Race Sim - Stopping analogue voltage / EMI problems

Discussion in 'Electronics & Electrics' started by Sn@Ke, Jan 14, 2020.

  1. Sn@Ke

    Sn@Ke Member

    Joined:
    Oct 5, 2003
    Messages:
    265
    Location:
    Sunshine Coast
    Hey all,

    First off, I know very little regarding electronics other than building PC's and the basics, so I'm not totally certain on my terminology, but I'll outline the symptoms and what we've tried and hopefully someone will be able to point me in the right direction :).

    So, I'm helping build a racing sim project and we've encountered an issue with our choice of steering wheel / base - it's a direct drive wheel (Fanatec DD1) it has really strong / realistic feedback, but this also means it's basically a big electric motor with the wheel attached and it seems to be putting out some crazy feedback, which is causing my other USB interfaces to freak out when the wheel is powered on (regardless if USB is plugged in or not).

    [​IMG]

    The problem:
    When the steering wheel is powered on, our brakes stop working all together (load cell on USB controller) and then we start getting random inputs from the shifter (second USB controller) where it will just randomly flicker other button inputs.
    Both the pedals and shifter run 'Leo Bodnar' Universal USB Interface Boards:
    http://www.leobodnar.com/shop/index.php?main_page=product_info&cPath=94&products_id=183
    (one controller on the pedals and a second one controls the shifter)

    The pedals / shifter work 100% perfect when the steering wheel is powered OFF, as soon as we power it on, the brake sensor stops, the potentiometers for brake and clutch start to jitter and the second controller for the shifter starts to jitter / flicker random buttons too.
    We contacted Leo Bodnar and they weren't very helpful, they just said it's analog voltage issue and not a problem with their product, which is true but not really giving me much to try out.
    We contacted the pedal / shifter manufacturer and they're aware of the problem with our setup and had success fixing their test rig by grounding the chassis, which we have done but no success.

    The whole setup is on an aluminium frame, made by "Simlab" and we have grounded the whole chassis to the mains power (electrician did this for us) so the frame is grounded.
    We have insulated the shifter / pedals as best we can using thick rubber from a car floor mat, we have also grounded each component too, which helps a bit but still not enough to use them.

    Album of photos to help description:
    https://imgur.com/a/BPlVrH8

    Things we have tried:
    • Spaced the steering wheel base off the frame with thick nylon washers so it's not touching.
    • Swapped bolts from steel to stainless non magnetic bolts.
    • Added thick nylon washers everywhere we could.
    • Ferrite cores on all USB cables, even smaller ones on the controller boards / wires to sensors.
    • We tried all different USB ports / combinations, but am sure it's not USB related as I can have the wheel base powered on and 1m away from the rig and everything works, as soon as I touch the wheel base and then touch the rig frame the USBs will get the feedback (via me bridging).
    • Run the PC off a separate power outlet to the wheel base / used extension cords so power is from different circuits.
    • Ghetto shielding around the USB controllers with tin foil / tin foil on the wheel base, attempted shielding etc.
    • Had electrician come and ground the chassis for us / made a cable we use.
    What DOES work:
    • Using the Fanatec brand racing pedals works fine, they're USB as well but have no issues with interference at all, I think they're better shielded from the issues we are having... But the owner is not a fan of these pedals as they don't have right feel to them for real racing.
    • If the problem pedal (sensitive ones) are a good distance from the rig, they work fine.
    We have also ordered an alternate USB controller from here: http://derekspearedesigns.com/12-bit-controllers.html
    Which maybe will have a better outcome.

    Any ideas / solutions would be awesome, we have spent MANY hours on this rig now and no success yet.
    We also have a second rig with similar parts that we use to troubleshoot but it seems the DD1 wheel really just puts out a heap of interference when it's running.
     
  2. aXis

    aXis Member

    Joined:
    Jun 27, 2001
    Messages:
    5,478
    Location:
    Kalgoorlie, WA
    Personally I would have grounded more items, not less, that way it wont matter if/when someone bridges it. Ideally you want to sink all of the radiated energy into the mains power ground, but preferably not your signal ground.

    Also maybe check that your mains earth is actually functioning well - back in the switchboard and at the external earth stake. The fact that one company was able to fix it but not you could be symptom of a house wiring issue.

    Another option is to plug some/all of the USB items into a USB isolator. Eg: https://www.adafruit.com/product/2107
     
    Sn@Ke likes this.
  3. OP
    OP
    Sn@Ke

    Sn@Ke Member

    Joined:
    Oct 5, 2003
    Messages:
    265
    Location:
    Sunshine Coast
    Thanks for the reply!

    Will get the electrician to come back out and check the mains power ground like you've suggested!
    We've also already tried grounding everything on the chassis of the sim, so maybe it is an issue with the building.

    The isolator idea might be good on the pedals / shifter USB cable too, I might buy a couple to test... at this point we are happy to try anything! The sim is pretty.

    Thanks again :)
     
  4. mtma

    mtma Member

    Joined:
    Aug 12, 2009
    Messages:
    5,245
    How is the wheel powered? Does it have AC going from the power-board to it or something else? Also then does its USB connect to the PC directly?

    The wheel is no doubt the aggressor. Foil shielding, connected to a ground works for airborne high frequency noise. If the actual noise is lower frequency inducted or conducted out through the connections to the device, the treatments are different. It may be that you need a power filter (example for visual depiction only) added as close as possible to the wheel to choke any power circuit noise it might still be emitting into the environment, as that generally will be the noisiest bit.

    A circuit or basic diagram of how things are connected to one another helps in working out how to treat individual issues.

    One diagnosis question I have is that if you connect the USB controllers to a laptop running from battery, do you experience the same level of degradation when the wheel is turned on?

    Is there anything on the bottom side of the Leo Bodnar controller? Another test I would do is bridge inputs with 10 to 100kohm resistors to their ground pins on the board and observe if there is any change in the behaviour of the spurious signalling. If there is no pullup or down it could be that spurious signalling could be expected as a matter of course.
     
    Last edited: Jan 15, 2020
    Sn@Ke likes this.
  5. merlin13

    merlin13 Member

    Joined:
    Apr 12, 2003
    Messages:
    1,954
    Location:
    back in BrizVegas
    I'd also be inclined to start poking around with a CRO on all the power and signal lines to see if you can isolate where the problem is appearing (and what the interference signal actually is, whether RF, EMI or both), then work on killing the interference path.
     
    Last edited: Jan 16, 2020
  6. Privatteer

    Privatteer Member

    Joined:
    Mar 19, 2006
    Messages:
    858
    I suspect it is a ground loop issue between the PC PSU and the separate PSU on the wheel.
    What size is its power supply, DC?

    As suggested above a signal isolator may work. Or powering it off the PC PSU if possible.
    It can also help noise issues sometime if you common the 2 negatives of different DC power supplies either directly or via ground.
     
    Last edited: Jan 19, 2020
  7. metalslaw

    metalslaw Member

    Joined:
    Feb 15, 2003
    Messages:
    220
    This may or may not help, but I thought I should make the suggestion.

    My Gigabtye board has whats called, "USB DAC-UP 2".

    Jump on here, search for "dac-up", https://www.gigabyte.com/au/Motherboard/GA-Z270X-Gaming-SOC-rev-10#kf
    This feature is on some of the latest boards as well. But, aside from the very clean noise profile, you can also add a little extra voltage to an output (0.1, 0.2, or 0.3v), which is good for any finicky usb devices that may suffer dropouts.

    A quick example, I had an old cycling computer that I left in a box too long. It's inbuilt battery would not charge from any regular power point, or usb cable. I upped the voltage to +0.3 on the dac-up 2 port, left it attached for a half hour, I reset it with a pin, and voila It turned on and charged. :)

    Anyway, just though I would suggest it, in case all other avenues are exhausted.

    Also, have you tested an addon pci usb board?
     
  8. Mjollnir

    Mjollnir Member

    Joined:
    Apr 20, 2002
    Messages:
    6,194
    Location:
    Melbourne, VIC3104
    A small Isolation transformer might help.
     
  9. OP
    OP
    Sn@Ke

    Sn@Ke Member

    Joined:
    Oct 5, 2003
    Messages:
    265
    Location:
    Sunshine Coast
    Thank you guys so much for the replies, some quick answers:
    • The wheel/base is running it's own power with a really beefy power brick, like xbox sized... This doesn't seem to matter if it's plugged into a different socket (other side of the house via extension cord) the USB is for signal inputs only.
    • The USB controllers (leo bodnar) are powered via the USB as well, so we have added ferrite cores to them, but I'm thinking a filter as mentioned about would be a great idea to test next.
    • The issue seems to be as soon as the wheel/base (steering wheel direct drive motor) touches anything metal on the chassis or even if you bridge it by touching it yourself and then touch the rig, it freaks out the other controllers, if there's an "air gap" you're good.
    • The USB control boards are mounting via standoffs you'd find on a PC motherboard tray, but we tried different mounts / shielding here too, we insulated with rubber, tried shielding with foil, earthed the boards from their common GND pin the the PC PSU (and wall socket earth) - no luck. Leo Bodnar support also informed me they don't recommend earthing their boards for whatever reason and prefer the insulation approach, but we tried both.
    Right now we have the electrician booked in to come back this afternoon and check the house / mains power earth, this house is brand new but it's a renovation of an old house, they apparently did update the whole power board and earth, the place is running cbus and all the fancy stuff, but I'm suspecting more and more it's an issue with the house wiring... Hoping the electrician can also do some testing on the rig to figure out exactly what the interference is so we can better treat it too. So that's step one, from here we'll order some filters if there's no luck... Will also bring the wheel to another location where we have a second race sim and see how it behaves here
     
