Thinking on how to reduce the power line noise and harmonics across my system. There is so much misinformation out there, like shielded power cords. Copper or aluminum foil does not shield magnetic waves, so totally BS. Even my good DIY inter-connection cables are just multi-layer copper strands, not high permutability steel. ( Belden stranded 75 Ohm)
Is there wire that is shielded against LF?
I can find thin nu-metal sheets I can use for inside box shielding, but surprised I can't find "utility" boxes. Only tiny ones for little audio or switcher transformers.
Does twisting a power cord, say 5 or 6 per-inch is possible with zip cord, reduce the magnetic radiation to any useable effect?
Is there a source for small toroid power transformers that are encased one can afford? Or even small e-core. Like a coupe amps, 36V ct
Getting new respect for remote supplies. Tempting to pull stuff apart and make a central rectified supply. Maybe a bigger switcher and some buck regulators to drop the outputs to what each board needs. Space is a pretty good shield too.
Is there wire that is shielded against LF?
I can find thin nu-metal sheets I can use for inside box shielding, but surprised I can't find "utility" boxes. Only tiny ones for little audio or switcher transformers.
Does twisting a power cord, say 5 or 6 per-inch is possible with zip cord, reduce the magnetic radiation to any useable effect?
Is there a source for small toroid power transformers that are encased one can afford? Or even small e-core. Like a coupe amps, 36V ct
Getting new respect for remote supplies. Tempting to pull stuff apart and make a central rectified supply. Maybe a bigger switcher and some buck regulators to drop the outputs to what each board needs. Space is a pretty good shield too.
Reduce power line noise with a mains filter, some of the best are fully enclosed chassis mounting ones.
I think this is an interesting topic...more than likely it’s been gone over many times...perhaps not a dedicated thread?
I have not pursued magnetic shielding other than building a remote power supply for a phono amp. Even then...it wasn’t realistic to place the supply box far away from the circuit box.
I build my own cases and find aluminum easy to work with and wood just as easy.
I almost always cover the inside of the wood with copper foil I buy at Home Depot (normally used under siding for exterior transitions). I accept whatever RFI or other protection it buys me.
I also separate my sensitive electronics inside my cases with aluminum barriers accepting whatever additional benefits it might give me.
Mumetal seems the only real solution available for magnetic shielding. You can buy it in sheets, but it is expensive as you pointed out. Plus, my understanding is over working it by machining or forming it to much can reduce its shielding properties. Additionally, any enclose you make is supposed to be sealed to my understanding.
I found this article very informative:
The Importance of EMI Shielding
It made me think about possibly trying a coating on my dividers that isolate my power supplies.
Home Depot has a high iron paint listed on their website. I am tempted to buy some to try coating a DIY enclosure for a transformer. I realize there are flaws to the idea. I don’t know the exact metal content of the paint, if scratched the shield is corrupted, it can oxidize so would have to be top coated etc. Still, I wonder if it might be worth a try.
https://www.homedepot.com/p/Modern-Masters-1-gal-Metal-Effects-Oxidizing-Iron-Paint-ME208GAL/207173271
In the past I have made DIY covers for external chassis mounted torroid transformers out of 4” PVC pipe caps and a few other bits I bonded with it to make an anchor system to hold it and the transformer down. I only ever lined it with copper foil, again accepting whatever shielding it might provide. I suppose I could try taking one and coating it. However, I’m not sure what testing I could do to establish if it was at all successful.
I have not pursued magnetic shielding other than building a remote power supply for a phono amp. Even then...it wasn’t realistic to place the supply box far away from the circuit box.
I build my own cases and find aluminum easy to work with and wood just as easy.
I almost always cover the inside of the wood with copper foil I buy at Home Depot (normally used under siding for exterior transitions). I accept whatever RFI or other protection it buys me.
I also separate my sensitive electronics inside my cases with aluminum barriers accepting whatever additional benefits it might give me.
Mumetal seems the only real solution available for magnetic shielding. You can buy it in sheets, but it is expensive as you pointed out. Plus, my understanding is over working it by machining or forming it to much can reduce its shielding properties. Additionally, any enclose you make is supposed to be sealed to my understanding.
I found this article very informative:
The Importance of EMI Shielding
It made me think about possibly trying a coating on my dividers that isolate my power supplies.
Home Depot has a high iron paint listed on their website. I am tempted to buy some to try coating a DIY enclosure for a transformer. I realize there are flaws to the idea. I don’t know the exact metal content of the paint, if scratched the shield is corrupted, it can oxidize so would have to be top coated etc. Still, I wonder if it might be worth a try.
https://www.homedepot.com/p/Modern-Masters-1-gal-Metal-Effects-Oxidizing-Iron-Paint-ME208GAL/207173271
In the past I have made DIY covers for external chassis mounted torroid transformers out of 4” PVC pipe caps and a few other bits I bonded with it to make an anchor system to hold it and the transformer down. I only ever lined it with copper foil, again accepting whatever shielding it might provide. I suppose I could try taking one and coating it. However, I’m not sure what testing I could do to establish if it was at all successful.
I’ve tried this one time and it worked pretty well for me.
I went to a Machine shop and asked them for a small bucket of iron shavings from one of their lathes and milling machines. Had to be iron. From a old engine block from pre 70’s
As fine as possible. I then mixed it into a thick paint, and used that to paint the inside of a old tube amp that I had. This was a super noisy amp, after the paint job, the noise was nearly gone. I also have one of those magnetic sweeping pens for finding noise and a/c currents. It went from beeping as soon as I got 1’ away from the amp , to nearly impossible to get it to beep at all
I went to a Machine shop and asked them for a small bucket of iron shavings from one of their lathes and milling machines. Had to be iron. From a old engine block from pre 70’s
As fine as possible. I then mixed it into a thick paint, and used that to paint the inside of a old tube amp that I had. This was a super noisy amp, after the paint job, the noise was nearly gone. I also have one of those magnetic sweeping pens for finding noise and a/c currents. It went from beeping as soon as I got 1’ away from the amp , to nearly impossible to get it to beep at all
It would be much easier to say something if you know your noise well. What exact frequency that noise is, etc. Do you have a spectrum of your system output?
Is this noise concentrated or is it distributed across a wide frequency range (white noise, etc)?
VoV, 60, 120, 180... Power line noise and not through the regulator, not ripple, but magnetically induced which is why aluminum and copper are of no use. They are fine for RF and I have that handled. This is not hash from the power line.
Looks like strips of mu-metal to put a band around a toroid transformer works to some extent. A secondary of copper will also help any radiated diode or secondary ringing. Just a divider in the case seems to be a common approach. I want to get my harmonics from the low 60's to below 100. Another solution is of course SMPS as a pre-reg so I don't have that big transformer there. My crossover is currently that way, but I need to convert it to a +/- tracking supply so going back to linear.
I have not opened it up yet, but my Shiit Asgard has more power harmonics than I Think it should. Kind of making me rethink if I should have bought the JSD Atom DAC and Preamp as they feed everything from remote 16V supplies.
Iron filings paint is an interesting idea. Might be worth some experimenting. Thanks.
I do remember something about mu-metal and bending it. I don't remember if annealing it helped. More to research.
Looks like strips of mu-metal to put a band around a toroid transformer works to some extent. A secondary of copper will also help any radiated diode or secondary ringing. Just a divider in the case seems to be a common approach. I want to get my harmonics from the low 60's to below 100. Another solution is of course SMPS as a pre-reg so I don't have that big transformer there. My crossover is currently that way, but I need to convert it to a +/- tracking supply so going back to linear.
I have not opened it up yet, but my Shiit Asgard has more power harmonics than I Think it should. Kind of making me rethink if I should have bought the JSD Atom DAC and Preamp as they feed everything from remote 16V supplies.
Iron filings paint is an interesting idea. Might be worth some experimenting. Thanks.
I do remember something about mu-metal and bending it. I don't remember if annealing it helped. More to research.
Seems there is a big market to the foil hat contingent so the amount of BS out there is as bad as speaker cables.
As far as I know, mu metal is the only real shield for LF magnetic. Twisting cable also helps a lot, but thats not shielding its more like common mode rejection in balanced inputs.
A belt of thick copper for toroids.
I saw it, claimed to be effective, but did not try.
There exist round Mu metal cases for toroids.
I saw it, claimed to be effective, but did not try.
There exist round Mu metal cases for toroids.
Last edited:
Yes,
* Mu-metal for LF, as in power and harmonics. Magnetic fields. 60, 120 etc. My main interest as I mentioned. Soft steel is not as good, but works some. Enough that if I built a remote supply, I would use a steel box. As I understand, not that effective when the transformer is 2 inches from the circuit.
* Copper and aluminum is for RF. Virtually useless for LF. So a copper band would be effective in shielding ringing caused by diode switching . A good idea for sure, but not a cure for LF. ( but so would a snubber, also a good idea as it knocks down noise being injected back into the mains)
*Twisting is not shielding, but it reduces the size of the radiated field. I think.
*Thicker the better.
I have only found manufactures who would design and built 10,000 cases, not any ones I can buy one of. Some for the tiny SMPS transformers or audio transformers, not for a 2A 36V toroid. I have not found a little "Bud Box" made of mu-metal. Maybe just not found them yet, I have not found any transformers that come pre-potted. I see many in some molded blue fin-things, but suspect that is plastic. No description so who knows. Looking at pictures of higher end stuff, I do see just flat shields between power supply and circuit boards. Transformer in my Nakamichi preamp is in a can.
MC, I whish a lot more people read your signature line! I might add, "every 10C is half life" or "135 kills"
Coffee is ready. More searching. I need to put in secondary lights over my work bench that are incandescent so I can turn off the LED and fluorescents when testing.
* Mu-metal for LF, as in power and harmonics. Magnetic fields. 60, 120 etc. My main interest as I mentioned. Soft steel is not as good, but works some. Enough that if I built a remote supply, I would use a steel box. As I understand, not that effective when the transformer is 2 inches from the circuit.
* Copper and aluminum is for RF. Virtually useless for LF. So a copper band would be effective in shielding ringing caused by diode switching . A good idea for sure, but not a cure for LF. ( but so would a snubber, also a good idea as it knocks down noise being injected back into the mains)
*Twisting is not shielding, but it reduces the size of the radiated field. I think.
*Thicker the better.
I have only found manufactures who would design and built 10,000 cases, not any ones I can buy one of. Some for the tiny SMPS transformers or audio transformers, not for a 2A 36V toroid. I have not found a little "Bud Box" made of mu-metal. Maybe just not found them yet, I have not found any transformers that come pre-potted. I see many in some molded blue fin-things, but suspect that is plastic. No description so who knows. Looking at pictures of higher end stuff, I do see just flat shields between power supply and circuit boards. Transformer in my Nakamichi preamp is in a can.
MC, I whish a lot more people read your signature line! I might add, "every 10C is half life" or "135 kills"
Coffee is ready. More searching. I need to put in secondary lights over my work bench that are incandescent so I can turn off the LED and fluorescents when testing.
Steel shield.
Capot de Blindage Acier pour Transformateur Torique 90x40mm - Audiophonics
1millimeter thick.
Capot de Blindage Acier pour Transformateur Torique 90x40mm - Audiophonics
1millimeter thick.
In my active filters I paid alot of attention to screening well, toroid belly band double sided pcb as screens, grounded of course, has it helped? Well soundcard testing was very nice, 112 db noisefloor which is the soundcard limit.
Removed windings from this to lower secondary voltage and made a belly band to
Removed windings from this to lower secondary voltage and made a belly band to
Attachments
Nice belt.
Copper belly band deserves thicker.
Thicker copper is better and best soldering to make a very best short circuit ring is a must.
I doubt the ground braided strap over the transformer is of any use.
Copper belly band deserves thicker.
Thicker copper is better and best soldering to make a very best short circuit ring is a must.
I doubt the ground braided strap over the transformer is of any use.
The toroid "belly band" is, in essence, one plate of a capacitor. The other plate is the transformer coil winding on its surface. (and, to be pedantic, you can calculate the value of the capacitance knowing the area of the belly band, its distance from the winding and the dielectric constant of the wire insulation). It's all discussed in "Grounding and Shielding Techniques in Instrumentation", by Ralph Morrison. What this means -- a current will flow from the outer layer of the transformer to the belly band in inverse proportion to the reactive impedance of the belly band.
Morrison goes on to say that unshielded or one-shield transformers may have a "preferred power cord orientation".
Twisting cables? "When conductor pairs are twisted along their length the resulting magnetic pickup is significantly reduced." (Morrison 8.11)
and keep cable runs as short as possible.
There is much more than that.
The "belly band" is a one turn coil that shorts stray fields.
It's effectiveness strongly relates to the thickness.
From a forum:
More, here:
Transformer shielding
Last edited:
The cheapest source of copper banded EI transformers is blown up equipment. For example Peavey PA amps often have copper banded transformers, and are a plague on the used market because they are too heavy to carry around to a different bar every night when class D amps weigh half as much. My PV-1.3k had a copper banded transformer, the PV-4c did not.
Sheilding in commercial equipment is often done by building a sheet steel cage around the AC parts, or around the high gain parts to prevent the magnetic flux from getting in. For example my dynaco PAS2 preamp has a steel cased transformer, also a separate metal box around the wiring to the high gain vacuum tubes. No line cord grounding in this 1959 design, it had a 2 pin AC plug.
The PV-1.3k has a steel bulkhead between the copper wrapped transformer and the high gain input boards. Also a sheet steel box around the AC cord entry, the rfi filters nearby.
A Peavey MMA-875t mixer amp does not have a copper band around the transformer, but it has a steel cage around it. The transformer is located near the 1 gain tone control boards. The high gain input boards are at the other end of the amp, in their own grounded steel box.
