I see a lot of general rules of thumb but not much technical information on the matter- leaving it a bit of a guessing game as to whether a two box build is required.
Usually this turns into a two box build of the “better safe than sorry” variety.
Is there accepted methodology for distances in relationship to transformer size and type?
What about for cans and covers?
How does epoxy sealing play into things?
Or horizontal versus vertical mounting?
How much can be contained from a simple toroidal can? Or a divider?
What thickness? What material ideally?
What’s the difference in stray flux for a given increase in VA?
To what degree does an electrostatic shield mitigate the issue?
Let’s use a more sensitive circuit as an example: a tube phono stage. How would one go about determining how to build a successful 1 box design that’s free from transformer noise? (Ideally, before it’s built- not trial and error which is a well established method).
Is there a simple way to measure effects; and to separate supply from transformer issues?
Is there some kind of spreadsheet or graph out there to determine distances?
Helpful articles / links on the topic I’ll post below:
Power Transformers for Audio Equipment | audioXpress
Usually this turns into a two box build of the “better safe than sorry” variety.
Is there accepted methodology for distances in relationship to transformer size and type?
What about for cans and covers?
How does epoxy sealing play into things?
Or horizontal versus vertical mounting?
How much can be contained from a simple toroidal can? Or a divider?
What thickness? What material ideally?
What’s the difference in stray flux for a given increase in VA?
To what degree does an electrostatic shield mitigate the issue?
Let’s use a more sensitive circuit as an example: a tube phono stage. How would one go about determining how to build a successful 1 box design that’s free from transformer noise? (Ideally, before it’s built- not trial and error which is a well established method).
Is there a simple way to measure effects; and to separate supply from transformer issues?
Is there some kind of spreadsheet or graph out there to determine distances?
Helpful articles / links on the topic I’ll post below:
Power Transformers for Audio Equipment | audioXpress
Cans and covers are mainly cosmetic. Distance is your friend, followed by orientation. Experimentation is the only solution; to do more than this you would need a careful full field simulation which would need a full transformer model - which you won't have, because the manufacturer won't have it either.
Really? Even a steel toroid cover such as this?
CA-400 Steel Cover - AnTek Products Corp
From description: These toroidal transformer covers are specially designed to prevent any magnetic field or electrical noise interfere with the sensitive components in equipment or high-end audio products.
Antek has always seemed reputable to me....
I would think steel could minimize effects just like physical distance?
No one has put together any measurement data on this topic? Not even the manufacturers here?
CA-400 Steel Cover - AnTek Products Corp
From description: These toroidal transformer covers are specially designed to prevent any magnetic field or electrical noise interfere with the sensitive components in equipment or high-end audio products.
Antek has always seemed reputable to me....
I would think steel could minimize effects just like physical distance?
No one has put together any measurement data on this topic? Not even the manufacturers here?
From this thread: Magnetic shielding with aluminum
“Aluminum works to absorb magnetic fields. 250 mils of aluminum is as effective as 6 mils of mu metal or about 100 mils of stainless all roughly equal 6dB of attenuation at 60 Hz. I didn’t see anything in the text that would lead me to believe multiple shields would have any unexpected effect with low impedance low frequency fields so stacking spaced plates should not be a problem. “
And
“At audio frequencies and up, aluminum and copper are better, because the field induces surface currents, reflecting rather than transmitting it.”
So it seems lots of divergent opinions on the topic with not much data.
Dead link in that thread with lots of useful info re-located here:
https://www.sealingdevices.com/documents/EMI Shielding Theory Gasket Design Guide of Chomerics.pdf
“Aluminum works to absorb magnetic fields. 250 mils of aluminum is as effective as 6 mils of mu metal or about 100 mils of stainless all roughly equal 6dB of attenuation at 60 Hz. I didn’t see anything in the text that would lead me to believe multiple shields would have any unexpected effect with low impedance low frequency fields so stacking spaced plates should not be a problem. “
And
“At audio frequencies and up, aluminum and copper are better, because the field induces surface currents, reflecting rather than transmitting it.”
So it seems lots of divergent opinions on the topic with not much data.
Dead link in that thread with lots of useful info re-located here:
https://www.sealingdevices.com/documents/EMI Shielding Theory Gasket Design Guide of Chomerics.pdf
Last edited:
Speaking of which I too have a question. Lets say I have a mono mono amp,
it’s going to be 2 chassis but one chassis will only house 2 EI transformers. In this context is there any specific oreintation that I can apply to minimize coupling between the transformers or rather will it affect performance regardless of orientation. I recalled reading in tube thread that mounting chokes in opposite direction will null magnetic coupling, wonder if this is applicable to trafos.
Thanks
it’s going to be 2 chassis but one chassis will only house 2 EI transformers. In this context is there any specific oreintation that I can apply to minimize coupling between the transformers or rather will it affect performance regardless of orientation. I recalled reading in tube thread that mounting chokes in opposite direction will null magnetic coupling, wonder if this is applicable to trafos.
Thanks
Perhaps because of your previous statement?
“which would need a full transformer model - which you won't have, because the manufacturer won't have it either.”
