Timeout, or abort, needed for proper trigger support

[drmcnelson]

spectrometer and system information

USB2000, etc - all models with external trigger capability

current problem

To adequately support triggered operation, the API to retrieve spectra, absolutely needs to have a timeout parameter, or at least a way to abort the intensities() call while it is waiting for a trigger.

[ap--]

Hi @drmcnelson,

This will be a bit more involved. It's possible to change the pyseabreeze backend to allow pass-through of a timeout value. For the cseabreeze backend these USB timeouts are hardcoded, so it'll involve more work to implement.

I will implement the functionality for pyseabreeze first. But I'm currently pretty overloaded with work, so this is a very low priority right now.

I'll update this issue when I start work on this. I can also give some guidance on how to implement this feature, if someone volunteers to work on a PR.

Cheers, Andreas 😃

[drmcnelson]

Yes, I saw where it would go in the pyseabreeze code. An abort could be a useful alternative for some use cases.

[ap--]

Note to self: It seems there's no easy hacky way to do this with the sync interface: https://libusb.sourceforge.io/api-1.0/libusb_io.html#sync

[drmcnelson]

Yes, it looks messy. The comment that they need to be called to handle events seems a little concerning.

I have been using the device in triggered mode for clocked data acquisitions. So I'm getting along with out it for the moment.

Thank you for looking into it.

[drmcnelson]

user space thread to implement a watchdog timer? (provided the overhead is small)

But again, I don't want to cause you extra work. It has been working very well as is, for my clocked data acquisitions.

At this point, I am pretty sure I would not change my code to use the timeout.