  10. aXis

    aXis Member

    Joined:
    Jun 27, 2001
    Messages:
    5,478
    Location:
    Kalgoorlie, WA
    Sounding like a voltage differential between your two systems. Can easily happen with two isolated power supplies that have a loose coupling via an alternate path - eg EMF supression capacitors. I bet if you measured with a multimeter between the wheel/base and the USB sheild you'd see some DC or AC voltages.

    The fixes are:
    1) Make sure everything has the same common negative (and preferably link that negative to ground). This means you have to tie the PC negative rail to the wheel base negative rail, possibly via the chassis.
    OR
    2) Make sure everything is fully galvanically isolated, like using the USB opto-isolators.
     
    Sn@Ke likes this.
  11. RP_Automotive

    RP_Automotive Member

    Joined:
    Apr 19, 2006
    Messages:
    204
    I think shielding, ferrite cores and isolators are a waste of time here. There is a fundamental problem causing the issues. What’s even weirder is that the factory/Fanatec USB pedals work fine but others don’t. Perhaps the wheel and pedal combo has some sort of grounding between both units, but that would be unusual.

    If you’re ok with a multimeter and are willing to pull some of the USB parts apart, I would get a multimeter on the voltage and ground lines and see what happens when you plug things in. You might see for example a stray voltage or a severe voltage drop. That will point you in the right direction and you can take it from there.
    There may even be some software for your PC that will log USB voltages so you can see what happens?
     
  12. OP
    OP
    Sn@Ke

    Sn@Ke Member

    Joined:
    Oct 5, 2003
    Messages:
    265
    Location:
    Sunshine Coast

    So we had the sparkie back out to look at the setup, I'm not there but was on the phone call, he pulled apart the power for the wheel and made a common negative with that and the PC to the chassis... No luck.
    Also he ran a new earth from outside into the setup, has a long lead to test house earths with, so ran that in and also double checked the mains power box and earth stake etc, all were OK.

    So no joy...
    I'm going to go back next week and have another play around, but in the meantime I've ordered some USB opto-isolators, I've got the adafruit one and another no name brand I've found on ebay, but express postage, so I'll do more testing when they arrive and I go back.
     
  13. aXis

    aXis Member

    Joined:
    Jun 27, 2001
    Messages:
    5,478
    Location:
    Kalgoorlie, WA
    Bugger!
     
  14. cvidler

    cvidler Member

    Joined:
    Jun 29, 2001
    Messages:
    13,098
    Location:
    Canberra
    that's not gonna help....

    you need electrically isolated hardware. You can get nylon (plastic) bolts. They may not be enough (strength) for actual use, but for testing, isolation of the wheel from the frame, they'll work.

    If indeed when the wheel is electrically isolated from the frame, then you need to figure out how to mount it permanently, with non-conductive hardware.

    So with nylon hardware, the nylon stand offs keeping the whole wheel isolated from the frame, I reckon will be your fix.
     
  15. OP
    OP
    Sn@Ke

    Sn@Ke Member

    Joined:
    Oct 5, 2003
    Messages:
    265
    Location:
    Sunshine Coast
    GOOD NEWS EVERYONE!

    Ok so, we had the sparkie come back out, we legit grounded and tested EVERYTHING, ran another ground into the house from outside, he even pulled apart the power supplies on the wheel and earthed that with the PC PSU and the Chassis, we used nylon bolts and nylon washers and tried to insulate all the things... around and around... no luck!

    Then the glorious $24 USB Isolator shows up from China, it looks terrible quality, didn't work out of the packet but fiddled with the switch got it working... anyway hope is down, there's no way this thing is gonna work... BAM fixes it.
    So we've ordered a couple more for good measure, but it worked!

    Thanks everyone for the replies and help, it's just taken forever to get my hands on a usb isolator, but can confirm that it totally fixed it.
    We are going to leave everything earthed but swap back to the stainless bolts, do a tidy up of wiring etc, then it's racing as usual.

    Out of all the support groups, hotlines and the manufacturers, OCAU were the only ones to suggest the isolator, so thanks again!
     
    metalslaw and Raptor_Eye like this.

Share This Page

Advertisement: