Re: Kicker overcurrent

[Michael Barnes <barnes@triumf.ca>]

Dear Brendan,

Thanks for the detailed email.I believe the information supports the
breakdown theory. The 112mA and -0.4kV are consistent with the breakdown
being on the output of the kicker stacks (as opposed to through two
stacks) too, i.e. 112mA*(50R+2500R+9*24R+17R)= 2783R*112mA = 310 V. Given
the resolution of the readout and the fact that, unless it is a perfect
short circuit to ground, there will be a voltage drop associated with an
arc the 310V is very similar to the 0.4kV readout.

The initial start-up not showing any current, until 2kV, also indicates
that something has broken down to ground as the voltage is ramped up.

I like Francoise's suggestion of firstly repeating the start-up of HV1
with the output HV plug removed from the beam-pipe socket and the HV plug
safely arranged in the cabinet. However I might be tempted to do this test
in DC mode (making sure that HV1 is set to negative polarity and that a DC
TTL high is going into the cabinet -- however it is sometimes necessary to
pulse a few times to reset everything, but this could be done with 0V on
the HV supply).

Hope this help,
Mike

> Dear Mike,
>   We double checked the connections and polarity and have found the
> following:
> 1) The NEG1 signal is connected to optical fiber that goes to the HV1-down
> cabinet.
> 2) The HV1 cabinet has negative polarity and HV2 has positive polarity.
> 3) We wish to clarify the exact nature of the overcurrent that occurred
> last night.
>    a) We began turning up HV1 and HV2 at the same time
>    b) There was no current drawn on either supply unit until HV1 reached
> about 2 kV.
>    c) Then HV1 current regulated and the display read -0.4 kV and 112 mA.
>    d) We ramped down the voltage to 0.
>    e) We tried ramping up but immediately saw an increasing current (as
> opposed to earlier, when there was no current until the breakdown.
>    f) We turned off the HV supplies, the MV1 supply as well as all of the
> fans.
>    g) We inspected inside cabinet HV1 and cleaned with compressed air (all
> cards).
>    h) We turned on the fans, MV1 and tried to ramp again.
>    i) At first there was no current drawn and we reached 2 kV.  Suddenly
> the current regulation light came on and the display read -0.4 kV and 112
> mA, as in step b.
>
> 4) With respect to the operating conditions last night,
>    a) HV1 and MV1 were on negative polarity
>    b) The frequency of the kick signal was quite low (<1 Hz) coming from
> our entrance counter, which was not seeing the beam as we were in the
> area.
>    c) The duration of the pulse was 26 microseconds, as determined by the
> muon on request logic documented by Peter.
>    d) We then inserted a fixed frequency pulser (between steps 3g and
> 3h) that had been used in previous tests with Francoise.  This pulser had
> a kick of 5 microseconds with a 1 kHZ rate.
>
> We will now go and inspect the kicker cabinet and connections for
> potential short circuits.  Please let us know if we can give more
> information concerning last evening.
>
> Regards,
>   Peter and Brendan
>
>
> On Tue, 29 Nov 2005, Michael Barnes wrote:
>
>> Dear Peter,
>>
>> Further to my email of earlier today, I have had a chance to further
>> think about the problem. I am now fairly well convinced that item "2"
>> and item "3" in my earlier email canNOT result in what you are seeing.
>> For example, for my item 3 a false signal back from the overcurrent
>> protection would result in the appropriate LED on the TTL splitter being
>> lit and the appropriate (pull-up or pull-down) stack would be turned
>> continuously: however this cannot result in the HV power supply seeing
>> an overcurrent unless there is a short circuit, e.g. on the output of
>> the stacks. A similar argument applies to my item "2", a faulty
>> interlock can trip the system and cause either the pull-up or pull-down
>> stack to be turned on; however, again, this cannot result in the HV
>> power supply seeing an overcurrent unless there is a short, e.g. on the
>> output of the stacks
>> Hence I am sure that there must be a breakdown problem  (or short
>> circuit) associated with HV1 cabinet (e.g. deflector plates, feed-thru,
>> output of stacks ....).
>> When you tried starting up the kicker last night, at voltages of say 1kV
>> on both HV power supplies were the currents drawn from the 2 HV supplies
>> more or less identical or was HV1 higher current than HV2?t. What were
>> the operating conditions (polarity of HV1 & MV1, approximate frequency
>> and pulse duration).
>>
>> Cheers,
>> Mike
>>
>> >Dear Peter,
>> >
>> >There sounds as thought there is a problem with an electrical breakdown
>> in
>> >one of the NEG cabinets: the 'NEG 1' LED identifies an over current
>> fault.
>> >However I cannot recall which cabinet 'NEG 1' is (i.e. HV1 or MV1,
>> >although I think it is probably HV1) -- it depends on how the fiber
>> optic
>> >cables have been connected back to the TRIUMF splitter box. NEG
>> indicates
>> >the bottom stack in the cabinet. What polarity is HV1 supply (negative
>> I
>> >assume)?.
>> >
>> >2kV is probably the lowest voltage that the over current circuit can
>> >indicate a fault (approx. 8A of current = 2kV/((19R*9)+17R)). But this
>> >fault current must be flowing to ground through only the one stack.
>> >
>> >I agree with Gary that this cannot be caused by a problem with a single
>> >card. Instead there is either:
>> >1) a conductor (e.g. wire or grounding rod) or a real fault/breakdown
>> near
>> >the output to the deflector plates;
>> >2) a problem with an interlock (which, depending on the interlock,
>> would
>> >cause the controls to switch on all the cards to short the stacks to
>> >ground) -- but I doubt this as the fault current must be flowing though
>> >only one stack for the fault to be indicated at only 2kV;
>> >3) A problem with the fiber optic signal from the over current circuit
>> in
>> >the lower stack.
>> >
>> >I will have to give some thought to the best way of diagnosing this --
>> so
>> >I will email you again later. In the meantime please let me know the
>> >polarity of HV1 and have a look for any obvious clearance problems.
>> >
>> >Good luck,
>> >Mike
>> >
>>
>>
>> Winter Peter wrote:
>>
>> > Dear Mike,
>> >
>> > this evening we were trying to start the kicker again after it wasn
>> > running on HV the last days. While trying to ramp the HV up, we had a
>> > current regulation on the HV1 side. We could go up until 2kV and then
>> > suddenly the current went to 112mA and the voltage dropped to 0.41kV.
>> > There was a red LED on the small black box at the 'NEG 1'. We reset
>> > the kicker and went up with the HV on the HV2 side without problems.
>> > Then we tried slowly to ramp the HV1 side again and at 1.4kV we could
>> > see the red LEDs being on. The current was ~75mA and therefore still
>> > below 112mA but higher than usual. We then noticed that on the lower
>> > stack of the cards in HV1 the card DOWN-13 didn show a red LED.
>> >
>> > With the online help of Francoise we took out this card and checked
>> > the resistance. It showed 1.6MOhm in one direction (similar to one of
>> > the spare cards). Then we put it back and bypassed the door interlock
>> > in order to measure the voltage on the resistor C13. That measurement
>> > gave 15.1 V while the two neighbouring cards had 14.9 and 15.06 V for
>> > this resistor. According to what I learned from Francoise, this card
>> > might not be a problem.
>> >
>> > Now we don want to switch on this half of the kicker before we have
>> > some advice what to do next. Do you have any suggestions / ideas for
>> > the high current? In principle we could exchange the card with a spare
>> > one and test the kicker again but we rather wait for your answer.
>> >
>> > Looking forward for any suggestions,
>> > best greetings
>> > Peter
>> >
>>
>
>