What magnetic shielding should I use?

[supro]

I have a circuit board (with chips, resistors, capacitors, etc) that I would like use inside of an MRI magnet (~5 T). Obviously the circuit board itself will cause interference and might not function properly, so I would like to enclose it in a box to prevent any interference. What should I use? There is a static magnetic field and also RF frequencies along with gradient coils.

 

[mgb_phys]

That's a lot of field to sheild against!
There is stuff called mu-metal (very high nickel steel alloy) that is used for sheilding instruments but magnetic sheilding is a tricky business, is there any way you can pout the circuit further away with only a sensor inside the field.

 

[berkeman]

I don't think you're going to be able to shield against 5T. You also need to check with the MRI device manufacturer about putting anything metal at all inside the MRI field. Sounds like a bad idea. Definitely you need to talk with the MRI machine manufacturer -- not with us.

 

[f95toli]

mu-metal won't work at 5T. Such a strong field would simply magnetize the mu-metal (it is after all ferromagnetic) and it would become a pretty good magnet. Not what you want.

The only way I know of to shield against that kind of field is to use compensation coils (used in e.g. adibatatic demagnetization refrigeration cryostats where the field are of the order of tesla) but that is hardly practical in your case.

However, are you sure the circuit won't work without shielding?
If you design it without low-impedance loops at the relevant frequencies (difficult, but perhaps possible) it should be relatively insensitive to magnetic fields as long as you only use "simple" components (although e.g. digital ICs will obviously not work).

 

[supro]

I'm not sure if the circuit will work in the magnet. For now the circuit board is basically an evaluation board shown here:

[Google 751272583ADXL204EB_0.pdf]

I was thinking the capacitors might interfere with the magnetic field. I'm not sure about the accelerometer yet but I'll give the manufacturer a call on Tuesday.

You think the solution is easier now that you guys now what's on the circuit board?

 

[mgb_phys]

That link just shows the raw accelerometer chip with a couple of caps.
As far as I remember those chips are just micro machined balances, they don't have a lot of extra circuitry in there.
The tricky bit is going to be reading it,any variation in the field of movement of the wires in the constant field is going to induce a current in the wires. Presumably if you interested in an accelerometer it's going to be moving!
I'm not sure simply braiding the leads together is going to be enough in a 5T field.

 

[supro]

That link just shows the raw accelerometer chip with a couple of caps.
As far as I remember those chips are just micro machined balances, they don't have a lot of extra circuitry in there.
The tricky bit is going to be reading it,any variation in the field of movement of the wires in the constant field is going to induce a current in the wires. Presumably if you interested in an accelerometer it's going to be moving!
I'm not sure simply braiding the leads together is going to be enough in a 5T field.

Thanks, so you think I would be better off with a digital accelerometer because AC current due to non-homogeneous magnetic field (or especially the gradients) will not effect it?

The accelerometer data sheet is here:
[Google ADXL204.pdf]

 

[mgb_phys]

You would need a digital accelerometer where the ADC circuit and micro can run in that sort of field. I think I would contact the makers of the accelerometer and ask them - these things get used in some extreme conditions.

 

[berkeman]

Thanks, so you think I would be better off with a digital accelerometer because AC current due to non-homogeneous magnetic field (or especially the gradients) will not effect it?

The accelerometer data sheet is here:
[Google ADXL204.pdf]

Sorry if you already said it and I missed it, but what exactly are you trying to do?

 

[supro]

Sorry if you already said it and I missed it, but what exactly are you trying to do?

I'm basically trying to detect any movement of the subject inside of an MRI magnet during the scan. The idea is to put 1 or more accelerometers under the subject's seat which has another seat underneath, connected by a sensitive spring. The accelerometers will detect the slightest movements (even breathing) and send a signal to the computer using a wire (fiber optic I'm guessing).

Another idea was to use the same seat (two surfaces separated by a spring) and have a coil connect to each surface. There is already AC current running through one coil and the other coil should also have AC current due to electromagnetic induction. A motion will compress or stretch the spring and cause a change in distance between the two coils which will cause a change in current of the receiving coil.

I figured it was better to go with the accelerometer because it might be more stable and can be used for wireless signal transfer later on. The problem with it is how to put the circuit board inside the magnet without causing a performance degradation or complete failure.

 

[marcusl]

Please DO NOT under any circumstances put any ferromagnetic material inside the bore. First, you cannot shield your components from a field that strong with any reasonable amount of iron, so the effort is bound to failure. Second, you will create a serious safety hazard. Past accidents within that industry are, unfortunately, numerous and grisly. Third, presence of your shield and even of small amounts of magnetic material on your circuit will irreparably distort the image (even "non-magnetic" tooth fillings cause bad distortions in high field imagers).

You need instead to use non-magnetic and field-tolerant components. Carbon film resistors are good, metal film (or metal wound) bad. Many components have ferromagnetic leads or finishes (the lead carriers of many IC's contain kovar or other magnetically undesirable metals). You can and should special order chip caps and resistors without nickel flashing on their soldering surfaces. Some op amps work in fields, others don't (large bias currents and offset voltages are common symptoms).

Whatever you do, make sure you work with someone truly knowledgeable about MRI. I suggest you call the company that supplied your instrument and work with their engineers. Most manufacturers of these experimental high field systems will support their customers with advice. Matters of and diagnostic efficacy and human safety are no places for amateurs.

 

[f95toli]

I'm basically trying to detect any movement of the subject inside of an MRI magnet during the scan. The idea is to put 1 or more accelerometers under the subject's seat which has another seat underneath, connected by a sensitive spring. The accelerometers will detect the slightest movements (even breathing) and send a signal to the computer using a wire (fiber optic I'm guessing).

Any reason why you can't use an all-optical solution? All the things you describe could presumably be done using interferometry, meaning you can put all the electronics at at safe distance.

 

[Gokul43201]

Alternatively, can you not use a purely passive sensor and run wires away from the magnet to the measurement circuitry?

 

[mgb_phys]

Alternatively, can you not use a purely passive sensor and run wires away from the magnet to the measurement circuitry?

It would be very difficult not to pick up any field on the wires from microphonics.

 

[Phrak]

Please DO NOT under any circumstances put any ferromagnetic material inside the bore. First, you cannot shield your components from a field that strong with any reasonable amount of iron, so the effort is bound to failure. Second, you will create a serious safety hazard.

Marcusl. I was hoping someone would bring that up. The magnetic field extends into the room as well so this goes for the environs as well, right?

 

[marcusl]

Am not sure what you are asking. The fields extend outside of the magnet either quite a distance or, in the case of modern "self shielded" MRI magnets, not very far.

 

[supro]

Please DO NOT under any circumstances put any ferromagnetic material inside the bore. First, you cannot shield your components from a field that strong with any reasonable amount of iron, so the effort is bound to failure. Second, you will create a serious safety hazard. Past accidents within that industry are, unfortunately, numerous and grisly. Third, presence of your shield and even of small amounts of magnetic material on your circuit will irreparably distort the image (even "non-magnetic" tooth fillings cause bad distortions in high field imagers).

You need instead to use non-magnetic and field-tolerant components. Carbon film resistors are good, metal film (or metal wound) bad. Many components have ferromagnetic leads or finishes (the lead carriers of many IC's contain kovar or other magnetically undesirable metals). You can and should special order chip caps and resistors without nickel flashing on their soldering surfaces. Some op amps work in fields, others don't (large bias currents and offset voltages are common symptoms).

Whatever you do, make sure you work with someone truly knowledgeable about MRI. I suggest you call the company that supplied your instrument and work with their engineers. Most manufacturers of these experimental high field systems will support their customers with advice. Matters of and diagnostic efficacy and human safety are no places for amateurs.

Very detailed, thanks.
I already contacted two companies that manufacture the accelerometers I'm interested in. Both use metallic caps and leads but they have non-magnetic material inside. I asked them if they will produce a special order with non-magnetic caps and I'm waiting for a reply. Otherwise, I think I'll be able to ask a 3rd party to replace the caps. I know a paper that does this.
As for the other components, I'm mostly concerned about the capacitors. Are there any that exist which can be used inside of a magnet this big? Is it even possible to have capacitors that work inside of a magnet?

I'm actually working in a team, being lead by a researcher who is very good with MRI. He just gave me the task of designing a motion detector and hence all this. At the end, we are all going to discuss it and only implement it if it's completely safe.

Any reason why you can't use an all-optical solution? All the things you describe could presumably be done using interferometry, meaning you can put all the electronics at at safe distance.

Hmm I just looked at interferometers and they look more complicated when it comes to their use inside of a magnet. Maybe I am looking at the wrong thing.
Could you please elaborate a little more?

Alternatively, can you not use a purely passive sensor and run wires away from the magnet to the measurement circuitry?

If this doesn't work out, I might have to switch back to using magnetically coupled coils where there are no circuits. I'm not sure about passive sensors though. I googled it and found passive infrared sensors which I think will not detect all the body movements.

 

[Phrak]

You might take an alternative approch that may or may not suite your requirements. optical telemetry, perhaps infared motion sensing.

 

[elec2050]

if you use any shield it will have some metal in it, and won't work with MRI. 5T is very high.

Rick