run on AC, it's doubtful you'll ever eliminate all hum and make the radio completely 'silent' no matter what you do. But maybe you could homebrew an outboard DC supply as you discussed earlier, just to run the radio chassis, while leaving the clock on AC. Just floating some muses.

[maxhifi]

Alright a plan is brewing. Rather than try and be 'clever', i'm going to go for brute force. Oldcoot has a very good point about the heaters. First I'll pick which radio is going to get the 'silent treatment', and then post back here tonight after 'surgery'. I was thinking about padding down the output too, but I'd rather treat the cause, not the symptoms.

[old_coot88]

If the cause of the `last vestige` of hum turns out to be from the heater string, it seems like heavily-filtered DC on the heaters would be the only 'cure'.

[wa2ise]

I've seen some radios use a special output transformer with an extra winding appended to the plate winding. This extra winding was about 15% of the turns of the plate to B+ winding. It's tube plate to B+ and then beyond that the 15% winding. This extra winding they passed the B+ current for the rest of the set. More importantly, the usually 1K resistor usually between the two filter caps is now fed by the 15% winding, and the other end of the 1K resistor has the 2nd filter cap. The idea is to balance the 1K resistor against the plate resistance of the 50C5 (10K) to get the hum ripple of the power supply to cancel, this cancellation happening inside the transformer, and thus little hum getting to the speaker. The far end of the 1K resistor having a cap to ground makes it look like a 1K load with respect to AC voltages.

More directly related to your radio project, you'd have to find a junker radio that used this special kind of transformer.

[TerrySmith]

One thing I ALWAYS do is use an isolation transformer, even for casual listening! Not only is it safer, it also filters out noise on the AC line.

[marty59]

Originally Posted by TerrySmith

One thing I ALWAYS do is use an isolation transformer, even for casual listening! Not only is it safer, it also filters out noise on the AC line.

Depending on the radio I have, it will either have a polarized plug or I mark one of the prongs with a sharpie so I ensure the chassis is at neutral/ground.