0
\$\begingroup\$

During a long-term, almost six month long, data acquisition in a proximity of scanning electron microscope I have found a significant source of magnetic field noise, which is altering the results. I have used the Spicer SC11 Magnetic Field Analysis system and focused on narrow-filtered RMS readings at 50 Hz. The readings are about 5 mG RMS, when the power was down the readings dropped to 0.3 mG RMS.

The problem is that the power line supplying the breaker panel is in the ceiling on the corridor next to the microscope and powers three precious devices drawing >5 kW each. Each consists of High voltage units, electromagnetic optics and supporting compressors and chillers.

The idea is to relocate the power line different way, which should reduce the EMF reading at the microscope. The microscope is supposed to be in a Farraday cage and the active EMF shielding is also installed in the lab.

Are there other options how to reduce 3 x 230 V @ 50Hz power line (IR) radiation to alter the results? Shieldings covering the cables, proper cable management, special cable arrangements, actual Farraday cage,etc.

So far I have found strategies how to protect wires and devices from the field outside. And some weirdos with aluminum hats. My target is to keep the field inside protecting the outside.

\$\endgroup\$
  • \$\begingroup\$ Why are you focussing on readings at 50 Hz - that's usually asking for trouble. Convince me why 50 Hz is your target measurment frequency. \$\endgroup\$ – Andy aka Nov 28 '18 at 19:40
  • \$\begingroup\$ Because I read the spectra and peaks are at 50 Hz and its harmonics. Also when the image is acqured slowly the straight vertical lines become sinusoidal. \$\endgroup\$ – Crowley Nov 28 '18 at 19:55
  • 2
    \$\begingroup\$ If the microscope is "supposed to be in a Faraday cage" why isn't it? \$\endgroup\$ – Solar Mike Nov 28 '18 at 20:10
  • \$\begingroup\$ @SolarMike Because contractor's contractor's contractor's contractor contracted someone who did not care at all during building. The whole room was supposed to be a Faraday cage and still I can make calls there... Seems some rural places here have better shielding.The cables were drawn different way than they are placed because it was cheaper... \$\endgroup\$ – Crowley Nov 28 '18 at 20:15
  • 1
    \$\begingroup\$ @Crowley should have been thoroughly tested before acceptance then... \$\endgroup\$ – Solar Mike Nov 28 '18 at 20:36
2
\$\begingroup\$

It seems to me that you need magnetic shielding if the electron microscope is affected. Here's what I would try:

  1. Use your gaussmeter to determine if the source of the magnetic is actually the power line. 5 milligauss seems like a lot if you are any distance away, because if the power cable contains all three phases, the net current in the cable should be near zero. The magnetic field contributions from the three phases should cancel each other if you are a reasonable distance away. It can be checked by measuring the field as you get closer to the power line; the field strength should be inversely proportional to the distance from the cable if the cable is the source of your field.

  2. If the cable is not your source, look at areas where the three conductors are not close together (junction boxes, switches, etc). Any individual conductor separated from the other conductor forms a loop which will produce a large local magnetic field. Your gaussmeter will help you check for these also. If possible, move these. If the power conductors are run on separate wires, have them bundled together.

  3. Check your furniture and structural fixtures. Steel structures, tables, etc can distort and locally concentrate the magnetic field in an area. Once again, your gaussmeter can help you here.

  4. If the power cable is your problem and the three conductors are close together, I would try buying some highly permeable magnetic shielding conduit (something like MuMetal). It is normally used to shield the cable from outside influences, and I have never tried using it in this manner. You could probably have a cable tray with a cover fabricated but I know there are manufacturers that provide flexible conduit. For this to work, you don't need to sheath the entire cable, just the length near your measurement for some distance in each direction. You can get a short length and use your gaussmeter to see if the field is attenuated, then make the decision whether to try it.

\$\endgroup\$
1
\$\begingroup\$

Loop area is the enemy here, reduce the loop areas (on both the generating and receiving sides) and you will reduce the coupling.

Twisting the wires forming the loops helps a LOT, for all that telling an electrical contractor that they need to twist the wires forming each circuit at a twist per inch tends to be a 'difficult' conversation (But it will make a meaningful difference). Do the same for the wiring between the controller and any deflection coils or such in the electron optics (Obviously there is not a lot you can do with the loop formed by the e- beam itself).

Steel conduit for all the power wiring will probably help some (as long as all the joints are electrically sound), but at 50Hz you are looking at quite a lot of skin depth in steel. Transformers should all have copper flux bands fitted as should things like compressor motors, and any VFD filters.

Sometimes a heavy steel plate six inches or so wide between the power services cable tray and the sensitive equipment can be surprisingly effective at shunting the flux from the existing wiring, and it is usually cheap to try.

I have never dealt with an electron microscope, but had this nightmare with an old flying spot film scanner.

Can you shield at least the low energy end of the accelerating column with some steel plate? There is probably a mu-metal shield already, but that stuff saturates very easily, the steel will take most of the flux leaving the mu-metal to do what it does best with the residual. My flying spot scanner had a mu metal shield, but adding an external heavy steel one still helped a lot.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.