Shielding power transformer or shielding power transformer + PSU PCB

[JohnRoberts]

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

it would just change the axis or direction of the hum leakage

[edit- if you can experiment with transformer location/orientation /edit]

JR

 

[abbey road d enfer]

So no good, I guess.

That's right.
I used the same sheet metal that was used for chassis, commonly called mild steel (less than 2% carbon).
Cold-rolled gives a better aspect than hot-rolled but there is no significant difference regarding magnetics.

 

[Audio1Man]

One of the tests on new samples of the toroid TX’s (also other types of TX) I use a plastic dynamic as a pickup connected to a voltmeter and search around the core to find MAX & MIN lobes.

As for the “SHORTED TURN PROBLEM, CENTER BOLT” I resolved this when I started working for BGW. They had a new amplifier in development that was using a toroid and it welded the top cover to the bolt. I took the toroid and filled the center with epoxy and then counterbored the center, added a flat washer, split lock washer, hex nut and shorter bolt. The amplifier went to the show, next show several other mfg’s had copied it.

Duke

 

[Matt Syson]

The external 'bands' used by a custom transformer manufacturer I used called it Grain Orientated Silicon Steel which is the strip they use for the core. If you like exercise (and cut fingers) you can unwind a burned out toroidal and get some from there. Cheap standard transformers use the least steel core they can get away with (cost) and run the magnetising current high so it is close to saturation. If the core dores start to saturate the radiated field gets much greater and if it gets into yopur audio path sounds nasty, like distorted mains hum rather than a gentle 50/60Hz 'sine wave'

 

[CJ]

that cold rolled sounds good.

offshore stuff usually uses non grain core, us give strip wound 014 sil gr or.

grain keeps the flux close due to increase perm

once lion out of cage, hard get him back in

 

[abbey road d enfer]

The external 'bands' used by a custom transformer manufacturer I used called it Grain Orientated Silicon Steel which is the strip they use for the core. I

GOSS has lower conductivity than standard mild steel. It is preferred for cores because it reduces eddy current losses, but for "shorted-turn shielding", it's counterproductive.

Cheap standard transformers use the least steel core they can get away with (cost) and run the magnetising current high so it is close to saturation. If the core dores start to saturate the radiated field gets much greater and if it gets into yopur audio path sounds nasty, like distorted mains hum rather than a gentle 50/60Hz 'sine wave'

When I was professional, I had the privilege of having custom-made xfmrs. I routinely asked the winders to use a 30VA rated core to produce 15VA xfmrs. Running xfmrs at 8000 gauss 0.8T makes them noticeably quieter.

 

[JohnRoberts]

When I was professional, I had the privilege of having custom-made xfmrs. I routinely asked the winders to use a 30VA rated core to produce 15VA xfmrs. Running xfmrs at 8000 gauss makes them noticeably quieter.

When I was "professional" using custom made transformers I didn't have the stones to ask for 2x core in my sundry high volume power transformers. My employer routinely used trainloads of power transformers and long before my time had "cheap" hammered into the transformer vendor's design ethos. Which is good because I am cheap too, but there were a small handful of SKUs where I got them to be a little more generous. I recall one big console power supply where the universal transformer needed to hang at 50 Hz as well as be low noise. I got them to make one heavy rock for that PS.

JR

 

[CJ]

Peavey used to weld base plates on their transformers, solid as heck but expensive, i bet it kept lam rattle down, i wondered about core loss but i guess no biggy otherwise they no do,

i will be gosh darn Duke, your right, everybody is using the epoxy in the middle trick,

looky here from the cavern club, #705-18729 but i bet John knew that,

no wait, not expensive to weld, that's a mig bead i can see the splatter, they can train anybody to squeeze the trigger, but if were talkin bout stackin dimes, i want 30 an hour, plus full bennies because i won't be able to drive at night,,

 

[abbey road d enfer]

So no good, I guess.

That's right.
I used the same sheet metal that was used for chassis, commonly called mild steel (less than 2% carbon).
Cold-rolled gives a better aspect than hot-rolled but there is no significant difference regarding magnetics.

 

[JohnRoberts]

Peavey used to weld base plates on their transformers, solid as heck but expensive, i bet it kept lam rattle down, i wondered about core loss but i guess no biggy otherwise they no do,

i will be gosh darn Duke, your right, everybody is using the epoxy in the middle trick,

looky here from the cavern club, #705-18729 but i bet John knew that,

no wait, not expensive to weld, that's a mig bead i can see the splatter, they can train anybody to squeeze the trigger, but if were talkin bout stackin dimes, i want 30 an hour, plus full bennies because i won't be able to drive at night,,View attachment 105459

like I said trainloads of heavy transformers...

About once a month the transformer vendor sent down an engineering team of maybe three guys. I remember the senior designer, they called him "beans" I never knew how he got that nickname but he really knew his stuff about transformers. I only used a handful of truly novel transformers and whenever I could I just used something that was already in the system. When I was in meetings with them I tried to listen and not embarrass myself by asking stupid questions.

JR

 

[warpie]

Is there any reasonably easy way (i.e not in an $1 million lab) to measure the magnetic field?. I mean an easy way other than listening tests. Maybe a 'scope with some DIY probes? Or even an audio interface + FFT software and some diy probes?

 

[Script]

Something with a microphone? Or guitar pickup ?

 

[ccaudle]

Is there any reasonably easy way (i.e not in an $1 million lab) to measure the magnetic field?. I mean an easy way other than listening tests. Maybe a 'scope with some DIY probes? Or even an audio interface + FFT software and some diy probes?

If you do not need to have calibration against a known level, but only need to make relative comparisons of different positions, you should be able to use a small coil of wire connected to an oscilloscope probe.
Signal and Noise Measurement Using Magnetic Field Probe (pdf)
Probing the magnetic field
Near field magnetic probes
EMI debugging with oscilloscopes
$10 DIY probe vs. $300 probe (YouTube eevblog)

 

[abbey road d enfer]

Is there any reasonably easy way (i.e not in an $1 million lab) to measure the magnetic field? I mean an easy way other than listening tests. Maybe a 'scope with some DIY probes? Or even an audio interface + FFT software and some diy probes?

Indeed, evaluating stray fields is relatively easy, but the big unknown is how your circuit is sensitive to them.
A few hints: Disturbance is proportional to loop area, so making the circuit as compact as possible is always a good thing, but if you attach long wires (or long PCB tracks) you're back to square one.
Sensitivity is position and angle dependant. Both the offending xfmr and the sensitive circuit.
A piece of magnetic material can shield, but it can also deviate flux in an unfavorable way.

 

[Audio1Man]

Is there any reasonably easy way (i.e not in an $1 million lab) to measure the magnetic field?. I mean an easy way other than listening tests. Maybe a 'scope with some DIY probes? Or even an audio interface + FFT software and some diy probes?

Look @ #23, cheep plastic dynamic mic

 

[JohnRoberts]

The true test is playing with transformer location/orientation in the presence of you actual circuit.

JR

 

[CJ]

run your I/O ports next to each other to minimize flux loop,

 

[NOON]

Transformer shielding in a Rotel RB-1000 power amp currently open on the bench.

 

[NOON]

Also found this guy mummified behind the front panel. Bug in the system or the secret of a great amplifier?

 

[Script]

I used the same sheet metal that was used for chassis

Didn't get the hint the first time. Could recycle indeed
Now need to find a way to cut and bend over here that's not too painful.
(Nearby car repair shop might have the tools needed).