Debugging EMI in the Simrig

4 min readOct 1, 2021

What a pain…

This write-up documents a bunch of measurements, hoping I can find a solution for the issues.

The Problem

I get sporadic USB disconnects, and Pimax blackouts. One hypothesis is EMI. Since I have two EMI filters and, in theory, all cables are shielded, I needed to do a bit of debugging.

Have done the debugging. Have no idea where to do from here.

The Rig

The following video has a walkthrough.

Four servo drivers (1 through 4) are from the SFX-150 setup that I bought used, and the last two servo drives (5 and 6) and their associated cables came with the LowRider platform.

The EMI filters

I have all six servos connected on a main (jumping from one to the other, and servo 6 is connected to an EMI filter, and then to the power strip). The jumper cables are shielded.

The other EMI filter is inside the ShakerBox. I have a power supply in there, and that power supply is connected to the LOAD side of the EMI filter, whereas the other side is connected to the power strip, of course.

I am not sure the EMI filters are working, or possibly I am measuring things incorrectly.

The above pic shows the measurement at the LOAD side of the servos’ EMI filter. About 300 V/m.

And this one is the other side of the filter. 227 V/m. Is that it? That’s less than 50% of “filtering”. When I’ve seen videos the values have jumped to almost nothing. Maybe the measurement is wrong?

Then there’s the ShakerBox EMI filter. The picture below shows the measurement roughly where the “input” to that filter is (I did not disassemble the box)

Which sounds reasonable… And yet the following picture is a measurement of the LINE side (the cable that connects the ShakerBox to the power strip), and notice the value!

The cables

I found a discrepancy here as well. All cables go from the servos to the actuators together, sort of. The four SFX ones separate from the LowRider ones after getting out of the Tuffiom enclosure and go their separate ways, as you can see below

I found marked differences when measuring the cables. In theory, all of them are shielded, and the EMI at the servo output is similar for both….so I am a bit at a loss.

First, these two images below show the measurement at the servo side for both #1 (SFX) and #6 (LowRider). Very similar both of them (as expected)

I took a number of measurements of the SFX cables (together and separately). In general, well below 100 (in the 60s). When I measured the whole bunch of cables I got 101 V/m. I guess this stuff adds up?

But then, when I measured the LowRider cables, I got substantially higher readings… as you can see below. Can it be that it’s because they are on the ground? Shouldn’t, but who knows… 300s…

An USB device, at its input

I wanted to see how much EMI was present at the input of one of my devices, so I picked one at random, and measured. Relatively decent 80 V/m

The rig, and the LowRider

Finally, I measured the large metal surfaces. The rig shows more than any SFX cable, but still relatively small. The LowRider, on the other hand, is in the 300s (which I suppose makes sense given the cables that connect to it)