matt_gsxr
Posts: 10
Joined: Thu Mar 13, 2014 3:55 pm

Pi in high magnetic fields

Thu Mar 13, 2014 4:56 pm

Hi there,

So I want to use a Rasp Pi interfaced to a rasp pi camera to do some monitoring in a high magnetic field.

I've got the camera set up (Noir model) on Rasp pi (B) and am broadcasting to another computer, so it is streaming nicely, and all is good using the scripts on the camera pages (http://www.raspberrypi.org/camera).

So I take the Rasp pi into the magnet area (long network cable and long USB power lead).
At around 1 Tesla the stream cuts out, but returns as I move away from the field. I'd like this to work at 7Tesla.

As this picks up again I am suspecting the ethernet port (you can get a similar effect by disconnecting the cable and reconnecting).

Any ideas on what I could change to get this working? I am suspecting that either the network port is dying on its own, or that there is a drop in voltage which is stopping the network port working temporarily.

Thanks,

Matt

p.s. From a safety perspective, the only significantly magnetic component on the board is the USB connector (easily desoldered), the camera is slightly magnetic but light.

User avatar
Burngate
Posts: 6483
Joined: Thu Sep 29, 2011 4:34 pm
Location: Berkshire UK Tralfamadore
Contact: Website

Re: Pi in high magnetic fields

Thu Mar 13, 2014 5:19 pm

About the only thing on the Pi which might be affected by an ordinary magnetic field is the RJ45 ethernet connector - but without measuring anything, I've no idea how a 1Tesla field would affect it, never mind 7 Tesla.
RJ45.png
RJ45.png (13.48 KiB) Viewed 7730 times
So we've got four transformers, each with a ferrite core. Put a large enough field across them, and the core will saturate (How large? Apparently 1 Tesla!) and the coupling will head towards zero.

User avatar
Burngate
Posts: 6483
Joined: Thu Sep 29, 2011 4:34 pm
Location: Berkshire UK Tralfamadore
Contact: Website

Re: Pi in high magnetic fields

Thu Mar 13, 2014 5:26 pm

Out of interest, what are you using the 7 Tesla field for? And the Pi? If you're allowed to tell us.
A quick <your favourite search engine name used as a verb> gave me 0.5 - 3 Tesla in MRI machines, with upto 60 Tesla for research.

jamesh
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 30118
Joined: Sat Jul 30, 2011 7:41 pm

Re: Pi in high magnetic fields

Thu Mar 13, 2014 5:32 pm

7 Tesla is a 'helluvalot'. I'm not surprised something fails. It not something the Raspi was designed to withstand!

I'm with the poster above - probably the magnetics in the ethernet connector, although the electrons in the chip will be all over the place.

Do you work at CERN on the LHC? That's the only place I can come up with that has magnetic fields that high.
Principal Software Engineer at Raspberry Pi (Trading) Ltd.
Working in the Applications Team.

PiGraham
Posts: 4874
Joined: Fri Jun 07, 2013 12:37 pm
Location: Waterlooville

Re: Pi in high magnetic fields

Thu Mar 13, 2014 5:37 pm

This is a static 7T field, yes?
Ethernet uses magnetic isolators. Could they be saturating the the strong field?
Could you put the Pi in a magnetically screened enclosure?
Perhaps the magnetics can be replaced with opto-isolators.

User avatar
Jim Manley
Posts: 1600
Joined: Thu Feb 23, 2012 8:41 pm
Location: SillyCon Valley, California, and Powell, Wyoming, USA, plus The Universe
Contact: Website

Re: Pi in high magnetic fields

Thu Mar 13, 2014 5:51 pm

There are ... ummm ... let's just say places that do research on very, very, very tiny things that can have enormous physical effects (including generating massive magnetic fields) that aren't directly related to particle physics research, but benefit from that research. Please state your answer in the form of a question, but I still may not be able to answer the question without someone coming out of hiding and killing each of us :lol:
The best things in life aren't things ... but, a Pi comes pretty darned close! :D
"Education is not the filling of a pail, but the lighting of a fire." -- W.B. Yeats
In theory, theory & practice are the same - in practice, they aren't!!!

PiGraham
Posts: 4874
Joined: Fri Jun 07, 2013 12:37 pm
Location: Waterlooville

Re: Pi in high magnetic fields

Thu Mar 13, 2014 6:19 pm

jamesh wrote:7 Tesla is a 'helluvalot'. I'm not surprised something fails. It not something the Raspi was designed to withstand!

I'm with the poster above - probably the magnetics in the ethernet connector, although the electrons in the chip will be all over the place.

