How I solved (?) the EMI problem in my SimRig

Multiple months of debugging, finally seem to be paying off

10 min readFeb 6, 2022

I have been into simracing for about two years, give or take. During all that time, I have been plagued by mysterious issues where devices would disconnect and reconnect at will, or would simply misbehave in other ways.

I have been debugging this issue forever, and I think I’ve gotten to a place where the system is stable and reliable. I am writing this post to hopefully save others from the trial&error approach I went through.

My Simrig

First, the video below has a quick walkthrough of my simulation rig. As you will see, I have many motors and devices going at all times, and these as it turn out might generate all kinds of interference.

Electro-Magnetic Interference (EMI)

While you do not need to become a master at it, an understanding of EMI might be useful. In short, the electronic stuff you have in your rig (your wheelbase, all those actuator motors, the servo-drives, the amp for your shakers, and more) can emit electro-magnetic waves which will then interfere with your nearby USB cables.

This Wikipedia article has details, and there are many videos on YouTube if you are interested. Interesting to note that all these waves can travel through the air, through the cables, through the power cables, etc. They suck, pretty much.

Symptoms that might point to EMI…

The list above is not exhaustive, and EMI might not be the only reason for these symptoms. However, if you see the behaviors below, it is possible that EMI is the culprit:

USB disconnects: apparently for no reason, and apparently uncorrelated with anything specific that you did, some USB devices would disconnect and reconnect. In my case, my wheelbase, as well as my VR headset (and at some point my pedals) were disconnected mid-race.
USB misbehavior: even when Windows shows the USB device as “connected” and healthy, it simply does not work. In my case, the TensionR belt system would show connected, green all the way, and yet it would not work/respond to commands.
Degraded signals: you might notice the interference itself, in the form of degraded signals. For example, you might hear crackling noises out of the speakers, or perhaps the view in your VR headset flickers or the tracking of the VR headset drifts away.

The tools I used to try to diagnose the problems…

USBs are not well handled by Windows. That’s the basis for much of our pain. Sure you could use Device Manager to do your debugging, but it would be painful, and slow.

Instead, I found the following tools very useful:

USBDeview (https://www.nirsoft.net/utils/usb_devices_view.html) — it lists all the USB connections that Windows knows about, and lets you order, search and manipulate them easily.
USBLogView (https://www.nirsoft.net/utils/usb_log_view.html) — keeps a log of all USB disconnects and reconnects. Priceless to see what went offline when, especially when you are in the middle of a race.

What has apparently solved the problem for me

Before going further, you need to know that what worked for me might not work for you. In addition, I’ve done a bunch of things over these two years, and perhaps not all these things are useful or needed at all. Hence, I am only sharing the “collection of things” that ultimately seems to have fixed my problems. Do note that I myself did not think that ALL these things would be needed. And yet, it was not until I did them all that the system became reliable.

Thus, without further ado, the list…

Outlets properly cabled and grounded

Seems simple, but even in a relatively new house, one of my outlets was cabled wrong. All you have to do is get a simple tester from Lowe’s, Home Depot, or the equivalent store in your area, to confirm your outlets are properly cabled. This one is the one I used.

Once you have the outlets sorted out, you need to worry about grounding. The goal is to have a common ground for all the systems that take part in your simrig. The best approach for this is to plug everything into one outlet, and then that outlet’s ground point becomes the common ground for everything you have, and ground loops are minimized or avoided altogether.

However, sometimes the electric load of all your stuff is too much to hang off of one outlet (you would be tripping the breaker constantly), and in such a case, you can connect the ground pins of the two different outlets together. Resist the temptation of just replacing your breaker at the box for a bigger one. The risk you run if you do that is that the cable inside your walls is probably not rated for the 30+ amps you want to run through it. It might overheat, catch fire, or simply malfunction.

I have a lot of motors as I have explained above, and thus one line is not enough, so I have everything connected to two different outlets. However, I am connecting the grounds using a couple of these. You should connect them like the sketch below (courtesy of Thanos).

In the sketch above, the red boxes represent the power strips that are each plugged into a different outlet/electric line in your house. You use the dummy plugs above to connect these power strips together.

Once you have gotten a common electrical ground, you need to ground the rig. This is not always needed (as a matter of fact, I do not currently have mine grounded) but sometimes it helps reduce static discharges. When you do this, I would suggest you use a specific cable for it, like this one. Finally, you need to make sure you connect the grounding cable to your rig properly, through the anodizing and into the core/metal part.

Shielded cables, everywhere. Good cables, everywhere

Again, this one seems like an obvious one, but it bears repeating. All your cables, in particular those that deal with motors, must be well shielded. Most motion systems today sell with shielded cables but check nonetheless. Get good USB cables. Not only shielded, but if they have ferrite cores at either end it does not hurt.

