Did your X570 motherboard already exhibit the USB disconnect issue with the VPForce controller? Not

Did your X570 motherboard already exhibit the USB disconnect issue with the VPForce controller? Not all X570 boards present this problem unless you're unlucky like me. I think there is chip to chip variation with AMD's B550/X570 chipsets. I wouldn't recommend using an external USB hub since the hub most likely has a switchplexer chip instead of a totally different USB controller chip. Most PCIe USB cards only uses PCIe 3.0 slot. Do you not have this slot available?

MLNL•1/3/23, 11:49 PM

I've had weekly usb dropouts with my SFX100 system and it now occurs to me this could be the problem. I am now trying to start up Walmis his script to scan the usb ports but i'm a total noob with Python and Powershell. Nope all PCIE slots are out. It's a config problem with this board. You can only choose PCIE-1+M1+M2 or PCIE-1+2 and some variations . But my startup disk is M1 and thats just the one wich can't be missed.

MMLNL I've had weekly usb dropouts with my SFX100 system and it now occurs to me this ...

HomeOfMomoOP•1/4/23, 12:05 AM

Don't bother running Walmis' Python script if you don't have VPforce controller. The script first looks for VPforce controller and bombards the controller with USB packets unless you know how to modify for your SFX100.

HHomeOfMomo Don't bother running Walmis' Python script if you don't have VPforce controller....

MLNL•1/4/23, 12:14 AM

Àh, that saves me a lot of hassle

No I'm not that skilled @ this moment to mod a script to the sfx system.

HHomeOfMomo Did your X570 motherboard already exhibit the USB disconnect issue with the VPFo...

walmis•1/4/23, 7:52 AM

Hey, I'm thinking of mailiing a board to you with a different MCU. Maybe it's got something to do with USB controller hardware on the MCU itself. And I can't reproduce this locally

Wwalmis Hey, I'm thinking of mailiing a board to you with a different MCU. Maybe it's go...

HomeOfMomoOP•1/4/23, 8:52 AM

Do you mean a different MCU unit from the same vendor/model as what I have or different MCU by a completely different vendor? I haven't had any issues once I used a PCIe USB card to talk to the VPforce controller I have.

HHomeOfMomo Do you mean a different MCU unit from the same vendor/model as what I have or di...

walmis•1/4/23, 8:55 AM

Yeah, I have used different stm32f103 compatible chips, so maybe there are some nuances that don't play nice with AMD USB chipsets

walmis•1/4/23, 8:56 AM

One more thing to try is to increase the USB interrupt priority (in the firmware)

HomeOfMomoOP•1/4/23, 8:56 AM

I just don't want you to waste money sending me another VPforce controller if the problem is to do with AMD's chipset.

walmis•1/4/23, 8:57 AM

But I'm interested in narrowing it down, whether the issue is some specific MCU + AMD combination

Wwalmis One more thing to try is to increase the USB interrupt priority (in the firmware...

HomeOfMomoOP•1/4/23, 9:03 AM

I'm not familiar with the specific MCU on my VPforce controller board. Do you happen to know if this MCU has a register setting to increase the differential driving strength (usually in the form of increasing current)? Sometimes driving hard can increase signal to interference/noise ratio. If there is, you can build a custom firmware for me to test out for the sake of curiosity.

ddrake1984•1/4/23, 1:33 PM

even if the setting for driving strength is turned off completely, the crash still occurs

ddrake1984•1/4/23, 1:34 PM

I'm not sold on that it's an issue with the forces applied

smitty•1/4/23, 2:16 PM

i'm getting this error too, i saw above maybe the

report_id

report_id

is incorrect? what should it be? Asking because i also have a X570 and every FFB program i am trying crashes in some way.

walmis•1/4/23, 2:19 PM

try report_id=106

walmis•1/4/23, 2:19 PM

uhh this sucks

walmis•1/4/23, 2:19 PM

need to find a workaround

smitty•1/4/23, 2:20 PM

same deal

smitty•1/4/23, 2:20 PM

is a PCI USB card in my future?

smitty•1/4/23, 2:23 PM

let me try restarting.. it may be stuck from a crashed airforce player

smitty•1/4/23, 2:25 PM

yep. 106 is firing away

smitty•1/4/23, 2:26 PM

oh wait it stopped

walmis•1/4/23, 2:26 PM

weird that it's 0ms

walmis•1/4/23, 2:27 PM

it should run at ~1 ms

smitty•1/4/23, 2:27 PM

most of them were

walmis•1/4/23, 2:27 PM

does the USB reconnect or drop out?

smitty•1/4/23, 2:28 PM

no, theres no usb disconnect sound or anything. your configurator still talks to it fine. but can't run hidtest anymore, instant error.

walmis•1/4/23, 2:29 PM

hmm, maybe the endpoint is in NAKed state... for some reason

smitty•1/4/23, 2:30 PM

gotta run for a little bit, have a meeting to connect to

walmis•1/4/23, 2:30 PM

roger that, perhaps if you have time we could make a remote desktop session, could try a few things

Ssmitty no, theres no usb disconnect sound or anything. your configurator still talks t...

HomeOfMomoOP•1/4/23, 5:54 PM

With Wireshark capture, I found that the problem was that the ACK from Endpoint responding to Host's Interrupt OUT got stuck in Host's buffer inside AMD's USB controller. This ACK was never subsequently properly processed and fed back to the host program in the PC. For programs like DCS, it does not allow for any non response from Endpoint, therefore USB hangs causing DCS itself to freeze until USB gets reset. VPforce configurator, on the other hand, doesn't seem to care whether it gets ACK or not so it continues to chug along. The fact that the ACK finally showed up after either we unplugged the USB cable or Walmis' hidtest program cleared itself after HIDException means that the ACK was properly sent by Endpoint (VPforce controller) but, for some reason, was never processed by the USB Host controller inside AMD's chipset.

walmis•1/4/23, 5:57 PM

I think I might be able to workaround this, @smitty981 ran some tests using a test firmware, and I found out the HID OUT endpoint gets stuck in NAK mode, so I will try to reset it back to active mode

Wwalmis I think I might be able to workaround this, @smitty981 ran some tests using a te...

HomeOfMomoOP•1/4/23, 6:17 PM

Good eyes! I didn't bother to look at the content in the ACK sent by Endpoint. The bad ACK has "IRP USBD_STATUS: USBD_STATUS_CANCELED (0xc0010000)" instead of "IRP USBD_STATUS: USBD_STATUS_SUCCESS (0x00000000)". This changes the whole dynamic. I wonder if it's possible to increase the driving strength of AMD's host USB differential driver...

walmis•1/4/23, 6:19 PM

I doubt that the signalling causes this, it's running at 12mbits, I have 33ohm series resistors on the board, so should be nice and clean

Wwalmis I doubt that the signalling causes this, it's running at 12mbits, I have 33ohm s...

HomeOfMomoOP•1/4/23, 6:31 PM

Then I have trouble understanding why there's never an USB issue when the 20V supply is shut off...

walmis•1/4/23, 6:32 PM

hmm, perhaps because there's no motor communication = less interrupt load

walmis•1/4/23, 6:33 PM

might be some race condition, that's amplified with AMD board

walmis•1/4/23, 6:34 PM

ahh, the wonderful world of embedded systems

smitty•1/4/23, 6:36 PM

ha, my rudder board had motor power and the joystick didnt.. hidtest still running on joystick board no errors

Wwalmis ahh, the wonderful world of embedded systems

HomeOfMomoOP•1/4/23, 6:37 PM

Haha! I don't envy sitting on a boiling plate... After all, AMD should have done something about it!

ddrake1984•1/5/23, 12:15 AM

how long do I leave the python run for?

ddrake1984•1/5/23, 12:16 AM

ddrake1984•1/5/23, 12:26 AM

I kept it running, loaded into a DCS module viggen, exited the module and the python script stopped working

ddrake1984•1/5/23, 12:26 AM

and now the script doesn't want to start

ddrake1984•1/5/23, 12:38 AM

whatever effect 7 is, is the one that hangs the system

ddrake1984•1/5/23, 12:52 AM

surprisingly, I was just in the spitfire and effect 7 was on/off continually, no issues, but as soon as I jumped into a mosquito, it happened pretty quickly

smitty•1/5/23, 5:53 AM

is there a way to change the singlemotor kit to the other axis? xpforce seems to want it on X