Power tube and speaker magnet proximity

[pjd3]

Just want to mention that it was good to come around to some info and experience on issues that can arise with Power tube and speaker magnet proximity.

In my current build I came face to face with placing a pair of EL34's that seemed to have no choice but to find themselves very close to the speaker magnet. Without knowing if or how much they would interfere with each other, there was a nagging drive to find out. While there seems to be a great range of affect in that situation, there was reported incidents of the proximity causing significant deterioration of the amps tone that was both confirm-able with varying the proximity and hearing the difference while also viewing the distortion of a sine wave applied to the input and viewing the output at the speaker. That was quite enough for me to consider that the relative smallness of my donor cabinet could introduce this if I wasn't cognizant of the possibility.

But, I took some advice to reposition the speaker baffle as far forward in the cabinet as possible while also understanding that the electro-magnetic profile from a speaker is not omni or symmetrical, and that a different 90 degree rotation of the speaker can and has minimized this effect to inaudibility and non-corruption of intended signal.
Plus, it seems like if all of these attempts still rendered an effected signal due to proximity, that there sill could be another option to install a shield of sorts, like a copper/zinc 770 alloy between the speaker and tubes and ground it.

I thought I'd create this thread incase anyone comes into a similar situation where chassis real estate is at a bit of a minimum, and your inclined to make the build work since the transformers, chassis and cabinet are very usable and money saving. That can be a big help when you still have 4 dependants and the only main house income! Not complaining, just making room for everything and everyone.
Thank you to everyone that passes on their info and experience for these builds. Its scary to think that there will be some tone sucking issue who's source is not at all apparent!

Best,
Phil D.

[xtian]

Hmm. Anyone have a nice big neodymium magnet and a scope? I'd like to see if you can actually deflect the electron beam like you can on a CRT tube. I'll try it later, but I only have tiny (<1cm) magnets

[Stevem]

It would be a interesting test out, but I think the electrons are traveling to short of a distance to be defected like they are in even a small length CRT in a scope.

I would also think that if electron’s where being manipulated that it would be more likely below a certain level of V+ .

Also let’s not forget that the speakers magnetized level is being changed as the energized voice coil is moving thru it.

Let us know what you find .

[pjd3]

Whatever speaker I end up using, I figure there will be at least an inch of clearance between the 2 EL34's and the speaker magnet. I'll probably be very inclined to inject a sine wave and compare the input to the amp speaker output as I vary the orientation of the chassis from way out, to all the way in, and see it there are any changes to the output both on the scope and listening. I hope not but, if there is, I'll try some of the reported resolves that others have mentioned in various posts on this forum.

and I'll be happy to report back when that time comes!

Best,
Phil D.

[martin manning]

A quick experiment tracing curves on a 6L6GC with and without the presence of a 2 x 3 cm neodymium magnet placed 17mm (one toothpaste cap) from the tube envelope, at approximately mid-plate height:
I tested with the magnet axis in-line with the electron flow (magnet axis aligned with the flanges), and perpendicular to the electron flow (90 degrees from the flanges), with both north and south poles placed nearest the tube. The magnetic field results in a rounding of the knee for all grid voltages, which is most evident when the magnet axis is perpendicular to the electron flow, with the south pole of the magnet nearest the tube (bottom image). It doesn't appear that there is any significant effect on the gm or ra, since the curves are not noticeably affected to the right of the knee. Moving the magnet closer would most likely increase the distortion of the curves, but I didn't want to risk physical damage to the tube.

[xtian]

The magnetic field results in a rounding of the knee for all grid voltages

You're on top of it, Martin! Is this effect something we could hear while playing?

[martin manning]

Is this effect something we could hear while playing?

The curves look a little more pentode like with the magnetic field present, but they're not nearly as rounded as a true pentode like an EL34. Clearly the magnetic field changes the curves to some extent, and so there will be some effect on the sound. My guess in this case is you wouldn't hear it.

Back to the original question, A speaker is constructed so that the magnetic field is mostly contained and directed across the voice coil, but there will obviously be some leakage (see image below from Wikimedia Commons https://commons.wikimedia.org/wiki/File ... ection.svg). The same is true of transformers, which power tubes are often near. One could measure distortion with the amp chassis mounted in the cab with the speaker vs. out on the bench and see if there is any significant difference.

[pjd3]

Now that was an ambitious follow up test to my post Martin. Kudos, nicely done and very telling of some effect regarding the proximity.

I don't have the level of software testing that you do but, when its time I'll be able to at least look at waveforms for any deviation at varying proximity. Should I detect any difference in the output waveform at closest proximity I would probably try rotating the speaker 90 degrees and see how those different positions effect the result.

It would sure be nice though if there was no or little difference with the chassis pulled out and away, or all the way in and buttoned down. As it is now, I would prefer a typically full depth ceramic speaker like the Scumback BM75 (5.125") but I might be fairly content with say, a NEO Creamback (Depth = 4.8"). The V30 that came with the amp was at a 5.3" depth and pushing the EL34's about a 1/4" back from vertical. There is a little more chassis room if I wanted to elongate the socket holes that I already bored, and I'll do that If I really have to but, don't consider that too optimal. On person with this issue was able to make the problem disappear with a single 90 degree rotation of the speaker.
I hope I get that kind of result!

Thank you again Martin, well shown on your part, I respect your tenacity!

Best,
Phil

[trobbins]

Maybe Martin has just shown how modders can install a mechanical 'tone' control - sort of like Theremin control.

[_Steve]

A quick experiment tracing curves on a 6L6GC with and without the presence of a 2 x 3 cm neodymium magnet placed 17mm (one toothpaste cap) from the tube envelope, at approximately mid-plate height:
I tested with the magnet axis in-line with the electron flow (magnet axis aligned with the flanges), and perpendicular to the electron flow (90 degrees from the flanges), with both north and south poles placed nearest the tube. The magnetic field results in a rounding of the knee for all grid voltages, which is most evident when the magnet axis is perpendicular to the electron flow, with the south pole of the magnet nearest the tube (bottom image). It doesn't appear that there is any significant effect on the gm or ra, since the curves are not noticeably affected to the right of the knee. Moving the magnet closer would most likely increase the distortion of the curves, but I didn't want to risk physical damage to the tube.

Wow I just have so much respect for people like you. Serious kudos!

With my extremely limited grasp of physics or anything vaguely intellectual I wonder if the original issue is more affected by a MOVING magnetic field than a static one? Current induction, right-hand rule etc...

[martin manning]

For sure, charged particles moving through a moving magnetic field will be affected by it. Maybe a magnetic tremolo is possible? I don't know how much the field around the speaker magnet is affected by the voice coil current, negligibly perhaps, but there could be some feedback happening there.

[_Steve]

I have actually thought about a magnetic tremolo in the past! Magnet on a motor next to a coil. Im not sure how sinusoidal it'd be