That's a lot of back EMF

[Red Squirrel]

You can demonstrate the principle of back EMF really easily.
All you need is a small induction motor (any small appliance such as a box fan works perfectly). Make sure it's turned on (of course).

Instead of plugging it into the wall, grab a 9V battery and touch the prongs of the plug on the battery terminals. Notice how the spark seems much bigger than it would be if you simply shorted out the terminals?

Now to show higher voltage without getting shocked, grab a NE-2 "pilot lamp" style neon lamp, like this one,

wrap each wire from the tiny glow lamp to a prong on the plug and repeat what you did earlier with the battery. Notice how the lamp flashes brightly? These require about 90V to reach point of ionization (giving off light) and (again) with the 9V battery alone, not possible.

Finally, if you want to actually feel the effects of this, grab the plug by the prongs while repeating what you did before. Nice shock and (again) no way you would feel this from the 9V battery alone.

Now imagine dealing with 100s of kVA of power and a huge electromagnet...

I was curious and tried this on a scope and a small coil I made a while back for magnetizing screwdrivers. I just put a screwdriver in it, I imagine the effects would be better if I found a nice metal slug to put in there. It's interesting how it oscillates in the 12v range for a while before it dips down in the negatives creating a quick jolt.




As a side note I also melted my probe ground wire a little, I connected everything and was going to double check my connections before I turn the PSU on but it was already on and I had the polarity reversed. Just glad the magic smoke I saw was from the wire and not the scope itself. I really need to get a non ground referenced PSU to prevent this sort of thing. D: To add insult to injury I was on the 5v rail, that one really packs a punch current wise.

 

[Mark R]

Any estimates on the peak current surge capacity out of something like that? Say I were to have a remote-control robot short out the ends of it with a length of a solid 3x3" aluminum billet....

You misunderstand. The current is flowing in the circuit anyway. The issue is that the magnetic field resists the change in current.

If we take a pretty extreme magnet (a real example, albeit a one-off ultra-extreme MRI prototype), then it has an operating current of about 1500 A. The inductance is about 400 H, which means that if you try to change the current at a rate of 1 A/s, then the magnetic field will push back at 400 V.

As far as the aluminium, if you used it to carry the magnet current, then it would get very hot (not melting hot, but too hot to touch) from the current. A 3"x3" bar of Al is no match for a 2 AWG superconducting wire.

If you tried to break the circuit, then the magnetic field would push back with enough voltage to keep the current flowing. This would typically mean, a big fat arc at the point of break. The voltage could be massive, many, many kV. In fact what would probably happen is that you end up burning out the magnet wire insulation from overvoltage, were it not for the fact that such magnets usually have protection diodes at every conceivable position, to try to limit any voltage spikes.

Of course, the problem is that the diodes introduce heat into the coolant, and by overheating the superconductor, you can break the superconductivity, where the magnet wire transitions to normal conductivity (and heat production) causing a chain reaction called a "quench". You end up explosively boiling a ton of liquid helium, blowing out a bunch of pressure relief valves. With low energy magnets, there usually isn't too much of a problem. But with extreme magnets, the energy can be enough to melt or weld the magnet wire.

When CERN first powered up the LHC, one of the magnets blew for this reason. A faulty solder joint on the superconducting wire triggered a magnet quench due to a hairline crack. The magnet was reduced to a pile of mangled and semi-melted scrap metal.

 

[Jeff7]

You misunderstand. The current is flowing in the circuit anyway. The issue is that the magnetic field resists the change in current.

Right....ok, oops. I was thinking of the spike through a protection diode on a relay's coils.

If we take a pretty extreme magnet (a real example, albeit a one-off ultra-extreme MRI prototype), then it has an operating current of about 1500 A. The inductance is about 400 H, which means that if you try to change the current at a rate of 1 A/s, then the magnetic field will push back at 400 V.

As far as the aluminium, if you used it to carry the magnet current, then it would get very hot (not melting hot, but too hot to touch) from the current. A 3"x3" bar of Al is no match for a 2 AWG superconducting wire.

If you tried to break the circuit, then the magnetic field would push back with enough voltage to keep the current flowing. This would typically mean, a big fat arc at the point of break. The voltage could be massive, many, many kV. In fact what would probably happen is that you end up burning out the magnet wire insulation from overvoltage, were it not for the fact that such magnets usually have protection diodes at every conceivable position, to try to limit any voltage spikes.

Of course, the problem is that the diodes introduce heat into the coolant, and by overheating the superconductor, you can break the superconductivity, where the magnet wire transitions to normal conductivity (and heat production) causing a chain reaction called a "quench". You end up explosively boiling a ton of liquid helium, blowing out a bunch of pressure relief valves. With low energy magnets, there usually isn't too much of a problem. But with extreme magnets, the energy can be enough to melt or weld the magnet wire.

When CERN first powered up the LHC, one of the magnets blew for this reason. A faulty solder joint on the superconducting wire triggered a magnet quench due to a hairline crack. The magnet was reduced to a pile of mangled and semi-melted scrap metal.

See, that could be interesting.
If you're nowhere near the magnet. And you can afford to buy a new one.

I think I've seen a Youtube video at some point of that happening.



Some day.....room temperature superconductors, with no need for all the fancy cooling systems. It would be a revolutionary advance for society. And I'm sure we'll get a lot more interesting/dangerous videos as a result.

 

[William Gaatjes]

I was curious and tried this on a scope and a small coil I made a while back for magnetizing screwdrivers. I just put a screwdriver in it, I imagine the effects would be better if I found a nice metal slug to put in there. It's interesting how it oscillates in the 12v range for a while before it dips down in the negatives creating a quick jolt.




As a side note I also melted my probe ground wire a little, I connected everything and was going to double check my connections before I turn the PSU on but it was already on and I had the polarity reversed. Just glad the magic smoke I saw was from the wire and not the scope itself. I really need to get a non ground referenced PSU to prevent this sort of thing. D: To add insult to injury I was on the 5v rail, that one really packs a punch current wise.

Best way is to get an isolation transformer for your scope to get galvanic isolation. You have to make sure you connect the ground first since your scope will be floating(ESD) but this way, you can never create a ground loop or fry your scope when connecting your probe ground to something that has potential with respect to the circuit ground. I use an isolation transformer permanently for the scope on my work. For my scope at home i have two old transformers where i connected the secondary sides together to get 230V in and out. Make sure the earth connections are not connected.

 

[SOFTengCOMPelec]

..DELETED...
..
.
..
..
..
Dangerous, unsafe advice DELETED. Not repeated, so the poster can than change it if they want.

http://electronics.stackexchange.com/questions/73991/why-do-we-need-an-isolation-transformer-to-connect-an-oscilloscope

You should never float a scope with an isolation transformer! This is reckless and dangerous advice from your professor, and he/she needs a reality check.

The accepted procedure for doing work that requires isolation is to ISOLATE THE UNIT UNDER TEST, NOT THE TEST EQUIPMENT.

Why?

It's much easier to remember that the unit under test is what's unsafe and needs cautious handling, not your oscilloscope
If you hook a communication cable up to your floating scope (USB, GPIB, RS232), guess what - it's NO LONGER FLOATING. (All of these cables have earth-referenced returns)
As soon as you connect that floating scope return to a potential, all of the exposed metal on the scope is now at that potential. Major shock hazard.
If you cannot float the unit under test, use an isolated differential probe to do your measurements, and keep both the UUT and scope earthed. No measurement is worth the safety risk.

 

[disappoint]

http://electronics.stackexchange.co...lation-transformer-to-connect-an-oscilloscope

If you cannot float the unit under test, use an isolated differential probe to do your measurements, and keep both the UUT and scope earthed. No measurement is worth the safety risk.

This is true, though isolated differential probes don't come cheap.

 

[SOFTengCOMPelec]

This is true, though isolated differential probes don't come cheap.

But there are alternatives, such as using two scope inputs, in differential mode.

Or using a scope (often a battery powered version), with isolated inputs, built in to the scope, already.

Example:
http://www.tek.com/oscilloscope/tps2000

TPS2000B Digital Storage Oscilloscope

Powerful productivity from bench to field.

