IOC: JASCO: Create IOC - Githubissues

ISISComputingGroup / IBEX

IOC: JASCO: Create IOC #3923

Closed John-Holt-Tessella closed 5 years ago

John-Holt-Tessella commented 5 years ago

As an instrument scientist I would like to control the JASCO pump so that I can do experiments with it.

The JASCO pump has been looked at and found to be ok to talk to see #3743. There are no nasty suprises as far as we can see.

As normal we need IOC, emmulator and OPI.

John-Holt-Tessella commented 5 years ago

This is a serial device, has about 15 commands. The only slight difficulty is you can send it programs which might be useful to users. I.e. pump this x minutes and this for y etc. I think we should probably expose this as a typical program with values you can set but it does make the IOC more complicated.

kjwoodsISIS commented 5 years ago

Please don't forget to estimate the points.

DominicOram commented 5 years ago

@FreddieAkeroyd: When is this needed and by which instrument?

aaron-long commented 5 years ago

IOC Device Generator fix

https://github.com/ISISComputingGroup/IBEX_device_generator/pull/10

Main PR

FreddieAkeroyd commented 5 years ago

Potential beamlines are:

OFFSPEC SURF POLREF LARMOR INTER

ThomasLohnert commented 5 years ago

Functionality looks good to me! Some minor points:

Would it be feasible / useful to validate the sum of the gradients similar to what I believe you did for the Knauer?
COMP:-related things are referred throughout the .db and tests as "Composition", whereas on this front panel, and in the wiki entry, it says "Component". Can you please make it consistent?
The device response is a bit odd when in RECSIM: Setting the flow rate sets the pump status to running, while start/stop/timed don't seem to work.
Starting the pump does not update the gradients in DEVSIM or RECSIM. My understanding is this is because the real device would get it from an internal file which is hard to simulate - is this about right?
Little bits of tidying up on these branches:
- [x] EPICS
- [x] ibex_gui
- [x] Device Generator

aaron-long commented 5 years ago

Thanks!

Re: changes

The device automatically calculates the composition sum to 100 and updates D automatically to ensure 100 sum. In the event of sum>100, the device sends an error message and uses the auto-calculated values rounded to 100. After speaking to the scientists this behaviour was expected and matched the old behaviour; we decided to leave it as it is without any additional logic.
I've updated to the wiki to reflect the correct wording of composition.
Updated the .db so that the RECSIM behaviour of FLOWRATE is correct.
Compositions now correctly update in both RECSIM/DEVSIM; the command was missing its time component.
Removed the old auto-name function from the device generator and cleaned up some of the unused auto fields from existing functions.

aaron-long commented 5 years ago

Rework: