[Dsg-hallc_controls] NMR - PLC Interface

[Brian Eng]

Hi Thia,

I'm not sure what Mike was referring to that is $30k (maybe replacing the older NMRs? additional probes?) I know the new unit itself is quite expensive at ~20k or so.

So I can only speak to what my suggestion was (the hundreds of dollars one). I'll try and elaborate some more...

This new NMR unit (from Metrolab) only has the option of either USB or Ethernet as its remote interface. Using USB requires a driver, so only the major OSes are supported, nothing for a PLC. The Ethernet interface has 2 different communication protocols it can use: either VXI-11 or MODBUS. Currently only VXI-11 is implemented by Metrolab.

If the MODBUS option was available in principal the PLC could directly talk to it. I would caution that this code will be quite complicated and honestly I would suggest using a MODBUS communication module for which there are a few vendors.

Tyler had earlier confirmed from Metrolab that MODBUS won't be available soon, they said for sure not this year and we should expect a more likely release date of 2020, that leaves VXI-11.

Unfortunately we haven't found a commercial solution for that, hence my suggestion to develop something ourselves. While I agree with Mike that this isn't ideal I still think that pursuing this option would be worthwhile. If nothing else we can put a board in the Hall for the upcoming runs in order to evaluate its reliability to maybe put into production on the next long shutdown?

If Mike/Steve are resolute about no external hardware/software being used then this project should be shelved as I don't see any way of using this piece of equipment with the PLC.

Hopefully that helped some? If not let me know if there are any questions.

--
Brian Eng
Detector Support Group
757-269-6018

----- Original Message -----
> From: "Cynthia Keppel" <keppel at jlab.org>
> To: "Mike Fowler" <fowler at jlab.org>, "Brian Eng" <beng at jlab.org>
> Cc: "dsg-hallc controls" <dsg-hallc_controls at jlab.org>, "Jack Segal" <segal at jlab.org>
> Sent: Friday, June 8, 2018 6:04:50 PM
> Subject: Re: [Dsg-hallc_controls] NMR - PLC Interface

> Hi Mike, Brian,
> 
> I think I need more input... will the cost for one option be the ~$70 x 2 for
> implementing the Python modules as Brian is suggesting, with the ~$30k for new
> NMR units the cost for a different option - or are these two the same thing?
> 
> Help!
> 
> Thanks,
> Thia
> 
> ----- Original Message -----
> From: "Mike Fowler" <fowler at jlab.org>
> To: "dsg-hallc controls" <dsg-hallc_controls at jlab.org>, "Cynthia Keppel"
> <keppel at jlab.org>
> Sent: Friday, June 8, 2018 4:22:38 PM
> Subject: Re: [Dsg-hallc_controls] NMR - PLC Interface
> 
> Hi,
> 
> It is my opinion that the need for this project should be reviewed and
> possibly delayed until such time that the NMR vendor has resolved the
> communication problem.
> 
> Currently the HMS dipole older NMR unit covers most of the dipole
> operating range, and regulation by current appears to be sufficient over
> that range. The SHMS Dipole is currently regulated by current which may
> be sufficient for the near future.
> 
> I believe having additional hardware and software external to the PLC
> that is subject to failure is not an ideal solution. Implementing the
> new NMR units for the SHMS also comes with an additional investment of
> about $30K.  Without a reliable solution I question the current need to
> proceed.
> 
> Steve and Thia please comment.
> 
> Regards,
> Mike
> 
> 
> 
> 
> On 6/8/2018 1:05 PM, Brian Eng wrote:
>> Since there is currently no direct way to interface the new NMR with the PLC
>> (MODBUS would work, but is at least 1 year away according to the vendor) and
>> none of the existing devices used by Halls B/D will work we're going to need to
>> develop something ourselves.
>>
>> Despite the communication issues discovered with the NMR (that the vendor is
>> aware of and working on) the proof-of-concept of using Python modules as a go
>> between the devices showed it could be a viable solution.
>>
>> The next step would be to order the hardware that would run this. The plan is to
>> use a small SBC that runs Linux, I was hoping to order through webstock but I
>> couldn't find any newer models. So we would need to use a PCard for the order.
>>
>> Each board/setup would be ~$70 and we'd like to order at least two of them.
>>
>> Is there someone in Hall C that normally handles PCard orders or should we just
>> place the order ourselves using COPS.PHALLC?
>>
>> Thanks
>> --
>> Brian Eng
>> Detector Support Group
>> 757-269-6018