For those USB cables that do not have ferrite cores, you can add them. Buying this pack from Amazon is relatively inexpensive. I have them and I have no idea if they help, but it is unlikely they hurt.

EMI filter for your servo-drive chain

You probably have a motion system, and its servo motors are drawing power from the outlet. I installed an EMI filter (this one) before I plug the chain to the outlet. In the picture below you can see it installed.

I have six servo drives, and thus I daisy-chained the 110v input. #1 is connected to #2, #2 to #3, and so on until finally #6 is connected to the EMI filter which in turn is connected to the outlet. The chart below shows how everything is connected in my system (module the cockpit grounding, which as I mentioned above, I have not done at the moment, but you should :-)

What if it all fits on one power line?

In the event that all the load does fit on one power line, the following diagrams and advice (courtesy of Thanos, once again) show how it can all be wired and connected.

In the diagram above, everything hangs off of one line, so you have by definition one grounding point. There’s the EMI filter, a switch, and the connection to the servos.

If you want to get even fancier about this, contact Thanos, but below you can see how you can package the filter, the switch, as well as a meter in one neat box, for a plug&play solution for your servo chain. This video explains how it is all connected.

Ferrites everywhere

Every cable that comes out of these servo drives is suspected of producing EM waves. The cables themselves are shielded, but at the point of connection, they are not. Hence, I decided to play it safe and install these green ferrites that you see in the picture below (these ones if you want to get them)

In the picture above, you will see a black and a white cable coming off the first two connections (from the top). Those are the power cables, which form a chain that goes from 1 to 2, 2 to 3…and 6 to the EMI filter. If you think about it, this is not as optimal as possible, for the cable that starts in servo drive 1, has to go all the way to 6 to finally reach the EMI filter. Therefore, whatever EM that is generated by #1 can travel freely until it gets to the filter. Thus, as currently shown, the system could still be improved by doing the following two things:

Twisting the black and white cables up until they are close to the connection point
Adding ferrites to the black and white power cables in each servo drive. (Update: I have done this, and added four green ferrites to each servo on these power cables.)

As you can see, the green ferrites have been liberally added to the cables. I have 14 ferrites total per servo-motor. It is quite possible (I would say certain at this point :) that fewer ferrites would have the same effect, but I’d rather err on the conservative side.

At one point I bought some ferrite cores by mistake. I mean, the ones I got were gigantic ones (I was intending to buy the green ones above but instead I got these). I could not return them, so I put them to good use, fitting the bunch of USB cables that go into the hub through them, as the picture below shows:

Pimax, wonderful Pimax…

One of the most sensitive pieces of equipment is my Pimax 8KX VR headset. I have been having problems with it pretty much since I first got it. Lots of effort with support had not yielded anything of substance.

In the end, the EMI solution might have helped. In any event, for the 8KX I have the cables running alone, unencumbered. The main cable that goes to the headset loops three times around a big ferrite core (this one), as you can see below.

What about Windows and your PC?

There is a lot to be said about the setup of Windows and the PC. First, is the physical connection. I have most of my USB devices connected to a powered USB Hub (this one), with the hub connected with a good USB 3.0 cable to the back of the PC.

Exceptions are the Thanos controller for motion and the TensionR belt system
The reason they are connected directly is that they are latency-sensitive.
Another device that perhaps deserves to bypass the hub is the wheelbase, but at the moment I have it connected through the hub.

Second, the drivers for Windows. I try to stay up to date, but I do not always succeed (as you will see below). I have not moved to Windows 11, and I am on 10 at the moment.

AMD and its USB woes

I happen to have an AMD X570 motherboard. As it turns out, the motherboard itself had issues with USB disconnects (EMI or not). There have been a couple of BIOS updates that purport to solve the problem. I am on the latest for my motherboard and things have improved. If you are on a B550 or X570 chipset-based motherboard, I strongly encourage you to try this first.

That said, I decided to buy a separate PCI-based USB card (this one), and I am running some of the connections through it, hoping to bypass the motherboard as much as possible.

Conclusion

It has now been several days without any major disconnects. Some minor disconnects have happened, but those are physical problems (e.g. the Aiologs sequential shifter disconnects on account of its USB connection being loose and the platform vibrating during a rally session)

In other words, seems to be stable. The green ferrites were the last addition. The stability extends to the Pimax, which has been rock solid for hours at a time (something I didn’t think was possible).

Of course, this is Windows, this is all very empirical and perhaps tomorrow I get a ton of disconnects, but I choose to remain optimistic. Hopefully some of this info is useful to you and saves you some time/money.

Good luck.

Acknowledgments

Throughout these two years, a ton of people have helped with their advice, experience, etc. Gary Tall, Thanos, Jeff Weaver, and many more. A couple of Facebook groups and Whatsapp gatherings have been key as well. Thanks to all.