Re: Kicker test - URGENT action required ?

[Brendan Kiburg <kiburg@npl.uiuc.edu>]

Dear Colleagues,

  I will inform you of our latest measurements, a new possibility that may
be causing the breakdown and then of our plans.  Please feel free to
criticize or weigh in such that we have as much feedback as possible.

----------Measurement----------
This evening we measured the resistance between the center pin in the
"fischer" socket on HV1 and ground.  We obtained 0.0 Ohms.  We measured
the resistance between the braid on the cabinet side of the "Fischer"
connector when it was pulled out of the socket and ground and obtained >20
MOhm.  We plugged in the fischer plug and measured the resistance
between the same location on the braid and ground and found 0.0 Ohms.

 We repeated this measurement on the MV1 cabinet.  With the fischer
connector pulled out we measured the resistance between the center pin of
the socket and ground to be >20.0 MOhm.  We found greate than 20 MOhm when
measuring from ground to the braid attached to the fischer connector,
regardless of whether it was plugged into the socket or not.  

>From this measurement, we concluded the problem is on the socket side of
HV1.  

--------New possibility--------------

Tonight while discussing the potential kicker work with Francoise on the
telephone, we discussed the possibility of a short circuit between the
metal collimators that were affixed to the beam pipe and the deflector
plates.  We are unsure if this would be physically long enough to make the
connection, but we should consider it.
   As possible evidence in favor of this model, we point to the strange
beam measurements made by both the mulan and mucap experiments that seemed
to produce a "split beam" effect.  That is to say, the beam profile
appeared to have a shadow under certain settings.  
   A possible inconsistency of this model is the initial failure mode that
we experienced.  Since we first ramped up to 2kV before the current
regulation kicked in, and subsequent measurements seemed to indicate a now
"permanent" ground condition, independent of the voltage, it may be
unlikely that the plate was causing the short.  It would seem more likely
a ground that was able to support 1.X kV but unable to support 2.0 kV
arced once or twice and the produced a more permanent path to ground.

   Still, I think we should consider this possibility when we look inside.

--------Plans-------------

After much discussion with the mucap collaboration, it seems clear that
solving this breakdown problem is a critical part of the current run.  We
would like to continue to take data while repairing the kicker if
possible.  Since tomorrow is Friday and we will not get external PSI
support during the weekend, we need to form our plans carefully to
facilitate adequate working time and to maximize beam measurements if we
remove the kicker.

Early morning:
 1) Bring the kicker to air.  This requires approximately 1 hour of
venting.
 2) We would investigate using the top and bottom flanges any
"obvious" problems regarding the connection from the Fischer socket to the
deflector plates, if we are able to see things.  Also, we would confirm
the location of the collimators.
 3) If the problem is very straightforward (like the
collimator plate has fallen down), then we could perform a test with the
short removed and verify the kicker works at air.  At this point we would
need to remove the plate and could probably arrange access via the bellows
that are upstream of the kicker.

 4) More likely, we will either be unable to correctly identify the
problem, or if we are able to identify the problem and need to resolder
some connection, we may need end access to actually repair it.

 5) At this point we would propose removing the kicker from PiE3 and
installing the beam pipe used at various points last year.  We would begin
pumping with the help of the PSI vacuum group and transport the kicker to
a work area.

 6) Peter and I would continue to investigate the problem Friday and
probably into the weekend.  This would allow us to either solve the
problem and return the kicker to the zone during next week or provide us
with an opportunity to work on it as long as necessary while continuing
the measurement at mucap.


The success of this program depends on the assistance of the PSI crew and
maybe some luck with respect to their schedules before the weekend.  We
clearly have a lot to gain by repairing the kicker as soon as possible and
returning it to the beamline in a running condition, so we have decided to
make that the top priority with respect to the ongoing mucap run.

Thank you all for your continued assistance.  We will keep you updated
with our progress and findings.

Best Regards,

Brendan



On Thu, 1 Dec 2005, Gary Wait wrote:

> Peter Winter wrote:
> 
> > Dear all,
> >
> > while Mike was sending us his last email, I was already inside the 
> > area to do the next tests. First I checked the optic fibre on card 
> > DN-13. For that I,
> > 1) I disconnected the fisher plug, cleaned it carefully with alcohol 
> > and also its counter part as much as possible. Then I taped it in a 
> > safe position
> > 2) I exchanged the optic fibres DN-13 and DN-12
> > 3) I ran with DC high TTL signal ramping up HV1 to ~3.4kV. ALL red 
> > LEDs on the lower stack in HV1 were on.
> > 4) I exchanged the optic fibres again to the normal status and it 
> > worked as well
> > 5) I changed to kicking mode (5mus @ 1kHz) and ramped HV1 and all 34 
> > red LEDs were visible at around 6-7kV.
> > It seems that the problem with the card is gone.
> >
> > Then I tried to run the kicker in DC with the cleaned fisher plug. 
> > While cleaning I was looking at it carefully but could not see any 
> > carbonized parts or black parts etc. I was also removing the white 
> > part of the insulation tube on the plug and cleaned the whole metal 
> > pin which makes the connection. The counter part connected to the 
> > plate was difficult to clean since it's inside and partly hidden by 
> > the metal side walls. After reinserting the plug, the current 
> > increased immediately when raising the HV on HV1.
> >
> > Due to Mikes last email I now turned off the HV on both cabinets since 
> > if I understood correct he's afraid that other parts might break down, 
> > too.
> >
> > So the question rises: When are we going to open the kicker to inspect 
> > it? I think we'll have a discussion here internally but proposals from 
> > all of you are welcome. For now I will go back to mu+ lower rate 
> > without kicker.
> > Peter
> >
> Peter
> I completely agree with Mikes analysis.
> 
> It appears that the fiber optic cable to DN-13 was not quite plugged in 
> all the way.
> 
> To determine whether the plug is breaking down or the socket is breaking 
> down. Here is a simple test  to try:
> interchange the plugs between HV1 and HV2 and see if the problem stays 
> in HV1 or changes to HV2.
> If the problem switches to the other cabinet then you know it is the plug.
> 
> Gary
>