[Dsg-hallb_magnets] Questions regarding digital filter for Hall B Magnet Fast-Daq

[Nicholas Sandoval]

Hi Tyler,

The whole magnet (VT22) is using a P29000 series iso-amp, the rest are using P29001 series.  Also, the dip settings (gain) for the coils are different than the splices.  This is the reason for the scaling.  I have not had a chance to look at Dwg -0183 yet, but perhaps this portion needs to be updated. 

Regarding the Iso-Amp noise you are correct, but we are in fact seeing closer to ~330mV on VT22, this is probably due to errors in my setup and perhaps a difference in the P29000 series vs the P29001 series as I only tested the P29001.  Fortunately, we do not have any PLC comparators that use the whole magnet for a controlled or fast dump.   

Yes, the cables are shielded on both input and output of the iso-amp.  We did have to add extra grounding straps for the DIN rails that the Iso-Amps reside on, which helped a little.  The biggest improvement we made to the installation was suggested by Fernando, which was to ground the negative output of the 24VDC DIN rail power supplies which feed the iso-amps.  This improvement was able to bring our installation noise level to similar levels as seen in the lab, both are now below the manufacturer's specification.

Best Regards,

Nick



----- Original Message -----
From: "Tyler Lemon" <tlemon at jlab.org>
To: "dsg-hallb magnets" <dsg-hallb_magnets at jlab.org>
Sent: Thursday, January 12, 2017 11:53:43 AM
Subject: Re: [Dsg-hallb_magnets] Questions regarding digital filter for Hall B	Magnet Fast-Daq

Hello Nick, 

My in-line reply and additional questions are below in blue. 

Hi Tyler, 

1. The noise is the same on all voltage taps, the problem arises when we multiply the noise for the scaled channels. The fix should be implemented on all channels so as to keep the timing the same. 
I can see on EPICS that the scaled voltage tap data is rather noisy. I also remember from the pre-power up check list that the scaled voltage taps were causing problems. 

2. Some of the voltage taps are scaled due to the nature of the magnet. The coils see much larger overall voltage so the settings on the iso-amp are configures such that we can see the full scale. This requires us to rescale on the software side to get a 1 to 1 ratio. 
This makes sense as Torus drawing B00000-09-00-0183 shows that all iso-amps are outputting voltage in range +/- 10V that can be read by the cRIO modules. 

However, the only iso-amp that is shown to be configured differently in the Torus drawing is the one for VT22_DAQ. Its iso-amp is configured for +/- 600V input; all others are configured for +/- 10 V input. If this is truly the case, is the x10 scaling on VT6, VT8, VT10, VT12, VT14, and VT16 even needed? 

3. We performed testing of the iso-amps installed in the hall and in my laboratory, both instances the output was very similar and less than the manufacturers specification. ~10mv RMS(spec), I see about 6mV on the O-scope for our installation and in the lab 
Could you clarify what this means? If it is that there is 6mV of noise from the iso-amp and we are scaling VT22 by 60, 6mV noise is scaled to 360mV. T hat could be the cause of the false trips. 

4. Currently the interlock trips @ 600mS after staying over the threshold(three samples). But, this is subject to change and has changes several times 
Does this mean that if there is a quick voltage spike, the interlock will not trip? 

5. The logged data is available through the fast daq viewer on the EPICs gui. It is also stored in ROOT. 60Hz is the noise we are most interested in removing. 
Could the 60Hz noise be due to incorrect grounding? 

Also, for my own curiosity, are the cables to and from the iso-amp shielded? Maybe this issue is similar to the Cryo-con unit grounding issue? 

If you would like please set up a meeting this morning or this afternoon and we can go though the data in more detail, and also I can show you how to access the root files if needed. 
I am able to access the root files. I will continue my investigation with this information. 

Best regards, 
Tyler 

---- Original Message ----- 
From: "Tyler Lemon" <tlemon at jlab.org> 
To: "dsg-hallb magnets" <dsg-hallb_magnets at jlab.org> 
Sent: Thursday, January 12, 2017 10:10:32 AM 
Subject: [Dsg-hallb_magnets] Questions regarding digital filter for Hall B Magnet Fast-Daq 

Hello Ruben and Renuka, 

I have been investigating how to add digital filters to the Fast-Daq programs and have the following questions: 


1. Are there noise issues on all voltage taps or just the ones that are being scaled? 
2. Why are some voltage taps scaled while others are not? 
3. Is there documentation for the iso-amps' grounding scheme? 
1. Looking at Solenoid drawing B00000-09-0683 for how the voltage taps are wired to the iso-amp and then to the cRIO, there is no indication that the iso-amps are grounded. 
4. How fast should the interlock trip after a voltage spike? 
5. Is there logged data of the voltage tap noise? 
1. For the filter, it would help to know if the noise is of a constant frequency or just white noise. 

In regards to applying and testing a filter in LabVIEW, there is a version of the Fast-Daq cRIO FPGA code almost ready for testing with lowpass filters applied to only the channels that are scaled. However, before testing this code, it would be good to have the above questions answered. 

Best regards, 
Tyler 


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


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