[Dsg-hallc_controls] NMR - PLC Interface

[Steve Lassiter]

Hi All, 


The need for NMR field regulation units: 
Setting the dipoles by current mode gives a regulation of 10ppm. There is also a long wait time at the end of current ramping for Eddy currents to decay. Using NMR field regulation provides regulation of the order 0.1 ppm, 100x times better than just current mode. The NMR also provides a much quicker stabilized field as the NMR can compensate for the Eddy current decay by continually adjusting the PSU current. Both of these factors are important in running experiments within Hall C. 


The need for standardization: 
Originally the HMS magnet controls started out as being part of the magnet contract and each magnet vendor had their own unique controls system. The magnet’s controls along with the spectrometer’s rotation and the Halls cryo controls were later combined into a single commercial available, redundant process controlled PLC. Repairs, modifications and adding new systems can be done by more engineers and techs other then just by a few system experts by using industry standard PLCs. Long term technical support and product availability over 20+ years are also key considerations. These PLCs and their components have been tested and nearly all (exception is controlnet) work in the Hall's environment. 



Why a new generation NMR unit: 
A few of the disadvantageous of the standard 2026 NMR unit is the need for a large number (7) of (large sized) probes to cover the field excitation range of the Spectrometer’s dipole magnets. These probes also need to be located in a region where the field gradients are tight in order for the unit to “lock” unto a signal. The demand for a flat field region and the volume required to house the 7 probes and not be a source of multiple scatter or be seen by the detectors is a challenge that has not been fully meet yet. The current configuration calls for up to six probes to be installed at one time. This limits the excitation range that an experiment can do before the probe assembly has to replace a lower range probe with a higher range probe. Reproducibility is comprised in doing this and finding the sweet spot of the magnet’s field is extremely hard at higher excitation as the Iron yoke becomes saturated. Also there is difficulty in replacing the probe bundle due to the highly restricted access to the area and being able to adjust the position of the probes while staying out of a high magnetic field. 
GMW has come out with a new generation NMR field probe and controller that covers the entire dynamic range of the Dipole magnet and is more tolerant of field gradients. A single unit was purchased to be tested by Hall C and the unit performed well with the exception of two features: changes to how the unit communicates to external devices and the lack of ability to control the PSU’s current directly. 


Path forward: 
Establishing a means of getting the new NMR unit to communicate with the PLC needs to be addressed and solved before the second problem is tackled. Using existing means of communication (RS323, MODBUS) is desirable as this has proven to be functional in the hash environment conditions within the hall. The vendor does not plan on implementing this option (in the forseeable future). Other means such as USB or Ethernet will have to be developed between the NMR and the PLC. After communications is established, then a means to get the NMR to regulate the PSU‘s current, either through the NMR or by a PLC routine has to be solved. Once these two problem areas are resolved, then a second unit and probe should be procured (~$30k) for the SHMS magnet. 

One fall back is to continue using the older models NMRs with limited probe/excitation capabilities. Another plan is the development of a external field shunt attached to the magnets to house the older NMR probes. 


Steven 
----- 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> 
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 
> _______________________________________________ 
> Dsg-hallc_controls mailing list 
> Dsg-hallc_controls at jlab.org 
> https://mailman.jlab.org/mailman/listinfo/dsg-hallc_controls 

_______________________________________________ 
Dsg-hallc_controls mailing list 
Dsg-hallc_controls at jlab.org 
https://mailman.jlab.org/mailman/listinfo/dsg-hallc_controls 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/dsg-hallc_controls/attachments/20180611/8a3397a2/attachment-0001.html>