eVOLVER Forum

GUI Errors after new installation

Hi all,

After a hard drive failure, I am reinstalling the evolver software and getting things and getting things up and running again.
I have installed the DPU, configured a router to connect the evolver and our PC, and installed the GUI. Now, I am trying to run new calibrations and I am getting errors when I try to use the desktop GUI.
First, when I open the GUI I get this error message:
Screenshot 2021-08-16 194936

Second, when I try to run a temp or OD calibration, the desktop GUI does not display any values from the sensors or record the values when I hit the “play” button:

When I open the calibration method on the evolver machine, however, it correctly displays values and appears to be working correctly.

Do folks have any thoughts about what might be happening and how to get the desktop GUI working?

Thank you.

Hi!

It seems to me that your GUI app is not able to reach your eVOLVER. You may want to double check your router configuration and your connections. Also, you can try to find your eVOLVER from the GUI using its IP and see what that gives you (see screen capture below).

About the first error, I’d say it should not affect the correct functioning of the app.

Hope it helps!

Hi Dani,

Thank you for the guidance and sorry for the late reply. As best I can tell, the router configuration looks fine


image

Additionally, when I try to access the evolver using the calibrate.py script it hangs here:
image

Overall, it seems like I am having problems similar to this thread:

Referencing this thread, I went ahead and updated the calibrations format but still have not been able to get the evolver to work properly with the GUI or dpu.

Is there anything I may be missing?

Thanks!

1 Like

Hi @santibk,

Can you show a screenshot of the calibrations file on the RPi? It should be at ~/evolver/evolver/calibrations.json. If this is empty or non-existent you will unfortunately need to upload the dummy calibrations file located here and then re-do your calibrations.

The calibration update script referenced in the other thread only works on a specific version of the old calibration files. We went through a couple of iterations before settling on the current version.

Thanks,
Zack

2 Likes

Hi Zack,

Here is a screenshot of the contents of the calibrations.json file (apologies for the blue text on blue background):

Additionally, when I navigate the GUI on the evolver’s touchscreen I am able able to see all of my calibrations there.

Santi

1 Like

Ok great, glad that isn’t the issue.

Next thing is can you show a screenshot of the logs? On the RPi:

sudo supervisorctl

then

tail -f evolver

You should see messages flowing up the screen every 20s or so.

Zack

@santibk and I spoke directly - it looks like there was a software mismatch between the evolver server code and the GUI/DPU. Hopefully updating the evolver server and arduino code will resolve the issue.

Guide for updating server.
Guide for updating arduinos.

1 Like

Update:

After updating the evolver server (conf.yml and all python files) and the arduinos. I took another shot at controlling the evolver via the GUI. This time, the calibration files show up but I cannot click on them or see the sensor readings from the machine

When I ssh into the RPi and look at the logs, they appear as follows:

Any ideas what I might want to try next?

Thanks,
Santi

I think there are two issues here:

  1. I think if you manually update the server, you might have to manually set the “active” calibration file. Otherwise, it will probably error out. You can find it on the calibrations.json example from the previous posts. Make sure it has a valid "active": true entry on the json that you want to be active.

  2. Seems like data isn’t properly getting to the Pi from one of the Arduinos (OD135) To properly communicate with the OD135 photodiode, seems like a longer delay would help. Here is instructions on how to do this. Pumps not working in chemostat mode - #5 by heinsz

Zack can correct me if I’m wrong! Good luck!

2 Likes

I think you’re right about the active cal not being set. Manually setting at least one of the fits to true for both temp and od should resolve that.

Pretty sure the od135 errors in the log are harmless - the od90 parameter is the one that matters. It’s a bit confusing as their setup is using 135 degree, but the arduino code for OD is handled in the OD90 sketch. The OD135 address is in the RS485_PD file for the dual LED/PD setup for wide range OD readings. We should probably just remove that parameter from the default conf.yml file on github since most people don’t have that setup, or think about renaming those files and addresses.

1 Like