AMIMagnetPowerSupplyProblems

Problems with AMI / Kepco magnet power supplies on the Bluefors LD250 Poseidon and Bluefors XLD fridges.

Problem 1 (Poseidon Kepco Slave Power Supply failed, during installation Feb 2020)

During installation the Kepco slave power supply for the Y? magnet would not turn on - it kept faulting. Flicking the slave mains power switch on and off every 2-3 minutes would eventually start it, but AMI recommend not to do this as it may damage the slave power supply.

A replacement power supply was shipped from BlueFors and installed.

Install date: XX/XX/XXXX. Serial E167142

Problem 2 (Poseidon Kepco Slave Power Supply failed, May 2020)

The Z-magnet Kepco slave power supply (serial E167154) started exhibiting similar problems to the above.

The issue was reported to AMI/BlueFors 5 May 2020. A replacement power supply (serial E167181) was delivered in late Jan/early Feb 2021 and installed on 8/2/21.

Problem 3 (Poseidon z-axis 430 Controller failed, 5 Jan 2021)

The model 430 magnet programmer (z-axis serial 430-19-014) is tripping the circuit breaker. After opening it and having an inspection inside we found two problems:

1. The power line filter is broken: “P” is not continuous.

2. The blue neutral cable coming from 230V AC to the AC DC converter is tightly clamped between the case and the box for temp sensor connection (photo attached). This damaged cable causes a short from 230V AC neutral to case/ground which can be quite dangerous.
The damaged cable can’t be seen without removing the power line filter and is clearly caused by tightening the box for the temp sensor. It is most likely what was causing the trip and could also be the reason why the filter is broken.

This was reported to AMI (RMA 3449) and a replacement was shipped immediately. The new unit is installed on 8/2/21 (serial 430-20-137). Corresponding current transducer is also replaced.

Problem 4 (Poseidon x-axis 430 programmer front panel failed, Feb 2021)

6 February 2021. The x-programmer, serial 430-19-059 is now not working. Constantly displaying "Error - key is inactive in this display" and beeping.

Daisy reported:

"This morning I found that the X-programmer of the Poseidon fridge is beeping as if buttons on the front panel are being pressed randomly (video attached Video1.mov). The X-magnet was staying at 0T with switch heater on (Z-magnet was sweeping between -1T and 1T).
This problem happened once last year with the Y-programmer when its slave PS was faulty but never happened again since we swapped the slave PS. However, this time the X slave PS is not showing any problem.
I tried to power cycle the system but it didn’t help (video after power cycle attached Video2.mov). The buttons on the front panel are not functioning properly either when this happens so I suspect some problem with the buttons and/or its circuit."

Reported to AMI/BlueFors 8 Feb 2021.

On 8/2/21 the programmer boots and displays "Turn on power supply Press enter to contiue" when powered on. The slave PS powers on with no fault, but the programmer shows no repsonse when enter is pressed.

Problem (installation error) (Poseidon current transducers, Feb 2021)

8 Feb 2021 When replacing the z programmer and its transducer Daisy found that the current transducers of all three magnets are connected in a different way from what AMI manual suggests.

According to the manual +ve lead need to go through the transducer in the right direction (indicated by arrows on the transducer). However, in our system, the -ve leads go through the transducer (also in the opposite direction).

This is the configuration the system came in - Krittika. Not sure if this is a problem - presumably the current in two leads should be the same?

Problem 6 (Poseidon y-axis 430 programmer front panel failed, April 2021)

9 April 2021. The y-programmer, serial 430-19-020 is now not working. Same error as the x-axis. Completely dead and doesn't respond to any button presses.
Reported to AMI on 16 April 2021. They are aware of issue, but will not have new front panels until end of April (email from Michael Diederich on 16-4-2021).

Problem 7 (Poseidon y-axis 430 programmer front panel failed, (again) March 2023)

24 March 2023. The y-axis front panel, serial 430-19-020 is not working (frozen keys). Same error as the x-axis and y-axis in 2021.
Reported to AMI on 28 March 2023. They are aware of issue and posted another fron panel ( Michael Diederich contact). This was replaced by J Hillier and is fully working again (3 April 2023). Michael Diederich stated that from now on, all front panels are provided by new supplier (should not fail?).

Problem 8 (Poseidon z-axis 430 programmer front panel failed, October 2023)

03 October 2023. The z-axis front panel is not working (frozen keys). Same error as previous 4 events.
Reported to AMI on 03 October 2023. They are going to post another fron panel ( Michael Diederich contact). This was replaced by J Hillier and is fully working again (20 October 2023)

Problem 9 (xLD Kepco slave power supply failed, November 2023)

27 November 2023. Kepco Slave power supply E167142 failed. This is a replacement for a previous power supply that had failed with the Poseidon fridge was installed (see problem 1). Same fault.
For some reason this power supply is presently installed on the xLD fridge - Aaquib/Scott can you please add details of why?

28th March 2023: As soon as the y-axis power supply (installed on xLD Fridge) was turned on there was an electrical arc - the power supply is broken.
Replaced the y-axis power supply unit with the one is Poseidon as there was no experiment running on it.

Problem 10 (xLD slave power supply unit: Fault light on. February 2024)

This problem also happened on 22nd November 2023. Scott power cycled the magnet on 12th December 2023 (instructions provided by the AMI support) and the fault light problem was fixed.

2nd February 2024. The fault light of the Y-axis (E167142) power supply is illuminated.
On 2nd February Aaquib reported that the fault light is illuminated again. During this period (when Scott power cycled the magnet on 12th December 2023 and 2nd February 2024) the magnet was not in use (not ramped or not held at constant magnetic field). The controller had to be taken out and connected to the power supply and the fault light disappeared. But when the the controller was put back in rack and switched on the current leads energized. Scott ramped it to zero and switched off the PSU and after switching it on the Fault light was illuminated again.

Problem 11 (Poseidon - z-axis PSU fault, January 2024)

08/01/2024? The fault light for z-axis magnet PSU (SN: E167181) is illuminated, field was only ramp to a maximum of 100mT.

06 February 2024 Solution: Grounded device and power cycled PSU. Fault still present, this led to problem 12.

Problem 12 (Poseidon x-axis PSU and z-axis 430 programmer front panel failed (again), February 2024)

06/02/2024 After power cycling all three magnet PSUs and controllers, the z-axis 430 controller (SN 430 20 137) displays frozen keys error and x-axis PSU (SN: E167191) now shows fault light. The controllers are known to fail due to being poorly manufactured (Note the z-axis front panel is new, 4 months old and from new AMI supplier). Two faulty PSUs is a significant set back!

06/02/2024 Solution: Pending with AMI/KEPCO

14/02/2024 Spare 430 controller front panel found, replaced by J Hillier. Front panel now functioning normally. No more spares left.

Problem 13 (Poseidon x-axis 430 controller front panel failed, November 2024)

12/11/2024 Replaced the x-axis 430 controller (SN 430 21 003) after it experienced the frozen key problem.

Problem 14 (Poseidon x-axis Kepco PSU failed, November 2024)

12/11/2024 Tried power cycling the x-axis PSU (SN: E167191) but fault light persists.

13/11/2024 Replaced the x-axis PSU with 'loaner' (SN: E165856) from AMI. Will likely need to get old PSU(s) repaired via Kepco, before returning the 'loaner'.