Note also transformer orientation can affect how flux affects the high gain stages.
I got rid of a 40 db hum problem in a herald RA-88a mixer by taking the transformer out of the case entirely. The transformer was steel wrapped, but it and the AC switch were right next to the 50x gain RIAA stage. Bad packaging. I replaced the transformer with a 18 vdc wall transformer, followed by a double pi filter in the box with RF killing toroid in series. Single pi filter only dropped the hum so much. Another 2 ohm resistor followed by another 1000 uf capacitor got the hum low enough I couldn't hear it.
Happy hum hunting.
Sheilding in commercial equipment is often done by building a sheet steel cage around the AC parts, or around the high gain parts to prevent the magnetic flux from getting in. For example my dynaco PAS2 preamp has a steel cased transformer, also a separate metal box around the wiring to the high gain vacuum tubes. No line cord grounding in this 1959 design, it had a 2 pin AC plug.
The PV-1.3k has a steel bulkhead between the copper wrapped transformer and the high gain input boards. Also a sheet steel box around the AC cord entry, the rfi filters nearby.
A Peavey MMA-875t mixer amp does not have a copper band around the transformer, but it has a steel cage around it. The transformer is located near the 1 gain tone control boards. The high gain input boards are at the other end of the amp, in their own grounded steel box.
Note also transformer orientation can affect how flux affects the high gain stages.
I got rid of a 40 db hum problem in a herald RA-88a mixer by taking the transformer out of the case entirely. The transformer was steel wrapped, but it and the AC switch were right next to the 50x gain RIAA stage. Bad packaging. I replaced the transformer with a 18 vdc wall transformer, followed by a double pi filter in the box with RF killing toroid in series. Single pi filter only dropped the hum so much. Another 2 ohm resistor followed by another 1000 uf capacitor got the hum low enough I couldn't hear it.
Happy hum hunting.
Last edited:
If you're seeing 120, 180, 240, 300 ... Hz, you're dealing with rectification crap and not "mains noise". Solution: Don't put your input stage (or wiring) near the rectifier. Distance beats shielding on price every day of the week. 🙂
Otherwise, you basically have two option: Attenuation and electrostatic shielding. A thick metal sheet will provide some attenuation at 50-60 Hz. The skin depth in steel is around 1 mm for 60 Hz if I recall correctly vs about 5 mm in aluminum, so use steel. This works well against electromagnetic coupling.
For inductive coupling, reduce the loop area of the inductor.
For capacitive coupling, electrostatic shielding is your friend. Think Faraday cage.
Depending on the noise source, it may be more feasible to shield the circuit being offended by the "noise". You can get shielding cans intended for soldering onto the circuit board. They're used a lot in the RF world, but would be effective in audio as well - at least for capacitive coupling.
Tom
Otherwise, you basically have two option: Attenuation and electrostatic shielding. A thick metal sheet will provide some attenuation at 50-60 Hz. The skin depth in steel is around 1 mm for 60 Hz if I recall correctly vs about 5 mm in aluminum, so use steel. This works well against electromagnetic coupling.
For inductive coupling, reduce the loop area of the inductor.
For capacitive coupling, electrostatic shielding is your friend. Think Faraday cage.
Depending on the noise source, it may be more feasible to shield the circuit being offended by the "noise". You can get shielding cans intended for soldering onto the circuit board. They're used a lot in the RF world, but would be effective in audio as well - at least for capacitive coupling.
Tom
Some confusion here. I am talking about H-field shielding, not E-field. E-field is easy.
Copper and aluminum block E-field. Mu-metal and that ilk are for H-field. Looks like I can get used and surplus sheets from e-bay not too terrible to play with.
Little 3/4 inch shield cubes are not of any help. As I said, I went looking for larger "bud-box" kinds of things Inches on a side. Can't find any. Steel and aluminum, fine, not mu-metal.
Yes, distance works, but what I am fighting, the supply is inside the case. Orientation of the transformer is not an option.
Ripping transformer and rectifier out may be the best solution, but I want to see what I can do in-place. I am getting some hints by looking at pictures of analyzers and scopes with the covers off.
Copper and aluminum block E-field. Mu-metal and that ilk are for H-field. Looks like I can get used and surplus sheets from e-bay not too terrible to play with.
Little 3/4 inch shield cubes are not of any help. As I said, I went looking for larger "bud-box" kinds of things Inches on a side. Can't find any. Steel and aluminum, fine, not mu-metal.
Yes, distance works, but what I am fighting, the supply is inside the case. Orientation of the transformer is not an option.
Ripping transformer and rectifier out may be the best solution, but I want to see what I can do in-place. I am getting some hints by looking at pictures of analyzers and scopes with the covers off.