-
-
Notifications
You must be signed in to change notification settings - Fork 130
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[BUG] QZ-Raspberry Pi 4 does not connect with Zwift on Windows 10 #3132
Comments
Are you able to run nrfconnect on pc to check how you see qz from rpi4 from both windows pc? |
I'm not familiar with nrfconnect, but yes. So install "nRF Connect Desktop" separately on the Windows 10 AND the Windows 11 PCs and then check their interactions with RPi4 QZ? |
Yes exactly. Theoritically we should see the same things
Roberto Viola
Software engineer and open source enthusiast
http://robertoviola.cloud
Il giorno sab 1 feb 2025 alle 21:01 yetanotherbrian <
***@***.***> ha scritto:
… Are you able to run nrfconnect on pc to check how you see qz from rpi4
from both windows pc?
I'm not familiar with nrfconnect, but yes.
So install "nRF Connect Desktop" separately on the Windows 10 *AND* the
Windows 11 PCs and then check their interactions with RPi4 QZ?
—
Reply to this email directly, view it on GitHub
<#3132 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAALYWAFHRCTA3IFVMUTC632NUR3BAVCNFSM6AAAAABWJSRTKCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMRZGA4TCNJRGU>
.
You are receiving this because you were assigned.Message ID:
***@***.***>
|
I've installed nRF Connect for Desktop v5.1.0 but I don't know what to do with it next. Do I need a Nordic dev kit to use it? I tried using the "Bluetooth Low Energy" module and the "Start scan" is grayed out, with "SELECT DEVICE" option at top telling me to "Connect a Nordic development kit to your computer". Does one of the modules work with any brand BT radio? I'm able to install the "nRF Connect for Mobile" on my iPhone just fine. |
I never used the nrfconnect on pc but should work as the ios one. I'm going to bed now |
I found a program called BLE Scout that was supposed to be able to connect to a device and show its characteristics. It worked fine on Windows 11 (pic below) however it caused the RPi4 to stop advertising the QZPI address when done with Windows 10. I didn't wait more than a minute to see if the RPi4 recovered, and just quit the application. Below are the QZ debug files for three attempts at getting the characteristics, first from the iPhone with nRF Connect, second with BLE Scout and Windows 10, and third with BLE Scout and Windows 11. I rebooted the RPi4 between each run, started the BLE analyzer after the BT icon in QZ went solid (as soon as I could anyways -- the Win11 machine is on the other side of the house so there was a delay), and then closed QZ with the GUI quit option as soon as I could. I diff'd the files a bit and see there is a GATT log entry for the iPhone and Win11 that isn't there for the Win10 PC. |
Ok I will check the log in the next days |
EDIT: debug logs are from starting QZ on the RPi4, waiting for it to connect to the KICKRv5, and then using the Windows "BLE Scout" program to try to Connect to the "QZPI" that was being advertised. Looking at the log files again, it seemed there might have been some noise from my Win11 PC being on the other side of the house and it not having a reliable connection to the KICKRv5. I took the Win11 PC out in the garage and captured the debug log from the RPi4 using both the onboard Intel Bluetooth radio and then separately using the TP-Link UB400 v1.6 Bluetooth adapter (and default/generic Microsoft drivers). Intel worked, UB400 failed, so I'm confident it has nothing to do with Windows 10 now. The RPi4 was restarted between each capture. ZIP of log also has my filtered version that made visual diff'ing a little easier. |
from the log i can't see anything unusual on the QZ side. the only thing that is very obvious is that on the FAIL one, no device is connected to the virtual interface of the raspberry at all. Stupid question: the UB400 is working fine for connecting zwift to your trainer directly? |
Not only does the UB400 fine work in the PC with Zwift alone, it works fine when using QZ on the iPhone beta (and release) apps. So problem somewhere in the Linux BT driver? I can rebuild code if there are debug levels that could be increased. The iPhone using QZ with the UB400 log is in the first message of the issue. |
ok with the BLE scout using the UB400 are you able to connect to QZPI? |
Yes, BLE Scout did not crash when attaching to QZPI. |
do you also see the same characteristic between running ble scout with intel and ub400? |
Disregard my last message. I was thinking you were asking about QZ on iOS-beta. I cannot Connect to the QZPI Characteristics with the UB400 using BLE Scout, however I can connect to it with the Intel BT. That is the only actional done in the debug log I attached -- let QZ start up and find my KICKR, then before doing anything else try using BLE Scout to connect to the QZPI name. |
that's the main thing. so you don't see the QZPI at all, or you can't connect to it? |
It is displayed, however BLE scout freezes when I try to connect to it.On Feb 3, 2025, at 7:34 AM, Roberto Viola ***@***.***> wrote:
that's the main thing. so you don't see the QZPI at all, or you can't connect to it?
—Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you authored the thread.Message ID: ***@***.***>
|
so it means that it can't connect somehow. i guess it's something about timing or similar |
I saw lots of compiler warnings when I built on the RPi 4 but assumed they were all to be expected. I don't know what the binary compatibility is between all the Pi's ... is there a binary you've built that I could try instead of my own? |
Yes it's in the release page here on github but I don't think it's the issue
Roberto Viola
Software engineer and open source enthusiast
http://robertoviola.cloud
Il giorno lun 3 feb 2025 alle 17:48 yetanotherbrian <
***@***.***> ha scritto:
… so it means that it can't connect somehow. i guess it's something about
timing or similar
I saw lots of compiler warnings when I built on the RPi 4 but assumed they
were all to be expected.
I don't know what the binary compatibility is between all the Pi's ... is
there a binary you've built that I could try instead of my own?
—
Reply to this email directly, view it on GitHub
<#3132 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAALYWHQIW4YEJ2GBTZLC332N6MUDAVCNFSM6AAAAABWJSRTKCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMZRGUZTIMJWHA>
.
You are receiving this because you were assigned.Message ID:
***@***.***>
|
I'm a noob when it comes to Linux, but I get "sudo: unable to execute./qdomyos-zwift: No such file or directory" when trying to run the release file. Sudo'ing to su doesn't make a difference (tho results in slightly different error message?). |
I see strings on qdomyos-zwift shows a "/lib/ld-linux-armhf.so.3", and I don't see that existing on my system. Again, Linux noob (beyond Unix in the 90s I have faint memories of. 🤣). Edit: but maybe that is a statically linked library. |
chmod a+x qdomyos-zwift
To put it executable
Roberto Viola
Software engineer and open source enthusiast
http://robertoviola.cloud
Il giorno lun 3 feb 2025 alle 18:32 yetanotherbrian <
***@***.***> ha scritto:
… Yes it's in the release page here on github but I don't think it's the
issue
I'm a noob when it comes to Linux, but I get "sudo: unable to
execute./qdomyos-zwift: No such file or directory" when trying to run the
release file. Sudo'ing to su doesn't make a difference (tho results in
slightly different error message?).
Screenshot.2025-02-03.092030.png (view on web)
<https://github.com/user-attachments/assets/d43054d5-2d7d-4ae0-b82e-88e0adc5b49b>
I see strings on qdomyos-zwift shows a "/lib/ld-linux-armhf.so.3", and I
don't see that existing on my system. Again, Linux noob (beyond Unix in the
90s I have faint memories of. 🤣).
—
Reply to this email directly, view it on GitHub
<#3132 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAALYWAZRJPEOTS5TP2LUFT2N6R3FAVCNFSM6AAAAABWJSRTKCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMZRGYZTOMJQG4>
.
You are receiving this because you were assigned.Message ID:
***@***.***>
|
Did you download the 32 or 64 bit? Your image is 32 or 64 bit? Must be the same |
I'm using "qdomyos-zwift" that is in |
This is the 32 bit version
https://github.com/cagnulein/qdomyos-zwift/actions/runs/13108575712/artifacts/2525637948
Roberto Viola
Software engineer and open source enthusiast
http://robertoviola.cloud
Il giorno lun 3 feb 2025 alle 18:44 yetanotherbrian <
***@***.***> ha scritto:
… Did you download the 32 or 64 bit? Your image is 32 or 64 bit? Must be the
same
I'm using "qdomyos-zwift" that is in
https://github.com/cagnulein/qdomyos-zwift/releases
—
Reply to this email directly, view it on GitHub
<#3132 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAALYWFM5IAEDVE5BIKNKU32N6TG5AVCNFSM6AAAAABWJSRTKCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMZRGY3DEMBVHA>
.
