[Halld-tagger] specs for microscope readout power supplies

[Richard Jones]

Fernando,

Our background study showed that we could continue to function for 10 years
of high-rate running without annealing, just allowing the dark rate to
increase at the rate that was measured in Hall B and allowing the supply
current to increase.  Under the zero-shielding scenario, just leaving the
electronics unshielded in the tagger hall, we estimated that we would reach
the point where we would need to do something (eg. anneal or replace the
SiPM's) after 10 years of high-rate running.  The usual assumptions were
applied (100 days of running per year, standard 20 micron diamond, 12 GeV
electrons at 2.2 microAmps).  With shielding, we expect to reduce this
figure by a factor of 5-10 (the simulation gave a factor around 8), so we
should be able to run for the duration of GlueX without annealing.

I agree with you that we can segment the supplies.  Each of them can be
split up to 6 ways, because there are 6 identical backplanes in the system.
 The specs we posted were for the total sum of all 6 backplanes.

-Richard J.


On Mon, Dec 9, 2013 at 9:52 AM, Fernando J Barbosa <barbosa at jlab.org> wrote:

>  Hi Richard,
>
> Certainly the rates in the tagger are higher but what is your planning for
> annealing the SiPMs? I am just curious given the higher rates.
>
> Regarding the power supplies, you mentioned the possibility of having
> three supplies instead of one for the whole system. Is it possible to
> consider further segmentation? One channel per backplane? Do you have a
> document that shows the whole system as installed? The issue is also
> related to the space available under the magnet.
>
> Best regards,
> Fernando
>
>
>
> On 12/6/2013 10:09 AM, Richard Jones wrote:
>
> Fernando,
>
>  These estimates for the BCal do not apply to the TAGM, where the rates
> are much higher.  We can set up a meeting to discuss this further if you
> feel we need it.  We worked out these rates back during the design phase
> for the readout electronics, and had them reviewed within the photon beam
> working group. Also keep in mind that the neutron flux is much higher in
> the tagger hall than in the experimental hall.  On the other hand, the TAGM
> can operate efficiently at much higher single-pixel rates than can a
> calorimeter because it is not producing an energy measurement, but only
> time, and the photon statistics are high, <n> = 350-400 within a 15ns
> window.  We are shielding these electronics, but we only expect to cut down
> the neutron flux by about a factor 10-20 in this way.  Alex Somov did the
> neutron rates and shielding studies for us, and can provide more details in
> this regard.
>
>  -Richard J.
>
>
> On Fri, Dec 6, 2013 at 9:04 AM, Fernando J Barbosa <barbosa at jlab.org>
> wrote:
>
>>  Hi Richard,
>>
>> In reference to the BCAL and taking Yi's measurements in Hall A regarding
>> radiation damage, we settled on a maximum of 10mA for 10 arrays or 160
>> cells (3mm x 3mm). This works out to 62.5uA per cell and your estimate is
>> higher by a factor greater than 10. What is your plan for annealing the
>> SiPMs? Frequency?
>>
>> Did you look into any supplies that might fit your requirements? I recall
>> that each backplane of the TAGM has a single bias supply input, correct?
>> This would be important in considering a multi-channel supply system.
>> Please send me your latest drawings on the TAGM system and its installation
>> in the Tagger hall.
>>
>> Best regards,
>> Fernando
>>
>>
>>
>> On 12/5/2013 10:50 PM, Richard Jones wrote:
>>
>> Hello Fernando,
>>
>>  I am spec'ing the Vbias supply capacity at 500mA to allow for the
>> maximum SiPM draw current that is permitted by the bias voltage supply
>> system.  Initially we expect the average current to be around 15 mA at full
>> intensity (2.2 uA electrons on 20 micron diamond) but that it will increase
>> gradually over time due to radiation damage.  To get the maximum current
>> spec, I imagine the worst possible circumstances under which the radiation
>> damage might accumulate faster than expected, and ask how much current
>> could we sustain across all channels before the resolution and efficiency
>> degrade significantly due to dark current.  The answer to that is
>> approximately 800uA per channel, which totals about 400mA from the supply.
>>  I rounded that up to 500mA.
>>
>>  -Richard Jones
>>
>>
>> On Thu, Dec 5, 2013 at 4:51 PM, Fernando J Barbosa <barbosa at jlab.org>
>> wrote:
>>
>>> Hi Richard,
>>>
>>> I looked at the specs and the bias supply current seems to be excessive
>>> at 500mA. What's the reason for this?
>>>
>>> Best regards,
>>> Fernando
>>>
>>>
>>>
>>> On 12/5/2013 3:33 PM, Richard Jones wrote:
>>>
>>>> Hello Alex,
>>>>
>>>> Here is a draft of the specs for the microscope readout dc power
>>>> supplies.  They can be a single unit for each of 3 levels, or a few
>>>> lower-capacity supplies in tandem with the same total output capacity.
>>>>  Please request justification for any of the specs.  One thing we do not
>>>> spec here, but think would be useful, is separate delivery and sense
>>>> terminals for each output.  We have separate pins on the backplane
>>>> connector for this purpose.
>>>>
>>>>
>>>> http://zeus.phys.uconn.edu/wiki/index.php/Microscope_Electronics#Power_Supply_Requirements
>>>>
>>>> -Richard Jones
>>>>
>>>
>>>
>>
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailman.jlab.org/pipermail/halld-tagger/attachments/20131209/1d7d5aef/attachment.html