Hi,
I'm building a voltage converter/RFI filter. Basically, this is a 2kVA torroid isolation transformer for powering some equipment I purchased in Europe.
The output of the transformer then feeds some cascaded RFI filters, which are located on a panel directly above the power transformer. Components use point-to-point wiring for the most part.
All of this is mounted in a 12x12x9 Hammond enclosure.
My question is, should the panels for mounting the RFI chokes be made from aluminum or some permeable metal (e.g., steel)? And why?
Thanks and regards,
Rob
I'm building a voltage converter/RFI filter. Basically, this is a 2kVA torroid isolation transformer for powering some equipment I purchased in Europe.
The output of the transformer then feeds some cascaded RFI filters, which are located on a panel directly above the power transformer. Components use point-to-point wiring for the most part.
All of this is mounted in a 12x12x9 Hammond enclosure.
My question is, should the panels for mounting the RFI chokes be made from aluminum or some permeable metal (e.g., steel)? And why?
Thanks and regards,
Rob
Hi
Yes you are correct.
I'm assuming the RFI choke you mention is actually a line filter module. They need to have a good solid electrical connection to shield chassis.
This is for suppression of common mode RFI and noise rejection.
Inside the filter there are two or more "Y capacitors" bypassing the hot and neutral to ground.
Yes you are correct.
I'm assuming the RFI choke you mention is actually a line filter module. They need to have a good solid electrical connection to shield chassis.
This is for suppression of common mode RFI and noise rejection.
Inside the filter there are two or more "Y capacitors" bypassing the hot and neutral to ground.
infinia said:
Thanks, but I didn't understand your response. Should the enclosure or chassis mounting plate be made out of aluminum or steel?
Some more details -- I didn't use a commercial line module. Instead, I saw a Monster power filter with a clear plastic cover and became inspired. This lead to some research, and I wound my own cores. Right now, there is one line filter and two common mode filters, all designed for minimal voltage drop at 60 Hz.
Thanks and regards,
Rob
weinstro said:
What about shields for RF performance on your designed filters?
Do you use x & Y capacitors? Your design is not going to be a see through chassis as the demo that inspired you on this endevour? Have you designed magnetics before. Have you designed power line filters before? A saturated core provides little filtering. Maybe there are many other factors you haven't condidered yet?
Maybe you should look for these as not DIY but get CSA UL approved parts. If you have answered no to any of my questions this could be a dangerous endevour. Check with your insurance before preceeding.
Just some general tips:
A magnetic material (e.g. steel, or one of the more exotic alloys like permalloy or "mu-metal") is useful for blocking magnetic fields.
Metals that conducting better (copper, aluminum, etc.) are better at blocking electrical fields, but not good at all at blocking magnetic fields.
A magnetic material (e.g. steel, or one of the more exotic alloys like permalloy or "mu-metal") is useful for blocking magnetic fields.
Metals that conducting better (copper, aluminum, etc.) are better at blocking electrical fields, but not good at all at blocking magnetic fields.
That's more or less my purpose of this thread. What about shields for the filters? And what about interaction between the isolation transformer and the RFI chokes? More on this below, though...
Yes
No, enclosed metal box. I can't see E-M fields, anyway.
Yes. Mostly for switchmode power supplies and variable frequency motor drives.
Nothing quite like this. But, I've followed the technical guides from Magnetics, Inc. I also was a practicing engineer in the power industry for a number of years.
Maybe. I believe I've done my homework here. But, there might be a million factors to consider. If I'm a genius, I've probably addressed half of them.
I have, for the most part. The "box" uses listed/recognized cabling, switches, fusing, XY caps, and receptacles. Insurance company (health or fire) couldn't care less, unless I do something like bypass the breakers in the distribution panel.
Anyway, I appreciate you trying to be helpful, but I wasn't looking to do a deep dive on every design aspect. But, allow me to distill my question into a simple form.
Because of the sizing of the enclosure, and the isolation transformer (Plitron 2kVA), the xfmr mounts in the bottom of the enclosure. The enclosure is made by Hammond, and has a steel body with aluminum front/rear panels. I then have a sheetmetal "deck" or subpanel directly above the xfmr where the RFI cores are mounted. DPDT on/off switches and panel mounted fuseholders are on the face. AC receptacles are on the rear. Now that the mental scenario is established, the question is should I make the subpanel from aluminum or steel?
Both have advantages and disadvantages:
Aluminum - easy to work with, can serve as a low z ground plane, expensive, soft, hardened alloys not readily available to the hobbyist, can't shield against induced noise.
Steel - hard to work with, dulls my tools, not the best conductor for chassis grounding, relatively cheap, strong, material permeability will help on inductive shielding.
From an RFI filtering point of view in 2 parts
a) steel deck holding differential filters.
b) Last stage of common mode filter should be completely shielded and as close to exit point as possible. Use thick copper foil or aluminum depends on how much performance you need. Are you filtering for FCC class B or the tougher TUV standard? Your call. Medical is toughest of all, usually requires a farady shield between pri. and secondary. Its better to stop noise at the source. So biggest differential filter would located before Xfmr pri.
hope this helps you on your new project.
regards,
a) steel deck holding differential filters.
b) Last stage of common mode filter should be completely shielded and as close to exit point as possible. Use thick copper foil or aluminum depends on how much performance you need. Are you filtering for FCC class B or the tougher TUV standard? Your call. Medical is toughest of all, usually requires a farady shield between pri. and secondary. Its better to stop noise at the source. So biggest differential filter would located before Xfmr pri.
hope this helps you on your new project.
regards,
A couple of general comments about shielding.
Electric fields are best shielded with materials that have as low a resistivity as practical.
Low resistivity materials are also good AC magnetic field attenuators via "Eddy currents" . Surprisingly, aluminum is quite good at attenuating magnetic fields above 1 KHz and even help at mains frequencies (the thicker the better).
Ferrous materials can be very good at attenuating magnetic fields, but their resistivity tends to be a higher than Al and thus not the best electric field or Eddy current shields.
Aluminum is a great compromise material for ac magnetic fields and for electric fields. Mains shielding will be so, so at the fundamental, but with each harmonic it gets better.
I am not an rfi expert, but my guess is that Al has the edge. The aluminum will be a good magnetic field shield at high frequencies and will afford some reduction at mains frequency, again, the thicker the better.
Also, my guess is that other factors such as how you do your grounding, orienting of components and component selection will have a bigger effect than the material differences.
My two cents.
Bill
Electric fields are best shielded with materials that have as low a resistivity as practical.
Low resistivity materials are also good AC magnetic field attenuators via "Eddy currents" . Surprisingly, aluminum is quite good at attenuating magnetic fields above 1 KHz and even help at mains frequencies (the thicker the better).
Ferrous materials can be very good at attenuating magnetic fields, but their resistivity tends to be a higher than Al and thus not the best electric field or Eddy current shields.
Aluminum is a great compromise material for ac magnetic fields and for electric fields. Mains shielding will be so, so at the fundamental, but with each harmonic it gets better.
I am not an rfi expert, but my guess is that Al has the edge. The aluminum will be a good magnetic field shield at high frequencies and will afford some reduction at mains frequency, again, the thicker the better.
Also, my guess is that other factors such as how you do your grounding, orienting of components and component selection will have a bigger effect than the material differences.
My two cents.
Bill
Infinia,
Thanks for your response.
OK!
If I understand correctly, you're recommending:
Diff filter 1>> Xfmr >> Diff filter 2>> Common Mode
with a shield on the common mode filter, and distance between the diff filters.
What I have right now is:
Xfmr >> Diff filter >> Common Mode 1 >> Common Mode 2
The Xfmr has a shield in it (medical grade).
Is this to avoid any mutual inductive coupling between the filters? Would mounting them orthogonal to one another be helpful?
I appreciate your insights. Thanks and regards,
Rob
Thanks for your response.
OK!
If I understand correctly, you're recommending:
Diff filter 1>> Xfmr >> Diff filter 2>> Common Mode
with a shield on the common mode filter, and distance between the diff filters.
What I have right now is:
Xfmr >> Diff filter >> Common Mode 1 >> Common Mode 2
The Xfmr has a shield in it (medical grade).
Is this to avoid any mutual inductive coupling between the filters? Would mounting them orthogonal to one another be helpful?
I appreciate your insights. Thanks and regards,
Rob
Yes I recommend 2 diff filters one before and after your isolation xfmer with last filter being common mode at the output (you can buy IEC panel jacks with built in shielded filters)
Medical grade xfmer will give good common mode rejection (that what their designed for). I know EI core designs offer up to 110 dB of isolation (seperated pri and sec windings with a copper sheild inbetween ) Not sure toroid designs can offer this.
Depending what kind of noise you are dealing with (worse being spikes with fast rise/fall time(dV/dT) times from a low impedance source) The problem is high noise current thru low loss Y caps is injected into chassis ground (if this happens then the battle is over) That is why you want a big series impedance before any low loss Y caps. The diff filters tend to add to this series impedance in the common mode measurement. Differential noise reduction is fairly straight forward (you need at least 2 stages because you can only get around 40 db for each stage) so physical seperation goal between stages is 40dB as well.
The physical limits for isolation are in the text "Reference Data for Radio Engineers". I hope I covered all your questions. So the main thing to avoid is Y caps at the input unless they are snubbers.
Medical grade xfmer will give good common mode rejection (that what their designed for). I know EI core designs offer up to 110 dB of isolation (seperated pri and sec windings with a copper sheild inbetween ) Not sure toroid designs can offer this.
Depending what kind of noise you are dealing with (worse being spikes with fast rise/fall time(dV/dT) times from a low impedance source) The problem is high noise current thru low loss Y caps is injected into chassis ground (if this happens then the battle is over) That is why you want a big series impedance before any low loss Y caps. The diff filters tend to add to this series impedance in the common mode measurement. Differential noise reduction is fairly straight forward (you need at least 2 stages because you can only get around 40 db for each stage) so physical seperation goal between stages is 40dB as well.
The physical limits for isolation are in the text "Reference Data for Radio Engineers". I hope I covered all your questions. So the main thing to avoid is Y caps at the input unless they are snubbers.
One more thing, Add a single common mode choke and MOV at the entry to your power conditioner. \This is for lightning protection (6000V for standard 20A distribution).
So your pwr cond. diagram would look something like this
in plug>fuse>Choke> MOV> diff filter >xfrmr pri >shieldgnd> xfrmr sec> diff filter> common mode> IEC conn
should work really great for clean power
have Fun
So your pwr cond. diagram would look something like this
in plug>fuse>Choke> MOV> diff filter >xfrmr pri >shieldgnd> xfrmr sec> diff filter> common mode> IEC conn
should work really great for clean power
have Fun
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.