[Halld-controls] Solenoid sub-coil inductances and the Danfysik quench detector

[Joshua Ballard]

Floyd,
That is correct -- the voltage taps for Coil 2 are not in the locations 
on the SLAC prints.  The first 2 pancakes of subcoil 2C are shared with 
subcoil 2B.

Did your whole-solenoid ANSYS model include this information?

Josh

On 9/19/2011 6:53 AM, Floyd Martin wrote:
> Good Morning Elliot,
>
> Upon reading this something sparked in my brain concerning my simulation.
>
> For certain Coil 4 needs to have subcoil C & D divided out to 
> represent the actual divisions.
>
> And then it seems like something in my brain is telling me either coil 
> 2 or 3's voltage taps aren't in the exact location as what's shown on 
> the SLAC prints.  Is this true or am I thinking of something else?
>
> Floyd
>
> On 9/15/2011 5:07 PM, Elliott Wolin wrote:
>> Hi,
>>
>> Calculated sub-coil inductances at 1500 A from Floyd for the 2-1-3-4 
>> coil configuration are (subsequent columns are for each sub-coil):
>>
>> coil 2 (total 4.15 H):    1.3260    0.6130    1.0042    1.2102   coil 
>> 1 (total 7.20 H):    1.4104    1.3824    0.5593    0.3664    
>> 1.0887    1.1580    1.2331   coil 3 (total 4.12 H):    1.6821    
>> 1.0953    0.6768    0.6624   coil 4 (total 9.61 H):    1.2362    8.3721
>>
>> Total all coils:  25.1 H
>>
>> We need fairly good balances to get the quench detector to work 
>> well.  Also, it is supposedly possible to balance non-adjacent 
>> sections of the coils, e.g. one should be able to balance coil 2 
>> against coil 3, even though coil 1 sits between them.  Note that the 
>> Danfysik QD has 4 channels.
>>
>> Considering entire coils, the only balance that works is coil 2 vs 
>> coil 3.  The others are too different.
>>
>> Thus to get additional good balances we'd have to add sub-coils 
>> together and not respect full coil boundaries.  I don't think this is 
>> a problem.  But we do want to ensure that every sub-coil is included 
>> in a balance somewhere...if not a quench in that sub-coil would not 
>> be directly noticed (although we'd see other problems, and the quench 
>> would likely propagate to adjacent sub-coils).
>>
>> So now we have a bit of a Chinese puzzle working out good balances 
>> and not skipping any sub-coils.  I assume QD channels can deal with 
>> overlapping measurements...Fernando, is this correct?  I.e. I assume 
>> that if e.g. sub-coil 1F is used in one QD channel, it can be used 
>> again in another.  Or will this introduce some sort of problematic 
>> cross-talk?
>>
>> Now to work out good balances...
>>
>> Comments?
>>