Do you work at CERN on the LHC? That's the only place I can come up with that has magnetic fields that high.
That's what I thought, but the LHC doesn't have the most powerful magnets.
Wikipedia wrote:4 T – strength of the superconducting magnet built around the CMS detector at CERN
8 T – the strength of LHC magnets.
11.75 T – the strength of INUMAC magnets, largest MRI scanner.
13 T – strength of the superconducting ITER magnet system
16 T – magnetic field strength required to levitate a frog, according to the 2000 Ig Nobel Prize in Physics.

User avatar
jbeale
Posts: 3893
Joined: Tue Nov 22, 2011 11:51 pm
Contact: Website

Re: Pi in high magnetic fields

Thu Mar 13, 2014 6:41 pm

Sounds like a fun project. Is there any purpose to placing the Pi itself in the high-field region? I assume you just need the camera to be there. How big is this magnet? You can use longer camera flex cables, and a ribbon cable adaptor* allows even longer lengths to put your Pi in a lower-field area. I had my camera working with a ribbon cable about 4 meters long (that is of course very much out of spec for the camera interface; do this at your own risk). I certainly haven't tried this near a MRI or similar installation. The high field may cause trouble with longer cables especially if there is any motion of the cable, or AC component to the field.

*http://www.bitwizard.nl/catalog/product ... anguage=en

If the ferrite cores in the ethernet connector really are the only problem, you could consider using a USB-wifi adaptor instead of ethernet. If you have to desolder the metal-shell USB jack for magnetic safety, you can still hardwire a USB dongle directly to the Pi.

User avatar
Burngate
Posts: 6483
Joined: Thu Sep 29, 2011 4:34 pm
Location: Berkshire UK Tralfamadore
Contact: Website

Re: Pi in high magnetic fields

Thu Mar 13, 2014 7:07 pm

PiGraham wrote:
Wikipedia wrote:4 T – strength of the superconducting magnet built around the CMS detector at CERN
8 T – the strength of LHC magnets.
11.75 T – the strength of INUMAC magnets, largest MRI scanner.
13 T – strength of the superconducting ITER magnet system
16 T – magnetic field strength required to levitate a frog, according to the 2000 Ig Nobel Prize in Physics.
You missed one: http://en.wikipedia.org/wiki/Magnetic_refrigeration ~1 T, with 3 T or greater for NDR

User avatar
Jim Manley
Posts: 1600
Joined: Thu Feb 23, 2012 8:41 pm
Location: SillyCon Valley, California, and Powell, Wyoming, USA, plus The Universe
Contact: Website

Re: Pi in high magnetic fields

Thu Mar 13, 2014 7:24 pm

Also, are we talking about a sustained magnetic field or one that's at that level for a brief period of time? Pulsed magnetic fields will induce currents in conductors, and if your Ethernet cable is positioned across a varying magnetic field, a substantial current can be generated. You could overwhelm the magnetic isolators in the Pi and make some semiconductor junctions very unhappy with a strong enough field.

BTW, the world's record for unclassified research is a 90 Tesla pulsed field achieved on June 22, 2011, at the High Magnetic Field Laboratory Dresden (HLD - Helmholtz-Zentrum Dresden-Rossendorf). It used a purpose-built coil that can generate a 50 Tesla field that was wound with a special copper alloy encased in an aramid fiber (e.g., Kevlar) jacket designed to withstand 10,000 atmospheres of pressure (pure copper wire would blow itself apart at around 25 Tesla - 100 Tesla would generate a pressure of around 40,000 atmospheres). That coil is, in turn, within a second coil consisting of twelve layers of copper wire that can generate 40 Tesla (it's in a plastic jacket that can withstand the 2,500 atmospheres generated in the larger coil). The combined field from the two coils is 90 Tesla and is only 16 mm wide through the throat of the coils, although Los Alamos National Lab has held a record for many years with a single coil that can generate 89 Tesla. These are used for materials science and superconductor research, as well as creating refrigerator magnets capable of destroying a home, kitchen, and refrigerator located inside, let alone the fridge door! :lol:
The best things in life aren't things ... but, a Pi comes pretty darned close! :D
"Education is not the filling of a pail, but the lighting of a fire." -- W.B. Yeats
In theory, theory & practice are the same - in practice, they aren't!!!

User avatar
cyrano
Posts: 718
Joined: Wed Dec 05, 2012 11:48 pm
Location: Belgium

Re: Pi in high magnetic fields

Thu Mar 13, 2014 7:54 pm

Burngate wrote:You missed one: http://en.wikipedia.org/wiki/Magnetic_refrigeration ~1 T, with 3 T or greater for NDR
I learned something entirely new today! Thanks for that!

matt_gsxr
Posts: 10
Joined: Thu Mar 13, 2014 3:55 pm

Re: Pi in high magnetic fields

Thu Mar 13, 2014 8:17 pm

Firstly, thanks for your helpful contributions. I suspected the ethernet connector (googling "magnet Rasp pi" throws this stuff up), although I had little to substantiate it.

Regarding the field strength issue. It is 7Tesla of static field (actually a little more in places). It is a research human MRI scanner that we get to play with (we call it work ;) ). More details here http://www.bbc.co.uk/news/health-19473136 (picture of magnet at around 3:14).

I don't want to extend the camera cable and put this thing out of the field because:
1) I have read that there is a limit to the cable length
2) The cable will transmit pick up RF (its an MRI scanner so lots of RF) which is undesirable

Long term I am probably to connect fibre optics to get the signals out, but before I started picking up those bits I wanted to check that what I have flies (it doesn't).

So, any solutions?
a) USB to ethernet adapter (how do avoid the same ferrite problem again?)
b) USB to fibre optic adapter (I like this plan, but can I find drivers for Rasp pi?)

Thanks for your help and ideas,

Matt

PiGraham
Posts: 4874
Joined: Fri Jun 07, 2013 12:37 pm
Location: Waterlooville

Re: Pi in high magnetic fields

Thu Mar 13, 2014 9:09 pm

Not my area of expertise, but maybe you could replace the RJ45 with a fibre module and run optical direct from the Pi's Ethernet differential logic signals (that go to the RJ45 connector/module now.
Not pin-compatible, because power supply lines are required, but worth a look.

User avatar
jbeale
Posts: 3893
Joined: Tue Nov 22, 2011 11:51 pm
Contact: Website

Re: Pi in high magnetic fields

Thu Mar 13, 2014 9:13 pm

A USB-Ethernet device seems workable. A standard passive USB cable can be 3 meters long, and much longer with an active cable (which has one or more built-in hubs to bost the signal). For a typical size diagnostic MRI scanner, I believe 3 meters away from the bore you're going to be well under 1 T in field strength; that allows you to put your ethernet connector with ferrites into the safe zone.

Just make sure you get a USB-Ethernet device that works with the Pi. The cheapest ebay device (Davicom / Kontron) does not; but it seems most others do: http://elinux.org/RPi_USB_Ethernet_adapters

If you want to spend > $1000, you can get a "USB 2.0 Fiber Extender" which does not need any driver, it appears to the Pi as a normal USB hub, but the data transport happens over fiber and this can go a long way away (100s of meters). For example: http://www.blackbox.com/Store/Detail.as ... fgodckgAog

At that price, of course, you're better off with a normal USB camera at the end of a long USB cable.

One other thing: if the MRI scanner is sensitive to RF interference itself, be aware that the R-Pi does generate some RF noise. It's usually not a problem, but your application is not usual.

matt_gsxr
Posts: 10
Joined: Thu Mar 13, 2014 3:55 pm

Re: Pi in high magnetic fields

Thu Mar 13, 2014 9:54 pm

What about simply swapping the ethernet socket for one without the ferrites?

hampi
Posts: 223
Joined: Fri May 31, 2013 11:29 am
Contact: Website

Re: Pi in high magnetic fields

Thu Mar 13, 2014 10:02 pm

The magnetoresistance might also change the properties of semiconductors at high fields, but I do not know if that is relevant here.

http://en.wikipedia.org/wiki/Magnetoresistance

PiGraham
Posts: 4874
Joined: Fri Jun 07, 2013 12:37 pm
Location: Waterlooville

Re: Pi in high magnetic fields

Thu Mar 13, 2014 10:21 pm

matt_gsxr wrote:What about simply swapping the ethernet socket for one without the ferrites?
I think the voltage isolation the magnetics provide might be an important factor in this application. Long wires direct to LVTTL pins in a 7T field would make me nervous!

matt_gsxr
Posts: 10
Joined: Thu Mar 13, 2014 3:55 pm

Re: Pi in high magnetic fields

Thu Mar 13, 2014 11:01 pm

Thanks

Will report back if I find anything worth posting.

Matt

jamesh
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 30118
Joined: Sat Jul 30, 2011 7:41 pm

Re: Pi in high magnetic fields

Fri Mar 14, 2014 9:25 am

matt_gsxr wrote:What about simply swapping the ethernet socket for one without the ferrites?
Ethernet won't work properly.
Principal Software Engineer at Raspberry Pi (Trading) Ltd.
Working in the Applications Team.

PiGraham
Posts: 4874
Joined: Fri Jun 07, 2013 12:37 pm
Location: Waterlooville

Re: Pi in high magnetic fields

Fri Mar 14, 2014 9:53 am

Ethernet can be connected without magnetics. Ti describe doing so here: http://www.ti.com/lit/an/slla327/slla327.pdf. Series capacitors are used for dc isolation and resistors handle biasing.

mikerr
Posts: 2828
Joined: Thu Jan 12, 2012 12:46 pm
Location: UK
Contact: Website

Re: Pi in high magnetic fields

Fri Mar 14, 2014 10:07 am

Remember the very first batch of Raspberry Pi in 2012 was delayed because it didn't have the magnetic ethernet sockets
(due to a factory error):

http://www.raspberrypi.org/archives/781

though whether it stopped ethernet working altogether, or just severly limited cable length I don't know.
Android app - Raspi Card Imager - download and image SD cards - No PC required !

User avatar
Jim Manley
Posts: 1600
Joined: Thu Feb 23, 2012 8:41 pm
Location: SillyCon Valley, California, and Powell, Wyoming, USA, plus The Universe
Contact: Website

Re: Pi in high magnetic fields

Fri Mar 14, 2014 10:13 am

matt_gsxr wrote:It is 7Tesla of static field (actually a little more in places). It is a research human MRI scanner that we get to play with (we call it work ;) ). More details here http://www.bbc.co.uk/news/health-19473136 (picture of magnet at around 3:14).
How the hell did you get the scanner to drop into building at π (3.14)??? :D

As an overgrown kid, I found the preps needed to drive the giant crane over the parking lot without crushing it the more interesting part of the video - talk about getting to play with really big trucks! :lol:

So, how do you generate an MRI image with a static magnetic field? Isn't the big, noisy rotating magnet in a conventional MRI machine moving to perform scanning, and therefore there's a big, honkin' rotating magnetic field permeating the space within the upwards of 800 tons of iron surrounding the MRI room? A household relative works for Varian, so I'm very familiar with your magnet (http://www.agilent.com/about/newsroom/p ... 10002.html).
The best things in life aren't things ... but, a Pi comes pretty darned close! :D
"Education is not the filling of a pail, but the lighting of a fire." -- W.B. Yeats
In theory, theory & practice are the same - in practice, they aren't!!!

gregeric
Posts: 1509
Joined: Mon Nov 28, 2011 10:08 am

Re: Pi in high magnetic fields

Fri Mar 14, 2014 10:33 am

Might be worth trying out a Model A Pi, & as jbeale suggested, hang a USB-ethernet converter at the other end of a long USB cable. There's one less chip on the A (the ethernet+USB hub) exposed to the strong magnetic field & any possible issues with it too.

User avatar
Burngate
Posts: 6483
Joined: Thu Sep 29, 2011 4:34 pm
Location: Berkshire UK Tralfamadore
Contact: Website

Re: Pi in high magnetic fields

Fri Mar 14, 2014 12:40 pm

mikerr wrote:Remember the very first batch of Raspberry Pi in 2012 was delayed because it didn't have the magnetic ethernet sockets
(due to a factory error):
http://www.raspberrypi.org/archives/781
though whether it stopped ethernet working altogether, or just severly limited cable length I don't know.
The way I understood it at the time, it wasn't the lack of magnetics per se., it was that txp, txn, rxp, rxn came out on different pins, and that there was nowt to maintain the DC offset at 3v3.

So, in theory, you could dismount the RJ45 and connect it on the end of a long shielded cable.

matt_gsxr
Posts: 10
Joined: Thu Mar 13, 2014 3:55 pm

Re: Pi in high magnetic fields

Fri Mar 14, 2014 1:06 pm

Jim Manley wrote:
matt_gsxr wrote:It is 7Tesla of static field (actually a little more in places). It is a research human MRI scanner that we get to play with (we call it work ;) ). More details here http://www.bbc.co.uk/news/health-19473136 (picture of magnet at around 3:14).
How the hell did you get the scanner to drop into building at π (3.14)??? :D

As an overgrown kid, I found the preps needed to drive the giant crane over the parking lot without crushing it the more interesting part of the video - talk about getting to play with really big trucks! :lol:

So, how do you generate an MRI image with a static magnetic field? Isn't the big, noisy rotating magnet in a conventional MRI machine moving to perform scanning, and therefore there's a big, honkin' rotating magnetic field permeating the space within the upwards of 800 tons of iron surrounding the MRI room? A household relative works for Varian, so I'm very familiar with your magnet (http://www.agilent.com/about/newsroom/p ... 10002.html).
Explaining how MRI works is too far off topic I think, but the magnetic field doesn't rotate. There are some switched magnetic field gradients that cause the noise, but these are tiny (<1/1000) perturbations in the magnetic field.

Sad about Varian (who were sold to Agilent) closing their manufacturing plant. Give my condolences to your relative.

Return to “Other projects”