All joking aside I seem to rember a discusion of this years ago.
The conclusion was that a thin metal shield like a tuna can won't
be verry effective as a shield. Besides it would hamper air flow
around the transformer and make it run hotter. And as was pointed out you should't have a lot of radiated fields from a toriod.
The conclusion was that a thin metal shield like a tuna can won't
be verry effective as a shield. Besides it would hamper air flow
around the transformer and make it run hotter. And as was pointed out you should't have a lot of radiated fields from a toriod.
Tuna cans are ideal for various things like holding cutting oil at the lathe. If you didn't do a good job of bypassing incoming RF on the mains cable, shielding the transformer might help contain it a bit. Or not. IMHO, toroids are over rated in their ability to magically contain stray magnetic flux. They still induce currents in nearby chassis metal, so they're not perfect, and magnetic shielding requires thicker metal and a full ring- just what you don't want with a toroid. The above mentioned heating is also an issue. I'd probably avoid the tuna can and rely on distance from sensitive circuitry, or go with a mu-metal shield supplied by the toroid manufacturer. OTOH, there's no harm in trying it. Just do some tests for temperature rise.
I was , I think, worrying about the short circuited winding. I was under the impression toroidal radiated more than other types of transformers and that was a good reason to put them in an outboard power supply. Are there any good posts on torodals? My searches here have had limited results. Thanks for the info and, the levity .
The metal of a typical tuna can has very low permability, which makes a lousy magnetic shield. Tuna cans work well from shielding tuna from the environment, not for magnetic shielding.
http://www.mushield.com/design-guide.shtml
As Eva has discussed in many previous posts, toriod emissions can be controlled by the way they are orientated when mounted on the circuit board.
E-core transformers can use a copper foil band (aka "hum band")installed around the outside of the transformer (parallel to the winding) and attached to ground that creates a current loop to attenuate emissions. However, since the external band is a current loop, it will reduce the efficiency and will decrease the maximum power load of the transformer.
A transformer with low leakage inductance should also help attenuate emissions. This is achived by keeping the close contact between the primary and secondary windings through the entire length of the transformer (ie bifilar, interleaving, etc). Unfortunately most pre-wound tranformers use a split bobbin to maximize isolation, but also maximizes inductance leakage, creating a noisy transformer.
BTW: Avoid eating tuna. Sea food is loaded with heavy metals cause by pollution. The US EPA now recommends limiting consumption of seafood meals to once a month.
http://www.mushield.com/design-guide.shtml
As Eva has discussed in many previous posts, toriod emissions can be controlled by the way they are orientated when mounted on the circuit board.
E-core transformers can use a copper foil band (aka "hum band")installed around the outside of the transformer (parallel to the winding) and attached to ground that creates a current loop to attenuate emissions. However, since the external band is a current loop, it will reduce the efficiency and will decrease the maximum power load of the transformer.
A transformer with low leakage inductance should also help attenuate emissions. This is achived by keeping the close contact between the primary and secondary windings through the entire length of the transformer (ie bifilar, interleaving, etc). Unfortunately most pre-wound tranformers use a split bobbin to maximize isolation, but also maximizes inductance leakage, creating a noisy transformer.
BTW: Avoid eating tuna. Sea food is loaded with heavy metals cause by pollution. The US EPA now recommends limiting consumption of seafood meals to once a month.
<snip
Post #10
“The metal of a typical tuna can has very low permeability, which makes a lousy magnetic shield. Tuna cans work well from shielding tuna from the environment, not for magnetic shielding.”
Thanks for that link, TechGuy. Seems that musheild is an informative site. It's going to take a while to go through eva's 4K+ posts but.. I'm digging. I discovered last night that I was looking at a potted toroidal transformer. Looked like a tuna can to me. So do I spec a non split bobbin wind? What about brass screen or something like that around the power supply? I was looking to reduce the foot print of the amplifier. And I am allergic to fish but I can eat canned tuna. (once in a while)
Post #10
“The metal of a typical tuna can has very low permeability, which makes a lousy magnetic shield. Tuna cans work well from shielding tuna from the environment, not for magnetic shielding.”
Thanks for that link, TechGuy. Seems that musheild is an informative site. It's going to take a while to go through eva's 4K+ posts but.. I'm digging. I discovered last night that I was looking at a potted toroidal transformer. Looked like a tuna can to me. So do I spec a non split bobbin wind? What about brass screen or something like that around the power supply? I was looking to reduce the foot print of the amplifier. And I am allergic to fish but I can eat canned tuna. (once in a while)
My understanding of toroid fields is that everything in theory is hunky dory...all perfectly contained, until the wire exits the toroid. Kinda like solar sunspots, magnetic fields....a point of eruption.