With up to 200 MHz bandwidth and 2 GS/s sample rate, the TPS2000B Digital Storage Oscilloscope Series allows you to safely make floating or differential measurements in a variety of challenging environments. With up to 4-isolated channels and a portable, battery-powered design, the TPS2000B Digital Storage Oscilloscope Series allows you to quickly and accurately tackle the tough challenges you face in the field of electronics and power systems – all designed with your safety in mind.

 

[skull]

You just connect the PSU to the coil terminals. Open the shorting switch (which is just a piece of superconducting wire mounted to a heater) - so you turn the heater on to break superconductivity, leaving it a regular high resistance wire.

Then you turn the current up on the PSU gradually. The magnet will produce a back EMF as the current rises, so you typically aim for 1-2 volts.

Once you reach nominal magnet current, you set constant current mode on the PSU, and as current is no longer changing the terminal voltage will fall to zero. Once this happens, there will be no leakage current through the superconducting switch (if V=0 then I=0). You can then turn off the switch heater, and allow superconductivity to recover in the switch.

You then start reducing current on the PSU, and the current from the magnet will start circulating through the switch. Once PSU current is zero, you disconnect.

The total amount of energy stored in the magnetic field varies with magnet design, but on a high end MRI, like the one just decommissioned it is around 5 kWh.

Some of the new higher field research MRI scanners going in have stored magnetic energy of about 30-50 kWh, with some of the extreme field scanners having up to 200 kWh of stored energy.

Am I understanding right that MRI magnets are charged 24/7 until you take them out of service? Does anything catastrophic ever happen with these things in the field?

 

[Brainonska511]

Am I understanding right that MRI magnets are charged 24/7 until you take them out of service? Does anything catastrophic ever happen with these things in the field?

Yep. MRI and NMR magnets, for the most part, are energized as long as they are stationary and in service. If they need to be moved or decommissioned, they are de-endergized and warmed up (remove the liquid helium and liquid nitrogen). However, there are a handful of low-field, small instruments that can be turned on/off, or use permanent magnets, and/or don't use any cryogens. (Though, I did see some work on a cryogen-free 9 T NMR powered by an electromagnet that could be turned on/off. I think it was cooled with cold helium gas and didn't require liquid helium, which can cost at least $10/L (but to give you an idea of helium costs: on our 14T magnet, we have to buy a 100 L tank every 4 months to top off the helium, and that's considered really good and efficient)).

But once they are energized and stable, nothing really catastrophic occurs in the magnet itself, as long as you keep the cryogens filled so that it remains superconductive. But there are other issues: the magnet will drift over time (so its field strength slowly changes with time), but this isn't a big problem. Generally, magnets have shim coils - small electromagnets that can be adjusted electronically - inside the bore of the magnet, which helps to create a homogeneous magnetic field through the sample of interest and can be set to change slowly with time to counter drift problems, which would be noticeable in longer experiments (eg: those greater than 24 hours). However, older, very high field magnets (early 2000s, 17 T to 22 T magnets) had more substantial drift problems, likely due to the technical difficulties in making such magnets and the fact that these were first-generation instruments. These drift problems could also be corrected with the shims, but it was something to be aware of if you were using the instrument, since shims can only correct so much.

 

[skull]

Yep. MRI and NMR magnets, for the most part, are energized as long as they are stationary and in service. If they need to be moved or decommissioned, they are de-endergized and warmed up (remove the liquid helium and liquid nitrogen). However, there are a handful of low-field, small instruments that can be turned on/off, or use permanent magnets, and/or don't use any cryogens. (Though, I did see some work on a cryogen-free 9 T NMR powered by an electromagnet that could be turned on/off. I think it was cooled with cold helium gas and didn't require liquid helium, which can cost at least $10/L (but to give you an idea of helium costs: on our 14T magnet, we have to buy a 100 L tank every 4 months to top off the helium, and that's considered really good and efficient)).

But once they are energized and stable, nothing really catastrophic occurs in the magnet itself, as long as you keep the cryogens filled so that it remains superconductive. But there are other issues: the magnet will drift over time (so its field strength slowly changes with time), but this isn't a big problem. Generally, magnets have shim coils - small electromagnets that can be adjusted electronically - inside the bore of the magnet, which helps to create a homogeneous magnetic field through the sample of interest and can be set to change slowly with time to counter drift problems, which would be noticeable in longer experiments (eg: those greater than 24 hours). However, older, very high field magnets (early 2000s, 17 T to 22 T magnets) had more substantial drift problems, likely due to the technical difficulties in making such magnets and the fact that these were first-generation instruments. These drift problems could also be corrected with the shims, but it was something to be aware of if you were using the instrument, since shims can only correct so much.

