Torroid? I've always heard that Torroid's don't throw the big EM field. Maybe you don't have to shield that one?
PCB shield mesh, or complete plane??? Well, you're going to have to investigate what kind of skin depth is necessary for 50/60Hz, to see if there is a shielding benefit at all. If there is a benefit and you do implement a PCB shield, I think a mesh would be more appropriate. The mesh greatly reduces eddy currents that could be generated by the magnetic flux cutting across a contiguous plane. I'm assuming you can generate eddy currents in copper. I don't know. You can, for sure, in aluminum.
Someone else should chime in on this. I've never examined stray fields coming off of a torroid.
PCB shield mesh, or complete plane??? Well, you're going to have to investigate what kind of skin depth is necessary for 50/60Hz, to see if there is a shielding benefit at all. If there is a benefit and you do implement a PCB shield, I think a mesh would be more appropriate. The mesh greatly reduces eddy currents that could be generated by the magnetic flux cutting across a contiguous plane. I'm assuming you can generate eddy currents in copper. I don't know. You can, for sure, in aluminum.
Someone else should chime in on this. I've never examined stray fields coming off of a torroid.
Hi,
I go with Nordic on this.
Orient the toroid to minimise any audible or measurable effects.
If the transformer lead outs are solid core, then swap them to flexible to enable easy rotation around the mounting bolt.
I have see it stated that there is a stronger field sent out in the direction (axis) of the hole through the toroid, but I cannot confirm this.
If this is the case then keeping the side of the toroid towards the pcb would avoid this potential problem.
Finally, the electrostatic field can be reduced by a non magnetic shield. This shield must completely enclose either the PCB or the transformer. A gap is just as bad as none.
The electromagnetic field can be reduced by a magnetic material and this material must be thick. I do not know if this shield needs to fully enclose.
I go with Nordic on this.
Orient the toroid to minimise any audible or measurable effects.
If the transformer lead outs are solid core, then swap them to flexible to enable easy rotation around the mounting bolt.
I have see it stated that there is a stronger field sent out in the direction (axis) of the hole through the toroid, but I cannot confirm this.
If this is the case then keeping the side of the toroid towards the pcb would avoid this potential problem.
Finally, the electrostatic field can be reduced by a non magnetic shield. This shield must completely enclose either the PCB or the transformer. A gap is just as bad as none.
The electromagnetic field can be reduced by a magnetic material and this material must be thick. I do not know if this shield needs to fully enclose.
Thanks for the suggestions.
To be honest, I designed the look first and ordered the case and the torroid accordingly. Only then started thinking about this shielding.
And yes, the torroid axis pinges the circuit in the heart
and turning it is no option.
If mesh is an option I can wrap the torroid in copper mesh connected to the metal chassis.
AndrewT: If a gap is just as bad as none then mesh is out of question, right? Or, the gap should be only 'small enough'?
I can do both: mesh AND board over each other but it seems to be an overkill.
What about those metal cups often seen in cd-players over the torroid? How are they connected to the chassis? They must be separated from the the bolt fixing the torroid, or else they would work as just another turn in the transformer. Not that I could get hold of such a thing, just wondering.
To be honest, I designed the look first and ordered the case and the torroid accordingly. Only then started thinking about this shielding.
And yes, the torroid axis pinges the circuit in the heart
and turning it is no option.If mesh is an option I can wrap the torroid in copper mesh connected to the metal chassis.
AndrewT: If a gap is just as bad as none then mesh is out of question, right? Or, the gap should be only 'small enough'?
I can do both: mesh AND board over each other but it seems to be an overkill.
What about those metal cups often seen in cd-players over the torroid? How are they connected to the chassis? They must be separated from the the bolt fixing the torroid, or else they would work as just another turn in the transformer. Not that I could get hold of such a thing, just wondering.
Hi,
if you are still considering copper or aluminium screening then you can only reduce electrostatic radiation.
These materials will have almost no attenuating effect on electromagnetic radiation.
Transformer hum is much more likely to be a magnetic field effect.
if you are still considering copper or aluminium screening then you can only reduce electrostatic radiation.
These materials will have almost no attenuating effect on electromagnetic radiation.
Transformer hum is much more likely to be a magnetic field effect.
a mesh will work, but I have no expertise in this field. EMI attenuation is a science for which expertise is hard won.
...
" ... any sort of shielding should be connected to the chassis ground, and nowhere else, right? ..."
Yes, although there may be a ground path from audio ground back to chassis ground (a commonly used topography), direct connection between the metal chassis and the shield plates or shield screens is the better procedure = isolating the primary side of the transformer (which references to the power line ground and/or common power line).
" ... any sort of shielding should be connected to the chassis ground, and nowhere else, right? ..."
Yes, although there may be a ground path from audio ground back to chassis ground (a commonly used topography), direct connection between the metal chassis and the shield plates or shield screens is the better procedure = isolating the primary side of the transformer (which references to the power line ground and/or common power line).
first, connect to ...
... connect to the chassis ground any and all shields ... this is safest and best for quality audio.
If at some point you feel the need or your technology dictates, you can run a jumper between the audio signal / pre-amp / amplifier input signal ground and your chassis ground(s).
It is axiomatic that ground loop problems develop when using a common signal ground and chassis ground = a signal ground of the "star" configuration is prefered, a common, independant central point for all signal grounds that is completely different and completely isolated from a "daisy chain" of chassis grounds.
Chassis grounds are meant for power protection isolation shields (usually meant to protect the power lines from the outside world and shield the power supply noise from the audio signals) ... audio signal ground is meant as a signal reference and signal return path (and rarely for power line protection and/or shielding ... as then the audio signal can be corrupted by power line noise).
... connect to the chassis ground any and all shields ... this is safest and best for quality audio.
If at some point you feel the need or your technology dictates, you can run a jumper between the audio signal / pre-amp / amplifier input signal ground and your chassis ground(s).
It is axiomatic that ground loop problems develop when using a common signal ground and chassis ground = a signal ground of the "star" configuration is prefered, a common, independant central point for all signal grounds that is completely different and completely isolated from a "daisy chain" of chassis grounds.
Chassis grounds are meant for power protection isolation shields (usually meant to protect the power lines from the outside world and shield the power supply noise from the audio signals) ... audio signal ground is meant as a signal reference and signal return path (and rarely for power line protection and/or shielding ... as then the audio signal can be corrupted by power line noise).
Hi,
I am just looking for the solution on the similar case, and I have learned that a copper plate can't shield a Magnetic field, only the electric field. You will need some magnetic material like mu-metal to attenuate the magnetic field. Please check the following link:
http://www.lessemf.com/mag-shld.html#276
Besides, the famous S&B TX-102 transformer is casted in the mu-metal to shield any external electromagnetic field.
Regards,
Adam
I am just looking for the solution on the similar case, and I have learned that a copper plate can't shield a Magnetic field, only the electric field. You will need some magnetic material like mu-metal to attenuate the magnetic field. Please check the following link:
http://www.lessemf.com/mag-shld.html#276
Besides, the famous S&B TX-102 transformer is casted in the mu-metal to shield any external electromagnetic field.
Regards,
Adam
True ...
" ... copper plate can't shield a Magnetic field, only the electric field. You will need some magnetic material like mu-metal to attenuate the magnetic field. ..."
True enough ... copper does the best job of shielding EMF ... but copper is very poor at shielding magnetic fields. ... But that's why we all try to use torrid transformers, because of the lower & more efficient configuration of the magnetic fields = less magnetic shielding required.
It is also true that the best comprimise for shielding both, EMF and magnetic fields, is the mu-metal material ... unless obtaining it is not cost effective for your project.
Look: you can spend your time & money trying to shield everything in the local, physical universe including the Earth's own magnetic field plus all the EMF in the neighborhood, or you can do your best to install decent, effective EMF shielding and then spend your time & resources of better circuit designs, designs that don't really need much shielding. Proper circuit design and layout covers a world of sins. The use of torridal transformers, good circuit board layout and component positioning, the use of shielded data lines and shielded low level signal lines ... etc ... can negate the need for any magnetic shielding as well as extra EMF shielding.
(Chasing that last 5% costs as much as getting the first 95% correct.)
" ... copper plate can't shield a Magnetic field, only the electric field. You will need some magnetic material like mu-metal to attenuate the magnetic field. ..."
True enough ... copper does the best job of shielding EMF ... but copper is very poor at shielding magnetic fields. ... But that's why we all try to use torrid transformers, because of the lower & more efficient configuration of the magnetic fields = less magnetic shielding required.
It is also true that the best comprimise for shielding both, EMF and magnetic fields, is the mu-metal material ... unless obtaining it is not cost effective for your project.
Look: you can spend your time & money trying to shield everything in the local, physical universe including the Earth's own magnetic field plus all the EMF in the neighborhood, or you can do your best to install decent, effective EMF shielding and then spend your time & resources of better circuit designs, designs that don't really need much shielding. Proper circuit design and layout covers a world of sins. The use of torridal transformers, good circuit board layout and component positioning, the use of shielded data lines and shielded low level signal lines ... etc ... can negate the need for any magnetic shielding as well as extra EMF shielding.
(Chasing that last 5% costs as much as getting the first 95% correct.)
Hi,
we are reaching a concensus that copper and other non magnetic shields do not reduce magnetic fields.
But, you do not have to fit very expensive mumetal shielding.
Thick magnetic steel will attenuate the magnetic field (not non magnetic stainless steel).
I have seen a thick ring of steel around the outside of toroids on two different units (a power amp and a pre-amp). The ring of steel almost closed the gap between the floor and lid of the chassis.
we are reaching a concensus that copper and other non magnetic shields do not reduce magnetic fields.
But, you do not have to fit very expensive mumetal shielding.
Thick magnetic steel will attenuate the magnetic field (not non magnetic stainless steel).
I have seen a thick ring of steel around the outside of toroids on two different units (a power amp and a pre-amp). The ring of steel almost closed the gap between the floor and lid of the chassis.
OK .... no problem-o
" ... we are reaching a concensus that copper and other non magnetic shields do not reduce magnetic fields. But, you do not have to fit very expensive mumetal shielding. Thick magnetic steel will attenuate the magnetic field (not non magnetic stainless steel). ..."
I don't have a problem with any of this ... except:
" Chasing that last 5% costs as much as getting the first 95% correct. "
... and ...
" Proper circuit design and layout covers a world of sins. The use of torridal transformers, good circuit board layout and component positioning, the use of shielded data lines and shielded low level signal lines ... etc ... can negate the need for any magnetic shielding as well as extra EMF shielding. "
... and ... it seems like a whole lot of extra effort to do both or either when the need may not out way the effort.
" ... we are reaching a concensus that copper and other non magnetic shields do not reduce magnetic fields. But, you do not have to fit very expensive mumetal shielding. Thick magnetic steel will attenuate the magnetic field (not non magnetic stainless steel). ..."
I don't have a problem with any of this ... except:
" Chasing that last 5% costs as much as getting the first 95% correct. "
... and ...
" Proper circuit design and layout covers a world of sins. The use of torridal transformers, good circuit board layout and component positioning, the use of shielded data lines and shielded low level signal lines ... etc ... can negate the need for any magnetic shielding as well as extra EMF shielding. "
... and ... it seems like a whole lot of extra effort to do both or either when the need may not out way the effort.
a big score ...
... I just made a great score:
I obtained a heavy duty, but broken, Crown 200W stereo amp for about US$50. I gutted the circuit board (now its a frisbee) and some other parts, keeping the torrid transformer, power switch & fuse, output to speaker connections & filters and some other stuff. The huge space inside the steel case allowed me to install two power supply filter boards and two 150W amplifier modules several inches away from the torrid transformer. Since all the new boards are mounted horizontally (parallel & in the same plane) to the magnetic lines of force of the torrid = no magnetic interference worries, mate = .
In fact since the unbalanced inputs to the amp(s) are also far away from the torrid, I could (but I won't) use unshielded input cable ...
(+/- 65 VDC PS rails through aussieamplifiers.com NX-150 amp modules = ~185 Watts into 8 ohms and less than 0.003% THD ... and the specs check out so far with the transformer unshielded from the rest ... )
... I just made a great score:
I obtained a heavy duty, but broken, Crown 200W stereo amp for about US$50. I gutted the circuit board (now its a frisbee) and some other parts, keeping the torrid transformer, power switch & fuse, output to speaker connections & filters and some other stuff. The huge space inside the steel case allowed me to install two power supply filter boards and two 150W amplifier modules several inches away from the torrid transformer. Since all the new boards are mounted horizontally (parallel & in the same plane) to the magnetic lines of force of the torrid = no magnetic interference worries, mate =
.In fact since the unbalanced inputs to the amp(s) are also far away from the torrid, I could (but I won't) use unshielded input cable ...
(+/- 65 VDC PS rails through aussieamplifiers.com NX-150 amp modules = ~185 Watts into 8 ohms and less than 0.003% THD ... and the specs check out so far with the transformer unshielded from the rest ... )
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.