Humbucking coils; I can't work them out.

This has had me puzzled for ages.

Take your average energized 'speaker, the field strength will have some hum on it as the field coil is used as the choke. This will modulate the audio output amplitude at 100Hz (UK) but give no audible hum if there is no signal.

Question is how does adding (subtracting?) a bit of voltage to the voice coil from from the humbucking winding reduce the hum from the field? As far as I can see all it would do is add hum when there is no output.

What am I missing?

[Leon Crampin]

The hum current (magnetic devices function with ampere-turns) is added antiphase to the induced hum current in the voice coil. If the relative phasing is reversed, you end up with twice as much hum.

Leon.

[lesmw0sec]

It is not adding hum, but by adding an equal and opposite field, cancels out or reduces the variations in the field. It cant add hum in the no signal case, any more than the original field variations produced hum (as you correctly stated in your first para!)

Les.

[Peter.N.]

I would concur. A lot of early radios had a half wave metal rectifier so would have had a 50hz hum.

Peter

[steve_gus]

Perhaps merlin is saying that under no signal conditions there is no current in the speaker, so adding an antiphase will not be cancelled by the original hum signal so you get just the antiphase hum?

I suspect that there is always a quiescent current through the final stage, which will have hum on it and which will be cancelled by the antiphase.

Steve.

but by adding an equal and opposite field, cancels out or reduces the variations in the field

That would work if it was the stationary field, the voice coil can't tell the difference between that and audio.

that under no signal conditions there is no current in the speaker, so adding an antiphase will not be cancelled by the original hum signal so you get just the antiphase hum

Yes, exactly.

[DangerMan]

Originally Posted by merlinmaxwell

That would work if it was the stationary field, the voice coil can't tell the difference between that and audio.

Yes, exactly.

Not quite.
If you imagine a speaker with a field that has a lot of ripple and an open circuit speech coil (i.e. not connected to anything), there will be no hum from the cone.
However, if you measure across the ends of the coil you will find an AC voltage due to the fluctuating field cutting the turns of the coil. If you now short out the speech coil, a current will be allowed to pass and the speaker will begin to hum. If you connect it to the secondary of an otherwise isolated output transformer, it will still hum.
The hum-bucking coil has the same number of turns as the speech coil, and so it picks up an equal voltage. This is added in anti-phase to that picked up by the speech coil, so the net voltage (and hence current and hence hum) is zero due to the hum in the field coil.

[Billy T]

Dynamic speakers use the field coil as a choke to aid in filtering the HT supply as well as a source of current to energise the speaker magnet, hence there is always a residual ripple on the field current and a consequential alternating field in the voice coil gap, in addition to the audio component.

The humbucking coil is connected in series/antiphase with the voice coil and automatically cancels the hum. An interesting factor here is that to a cetain extent it is self-correcting i.e. if the filter caps are drying out, the hum in the speaker coil increases and the antiphase bucking field increases proportionally, so you don't always realise that there has been a deterioration in the caps.

There is another form of hum reduction via the speaker (but not bucking per se) that I have never actually seen myself, and that is the inclusion of a copper disk (called a shading ring) between the field coil and the voice coil. The eddy currents in the disk act to prevent the fields from reaching the voice coil.

When I was an apprentice, a common trap for young players was not connecting humbucking coil, which made the hum bad, or connecting it up in reverse, which made the hum twice as bad!

The humbucking coils I have seen have always had less turns than the voice coil, being matched to the quiescent state, and I think also that their positioning possibly has some influence on the number of turns required.

If a humbucking coil has gone OC, you can put a WW pot across the filament supply, create a pseudo centre tap with WW resistors (if the transformer filament supply hasn't got one) and feed one end ot the voice coil from the centre tap, and the other end of the o/p transformer secondary from the wiper of the pot, then adjust for minimum hum.

Cheers

Billy

[kalee20]

Originally Posted by Billy T

An interesting factor here is that to a cetain extent it is self-correcting i.e. if the filter caps are drying out, the hum in the speaker coil increases and the antiphase bucking field increases proportionally, so you don't always realise that there has been a deterioration in the caps.

Ì can't accept that - if the reservoir capacitor has 50V of ripple and the smoothing capacitor is good and big, it may have 0.5V ripple. Then the field coil will have 50V-ish of ripple voltage across it.

The AF stages will be fed from an HT source with very low ripple so will be hum-free. The loudspeaker will generate its own hum but this will be cancelled by the humbucking coil.

If the smoothing capacitor dries out, then potentially it could degrade so that there is 5V ripple across it. The field coil will now have between 45V and 55V ripple across it (depending on phase relationships) - not a great change, so the induced hum and cancellation will stay about the same - but the AF stages will be fed from an HT source with ten times the ripple so the audio at the output transformer secondary will be 10 times as hummy. And as cancellation will not have changed by anything like as much, you'll notice the difference.

[John_BS]

Very clear explanation from DangerMan: replies such as this should be archived in a "Help" section.

John

[Phil G4SPZ]

This is an interesting debate.

I have only seen a handful of energised loudspeakers, but seem to remember that the humbucking coil consisted of a relatively small number of large diameter turns, wound at one end of the field coil. It's therefore unlikely that the number of turns would have equalled the number in the voice coil; all that is needed is to ensure that the net current at hum frequency passing through the voice coil is zero.

Is it possible that humbucking windings were designed not only to cancel the hum generated within the speaker itself, but also any additional hum arriving along with the audio from the output stage due to imperfect smoothing? I can't see why this technique could not also have been used to reduce the need for large electrolytics and enable a relatively high level of ripple to be tolerated on the HT line feeding the output stage. Does anyone know whether energised loudspeakers were designed and made specifically to match particular radios, or did set designers simply pick a speaker from the manufacturer's catalogue?

From Dangerman...

If you imagine a speaker with a field that has a lot of ripple and an open circuit speech coil (i.e. not connected to anything), there will be no hum from the cone.
However, if you measure across the ends of the coil you will find an AC voltage due to the fluctuating field cutting the turns of the coil. If you now short out the speech coil, a current will be allowed to pass and the speaker will begin to hum.

Brilliant, why didn't I spot that, bloomin' obvious when you know!

[Billy T]

Originally Posted by kalee20

Ì can't accept that - if the reservoir capacitor has 50V of ripple and the smoothing capacitor is good and big, it may have 0.5V ripple. Then the field coil will have 50V-ish of ripple voltage across it.

I'm not sure what it is you don't accept, but the hum minimisation does occur. The examples you quote are perhaps a bit extreme and maybe outside the envelope, but for your average old set back then with liquid electrolyte filter caps a little past their prime but not dead yet, the increase in ripple on the field coil was matched by the humbucking, with the net result being a lot less hum than you might expect to get had you attached a PM speaker, which we often did in the workshop if only the chassis came in.

Ripple on the bucking coil being directly proportional to, and antiphase with the field ripple, that source of hum was removed. Other sources may remain, but that was not the point.

I was admonished for fitting unnecessary components to make it quiet and the boss told me that the level of hum would be fine once back in its cabinet.

I was seeing this in the late '60s, and I learned something else about EM speakers at that time too, when I slid a speaker over the edge of the bench to get some bass response to get a better feel for the hum level, then took hold of the volume control shaft. The speaker field coil had gone to ground, but since the speaker was on a wooden baffleboard in the cabinet there was no evidence of that problem, at least not until I found it! The speaker went one way, the chassis the other, and I hit the roof. From that day on I always checked for volts from the speaker to ground before handling them!

I can still see it, hear it and feel it to this day, including the bollocking from the boss.

Cheers

Billy

[lesmw0sec]

[QUOTE=Billy T;546470]

Originally Posted by kalee20

I can still see it, hear it and feel it to this day, including the bollocking from the boss.

Cheers

Billy

Happy memories - in my case it was a mains/battery portable which I had my hands in, with the result that the set went flying across the room, fortunately not damaged. My boss's only comment was a dry suggestion that I might be better off finding the fault just using the batteries!

Les.

[crackle]

This set has a humbucking coil, and a hum adjusting resistor as described by Billy.
To be honest I found the humbucking coil (I had to reconstruct the field coil)made little difference whichever way I connected it. The resistor made a distinct difference.
Mike

[Herald1360]

That resistor is a classic "humdinger" setup- used to centre tap (ie balance) the cathode (filament) of the directly heated valve(s) to earth. Without it there would be a lot of hum from the ac heater supply directly connected to the valve cathode(s).

It doesn't have anything to do with the humbucking coil arrangement which operates independently.

[crackle]

Hi Chris
I realise it is not directly related to the hunbucking coil on the speaker field winding, but can you adjust the resistor to counteract any hum induced in the speach coil by the field winding ripple.?

[DangerMan]

If you imagine the pot in the filament set to one end, that means one end of the filament has a steady DC bias from the bypassed cathode resistor and the other end of the filament has this DC with the AC heater voltage superimposed. This is just the same as imposing an AC feed onto the grid at this end but not at the other (which is clearly not actually possible).
The effect is that you get mains frequency hum.

If you turn the pot to the other end, you get mains hum again but now the phase is reversed.

The idea is to adjust to a point in between, where at one end of the filament current increases but at the other end current decreases by an equal amount. Note that if you don't get it right, the hum is still at 50Hz. If you do, the mains hum is zero due to the filament heating voltage.

Any hum from the speaker's field will be at the HT ripple frequency. For a full wave rectifier this is twice mains frequency.

In this case, no matter what you do you won't be able to balance a 100Hz hum by injecting a 50Hz ripple no matter what phase it is in.

So the answer to Crackle's question has to be "no".

Even if the rectifier is half wave, the field ripple will not be sinusoidal, and unlikely to be nullable by any amount of fiddling with the hum balance pot. Whether you can make a useful improvement... who knows. There are also phase shifts to be considered.

A correctly designed speaker will have a hum bucking coil which has the correct number of turns to provide for complete hum nulling of field ripple: the exact number of turns will depend on the exact magnetic circuit and how much of the coil is actually inside all of the gap all of the time. This affects speaker sensitivity and linearity (which is why I left this bit out of the "ideal" explanation !). Needless to say, it's no surprise compromises and cheapening happened in the real world.

HTH

Pete

[DangerMan]

In the cold light of day, there is of course another hum mechanism from the filament, which does modulate the valve at twice mains frequency. This is the actual heating of the filament which is greatest at both peak positive and negative voltage and will be audible from a direct heated valve in a single ended output stage.
Fortunately, push-pull will cancel this out, but the hum balance pot won't.