Do they need a constant feed of power or they just can't be discharged? I've heard horror stories about hospital generators not being tested regularly and not kicking in when they need to.

 

[Brainonska511]

Do they need a constant feed of power or they just can't be discharged? I've heard horror stories about hospital generators not being tested regularly and not kicking in when they need to.

No. The magnet itself doesn't need a constant feed of power. Once it's energized, it's no longer attached to any wires and it's effectively a self-contained system. The room temperature shim coils will need constant power to work, but its no big deal if the power goes out. The only worries with power outages would be traditional stuff: potential damage to expensive consoles, which are used to generate and record the radio frequencies used during an NMR/MRI experiment, loss of cooling gas for samples that need to be kept cold, having to manually extract a patient that's on a motorized MRI bed...

 

[Mark R]

Do they need a constant feed of power or they just can't be discharged? I've heard horror stories about hospital generators not being tested regularly and not kicking in when they need to.

No, once the magnet is charged, and the internal superconducting shorting switch is closed, then it will stay energised indefinitely. No power source needed.

The only issue is that it has to be kept cold, so that superconductivity is maintained.

The old way of doing it, was to immerse the magnet in a liquid helium bath (often hundreds of gallons for an MRI). This was then insulated, and that chamber was then placed in a bath of liquid nitrogen. However, heat would get in, and the helium would slowly evaporate, needing regular top ups, maybe 100 litres per year.

For about the last 20 years, refrigeration and insulation technology has improved, so that an insulated liquid helium bath is used, and a cryo-cooler has a small "cold finger" in the space above the helium bath. This causes some of the evaporated helium to recondense, massively extending the time between helium top ups to maybe every 3-5 years.

About 10 years ago, further refrigeration improvements mean that you can now get "zero boil off", where under normal circumstances, there is no helium loss. At end of life, the magnet is discharged in a controlled fashion, and shipped back to the factory, where the helium is transferred to a new magnet. The disadvantage is the energy consumption of the refrigerator. A typical ZBO MRI will need a 3 BTU/hour chiller - because of the temperature required, this type of chiller needs about 20-30 kW of electrical power and needs to run 24/7, so energy costs are substantial. In the event of power failure, the helium bath keeps the magnet cold, and the only side effect is that during the power failure helium will start evaporating at about 0.01 litres per hour (which is pretty much irrelevant, as you would need to lose about 100 litres to compromise the operation of the magnet).

The latest technology is "zero cryogen" magnets. Instead of sitting in a helium bath, the magnet coils sit in a vacuum chamber, and are thermally mounted to copper bars. These copper bars are then thermally attached directly to a high performance refrigerator. These avoid the problem of explosive boiling during quench and require very little helium (just a few grams for the refrigerant charge for the refrigerator, rather than several hundred kilograms for a helium bath). The disadvantage, is that a power failure or refrigerator failure can lead to quench within hours or days, and in the event of quench the heat will stay in the magnet, and it could take a week or more for the refrigeration system to get it back down to operating temperature, before the magnet is ready for recharging.

 

[SOFTengCOMPelec]

How easy is it to bring YOUR electronics, into the MRI environment, given the HUGE magnetic field.

I.e. Is there not a risk that it will draw (suitably magnetic) metals towards itself, increasingly powerfully, as it gets closer and closer ?

Similarly, is there not a danger that the huge magnetic fields, won't induce emfs (voltages, when you move your electronics in that field), sufficient to adversely affect its operation, or even break it ?

Presumably, part of the answer might be that you use very long, thick wires, to do everything, well clear of the MRI, itself.

I've heard/seen funny stories, where people who did not realize about the huge magnetic fields, of MRIs. Causing huge metal objects to violently hit the innards, of the MRI.

Also, next time you take your forum picture. Please keep the camera, well away from the MRI machine (joke).

