add example for 2-axis piezo controller in device_configs/user_template.yaml #204

2026-05-07T14:45:12+02:00

diaz commented

2026-05-07 14:45:12 +02:00

I was testing the devices/npoint.py ophyd to control the 3-axis and the 2-axis nPoint controllers with the help of Klaus. We found that the current way to change the settling time of the piezo is:

dev.npx.settling_time.put(0.08)

And this can be used to get the current value:

dev.npx.settling_time.get()

I checked that 2 axes form 2 different controllers can be added simultaneously in the yaml file. I have thus added an example to configure an axis from the 2-axis controller in the device_configs/user_template.yaml file. The main difference with the 3-axis controller is the hostname and the limits of the motion.

We think that they way it works is that when connecting to each controller, it reads the corresponding range directly from the controller, so that the correct conversion from voltage to distance in microns can be done. Perhaps we should check this once with a microscope, I will discuss this with @holler next week.

Perhaps another thing to report is that I typically had to load the config file in bec more than once until it worked... I found this a bit strange but as long as it works I am happy.

Thanks for your patience @wakonig_k

I was testing the devices/npoint.py ophyd to control the 3-axis and the 2-axis nPoint controllers with the help of Klaus. We found that the current way to change the settling time of the piezo is: dev.npx.settling_time.put(0.08) And this can be used to get the current value: dev.npx.settling_time.get() I checked that 2 axes form 2 different controllers can be added simultaneously in the yaml file. I have thus added an example to configure an axis from the 2-axis controller in the device_configs/user_template.yaml file. The main difference with the 3-axis controller is the hostname and the limits of the motion. We think that they way it works is that when connecting to each controller, it reads the corresponding range directly from the controller, so that the correct conversion from voltage to distance in microns can be done. Perhaps we should check this once with a microscope, I will discuss this with @holler next week. Perhaps another thing to report is that I typically had to load the config file in bec more than once until it worked... I found this a bit strange but as long as it works I am happy. Thanks for your patience @wakonig_k

diaz added 1 commit 2026-05-07 14:45:12 +02:00

add example for 2-axis piezo controller in devoces/user_template.yaml

CI for csaxs_bec / test (push) Successful in 1m31s

Details

CI for csaxs_bec / test (pull_request) Successful in 1m30s

Details

3ea4e9195a

diaz requested review from appel_c 2026-05-07 14:45:12 +02:00

diaz requested review from holler 2026-05-07 14:45:12 +02:00

diaz requested review from wakonig_k 2026-05-07 14:45:12 +02:00

wakonig_k commented

2026-05-07 17:08:45 +02:00

@diaz

Perhaps another thing to report is that I typically had to load the config file in bec more than once until it worked... I found this a bit strange but as long as it works I am happy.

Was it complaining that it could not connect? Or any other error message?

@diaz >Perhaps another thing to report is that I typically had to load the config file in bec more than once until it worked... I found this a bit strange but as long as it works I am happy. Was it complaining that it could not connect? Or any other error message?

diaz commented

2026-05-07 17:19:43 +02:00

Yes, I think the complaints were about the connection. Typically, the first time it would complain, and the second time it was then asking weather to load all new parameters from the configuration, or to keep the values from previous ones (the question that it asks with 3 different possible answers)

wakonig_k commented

2026-05-07 17:47:38 +02:00

From the logs, it looks like it was able to connect but the initial read that it performs afterwards fails with a timeout. Maybe the controller needs more time after opening the connection. We could try to repeat the same readout 2-3 times, giving us a total timeout of 6 s (2 s per try) before giving up.

diaz commented

2026-05-08 09:12:39 +02:00

Yes, we could try that, thanks! My impression is that it needs longer time to establish the connection, indeed. I also noticed that it took a while from physically connecting to network (plug ethernet cable and turn controller on) until the controller was detected in PacketFence. It needs its time...

wakonig_k commented

2026-05-08 09:14:36 +02:00

okay, I will add the retry to this branch

wakonig_k added 1 commit 2026-05-08 09:22:37 +02:00

feat(npoint): add retry to initial readout and improve error message

CI for csaxs_bec / test (push) Successful in 1m35s

Details

CI for csaxs_bec / test (pull_request) Successful in 1m42s

Details

9b09d5a64d

wakonig_k commented

2026-05-08 09:24:08 +02:00

It will try it now 5 times with a 1s delay in between. Let's see if that "fixes" it

👍 1

wakonig_k commented

2026-05-08 09:31:05 +02:00

Regarding "settling_time" and "settle_time": We've had a discussion with Controls yesterday and we agreed that we would override the default ophyd property "settle_time" with what we currently have in "settling_time". This should make it more consistent: configurations are signals.

If you want, we can make the change already here for the NPoint. It is merely renaming

    settling_time = Cpt(Signal, value=0.1, kind="config")

to

    settle_time = Cpt(Signal, value=0.1, kind="config")

Regarding "settling_time" and "settle_time": We've had a discussion with Controls yesterday and we agreed that we would override the default ophyd property "settle_time" with what we currently have in "settling_time". This should make it more consistent: configurations are signals. If you want, we can make the change already here for the NPoint. It is merely renaming ```python settling_time = Cpt(Signal, value=0.1, kind="config") ``` to ```python settle_time = Cpt(Signal, value=0.1, kind="config") ```

diaz commented

2026-05-12 10:04:18 +02:00

Hi Klaus. I am not sure I completely understand. Would this mean that we can set the settling time of the piezo in the configuration? How?

wakonig_k commented

2026-05-12 13:30:22 +02:00

Yes, this would be a side-effect once merged. It would then behave like a normal signal, allowing you to do something like

npx:
  description: nPoint x axis on the big (3-axis) npoint controller
  deviceClass: csaxs_bec.devices.npoint.npoint.NPointAxis
  deviceConfig:
    axis_Id: A
    host: "nPoint000003.psi.ch"
    limits:
    - -50
    - 50
    port: 23
    sign: 1
    settle_time: 0.08
  enabled: true
  onFailure: buffer
  readOnly: false
  readoutPriority: baseline
  connectionTimeout: 20
  deviceTags:
  - npoint

If you want, I can make the required changes

Yes, this would be a side-effect once merged. It would then behave like a normal signal, allowing you to do something like ```yaml npx: description: nPoint x axis on the big (3-axis) npoint controller deviceClass: csaxs_bec.devices.npoint.npoint.NPointAxis deviceConfig: axis_Id: A host: "nPoint000003.psi.ch" limits: - -50 - 50 port: 23 sign: 1 settle_time: 0.08 enabled: true onFailure: buffer readOnly: false readoutPriority: baseline connectionTimeout: 20 deviceTags: - npoint ``` If you want, I can make the required changes

diaz commented

2026-05-12 17:07:03 +02:00

This sounds great, Klaus, thank you!
Should I test the branch before merging? - Did you make a branch?

This sounds great, Klaus, thank you! Should I test the branch before merging? - Did you make a branch?

wakonig_k added 1 commit 2026-05-12 19:36:44 +02:00

fix: rename settling_time to settle_time to override ophyd property

CI for csaxs_bec / test (pull_request) Successful in 1m28s

Details

CI for csaxs_bec / test (push) Successful in 1m28s

Details

40e1efa329

wakonig_k force-pushed 2axis_npoint_controller from acdc3728b9 to 40e1efa329

2026-05-12 19:36:44 +02:00

Compare

wakonig_k commented

2026-05-12 19:41:32 +02:00

This sounds great, Klaus, thank you!
Should I test the branch before merging? - Did you make a branch?

I pushed it now to this branch. We can merge it and test it the next time we find the opportunity. I would be particularly interested in the startup behavior (as discussed above) and whether the repeated readout attempt improves the stability.

Ana, feel free to merge :)

> This sounds great, Klaus, thank you! > Should I test the branch before merging? - Did you make a branch? I pushed it now to this branch. We can merge it and test it the next time we find the opportunity. I would be particularly interested in the startup behavior (as discussed above) and whether the repeated readout attempt improves the stability. Ana, feel free to merge :)

diaz commented

2026-05-13 08:45:29 +02:00

OK, thanks!
I am trying to figure out how to merge. I have button here that says: "rebase, then fast-forward"
And a message:
"The merge will not be signed as the base commit is not signed.
This pull request can be merged automatically"
But I don't see a clear button saying "merge"

OK, thanks! I am trying to figure out how to merge. I have button here that says: "rebase, then fast-forward" And a message: "The merge will not be signed as the base commit is not signed. This pull request can be merged automatically" But I don't see a clear button saying "merge"

wakonig_k commented

2026-05-13 08:58:11 +02:00

You can press the rebase button (see https://bec.readthedocs.io/latest/how-to/git/merge-changes-to-main.html). Rebasing is similar to merge but puts the commits on top of the main branch, not just mingles them in between.

diaz commented

2026-05-13 09:48:14 +02:00

OK, now I was reading the instructions and I think it is clear. Thank you. It's just that the buttons have nothing to do with logic or intuition... I guess I need to learn all this
I have a chance to test the piezos today, so I am going to test on the branch, maybe add the settling time in the config template, and I will then merge

OK, now I was reading the instructions and I think it is clear. Thank you. It's just that the buttons have nothing to do with logic or intuition... I guess I need to learn all this I have a chance to test the piezos today, so I am going to test on the branch, maybe add the settling time in the config template, and I will then merge

wakonig_k commented

2026-05-13 10:02:56 +02:00

OK, now I was reading the instructions and I think it is clear. Thank you.

That's great to hear!

[...] maybe add the settling time in the config template

Please note that it is now called settle_time, not settling_time

>OK, now I was reading the instructions and I think it is clear. Thank you. That's great to hear! > [...] maybe add the settling time in the config template Please note that it is now called `settle_time`, not `settling_time`

bec_bl_push-gitea added 1 commit 2026-05-13 16:06:43 +02:00

add settle time to npoint user template a4d8c570e1

wakonig_k added 1 commit 2026-05-13 17:09:39 +02:00

fix(npoint): retry on wait_for_connection

CI for csaxs_bec / test (pull_request) Successful in 1m31s

Details

Read the Docs Deploy Trigger / trigger-rtd-webhook (push) Successful in 2s

Details

CI for csaxs_bec / test (push) Successful in 1m33s

Details

abca4364bd

wakonig_k force-pushed 2axis_npoint_controller from 1c4165f301 to abca4364bd

2026-05-13 17:09:39 +02:00

Compare

diaz merged commit abca4364bd into main

2026-05-13 17:21:45 +02:00

diaz deleted branch 2axis_npoint_controller

2026-05-13 17:21:45 +02:00

diaz commented

2026-05-13 17:22:11 +02:00

Merged!
Thanks for your help, Klaus!

Merged! Thanks for your help, Klaus!

Sign in to join this conversation.

No Reviewers

appel_c

holler

wakonig_k

3 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: bec/csaxs_bec#204