Cable and equipment shielding

[logjam]

I've been trying to learn more about shielding equipment and cables.
I've bought some books and read some archived posts but haven't found
anything like what I'm trying to do.

I have an analog sensor with a sampling rate of 1-2MHz and 12-14bit.
The cable we use has an overall shield and each of the twisted pairs
are also shielded. I am getting what I assume is EM interference.
When the system is used near a magnetron at 6-15MeV we get a very
small but annoying amount of noise. About 150mV positive only spikes.

We have simulated the noise using a TIG welder's high voltage arc
start feature. When not welding, just letting the torch point
straight up we get a HUGE amount of noise, much more than we get near
the magnetron. What ever type of noise this is, it can be picked up
by an ionization chamber quite well. Our only choice is to use this
source of interference as a test of our equipment before we travel to
the customer's site. When we touch the source of the HV, turning our
bodies into antennas I guess, the sensor quits working completely. In
fact, a motion control system in the next room crashed and started
free running into its hard stops!

From what I understand the power is supplied to the magnetrons via a
coaxial at 10,000v 1000-10,000A in extremely short bursts at a maximum
of 200-250 pulses per second. There are then transformers which step
that up and finally something magic happens in the magnetron and we
get 6-15MeV.

The high voltage pulse cables (10,000v) are extremely well shielded.
They are running in the same cable tray with some unrelated BNC cables
(for about 110 feet) and with a 5vp-p signal there is no
interference. That leads me to think that the interference is coming
from the destination. We can only modify our equipment, so I am
looking into what it will take to shield it from everything possible.

My biggest question is what to do with the cable shield. Which end do
we connect it to? I've read in the past that you should connect the
shield to the end where you care about the signal MOST. Other people
suggested you connect the shield on the opposite end of where you care
about the signal. Should it be connected on both ends?

Thanks : )

Grant

 

[Mark]

I've been trying to learn more about shielding equipment and cables.
I've bought some books and read some archived posts but haven't found
anything like what I'm trying to do.

I have an analog sensor with a sampling rate of 1-2MHz and 12-14bit.
The cable we use has an overall shield and each of the  twisted pairs
are also shielded.  I am getting what I assume is EM interference.
When the system is used near a magnetron at 6-15MeV we get a very
small but annoying amount of noise.  About 150mV positive only spikes.

We have simulated the noise using a TIG welder's high voltage arc
start feature.  When not welding, just letting the torch point
straight up we get a HUGE amount of noise, much more than we get near
the magnetron.  What ever type of noise this is, it can be picked up
by an ionization chamber quite well.  Our only choice is to use this
source of interference as a test of our equipment before we travel to
the customer's site.  When we touch the source of the HV, turning our
bodies into antennas I guess, the sensor quits working completely.  In
fact, a motion control system in the next room crashed and started
free running into its hard stops!

From what I understand the power is supplied to the magnetrons via a
coaxial at 10,000v 1000-10,000A in extremely short bursts at a maximum
of 200-250 pulses per second.  There are then transformers which step
that up and finally something magic happens in the magnetron and we
get 6-15MeV.

The high voltage pulse cables (10,000v) are extremely well shielded.
They are running in the same cable tray with some unrelated BNC cables
(for about 110 feet) and with a 5vp-p signal there is no
interference.  That leads me to think that the interference is coming
from the destination.  We can only modify our equipment, so I am
looking into what it will take to shield it from everything possible.

My biggest question is what to do with the cable shield.  Which end do
we connect it to?  I've read in the past that you should connect the
shield to the end where you care about the signal MOST.  Other people
suggested you connect the shield on the opposite end of where you care
about the signal.  Should it be connected on both ends?

Thanks  : )

Grant

connect the sensor cable shield to the sensor at the sensor end and
connect the cable shield to your test equipment at the test equiment
end but.... isolate the sensor and cable grounds from the other
system.

keep your sensor and test system electrically isolated from the other
system except AT ONE PLACE only. This will usually be at the 3rd
prong of the power cord.

You want your system to have its own totally enclosed shield that
doesn't connect to the other system except at one place only.

This will eliminate ground loops.

If your sensor and cables are all well shielded and you eliminate the
ground loops and you still have a problem then we have to dig
further.

If your sensor MUST be electrically connected directly to the other
equipment for some logistics reason, then the problem becomes MUCH
more difficult.


Mark