You are receiving this because you were assigned.Message ID:
***@***.***>
|
I believe I need the 64 bit? I'm using the default Raspberry Pi Desktop install from the Pi Imager. My "uname -a" in screen pic above says it is "aarch64". |
Yes I disabled auto build 64 bit because of bug of github actions in the
last month
Roberto Viola
Software engineer and open source enthusiast
http://robertoviola.cloud
Il giorno lun 3 feb 2025 alle 18:49 yetanotherbrian <
***@***.***> ha scritto:
… This is the 32 bit version
https://github.com/cagnulein/qdomyos-zwift/actions/runs/13108575712/artifacts/2525637948
I believe I need the 64 bit? I'm using the default Raspberry Pi Desktop
install from the Pi Imager.
My "uname -a" in screen pic above says it is "aarch64".
—
Reply to this email directly, view it on GitHub
<#3132 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAALYWCFNPJVTDBNHPIDJRD2N6T37AVCNFSM6AAAAABWJSRTKCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMZRGY3TGOJUHE>
.
You are receiving this because you were assigned.Message ID:
***@***.***>
|
Would it be worth trying to install the 32-bit Desktop on an SD card and trying the release build? I would like to get as close to a known good (that you are aware of) state as I can. The RPi4 w/ my 64-bit QZ works fine with the UB400 BT adapter when using Win11 though. |
As I said I don't think it's the cause
We can try to simplify the signal from qz to check when it starts to
connect.
I can do this in the next days
Roberto Viola
Software engineer and open source enthusiast
http://robertoviola.cloud
Il giorno lun 3 feb 2025 alle 18:56 yetanotherbrian <
***@***.***> ha scritto:
… Yes I disabled auto build 64 bit because of bug of github actions in the
last month
Would it be worth trying to install the 32-bit Desktop on an SD card and
trying the release build? I would like to get as close to a known good
(that you are aware of) state as I can.
The RPi4 w/ my 64-bit QZ works fine with the UB400 BT adapter when using
Win11 though.
—
Reply to this email directly, view it on GitHub
<#3132 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAALYWHHL43WRSUV4EOXZBD2N6UT5AVCNFSM6AAAAABWJSRTKCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMZRGY4DOMZTG4>
.
You are receiving this because you were assigned.Message ID:
***@***.***>
|
Ok, thanks! |
one question: did you try to toggle "Relaxed Bluetooth for mad devices" in
the experimental settings in the QZPI and restarting QZ?
Roberto Viola
Software engineer and open source enthusiast
http://robertoviola.cloud
Il giorno lun 3 feb 2025 alle ore 19:25 yetanotherbrian <
***@***.***> ha scritto:
… As I said I don't think it's the cause
We can try to simplify the signal from qz to check when it starts to
connect.
I can do this in the next days
Ok, thanks!
—
Reply to this email directly, view it on GitHub
<#3132 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAALYWFAB6P7O4HSNGXRSUL2N6YBFAVCNFSM6AAAAABWJSRTKCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMZRG42DONJUGE>
.
You are receiving this because you were assigned.Message ID:
***@***.***>
|
I had not previously tried. I've done so now -- no change in the Win10 behavior. Now probing QZPI from either nRF Mobile (iPhone 12) or AE Scout (Win11) will cause the QZPI device name to stop being advertised/detected. Previously only probing with the AE Scout on Win10 did that. |
Ok so probably the right spot? I mean changing it seems to extend the issue?
Roberto Viola
Software engineer and open source enthusiast
http://robertoviola.cloud
Il giorno mar 4 feb 2025 alle 17:15 yetanotherbrian <
***@***.***> ha scritto:
… one question: did you try to toggle "Relaxed Bluetooth for mad devices" in
the experimental settings in the QZPI and restarting QZ?
I had not previously tried.
I've done so now -- no change in the Win10 behavior. Now probing QZPI from
either nRF Mobile (iPhone 12) or AE Scout (Win11) will cause the QZPI
device name to stop being advertised/detected. Previously only probing with
the AE Scout on Win10 did that.
—
Reply to this email directly, view it on GitHub
<#3132 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAALYWFUSBOG75GR55BFNKL2ODRRTAVCNFSM6AAAAABWJSRTKCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMZUGQ2DKNBRGI>
.
You are receiving this because you were assigned.Message ID:
***@***.***>
|
Yes, I would say that makes it worse. |
Ok perfect so I will try to reduce that value layer today
Roberto Viola
Software engineer and open source enthusiast
http://robertoviola.cloud
Il giorno mar 4 feb 2025 alle 17:17 yetanotherbrian <
***@***.***> ha scritto:
… Ok so probably the right spot? I mean changing it seems to extend the
issue?