Upon the exiting (pigtails) of the toroid is where the fields are the greatest...if that is indeed the case, and I believe it is, you might want to be carefull where these exit..don't be having signal runs near these 'pigtails'.
With attention to routing and perhaps clocking of the toroid, fields will be very low from a toroid.
____________________________________________Rick.....
Upon the exiting (pigtails) of the toroid is where the fields are the greatest...if that is indeed the case, and I believe it is, you might want to be carefull where these exit..don't be having signal runs near these 'pigtails'.
With attention to routing and perhaps clocking of the toroid, fields will be very low from a toroid.
____________________________________________Rick.....
I would imagine Brass would work, but it will be less efficient in attentuating emissions since Brass isn't as good as a conductor as Copper. I recall seeing transformers using a Silver foil band since Silver has an even lower resistance than copper, but that might be a sales gimmick since the resistance difference between Silver and copper is small.
If transformer size is an issue, you can decrease the core size by increasing the switching frequency. The higher the switching frequency the smaller the core can be used to deliver the same power output. Althought at very high frequencies (500 Khz+), switching losses, winding capacitance, and wire skin effect can reduce maxium power output ratings. High switching frequencies also remove noise in the audiable frequency range. its also much easier to filter out High frequency noise because you can use much smaller value filter components (ie output caps and inductors). The only issue is that caps can have higher ESR (Equivelent Series Resistance) at higher frequencies, which reduce their efficiency of attenuating noise.
If transformer size is an issue, you can decrease the core size by increasing the switching frequency. The higher the switching frequency the smaller the core can be used to deliver the same power output. Althought at very high frequencies (500 Khz+), switching losses, winding capacitance, and wire skin effect can reduce maxium power output ratings. High switching frequencies also remove noise in the audiable frequency range. its also much easier to filter out High frequency noise because you can use much smaller value filter components (ie output caps and inductors). The only issue is that caps can have higher ESR (Equivelent Series Resistance) at higher frequencies, which reduce their efficiency of attenuating noise.
My opinion on this:TechGuy said:
I strongly prefer xformer designs with low capacitive coupling (split bobbins, R-cores, toroids with non-overlapping sectoral windings), since magnetic emissions can be handled (by shielding and/or strategic placement) while there is practically almost no way to lower the coupling capacitance which often results in ground currents between gear, not good at all (unless you have xformer inputs to block them). Plus, a higher xformer bandwidth than necessary is also something you don't want in a regular, non-switching power supply because this only allows more noise and hash to leak through, in both directions (primary<-->secondary).
- Klaus
Transformer capacitance coupling can be attenuated using a electrostatic shield between the primary and secondary. An electrostatic sheild is a layer of copper foil that is placed between the primary and secondary windings and is connected to ground (but not connected as a loop).
Coupling capacitance can also be reduced by using a lower number of primary and secondary turns and driving the primary at a higher switching frequency to compensate for the lower number of turns.
Noise from the primary to secondary is going to transferred anyway even with low transformer coupling capacitance, especially when the secondary draws a load.
I suppose your point makes sense for linear power supplies using 50/60 HZ as long as the input supply hasn't been squared up significantly. Since a Linear transformer has a very large number of turns required to operate at 50/60 hz, and the coupling capacitance would be significantly higher than a Switch Mode PS.
Coupling capacitance can also be reduced by using a lower number of primary and secondary turns and driving the primary at a higher switching frequency to compensate for the lower number of turns.
Noise from the primary to secondary is going to transferred anyway even with low transformer coupling capacitance, especially when the secondary draws a load.
I suppose your point makes sense for linear power supplies using 50/60 HZ as long as the input supply hasn't been squared up significantly. Since a Linear transformer has a very large number of turns required to operate at 50/60 hz, and the coupling capacitance would be significantly higher than a Switch Mode PS.
I brought up the tuna can as a shield, thinking I might reduce emi or not induce other problems with the toroidal when mounted in a small chassis made of hardwood. I wanted to keep the size to about 7x7 inches so that I could use a salvaged bronze embossing die as a heat sink and cover. The bottom would be aluminum with returns on 2 ends so I'd have a hard surface for connectors.
I'm inclined at this point to make a separate power supply. I've seen many advocates for this method.
I looked at the transformers I had on hand and both had the electrostatic shields.
Thanks for the replies I'm trying to keep up!
I'm inclined at this point to make a separate power supply. I've seen many advocates for this method.
I looked at the transformers I had on hand and both had the electrostatic shields.
Thanks for the replies I'm trying to keep up!
Tuna cans etal
I think the tuna can at least looks better
http://www.diyaudio.com/forums/showthread.php?threadid=121228&perpage=25&pagenumber=152:
I think the tuna can at least looks better
http://www.diyaudio.com/forums/showthread.php?threadid=121228&perpage=25&pagenumber=152:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.