I've recently pulled out a Healling radio I restored and recapped many years ago - you'll see it in my battery valve power supply posting in General Discussion.

I never was very happy with the sound of this radio which is why it's languished in a plastic tub in the garage. As well as not being very loud before clipping, there is a strange IM distortion on bass tones. It sounds like the kind of distortion you get from a diode detector when the AC load predominates. (I checked for that and all is well...)

Then, looking at the picture I took for the power supply posting, the idea dawned - could it be the field from the speaker is distorting the electron path in the 1S5 and the 3V4?

So I pulled the chassis to clear the speaker from the chassis and Presto! Radio is clean and loud!

Could it be that such a glaring design flaw was been around all that time and gone un-noticed?

So now you should be able to build a shield to stop the problem.

Perhaps there may be a shield missing most likely from the 1S5? Generally, Healings are a pretty well designed set, the battery valve one I have is very sensitive.

Or else there could something else affecting the radio's performance when it is in the case.

Early B/W television receivers had magnets on the CRT to adjust beam deflection.
But I can't believe magnetism would contribute to this audio problem.
Maybe a internal structural fault within a valve could cause this effect.
There are plenty of radio receivers in which the speaker magnet virtually touches the output valve, with no problems.
Ian, I think there is another answer to this problem. If interesting all the same.
JJ

Hi Ian, is the fault repeatable, that is if you put the magnet back into proximity does the distortion return?

Fascinating if it does..

Is the speaker the original type?
Or was it fitted with a capped magnet speaker?

Fred.

An addendum to my last post.

Swung down the back cover on my Healing 503B and there is not a shield on the 1S5. Both this valve and the 3V4 audio output are barely a cm away from the speaker magnet.

Think there must be some other problem with your set.

Does it have a loop or ferrite rod as an aerial ? Probably something there that may be affecting the performance of the set when the chassis is removed.

The 503B is very directional at times and has to be moved around to get best signal.

If some spare valves are available try a valve swap just in case there is a odd faulty valve.

is the fault repeatable, that is if you put the magnet back into proximity does the distortion return?

That's the first thing I'd try, and I'd try other speaker magnets, too.

I haven't heard of such an issue before, nor can I find authoritative references to it online.

You can have problems like that and I am sure that I have seen it written up in books. The magnet will impinge on the electron path in the same way it does in a CRT.

You also have to be seriously careful replacing speakers with those having extremely strong modern magnets. I actually had to put a shield between a power transformer and the speaker in a Kriesler (I think) to stop a war between the two.

I am sure that I have seen it written up in books

I guess if I trawled through my library I might find mention of it, but if it were a significant design issue then I think we'd have heard about it before this.

Like Fred, I want to know if it's repeatable with that Healing set.

Yes, it's repeatable. You can move the speaker magnet near the 1S5 and the distortion (it is subtle and depends on the program) returns.

Same goes for the 3V4. There is definitely more output when the field magnet isn't hard up against the valve.

The effect in both cases is not dramatic, what aroused my curiosity is that, with an 8" speaker and good-sized transformer, I would have expected this radio to sound better than it did, better than other radios in my collection that have smaller speakers.

Next, I should measure the parameters (such as current draw) with and without the magnet.

There is definitely more output when the field magnet isn't hard up against the valve.

Interesting. Possibly more of an issue with some low voltage valves?

Here is a slightly blurred photo of my 503B and the 1S5 is only 10mm from the speaker magnet.

As well had a look at my AWA 500MY Radiolette and the speaker magnet is only about 5mm away from the 1S5 although this speaker magnet is a lot smaller than that of the Healing.

The Healing has quite a powerful magnet and the valves on mine are hard up against it. Simplex, do you think the audio performance of your 503B is not quite as good as you'd think it should be?

Yes GTC I would expect that a battery valve would be more affected. The electron beam is not as "stiff" as when you have a higher voltage.

If someone has a valve tester that can test gm of battery valves it would be interesting to see how a magnet affects it.

Ian here is a wild guess.
The 1S5 is a diode pentode and I guess the diode is the audio detector.

Distortion of rectification transfer curve by the magnetic field?
Run a sine modulated RF signal into the set and watch the audio sine wave from the output valve clip/distort with the presence of the magnet.

Fred.

If the anode was steel you'd think it would block the magnetic field within. But maybe the anode on battery valves is aluminium... I don't know. Normally you can tell because a good magnet can pick up a valve off the table, or at least make it roll around.