Cathode High Voltage - Experts Only

Changing the Cathode High Voltage

Starting the expert GUI

To bring up the expert GUI, kinit to get your kerberos credentials. Then, to log in through your own account :

ssh -X -Y

Or to log in through the common ubooneshift (you must be listed in the krb5 login file) :

From the DAQ node, log into the Glomation in the HV Rack : (The password is available to HV experts only, from other HV experts or from the slow controls experts.)
ssh -X -Y uboonedaq@

To run the GUI script :


If you want the commands recorded in a log file, then instead enter : (give the log file name the appropriate date in the format YYYYMMDD)
python >& RampLogs/ramplog_YYYYMMDD.txt

The expert GUI window should appear, similar to the image below. NOTE that the values in this image are probably not current.

Using the Expert GUI

The GUI has three function areas, Program, Readback, and Message Box.
Clicking on the orange buttons sends a command to the HV supply.
The Send Program button will tell the HV supply to change its settings to the values in the white boxes under the Program heading, as circled in the image above. The values in those boxes may or may not be what you want - always check the values before clicking Send Program

Set voltage: PRESENT VALUE INITIALLY - the value the supply is at NOW!
Set current: 40
HV Enable: 1
Set critical voltage: 70000
Ramp step 1: 40
Ramp step 2: 40
Ramp down step: 40

IF RAMPING DOWN, and BEFORE sending a command, enter the present voltage value in the set voltage window (otherwise the supply will try to go immediately to 0 when you press Send Program).
IF RAMPING UP, and before sending a command, enter zero in the set voltage window.
If the HV supply has tripped, then you need to CLEAR TRIPS first.
If HV Enable is initially set to 0, then send a command to set it to 1 before you change the voltage value.

The SEND PROGRAM button is disabled (white) while the present command is being executed.
The next command can be sent after the button returns to yellow/orange

While ramping up / down, we watch the Pick-Off value, and wait for this to stabilize; this can take a few seconds beyond when the next command can be sent. Stabilize means stable to +/- 0.05V. A record of the pick-off voltage value during three ramp-ups, at each 1kV set voltage, is in a file linked on this page, below.

Please record the values of the pick-off point voltage before moving onto the next step.

HV Drift Rack Power Down Procedure

  • Ensure that the Glassman is at 0 V. One can see this on the front panel. If the power happens to be off, the voltage is 0.

At this point, if a power outage occurs, it is ok from a HV health standpoint. The remaining steps are additions for a more graceful shutdown.

  • Turn off the Glassman by pushing the orange on/off button on the front of the supply.
  • Turn off the slow controls box by pulling the on/off switch on the front left side of the box down.
  • RPS can be turned off from the back of unit in the rack.

When you bring things back up after the rack has been powered down, the Keithley monitor will not be in the right state.
You must restore settings before you begin ramping up the HV. Do this by loading the .pvs file in SlowMon and right-clicking in any column to choose "Restore settings".

The .pvs file is attached at the bottom of this page, in case you cannot find it from the SlowMon GUI.

Response to a TPC High Voltage Power Supply Trip

Stop. If this occurs during ramping, stop. Turn the voltage on the power supply to zero if it is not already there. Turn off power to the supply.

Discuss with experiment leaders.

If it occurs after running for some time, also stop. Turn the voltage to zero volts if it is not already there. Discuss with the experiment leaders and/or spokespersons.

The first question is if there was actually a spark in the cryostat. We have a tripping supply meaning that if we draw above a set current level, e.g. during a significant discharge, the supply's output goes to zero volts. The supply can also go to zero volts due to power glitches in the unit. This is rare and has been seen only once during an electrical storm. The supply can also behave sub-optimally in extreme heat and humid conditions; such conditions should never be seen at LArTF during normal operations. Real sparks have a fast signal, and without the additional current monitoring and scope, we will not see them.

If we assume there was a spark, it is proposed that we first check out the health of the other subsystems.

  • Controls group: Notify the controls/cryo group. Ask if they see anything anomalous.
  • Electronics: Perform noise runs. Verify that none of the channels are damaged.
  • Light Collection: Ask the light collection group to check their read back.
  • Laser/Purity monitor: These systems are unlikely to be damaged, but they should verify read back.
  • Field cage: Some previous experiments have broken field cage components with high voltage events. MicroBooNE has learned from this and we not only installed more robust resistors, but also varistors for protection. Here, a 5~kV ramp can be performed and the current draw can be measured from the pick off point to check for a change in current draw indicative of a change in resistance due to a broken resistor. Discharges are not favored along the FR4 supports of the TPC.
  • HV: The feedthrough is likely not damaged. If there is reason to believe it or the cable is damaged, we have spares.

If all of the systems pass check out, one should consider the length of time the system was at the previous voltage before tripping and how many trips the experiment will accept.

Location of Spares

We have spares of all of the contributed HV components.

  • Power supply - the PAB.
  • Glassman cable - the PAB.
  • Filter pot - the PAB.
  • Dielectric Sciences cable - the PAB, MTest, the DAB cage.
  • Feedthrough - MiniBooNE building. It is in a cylinder. It is labelled.
  • Spring tip - the PAB.