Yes, I would say that makes it worse.
—
Reply to this email directly, view it on GitHub
<#3132 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAALYWGNNIOGW4KLVUE33Q32ODRZXAVCNFSM6AAAAABWJSRTKCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMZUGQ2TCMJQGE>
.
You are receiving this because you were assigned.Message ID:
***@***.***>
|
Thanks. I will be around the systems all day (at least during GMT+1) if you have anything you'd like me to try. |
i created this, pull it and build it
#3146
let's see if it makes any difference. leave the mad device setting as
default
Roberto Viola
Software engineer and open source enthusiast
http://robertoviola.cloud
Il giorno mar 4 feb 2025 alle ore 17:23 yetanotherbrian <
***@***.***> ha scritto:
… Ok perfect so I will try to reduce that value layer today
Thanks.
I will be around the systems all day (at least during GMT+1) if you have
anything you'd like me to try.
—
Reply to this email directly, view it on GitHub
<#3132 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAALYWBTMD43HOKQLCTZDTD2ODSQ5AVCNFSM6AAAAABWJSRTKCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMZUGQ3DMNBYGI>
.
You are receiving this because you were assigned.Message ID:
***@***.***>
|
My noob'ness strikes again. I haven't seriously used a versioning system since ClearCase. What exactly do I need to do with git from the command line to get that branch (?). I am still on the cloned repository. Thanks. |
If you missed it, I don't know what I'm doing with git. I've tried the below, but still unsure how to pull it. |
I diff'd the virtualbike.cpp file on the repo and saw the changes to pars.setInterval. I modified my file and am building with that. I still need to learn the proper way to pull a dev branch/changes though. |
Unclear how to check out the branch, it looked like the change was 2x instances of pars.setInterval(100,100) changing to 10,10, and I made the change to my local file and built it. The QZPI appearing felt quicker to pop up in the Bluetooth monitors, but I don't think there was a change in behavior. The BLE Scout locked up when trying to Connect to the QZPI name/service from the Win10/UB400 adapter but continued to work fine from Win11/Intel and iOS. Attached are two zips, one with the relaxed mode enabled, the other with relaxed mode disabled. I wasn't sure what was meant by default. |
ok so i guess it's a dead end from the QZPI side. |
I tried the vendor's drivers in the very beginning when I was only logging the Zwift app showing but not maintaining a connection to the QZPI name. When I didn't see a difference from the generic Microsoft drivers I switched back to Microsoft's. I have downloaded some of the Bluetooth SIG docs recommended on the BlueZ.org site and will look through the code. I think leave the issue open for awhile and I'll come back later to provide an update if I've found the issue or gave up. In the meantime I'll use the iOS beta with the Win10+UB400+Zwift. |
Describe the bug
QZ built and running on fresh Raspberry Pi 4 Desktop does not connect to a fresh install of Zwift on Windows 10 using a TP-Link UB400 v1.6 Bluetooth 4.0 adapter (debug log "1--"). The QZ-RPi 4 DOES WORK with Zwift on iOS (debug log "2--") as well as a Windows 11 PC with embedded Intel Bluetooth.
The Zwift/Win10/UB400 DOES work fine with the QZ-iOS beta build (debug log "3--") so I don't believe Win10 is the entire issue.
To Reproduce
Expected behavior
Expect to see "QZPI" as available device. SOMETIMES it will appear and it can be selected. Selecting "RESISTANCE" on Paired Devices will not show QZPI as expected and will show the "Wahoo KICKR 0000 33" (virtual?) device and/or the regular "Wahoo KICKR C1E2 96" device. While on the RESISTANCE screen the POWER screen will unpair in the background (turn orange) and switching back to it the QZPI device will show "No Signal" and cannot be selected/will not be displayed if close/reopen/restart pairing screens.
Desktop (please complete the following information):
Windows 10, latest updates
newly bought TP-LINK UB400 v1.6 Bluetooth adapter. Have also tried with ~4 year old UB400 v1.0 Bluetooth adapter, and some other random ~4 year old'ish adapter.
Raspberry Pi 4, fresh install image updated/upgraded with git repository cloned on Jan 31.
Smartphone (please complete the following information):
iPhone 12 Pro Max,
iOS v18.1.1
QZ beta release
Append a debug log
QZ-RPi4-to-Win10-Zwift-Debug.zip
The text was updated successfully, but these errors were encountered: