VectorFieldSystemMagnetQuenches

Here we document all the magnet quenches that have happened to the vector field system.

Magnet Quenches Log

1. 27-Nov-2011 12:00 DAQ mode

Magnet quench happened right after sweeping Bz to -0.5T at rate 0.1T/min when Bx=4.5T in Vector(DAQ) mode. 5T magnet temperature went over 7K before the magnet quench.

Possible reason: Ramping Bz too fast and warmed up the 5T magnet. As soon as the magnet temperature is above 7K, the "Quench Detect.vi" will ramp all the fields down to 0 in 5 mins, which suggests it's not a real physical quench but a "software quench".

2. 27-Nov-2011 15:55 DAQ mode

Magnet quenched happened again when Bx=4.5T and Bz=-0.5T and running a sweep with Bz target -0.5T (also vector mode). 5T magnet warmed up to over 90K(Opp Instr Port) and the fridge also crushed due to this large amount of heat load.

NOTE: Re-tested the sweep which has a target the same as the current field and which was running when the magnet quench happened, no problems were caused - Sweep B vi stopped normally after taking a few data points.

3. 7 May 2012 DAQ mode

Giordano Scapucci writes:

Rotating the XZ field of 0.9T at rate ... give details caused ... to happen.

Afterwards restarted the software, retrained the magnets (details).

Problem probably caused by ...

4. 3 AUG 2012 DAQ mode

B=0.4T, rotationg field in x-y plane. The Y-magnet quenchd at ~ 16:30. The Y magnet PSU reads:

"Quench trip at 0.000 Tesla"

The quench is attributed to the spike on 9T magnet temperature. After the quench, there was no error message given by the vi. The software cannot control the X and Y PSU (Z magnet is fine). The computer sweep says it is ramping the field, but magnet PSU shows 0T and 0 current.

Restart the Labview and restart the computer did not help. After restart PS, X and Y magnet work fine.

Possible reason: There are something wrong with Keithley 2700 or the cable connected to the magnet themometer.

5. 7 AUG 2012 DAQ mode

B=0.4T, rotationg field in x-y-z plane. The X-magnet quenchd at ~ 6:30 am. The vi and PSU did not give any error message. According to the log, there were large peaks (>30K)on 5T and 1T switch. But the magnet temperature were below 7K.

After the quench, the software cannot control PSU as 3 AUg 2012.

Alex called the company who wrote the control software.

• The quench-detect is always on (no matter whether you are in DAQ or USB mode)
• The quench wil be detected by PSU when the current through the magnet is decreasing and the voltage across the magnet increase (ΔV>5V). That will lead to a harware quench, which will display '' quench trip at xxx T '' on the PSU front panel.
• If the current through the magnet is decreasing and ΔV>7V, the PSU will switch off the power supply, which supplies power to the power block transistor. The power block transistor deliever current to the magnet. If that happens, the power cannot be sent to the magnet. But the PSU may still appears to be working.
• To reset the switch, the only way is restarting the PSU.
• The PSU has some problem in changing the field direction. If you go 1T to -1T, the PSU may misbehave. So whenever you want to change the field direction in your sweep, it is highly recommonded to put B=0 into your steps.

6. 15 JAN 2014 DAQ mode

Sweeping B in polar coordinates (rather tan the usual cartesian). Initial conditions: R=0.9T, Theta=90deg, Phi=0deg. After completing the XY plane full sweep of theta from 0deg to 360deg (ie targets were R=0.9T, Theta=90deg, Phi=360deg), software finished at Phi=359.958deg.

After that a sweep was set in polar coordinates to bring the field to zero ie. R=0T, Theta=90deg, Phi=360deg (ie. the only number which changed was R, the angles were kept the same in an attempt to keep the software from recalculating a backwards path). At this point the only magnet that should be swept to get back to 0T is Bx.

However during this sweep to zero the software calculated a spiral path that swept By on top of Bx!! The sweep rate was much too high to cope with the heat load from both magnets, and both magnets reached a temperature >7K and quenched (see picture below).

Recommendation - for any sweep to 0T in DAQ mode to finish off an experiment, use the Cartesian coordinate system instead of polar!!

The following day after this event, all 3 magnets were subsequently trained in a series of sweeps through 0T.

A few mini-quenches were present.

7. 7 May 2014 DAC mode

The magnetic quenched when Bz = 3.5T and Bx=By=0T in polar coordinates. A hardware quench was triggered when setting Phi from 0 to 180 at Theta = 0.

8. 22 May 2014 DAC mode

The magnet quenched at Bz ~ 3.2 T, Bx ~ 1.5T at ~7:40pm. The Bx and Bz went to 0 very fast and the PSU for Bx and Bz displayed "Quench Trip at 0.00 Amps".

The magnet is NOT sweeping. The cooling water in temperature is normal and no other problem has been found on the compressor. But according to the lab logbook, the small-cryostat's compressor in Jack's office was switch off at the same time. So maybe the switching off of that small compressor triggered the quench.

Temperature of the magnet

Field

9. 23 May 2014 DAC mode

A hardware quench was triggered when setting to persistent mode.The 9T magnet PSU displays "Quench Trip at 0.00 Amp". The 5T front panel display is normal.

• All the switch heater was switched off and the temperature <5K before the software decrease the current.
• It seems the 9T magnet is quenched before the software decrease the current.
• The quench of 9T magnet triggered the quench of the 5T magnet
• In the condition of hardware quench, the software cannot record the correct magnet value so the Bx and Bz is not the actual field. They are only indications of when the software decrease the lead current.

10. 18 Jan 2016 - 17:30 DAQ Mode

Quench of the Y magnet (1T) occured at ~0.9T at a sweep rate of 0.1T/min. No quench message was displayed on the magnet PSU. Probably caused by magnets getting too hot, temps peaked at ~13K.

11. 18 May 2020 - 13:00 DAC mode

We replaced two TRACO AC/DC power modules (TML 15105) inside the x- and y- magnet power supplies.

Then we ramped y-magnet up to 0.5 T for 400 s. The magnet reached 0.5 T without a problem. But the temperatures slowly increased even after the magnet reaches 0.5 T. The magnet eventually quenched and the magnet swept back to zero rapidly. Below are the log files:

<- Temperature after the quench.

We can see that magnet temperature has increased even though the magnet reached the target (0.5 T). It looks like the magnet quenched when one of the magnet temperature hits 7 Kelvin.

Magnet Training after Quench

Trainging in DAQ mode

After the big quench on 27/11/11, the 5T magnet experienced a lot of small quenches - temperature spikes shown below even when it was at zero-field and only either the 9T or 1T magnet was ramping.

Ramp the 5T magnet to full field slowly:

• First it had small quenches even at very slow sweeping rates;
• The amplitude of the temperature spikes got smaller;
• Ramped it to 1T and brought it back to zero at rate 0.05T/min(medium-fast rate);
• Then the magnet became much more stable and reached 5T without any temperature spikes.

But temperature spikes in the 9T magnet can still be observed every now and then when it reaches a field around -0.3T as shown by the following figure.

Small Quenches in USB mode

When we switched to USB the first time after the quench, the 9T and 5T magnets experiened a series of small quenches while sweeping the field.

Problems with the DAQ Magnet Control software - Small Quenches

We found that the status of the magnets changes every time stopping and restarting the "Vector Magnet Control.vi". And sometimes the difference can be as large as 15mT and cause small quenches of the magnets.

Access Control

Previously the default template was set up so that all pages were visible by the whole world, and we relied upon authors to use access control to limit access to sensitive pages. That hasn't worked, so now pages are by default only visible to QED group members. Delete the following line if this page is meant to be visible by the whole world. (*) Set ALLOWTOPICVIEW = QedGroup

Tags

create new tag

Comments