Shielding power transformer or shielding power transformer + PSU PCB

[wrentema]

Hi.
Just a quick question. I'm building the Fripholm TG1 zener limiter. I have some aluminium strips lying around I could use for shielding. And was wondering if it's better to shield...:

(A) ...the whole power supply (incl transformer and PSU PCB)
or
(B) ...if I should shield just the toroidal power transformer.

I see both. And was wondering if there's a "better" version.

It's 1mm thick Aluminium. Should be fine for shielding right? I know MU is better but money..

(added a little drawing on my STA Level build.. Just to show what I mean). The TG1 PSU is way smaller.

Thanks!

 

[CJ]

Toroidal should not need a shield as all the flux cancels out.

If you do try a shield I would use steel

 

[warpie]

Toroidal should not need a shield as all the flux cancels out.

If you do try a shield I would use steel

Interesting. I've seen many designs where the toroidals are shielded.

 

[JohnRoberts]

In theory toroids have contained magnetic fields, but windings are not perfect so there can be magnetic lobes from these discontinuities. Can you experiment with the actual transformer?

JR

 

[klem]

Toroidal should not need a shield as all the flux cancels out.

If you do try a shield I would use steel

my api 5500 got a couple db quieter when i rotated the PSU toroidal a bit, but the noise floor dropped 11dBFS* when I put it in its own remote enclosure.
*i don't remember how i had my lynx converter calibrated for dBFS conversion.

agreed about steel - a poor man's version of mu that i've always gone by is just a tin can with some kentucky chrome. that made my send n blend build a bit quieter as well, i don't remember the values offhand like the API. once i'm done with some tube mic PSUs I may try some mu strips, the PSUs are just too damn pretty to get my aforementioned dirty but effective shielding treatment.

 

[dmp]

Aluminum has a very low permeability so does not shield magnetic interference well. It may be used for emf shielding in some applications.
For shielding around a transformer use a high permeability material.
It needs to be a complete path as the magnetic flux must complete a path around the toroid, so a circular shape is better than a wall.

 

[abbey road d enfer]

Also remember. two screens of thickness x are more efficient than one of thickness 2x. They don't need to be separated by much, like non-conductive material (air, plastic, cardboard...)of similar thickness.
I've often used shields made out of the same material than the chassis, some times thicker.
As others have mentioned, aluminium is not good at shielding magnetic flux, neither are copper, or many variants of stainless steel, or gold, as seen in some lunatic HiFi systems.

 

[wrentema]

already some great info! thanks. Will get some steel strips. And will just test and see if it makes a difference.

 

[warpie]

already some great info! thanks. Will get some steel strips. And will just test and see if it makes a difference.

Let us know!

 

[dmp]

Before you go to that trouble, just compare the measured noise (as a function of frequency) with the transformer in and out of the case. It looks like you could snake the wires through the holes in the side. If you have one, you could also use a second power supply to power the gear from a few feet away, so the transformer can be turned on and off in the case to compare the noise.
If you do see an increase in the noise @ 60/120Hz with the transformer in the case, try rotating it to see if it can be nulled.

 

[wrentema]

Before you go to that trouble, just compare the measured noise (as a function of frequency) with the transformer in and out of the case. It looks like you could snake the wires through the holes in the side. If you have one, you could also use a second power supply to power the gear from a few feet away, so the transformer can be turned on and off in the case to compare the noise.
If you do see an increase in the noise @ 60/120Hz with the transformer in the case, try rotating it to see if it can be nulled.

Will try! Just don't love external power supplies for practical reasons. But will try all things above and I'll report. It's a slow build so will take a few weeks maybe.

 

[dmp]

Just don't love external power supplies for practical reasons.

Don't BUILD it with an external PSU, TEST it with an external PSU.
Doing this method with preamps has helped me get to the place where I know an internal PSU performs as well as an external.

 

[Matt Syson]

Torodals are normally quite reasonable for radiation but the 'cheap' standard stock tend to run a high magnetising current so are nearer saturation which makes the radiated field considerably worse. The toroid is mostly 'OK' but the lead out wires create a discontinuity of radiated field so turning them will usually allow a bit of 'fine tuning'. Adding a 'goss band' (grain orientated silicon steel) like is used for the actual core wrapped carefully around the toroid helps 'mop up' stray field and might be as good as a full shielding can which would also trap the heat, although for a preamp you should not be creating much heat. You shoud test your unit with the transformer temporarily a decent (several feet) away so you know how 'good' (hum free) your design should be. Then try introducing the transformer to where you hope to put it.

 

[Audio1Man]

Like John said: The Torid has BIG LOBES where the lead wires enter/exit (non uniform spacing). Try to place the leads away from any pickup circuity loops. Many Toridal transformers use a shorting ring (Copper turn) around the outer core. By using a bigger core, the field can be reduced.

Duke

 

[neil.johnson]

I always get nervous with the words "shorting ring" anywhere near toroidals! You DO NOT want to short the retaining bolt, e.g., by the top half of the case touching it.

Neil

 

[JohnRoberts]

Like John said: The Torid has BIG LOBES where the lead wires enter/exit (non uniform spacing). Try to place the leads away from any pickup circuity loops. Many Toridal transformers use a shorting ring (Copper turn) around the outer core. By using a bigger core, the field can be reduced.

Duke

In my experience the hum from a decent toroidal transformer is pretty well managed by rotating it until the hum lobe is pointed somewhere harmless and away from sensitive circuitry.

I always get nervous with the words "shorting ring" anywhere near toroidals! You DO NOT want to short the retaining bolt, e.g., by the top half of the case touching it.

Neil

I have seen the one turn shorting windings (thin copper foil) used around transformers with other form factors (like EI, etc) and afaik they are just shunting the leakage magnetic flux to itself, so not much power to speak of. I suspect grounding the foil could add some electrostatic shielding, but I am not the transformer expert around here.

JR

 

[Script]

As others have mentioned, aluminium is not good at shielding magnetic flux, neither are copper, or many variants of stainless steel

What kind of steel would be good for this? Or where could one get it ?
I see generic stainless steel foil strip, for example, of say 0.5x40x1000mm, not expensive but also not that cheap -- especially if it's not really effective.

 

[CJ]

mu metal if your pockets are deep, but i think regular old hot rolled 20 ga would be fine.

 

[Script]

What I had found is cold-rolled and called SAE 304

(SUS304 in Japan, 18/8 in cookware industry, and European norm 1.4301).
Contains both chromium (between 18% and 20%) and nickel (between 8% and 10.5%), it seems.

So no good, I guess.

 

[warpie]

Would mounting the toroidal vertically (i.e. perpendicular to the PCB) help?