?
You have not provided a single number either when asserting your claims, mind you.
I’m asking for some kind of technical data, which could inform rules of thumb regarding this sort of thing.
I’m not saying Antek is correct, just providing an example and why the issue is rather obscured and unclear for the average DIYer.
From what I gather it doesn’t seem like anyone has any reliable data one way or the other in reference to this issue for our purposes. Or it’s specific to a given build or instance and doesn’t provide insight for broader utility.
Or if someone does, it’s not clear to me.
This seems unusual in that capacitor sound and other fringe topics are beaten to death, yet this is an issue people have to deal with repeatedly in builds and has more of a scientifically accepted and measurable effect.
Last edited:
As usual it depends; any advice is relative as real transformers (and manufacturing) vary a lot.
Concentrating on the toroidal transformer it depends on the size, the symmetry of the windings and any saturation effects.
As said before, distance and orientation are crucial points, as (good winding practice assumed) the wire entries are usually the points/section where mainly external fields emerge from the toroidal.
The shielding belt made from core metal is effective but will obviously not help if the transformer is mounted vertically.
A fully closed can made from mumetall is able to deliver a strong attenuation but care is still needed in orientation as the wire entry points are potentially weak points. Unfortunately mumetall cans are quite expensive so if are able to do a diy can made from the thick alumium/copper that you´ve mention it will provide the same attenuation but maybe at lower costs.
A crucial point is to avoid saturation of the toroidal core - especially due to mains dc - as the external stray field will be much higher and will contain higher harmonics so coupling to external receivers will be much stronger.
I don´t know if you have access to an oscilloscope or something for spectrum analysis, if so i´d recommend to build some of the diy sniffing probes for EM fields to find out what a specific toroidal (under real conditions means with rectifiers, capacitors and load) will show wrt to stray fields.
Does a snubber reduce mechanical noise too?
Yes, I've done something similar that helps reduce coupling to the chassis. It's one transformer in a chassis of a number of remote power supplies that's a bit noisy, the fact it's in the corner of my room on the floor probably doesn't help.
Watch yourself for "shorted turn"
In my chassis, I use a piece of circuit board for a few purposes. One, being a support structurally for my mainboard pcb, and another as a "shield" between my transformers and the amp mainboard. Initially I had the center mounting bolt attached both to the chassis and to the "shield", resulting in a shorted turn. It ran soooo hot lol. I ended up needing to detach the center mounting bolt from my shield, however, the shield is still attached to ground and functioning perfectly as far as I can tell. As a note, I also use the Jensen line level transformers mounted towards the opposite side of the chassis about 10in away and experience zero hum or noise on my amplifiers. I'd imagine that if there were an issue, I'd be towards the front of the line in experiencing such, and I do not.
I would imagine that you could make yourself some sort of plate for the top of the transformer, some sort of large washer type plate and check your results, just watch for shorted turns.
S
In my chassis, I use a piece of circuit board for a few purposes. One, being a support structurally for my mainboard pcb, and another as a "shield" between my transformers and the amp mainboard. Initially I had the center mounting bolt attached both to the chassis and to the "shield", resulting in a shorted turn. It ran soooo hot lol. I ended up needing to detach the center mounting bolt from my shield, however, the shield is still attached to ground and functioning perfectly as far as I can tell. As a note, I also use the Jensen line level transformers mounted towards the opposite side of the chassis about 10in away and experience zero hum or noise on my amplifiers. I'd imagine that if there were an issue, I'd be towards the front of the line in experiencing such, and I do not.
I would imagine that you could make yourself some sort of plate for the top of the transformer, some sort of large washer type plate and check your results, just watch for shorted turns.
S
Addressing some more question from your opening post:
-) epoxy and other potting material do help, if the winding quality isn´t as good as it could be and therefore the windings are somewhat loose. It does not help against humming noise caused be any motion of the core.
Mechanical vibration energy transfer could even be better (making the humming more obvious) in case of solid potting material, so providing insulating material is a good idea.
-) Noise from the core is a matter of core production quality and saturation effects. Blocking any dc from the mains helps in the saturation department, quality of the core can be roughly evaluated by measuring the idling current without load.
-) the topic of heat was mentioned already, which is a valid concern, but usually not for home usage of power amplifiers
-) i think it was already mentioned too; interference is a matter depending on two variables (to keep things easy) the strength of the emitting source and the vulnerability of the receiver, so keeping any potential receiving area as small as possible will help a lot too
-) knowing about the emitted fields from a toroidal isn´t unfortunately possible in most cases, as DF96 already pointed out. Transformer made according to the same specs wrt to power, current, but made from different manufacturers can be quite different in their selfnoise and emitted fields
-) a brute force approach will imo help if hunting for the lowest possible interference, using rubber-bonded metal (or other damping material for mechanical insulation) and combining different canning options; first the belt made from core material, then a mu-metal can enclosing the transformer totally but made larger so that there is some distance between transformer and can inside and finally another enclosure made form the thick aluminium that you´ve mentioned before
-) combine that with orientation aiming for the lowest emitting in direction of the vulnerable parts of the circuits
-) remember that you only can provide attenuation for the emitted fields; beside thinking about the receiving area (made preferable as small as possible) any positive effect depends also on the gain of the receiving circuit
Epoxy sealing in general does nothing if it's being used to hold a cover. If it's potting material, it won't change magfield at all.
A conductive cover will try to exclude time varying magnetic field by the generation of eddy currents. the more conductive, the more shielding. Copper better than aluminum. A superconductor will totally exclude field penetration, but is a bit difficult to do in a living room..
A steel cover will do two things. First, it will have eddy currents and the resultant field will exclude some of the field. Second, it will try to route the magnetic flux through the cover material. It is basically trying to short the air path (by lowering the reluctance where the cover is.)
Using multiple steel paths will work quite well.
Mu metal can certainly work quite well as shielding. However, it will tend to saturate at very low fields, so will be of limited use if the transformer is leaking lots of flux. Typically, the cheaper units will have more flux escaping because the iron is being worked too hard. Toroids do not suffer that as much, but they will broadcast the parasitic inductance of the windings.
As to using steel as a boundary between circuits, there have been reported instances where a time varying flux being blocked by a magnetic wall can actually modulate the permeability of the wall, thereby causing a non linear mixing of signals, multiplication to be exact.
That said, many of the suggestions already given are quite good.
John
It is a difficult issue to quantify indeed - which is the frustration that appears to have started the thread.
As DF96 alluded, there has been effort put in to simulate the field around transformers - I have made an effort to search for such reports over the years. Most reports are (obviously) related to three phase transformers and practical issues that they face, especially for larger power distribution devices, like how to manage clamping and a nearby enclosure. Some reports have started with a simpler single phase transformer structure model in order to eek out particular issues like dynamic leakage inductance effects that affect inrush current. Most reports don't show simulation plots that are of much insight from a learning curve perspective. One simple simulation plot I located is in Fig 27 of the linked simple article on the topic of hum:
https://www.dalmura.com.au/static/Hum%20article.pdf
Some books show insightful external field diagrams - such as from a book of Nave's that I have - purloined in to Fig 29 of that article.
There are on-line books around that indicate the field disturbance introduced by shielding, especially for sensitive devices like microphone transformers. Lee's 1995 scanned book (Electronic Transformers and Circuits) shows such an example in Fig.153, and that book is a great tome of information and insight. http://www.tubebooks.org/Books/Lee_1955_Electronic_Transformers_and_Circuits.pdf
On the other side of the fence is how to make circuitry less susceptible to external fields.
Thousands of amplifiers have been able to suppress hum down to what many consider to be negligible levels. Which begets the issue of how do you practically go about that when starting a fresh project, even when following an existing schematic, or even with a kit of parts with a layout description, or even a kit with layout description and diagrammatic wiring layout as well. Not all diy amplifiers turn out to have the same hum performance.
The first and biggest hassle is to determine how to measure hum and how to differentiate or eliminate or localise hum contributions - of which transformer related layout and shielding is but just one contribution.
Apart from reading up widely on the topic, and trying to appreciate better how transformer hum may or may not be contributing to an output hum signal, there is nothing like practical testing in an amplifier using a step-by-step change approach and a means of measuring hum that is quantifiable and insightful.
As DF96 alluded, there has been effort put in to simulate the field around transformers - I have made an effort to search for such reports over the years. Most reports are (obviously) related to three phase transformers and practical issues that they face, especially for larger power distribution devices, like how to manage clamping and a nearby enclosure. Some reports have started with a simpler single phase transformer structure model in order to eek out particular issues like dynamic leakage inductance effects that affect inrush current. Most reports don't show simulation plots that are of much insight from a learning curve perspective. One simple simulation plot I located is in Fig 27 of the linked simple article on the topic of hum:
https://www.dalmura.com.au/static/Hum%20article.pdf
Some books show insightful external field diagrams - such as from a book of Nave's that I have - purloined in to Fig 29 of that article.
There are on-line books around that indicate the field disturbance introduced by shielding, especially for sensitive devices like microphone transformers. Lee's 1995 scanned book (Electronic Transformers and Circuits) shows such an example in Fig.153, and that book is a great tome of information and insight. http://www.tubebooks.org/Books/Lee_1955_Electronic_Transformers_and_Circuits.pdf
On the other side of the fence is how to make circuitry less susceptible to external fields.
Thousands of amplifiers have been able to suppress hum down to what many consider to be negligible levels. Which begets the issue of how do you practically go about that when starting a fresh project, even when following an existing schematic, or even with a kit of parts with a layout description, or even a kit with layout description and diagrammatic wiring layout as well. Not all diy amplifiers turn out to have the same hum performance.
The first and biggest hassle is to determine how to measure hum and how to differentiate or eliminate or localise hum contributions - of which transformer related layout and shielding is but just one contribution.
Apart from reading up widely on the topic, and trying to appreciate better how transformer hum may or may not be contributing to an output hum signal, there is nothing like practical testing in an amplifier using a step-by-step change approach and a means of measuring hum that is quantifiable and insightful.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.