[dsg-svt] Location of SVT sensitive components on the service cart

[Yuri Gotra]

Hello Chris,

In the SEE presentation I wanted to stress a few points which we need to 
take into account:

We had a recommendation from the ERR which we need to respond. Proper 
response would be not only to do the test which we did but to do our 
best to identify and protect the sensitive components of the SVT system 
by distance and shielding.

Some thoughts:

- SEE could cause substantial operational issues and downtime

- Even detailed simulation of radiation environment or SEE won't 
guarantee that there won't be SEE issues

- The worst damage is expected from neutrons, it would be wise to put 
shielding upstream of the SVT service cart, maybe close to the collimators

- All the crates are already placed at the maximum distance from the 
beam pipe defined by the cable length

- There are still components (switches and UPS) which are sensitive to 
the radiation but could be moved away from the beamline

- SEE could depend on the dose and we don't have a way to test whether 
these switches and UPS will survive the CLAS12 environment

- The safe approach would be to move the switches and UPS radially to 
the SVT racks

- Yes, this would require some expenses but the SEE consequences are 
expensive too

- There is no guarantee that moving the switches and UPS to the racks 
will protect them from SEE but the doses will be lower

- Remote reset of the switches is a good idea, we need to implement this



On 3/3/17 4:20 PM, Chris Cuevas wrote:
> Hi Brian,
>
> Nice SEE report but I did not read any mention of failures noted for 
> the network switch gear during KPP or during other periods. I also am 
> concerned about the small Ethernet magnetics in the switches that are 
> probably more susceptible to magnetic fields. Ive seen the fringe 
> field models for the Soleniod, so maybe moving them radially away from 
> the beam pipe will be adequate. Moving the network switches to the 
> deck 1 space frame racks is easy enough, and you are right about 
> buy/provide the network cable to run from the cart to the space frame 
> rack.  A bundle of 16-20 RJ45 cables is not too bad.  We should also 
> invest in a network controlled AC power strip in the space frame rack 
> that would allow remote reset for these network switches from the 
> counting house or other remote control.  Ill check with Sergey because 
> there may already be one in the space frame rack that holds the 
> network gear.
>
> The UPS relocation is a more challenging job, and I still wonder why 
> these are needed.  We are on the topic of radiation 
> sensitivity/susceptibility and I know these UPS are always run in 
> inverting mode, so the switching devices internal are most likely 
> power FET, which are known to be susceptible to radiation damage. 
> Something to think about.
>
> Relocating the UPS to the space frame racks will require a new array 
> of conduits be installed to power devices on the SVT cart because you 
> cannot run AC power cords in cable tray.  The new clean power 
> installation for the cart should be more than adequate and the smart 
> distribution strip can be monitored. I do not recall if the output 
> ports can be switched, but if yes, it would be another way to reset 
> devices on the cart that do not have remote reset capability.
>
> Keep me posted,
>
> -Chris
>
> ~~~~~~~~~~~~~
>
>
> On 3/3/2017 1:49 PM, Brian Eng wrote:
>> Hall B will need to buy/provide the network cable and connectors to 
>> run 16 cables instead of the 2 that were already run if the switches 
>> need to be moved off the cart.
>>
>> There would also need to be some coordination with fast electronics 
>> for the UPSes to be moved since none of the AC power cables would reach.
>>
>> Personally I'd prefer to move (radially away from the beam) the 
>> equipment and/or shield it before moving it off the cart.
>>
>>> On Mar 3, 2017, at 1:23 PM, Yuri Gotra <gotra at jlab.org> wrote:
>>>
>>> https://userweb.jlab.org/~gotra/svt/mtg/tcb/2017_03_03_err_tcb.pdf
>>>
>>>
>>> On 3/3/17 11:12 AM, Brian Eng wrote:
>>>> Hi Yuri,
>>>>
>>>> We can move the switches (and maybe some of the UPSes [this might 
>>>> require cart modifications by Hall B Engineering]) to the bottom 
>>>> corners of the cart, but moving them to the racks would be quite a 
>>>> big change (and a lot more time/money to implement).
>>>>
>>>> Is there any actual reason to think the corner locations on the 
>>>> cart are to high of a dose for the equipment to be placed?
>>>> -- 
>>>> Brian Eng
>>>> Detector Support Group
>>>> 757-269-6018
>>>>
>>>>> On Mar 3, 2017, at 10:32 AM, Yuri Gotra <gotra at jlab.org> wrote:
>>>>>
>>>>> Please review the current location of the SVT system components 
>>>>> which are sensitive to radiation and consider moving what is 
>>>>> possible from the service cart to the SVT rack further away from 
>>>>> the beam. Moving the network switches would be desirable. Maybe 
>>>>> the UPS too. Lowering the accumulated dose would increase 
>>>>> reliability of the component and reduce the probability of the 
>>>>> detector down time. Thank you
>>>>>
>>> <gotra.vcf>
>
>
> _______________________________________________
> dsg-svt mailing list
> dsg-svt at jlab.org
> https://mailman.jlab.org/mailman/listinfo/dsg-svt

-------------- next part --------------
A non-text attachment was scrubbed...
Name: gotra.vcf
Type: text/x-vcard
Size: 102 bytes
Desc: not available
URL: <https://mailman.jlab.org/pipermail/dsg-hallb_svt/attachments/20170306/633710b4/attachment-0002.vcf>