 

[William Gaatjes]

http://electronics.stackexchange.com/questions/73991/why-do-we-need-an-isolation-transformer-to-connect-an-oscilloscope

Of course you are partially right. When the device under test is not carrying hazardous voltages, this is not an issue, and it breaks ground loops. But i should have mentioned that when i measure on a device that carries hazardous voltages, i make sure i never create a current loop. When i measure on a device under test that is connected to the mains voltage, i always use an isolation transformer for that device. A differential probe is indeed the best solution.
Short story is that i use usually two isolation transformers when needed and never had any issue.

 

[William Gaatjes]

But there are alternatives, such as using two scope inputs, in differential mode.

This i have done often.

 

[SOFTengCOMPelec]

Of course you are partially right. When the device under test is not carrying hazardous voltages, this is not an issue, and it breaks ground loops. But i should have mentioned that when i measure on a device that carries hazardous voltages, i make sure i never create a current loop. When i measure on a device under test that is connected to the mains voltage, i always use an isolation transformer for that device. A differential probe is indeed the best solution.
Short story is that i use usually two isolation transformers when needed and never had any issue.

I'm glad you at least partially agree.

People with all sorts of electronics experience levels could/would be reading these forums. So the worst case could be that someone who knows little, could try it, and end up hurting themselves. That is why I mentioned it.

It is all too easy to accidentally connect to a much higher voltage than you realize (I've done this, on at least one occasion). If your equipment is not properly grounded, (e.g. Because of isolation transformers or removing the ground connection), it could be very serious (Electrocution).

 

[SOFTengCOMPelec]

This i have done often.

The scope grounds, are a real pain in the neck, when you are taking measurements. I agree. You want to measure the voltage between two points, which are neither of them at ground potential. Hence needing solutions to this issue, such as isolated inputs or differential mode on the scope, as you just mentioned.

Anyway, I'd better let this thread get back on track.

 

[William Gaatjes]

I'm glad you at least partially agree.

People with all sorts of electronics experience levels could/would be reading these forums. So the worst case could be that someone who knows little, could try it, and end up hurting themselves. That is why I mentioned it.

It is all too easy to accidentally connect to a much higher voltage than you realize (I've done this, on at least one occasion). If your equipment is not properly grounded, (e.g. Because of isolation transformers or removing the ground connection), it could be very serious (Electrocution).

Well, of course you are right, i should have mentioned when it is save to do so. I had the text in mind from Red Squirrel and did not think about hazardous voltages or measuring directly on mains connected circuits (Which i do not without an isolation transformer). I am happy you pointed out that it can be dangerous when used unwise.
I think the most common problem is people making ground loops or connecting the ground of the scope to a voltage rail and creating a short circuit in low voltage applications.

 

[William Gaatjes]

You just connect the PSU to the coil terminals. Open the shorting switch (which is just a piece of superconducting wire mounted to a heater) - so you turn the heater on to break superconductivity, leaving it a regular high resistance wire.

Then you turn the current up on the PSU gradually. The magnet will produce a back EMF as the current rises, so you typically aim for 1-2 volts.

Once you reach nominal magnet current, you set constant current mode on the PSU, and as current is no longer changing the terminal voltage will fall to zero. Once this happens, there will be no leakage current through the superconducting switch (if V=0 then I=0). You can then turn off the switch heater, and allow superconductivity to recover in the switch.

You then start reducing current on the PSU, and the current from the magnet will start circulating through the switch. Once PSU current is zero, you disconnect.

The total amount of energy stored in the magnetic field varies with magnet design, but on a high end MRI, like the one just decommissioned it is around 5 kWh.

Some of the new higher field research MRI scanners going in have stored magnetic energy of about 30-50 kWh, with some of the extreme field scanners having up to 200 kWh of stored energy.

That is a lot of energy.
If i would take the 5kWh :
If i am not mistaken, that is 5000 * 3600000 = 18GJ.
W= 1/2 *L*(I*I)

If the coil would be for example 100H that would be a current flowing of
18.973 A., almost 19kA.
I have to be making a calculation error. That cannot be right.
But then again, I do not know the inductance and it is super conducting.

And does the PSU to charge the coil really have to be able to deliver that current ? That would be a massive psu, with lot of diodes in parallel.

EDIT :
How does current density work out for superconducting wire ?
How thick is the wire of such a coil ?
Can it really be that high ?

 

[SOFTengCOMPelec]

Well, of course you are right, i should have mentioned when it is save to do so. I had the text in mind from Red Squirrel and did not think about hazardous voltages or measuring directly on mains connected circuits (Which i do not without an isolation transformer). I am happy you pointed out that it can be dangerous when used unwise.
I think the most common problem is people making ground loops or connecting the ground of the scope to a voltage rail and creating a short circuit in low voltage applications.

Don't worry. I've sometimes made posts, which on later reflection, had to be redacted or modified. E.g. Because it could be used unsafely.

It can be very difficult, to know where the best places are to ground your probe(s), on circuits. But it comes, with experience. Different engineers, seem to have different ideas, on where these best places are.

The super conductors (on the MRI scanner), must be fascinating to probe about with. Since the electrical potential, should be identical, at all places on the wire (since its resistance is zero).

I bet if I did that, I would find a way of accidentally causing the entire hundreds of Kgs of Helium, to almost instantly boil/vaporize.

 

[William Gaatjes]

Don't worry. I've sometimes made posts, which on later reflection, had to be redacted or modified. E.g. Because it could be used unsafely.

It can be very difficult, to know where the best places are to ground your probe(s), on circuits. But it comes, with experience. Different engineers, seem to have different ideas, on where these best places are.

The super conductors (on the MRI scanner), must be fascinating to probe about with. Since the electrical potential, should be identical, at all places on the wire (since its resistance is zero).

I bet if I did that, I would find a way of accidentally causing the entire hundreds of Kgs of Helium, to almost instantly boil/vaporize.

It sure is fascinating. MRI scanners.

I just looked up a diode at vishay. It does 3800A. A few of those parallel should work. But i do not think they are fast and i would think that the PSU in a MRI scanner would be a switching type : SMPS, since efficiency would be a lot better. Then again, they are advertised for use in converters.

http://www.vishay.com/diodes/med-high-diodes/on-state-current-gt-3000-a/

EDIT:
This one is rated up to 990A, a fast recovery diode.

http://www.vishay.com/diodes/list/product-93182/

 

[SOFTengCOMPelec]

It sure is fascinating. MRI scanners.

I just looked up a diode at vishay. It does 3800A. A few of those parallel should work. But i do not think they are fast and i would think that the PSU in a MRI scanner would be a switching type : SMPS, since efficiency would be a lot better. Then again, they are advertised for use in converters.

http://www.vishay.com/diodes/med-high-diodes/on-state-current-gt-3000-a/

EDIT:
This one is rated up to 990A, a fast recovery diode.

http://www.vishay.com/diodes/list/product-93182/

Yes, it would be REAL fun, designing the power supply (probably SMPS), for an MRI scanner. The second diode you linked to, seems to ONLY have a voltage drop (typically) of around 2 volts, at 1000 Amps. That is impressive. Most diodes, I have ever encountered, would probably explode, at those sort of currents.
I'm amazed that the diode is so small (relatively), given the huge currents that they can take.
It would probably be great fun, messing around with a 1000..10,000 Amps, SMPS. It could probably make things like coins, glow RED hot. Etc etc.
Safe ... NO ... Fun ... YES.

Given the HUGE power rating of such a SMPS, it would need careful design, to make sure that it handles overloads/over-voltage/over-temperature etc, safely and quickly enough, to avoid possible dangers.

SMPS designs, seems to have its own set of specialized engineers, who know how to design such things, well. It seems to get especially specialized, when factors, such as power factor correction, come into play.

 

[Brainonska511]

How easy is it to bring YOUR electronics, into the MRI environment, given the HUGE magnetic field.

I.e. Is there not a risk that it will draw (suitably magnetic) metals towards itself, increasingly powerfully, as it gets closer and closer ?

Similarly, is there not a danger that the huge magnetic fields, won't induce emfs (voltages, when you move your electronics in that field), sufficient to adversely affect its operation, or even break it ?

Presumably, part of the answer might be that you use very long, thick wires, to do everything, well clear of the MRI, itself.

I've heard/seen funny stories, where people who did not realize about the huge magnetic fields, of MRIs. Causing huge metal objects to violently hit the innards, of the MRI.

