Remove Toptica adapter

[BenediktBurger]

Remove the Toptica Adapter and make the ibeamSmart depending on the VISAAdapter alone.
Takes hints from original PR #352 and the new issue #816.

The tests require empty strings, which are added to the ProtocolAdapter in #818

I added a few tests from the examples @dkriegner gave in #352.

Fixes #816

[msmttchr]

Few comment from my side. Thanks

[BenediktBurger]

Thanks @msmttchr for your Tipps.
My possibilities to refractor the instrument are limited, as I do not know it.

[msmttchr]

Yes, I meant this. Thanks Il Dom 8 Gen 2023, 14:28 Benedikt Moneke ***@***.***> ha scritto:

…

***@***.**** commented on this pull request. ------------------------------ In pymeasure/instruments/toptica/ibeamsmart.py <#819 (comment)>: > includeSCPI=False, + asrl={'baud_rate': 115200}, You mean, I should give the baud rate directly? As I don't see, where it is redundant (double). — Reply to this email directly, view it on GitHub <#819 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AJAXVWSTC2XDFW4OVPC22S3WRK6JNANCNFSM6AAAAAATQ5CGSM> . You are receiving this because you commented.Message ID: ***@***.***>

[BenediktBurger]

@waveman68 could you please test this implementation with your device?

Does anyone of you, @dkriegner or @waveman68 have a manual and could help me with the protocol tests (just give me some examples)?

[dkriegner]

The manual was delivered as printout. I will ask to get it for you. I'll come back to you.

Edit: The manual was shared with @bmoneke by email

[dkriegner]

Thanks for all the work. If you give sufficient time, I will instruct someone (remotely) to test the new code with the hardware.

[BenediktBurger]

Thanks @dkriegner , it is not urgent. Whenever you (and your acquantance) have time, is fine.

[dkriegner]

I can confirm that these changes fix issue #816.

There are some remaining issues which need to be discussed. Some of them only are important if a translator device (see #815 for a more detailed description) are used. But some are also completely generic.

[BenediktBurger]

Thanks for the information regarding the flush.
I don't know yet, how to do it properly (best would be to not use it).

[dkriegner]

I think a flush method is very important to be able to clean up after an error. It's a pity that pyvisa does not implement it for all connection methods.
Since even on the NI webpage they recommend using read to flush a TCPIP socket in Labview I assume there is a technical reason for that.

If you would not want to implement a fallback, the colleagues will keep some local code changes. In the end this is again only an issue when using a TCPIP translator. When using via the serial device, the existing flush_read_buffer is sufficient and the error handling is only needed for your unittests. But likely this should then be mentioned as comment. One could also mention that because of this using any Adapter defined outside is unlikely to work (even SerialAdapter would likely fail) since the communication is screwed up after the __init__

[BenediktBurger]

What do you think of a flush method based on read?
That would be the most complete instrument implementation. And basically, we strive to just use write and read (or their bytes variants) as these are universal to all connection types (and adapters).

[dkriegner]

yes, something based on read (repeated read as shown in my comment above or something equivalent) with the timeout set to minimum seems sensible. Not sure if the string conversion in read could make troubles and the byte equivalent would not be better ?!

[BenediktBurger]

I raised an issue #834 for that missing flush method in pyvisa-py as it does not pertain to this instrument itself, but to the VISAAdapter (and more properly to pyvisa-py).

From my side, the code is ready and any maintainer may merge it.

[BenediktBurger]

@bilderbuchi you might merge at will.

[bilderbuchi]

Thanks!