Also, next time you take your forum picture. Please keep the camera, well away from the MRI machine (joke).

For modern magnets, like the NMR we have in my lab, they are well shielded (see: http://chem.wayne.edu/lumigen/nmr-instruments.html the top one is like the one we have - 14 T). The 5 Gauss line, which could be thought of as the line where you shouldn't bring credit cards past, if you want them to stay alive, is within the legs of the magnet. The only direction shielding isn't as good is the z axis of the magnet (basically, collinear with the bore). All the electronics in the room are within 10-15 feet of the magnet, or even a little closer (computer, console, cooling unit, pre-amps). As for running cables, we use 3/4" coaxial cables.

For an MRI, this can be a bit more problematic, since people lie on beds that are collinear with the Z-axis and there is a much bigger bore to deal with, but even with this hurdle, you can still have some things fairly close - worst case, you use an appropriate run of coaxial cabling to keep it at a safe distance. The strongest part of the field is going to be inside the magnet, where you'll never place electronics beyond the probe.

Older high-field magnets (like a 22 T), you also had issues about being too close with metal, especially when shielding wasn't as good. In those cases, you'd have lines painted on the ground for 5 Gauss and 10 Gauss limits, the latter being indicative of no magnetic tools since they could fly off a cart if you got too close. Those lines could be ~10-15 feet from the center of the magnet. Even still, neither getting cabling to them to connect the probe nor running coaxial cable from a directional coupler to an oscilloscope was a problem.

 

[OverVolt]

What I imagine this sounds like:

https://www.youtube.com/watch?v=sfU0HKXXfiU

 

[SOFTengCOMPelec]

For modern magnets, like the NMR we have in my lab, they are well shielded (see: http://chem.wayne.edu/lumigen/nmr-instruments.html the top one is like the one we have - 14 T). The 5 Gauss line, which could be thought of as the line where you shouldn't bring credit cards past, if you want them to stay alive, is within the legs of the magnet. The only direction shielding isn't as good is the z axis of the magnet (basically, collinear with the bore). All the electronics in the room are within 10-15 feet of the magnet, or even a little closer (computer, console, cooling unit, pre-amps). As for running cables, we use 3/4" coaxial cables.

For an MRI, this can be a bit more problematic, since people lie on beds that are collinear with the Z-axis and there is a much bigger bore to deal with, but even with this hurdle, you can still have some things fairly close - worst case, you use an appropriate run of coaxial cabling to keep it at a safe distance. The strongest part of the field is going to be inside the magnet, where you'll never place electronics beyond the probe.

Older high-field magnets (like a 22 T), you also had issues about being too close with metal, especially when shielding wasn't as good. In those cases, you'd have lines painted on the ground for 5 Gauss and 10 Gauss limits, the latter being indicative of no magnetic tools since they could fly off a cart if you got too close. Those lines could be ~10-15 feet from the center of the magnet. Even still, neither getting cabling to them to connect the probe nor running coaxial cable from a directional coupler to an oscilloscope was a problem.

Thanks for the extensive explanation and link. That has helped my improve my understanding a lot.

I got mixed up. I incorrectly thought that magnetic shielding was ONLY possible, with moving/changing magnetic fields. But of course magnetic patterns can be shielded, by guiding the magnetic flux through suitable materials.

Worse still. I would have expected it to be the Physics department. So I am feeling even more foolish, as it is the Chemistry department.

I was initially confused by the mention of 600 Megahertz. Because I thought it was low frequency, for the magnet. But I assume the extra coil(s), which allow high frequency modulation of the magnetic field, are what the 600 MHz are talking about. Presumably the 600 MHz is the "excitation" signal (Not sure of proper terminology, as I am NOT an MRI guru). Which makes the radio signals, become emitted from the sample(s), as they fall back to where the electrons were originally orbiting from.

So because the magnetic fields are reasonably shielded, the electronics (and cabling), are relatively straight forward. Except as you go inside the main part of the unit, which then has a huge magnetic field.

Before this thread, I thought/assumed that the huge magnet (of an MRI), was done by using HUGE permanent magnets, such as Neodymium Magnets.
I never realized it was done, by creating huge electro-magnets, and super-conductors.