On 2024-04-22 around 3:30 pm it was discovered that the crane's main hoist was in a faulted state. The operator (Adam Newsome) experienced the fault after performing a safety reset and attempting to lower the hoist. There were no visible signs of any issues, mechanically. This fault had not been previously reported by any other crane operators. The crane was previously unused, sitting idle. Note: the auxiliary hoists appeared to function normally.

Fault messages on the HMI indicated:

"135. Main Hoist West Drum North Motor Drive Fault (1000VFD)"
"104. Main Hoist East Drum South Motor Drive Fault (800VFD)"
"103. Main Hoist East Drum North Motor Drive Fault (700VFD)"
"138. Main Hoist West Drum South Motor Drive Not Ready (900VFD)"
"136. Main Hoist West Drum South Motor Drive Fault (900VFD)"

It is clear from this that there is either some sort of common issue seen across all these main hoist VFDs, or one issue with one of them which caused a cascading series of faults.

On 2024-04-23, upon investigation of these faults by going online with the PLC to determine fault logic, it was clear that all faults were indeed present, but there was no obvious indication as to what it was. Inspection of the VFDs in the control panel showed that all of the aforementioned drives displayed a fault code. This was fault code 4 which indicates DC bus undervoltage.

After researching this fault online, it appears this is typically caused by an issue with the input mains supply (480VAC @ 3 phase in this case), or by the drive's input DC filter/buffer circuitry. When measuring the DC bus voltage, it is expected for it to be approximately 1.414 times the AC supply voltage. In this case, 1.414*480 = 678 VDC. The first troubleshooting step was to measure to confirm the DC bus voltage on a known working and not working drive.

The bus voltage was probed on drive 500VFD which, based on lack of fault message, was expected to be functioning properly. With the safety off due to E-stop condition, the voltage was nearly zero. When a safety reset was pressed to enable the drive, the voltage changed to 690 VDC. This is close enough to the expected 678 VDC. The voltage did not change, even when the fault message for the other drives appeared again.

Next, the bus voltage on two known faulted drives, 900VFD and 1000VFD, was probed. In both cases, after the safety reset, the bus voltage started at 690 VDC but then slowly dropped down towards nearly zero (somewhere around 18 V) over a period of 3-5 seconds. It was during this transition when the voltage dropped off that the undervoltage condition became true, which triggered the fault.

As of 2024-04-24, the root cause of this issue is unknown. It will be investigated further and this e-log will be updated when a solution is found.

Update 2024-04-24: the root cause has been identified. The undervoltage fault was just a symptom as a result of the input contactors for these drives switching off, causing the DC bus voltage to drop slowly due to capacitance in the input filtering circuit. The reason for the input contactors switching off is attributed to safety signals dropping out due a fault observed specifically on 900VFD. When looking at the drive itself, a fault with code 4030 was displayed: Enc1 open wire.
This fault implies the encoder may be disconnected. The encoder wiring was checked: OK. The encoder cable was swapped with a unknown working one from an adjacent drive - the issue remained with 900VFD, which suggested the problem is with the drive itself. The encoder board (20-750-DENC-1) was swapped with an adjacent drive and the problem followed the board. The encoder board was then switched with a brand new spare (note: jumpers needed to be set!).
Upon power-cycling the system, the fault did not persist. Therefore, it is suspected that the encoder board had failed (in fact this happened previously with the same drive - see e-log 42). Upon inspection, one of the capacitors on the board appeared to be cracked - this will be investigated further to see if replacing it fixes the problem.

A spare encoder board will be ordered. This issue should be monitored in the future - it seems as if the drive itself is perhaps causing the encoder boards to fail.

Note: upon powering the system on again, another issue was noticed: 500VFD shows "drive not ready" fault. This is not displayed on the HMI though - it was just not possible to reset the safety system and this was only discovered from going online with the PLC. This will be investigated further.

Update 2024-04-26: the 500VFD "drive not ready fault" was investigated. It was determined that the SP+ (safety power +) signal wire was loose, which meant the safety signals to the drive were not getting through, causing it to remain in a "not ready" state. The wiring issue was corrected. Performing a safety reset resulted in successful drive enables across all drives. However, upon attempting to move the crane, even though all safety signals were green and it appeared to be able to move, it did not.
It turned out that there was a crane pendant fault: "Fault 305. Radio Control Receiver in Fault (3700RC)". This is exactly what happened previously in a similar situation (see e-log 55). The steps mentioned in that e-log were followed (power cycle, reconnect antenna/connector). Upon powering back up, everything worked as normal. The crane was tested in local mode - all three hoists up/down (main hoist run to upper limit), all trolley travels in each direction, and bridge travel in each direction.

As of now, everything is operating normally and all issues are considered to be resolved. A spare encoder board has been ordered.

The main trolley's E/W motion was observed to be slowed on 2024-02-13.
Upon investigation, the following faults were present:
- 200. Rotating Hook Velocity Difference between SSI Encoder and Drive Speed Feedback
- 82. Main Hoist Slack Rope

It is suspected that these faults are irrelevant in this situation. Upon inspection of the main trolley status screen, it was evident that both the SLOW EAST and SLOW WEST indicators were on (which matches observations). Clicking one of these status icons showed "Main Trolley Encoder Not Homed 2253ENC".

Logging in as an administrator allowed for homing the encoder.

The trolley was moved all the way East, to the limit, and the encoder was homed at position 0.389m (encoder count = 6886940). Note: the alignment arrows were not used as this would have required gaining access to the crane which was difficult at the time. It is suggested to properly home it in the future.

The E/W motions were fully tested to confirm proper slow down and stopping at the expected positions. The crane appears to be operational as of 2024-02-14.

 

On August 1, 2023 HC1 was pumped down to -1200 Pa (gauge) and then valved off to test the leak rate after turntables had been sealed with backer rod in the gaps.

The sealing with backer rod allowed the cell to reach the test pressure of 1000 Pa. The cell pressure equalized within 30-40 seconds which is an improvement but still not sufficient to meet the goals we have set.

Next step is to pump down to -500 Pa and use smoke generator and perform thorough investigation to identify any leaks.

It was observed recently that the main hoist's wires seemed skewed, and there was occasionally some noise heard when raising and lowering the hook block. It was suspected that this was due to wires having jumped in their tracks at some point. This is typically caused by side loading. It is possible that wire wrap skew occurred during the incident outlined in e-log 46. 

Gordon Crane was contacted for a service call. Crane inspector Ali from Gordon Crane was supervised/assisted by Maico Dalla Valle. They lowered the hoist down into the target pit and supported it on saw horses, then let the wires run all the way out. Ali confirmed that there was a wire wrapping issue, and re-seated the wires properly. The crane is now functioning normally. It is recommended to re-apply lubrication as soon as possible.  

Note: Ali from Gordon Crane has advised that it is acceptable for some degree of wire skew to occur - they can handle the abrasion if there's a small angle between them. However, wires should not be rubbing across each other at a large angle (ex. more than approx. 10 degrees). 

Today while operating the crane with load, the crane stopped moving. Upon investigating, all safety signals appeared to be OK. Fault messages regarding drive faults for various VFDs were present, which seemed to stem from an ethernet adapter card fault (see attached photo). It is unknown how this occurred, but suspected that it was due to a brief blip in network communications or the external drive power supply control signal. It is worth noting that the diesel generator tests took place today, and there could have been some affect from a power surge because of this.

The crane was power-cycled using the main disconnect, and safety system reset. The faults disappeared. The crane was operational again.

If this issue occurs again, it should be investigated more thoroughly. It would be worth checking if there were any generator tests or other things which could cause some sort of power surge.

There was an issue with a 20 A circuit breaker used to control motor assist functions for the telemanipulators (specifically, the rightmost CB in the panel for the right (East) manipulator set). This circuit breaker was moved to the panel for the left (West) manipulator set in the rightmost position, and then subsequently replaced with a new one, which is functioning correctly. It is suspected that the former CB which was not working was calibrated towards the low end from the factory, based on discussions with the manufacturer and extensive testing of multiple scenarios.

A spare pendant for the overhead crane was procured from COH. The pendant's functionality is the same as the original one's. The pendant was tested today by Adam Newsome - all functions behave as expected.

Important - to change to the spare pendant in the case of failure of the original one, insert a battery into it, and transfer over the small black transmitter card shown in the attached picture. The pendant will not pair with the receiver without this card. There is nothing else that needs to be set up for it to work.

The spare pendant is kept in a labeled box on top of the electrical cabinets located in the crane control room (L1 floor).

An issue was observed in which the pendant did not work to control the crane. This is the second time in recent memory this has happened. See e-log 55 for details. The same sequence of steps outlined in e-log 55 was followed, and the pendant control was re-established. It is still unclear what the root cause of this issue is, but it is suspected that the crane pendant loses communication with the receiver for some reason, and even though communication can be re-established, the reset button on the crane control console does not clear the fault message, and the system must be power-cycled in order to fully reset. This is just speculation based on what is observed. Further investigation required.

It was observed today that the crane's west light was not functioning.

[edit 2023-01-03] The crane was powered back on this morning - there do not appear to be any faults on the HMI and it seems that the crane was unaffected by the power outage.

The ARIEL target hall overhead crane was powered off for the holidays. In the new year, throw the main disconnect in the crane control room to power back on.

In preparing to troubleshoot the issue described in e-log #54, the crane was power cycled, and the pendant battery removed, to get to a "fresh start" state. Prior to this, the crane functioned normally aside from the incorrect E/W direction control on the North Aux Hoist.

Upon powering the crane back on, the pendant did not work to control the crane in any mode. The HMI had a fault present which read "305. Radio Control Receiver in Fault (3700RC)". This was accompanied by a red light in the safety control relay status summary screen, for "Radio Control Converter Communication Fault With Receiver (3714CONV)".

If the crane was placed in remote mode, the safety control relay status summary all turned green, and the reset button could be pressed, enabling remote mode. This means the issue was specific to the pendant/receiver communication specifically.

The receiver (in the cabinet in the control room) was inspected and the power status LED was on, the RF LED was off (which means there is no issue, OK), and the "OK to receive signal" LED was on - this indicates that the receiver was functioning as expected.

To further troubleshoot the issue - the crane was powered off, the receiver's antenna was disconnected and reconnected, and the multipin connector for the receiver was disconnected and reconnected. The system was powered back on, and the reset button was pressed. The fault message and the safety relay issue were cleared, and the pendant functioned again as normally expected (aside from the persistent issue from e-log #54).

It is unclear exactly what the issue was and how it was resolved, but could have been related to either a loose connector, or power cycling and resetting the system.

Update 2022-10-06: During routine inspection of the crane, Tom Kauss investigated the direction swap issue. It was determined that there was likely an issue with the initial installation of the new VFD back in March 2022, and that it was unnoticed (directional control wires crossed). The issue has been resolved by Tom Kauss and is no longer a concern. 

It was observed during operation today (Aug. 23, 2022) that the motion axes were reversed for East/West travel on the North Aux Hoist. i.e. when pressing the joystick on the local pendant in the East direction, the hoist traveled West, and vice-versa. This behaviour was not noticed (or at least not reported) before. It could potentially be related to the recent replacement which was done in March 2022 (https://elog.triumf.ca/TIS/RH-ARIEL/52), although the motion was checked after replacement. It is possible that an oversight was made in the testing at that time though.

The controls should be checked and the issue should be rectified as soon as possible. The Aux Hoist controls are not run through the PLC, but rather handled directly within the hoist's control box. It is recommended to go online with the PLC and check if there is any hint as to why this is occurring, before making any modifications. If no info can be gained, it is recommended to try swapping the VFDs for East and West motion, and check if that rectifies the issue. It is possible that the VFDs are programmed specifically for their direction output, and they simply need to be swapped. 

The position limit modification described in e-log 49 has been reverted. Currently the South soft limit is set back to 0.36m.

The limit was tested and confirmed (craned stopped at 0.33m position which is acceptable due to coasting). The entire bridge range of motion was checked to ensure there were no adverse effects - all is OK. The South bumpers need to be reinstalled - this will take place in the coming weeks.

Changes noted in crane log book.

Relevant work permit: A2022-05-30-3

Edit (2022-06-08): the bumpers have been re-installed.

The North Aux Hoist VFD which had failed (e-log 50) was replaced today by Tom Kauss and Adam Newsome on Work Permit A2022-03-15-1.

The hoist's motions were all tested - it is functioning normally again.

Naimat Khan noticed after using the crane that some grease dropped onto a spreader bar, and believes it is a leak from the main hoist (see attached photo). The hoist should be inspected.

Edit: Dan McDonald notes that this has been observed since the crane was installed and is likely due to excess grease dropping out. Perhaps not a major concern.

An issue was observed in which the North Aux hoist failed to travel West when commanded to, but the East motion was functioning. The up/down motion was also functioning. The North Aux Hoist Trolley failed during regular usage, seemingly randomly, when it was about 70% towards the West end of travel. Note that the West motion VFD inside the North Aux Hoist is the one previously taken from the South Aux Hoist (the same issue with the North Aux Hoist was observed last year, and the VFD was swapped, and a new one was installed in the South Aux Hoist).

A replacement VFD (Delta TDN007E1100WM0) will be ordered and installed, and this e-log will be updated.

Edit 2022-03-15: the replacement VFD has been installed and the system is operational. See e-log 52.

Edit: as of May 30, 2022, the modification has been reversed, and the crane is functioning normally as per the manufacturer's limits. 

Modifications have been made to the crane to allow for the South Aux Hoist to reach closer to the South wall during the ARIEL Hot Cell installation period. These modifications are temporary, and were considered to be a safer and simpler alternative to the modification proposed by the hot cell contractor workers, which involved installing a sort of come-along on the wall and pulling the hoist closer to obtain the desired reach. The following modifications were made:

1. Removal of the bumpers on the south side (T. Kauss). There are no bumpers on the South side of the crane’s bridge currently.

2. Modification of the PLC program to reduce the software limit by approximately 4 inches (A. Newsome). The previous soft limit was set at 0.36 m and it was reduced to 0.26 m (~3.9 inch reduction).

 These changes have been tested – the crane stops in the expected position. However, it is very close to the wall now. If you are using the crane anywhere near the South wall: exercise extreme caution. The crane automatically slows down when approaching the South wall, and there is plenty of time for the operator to react and further slow the motion by only holding the joystick down at about 20% speed. I strongly recommend operating at about 20% speed when close to the wall (on top of the automatic speed reduction I mentioned).  

• These changes are based on Work Permit A2022-01-06-1
• Changes have been noted in the crane’s logbook

See attached screenshot of PLC program indicating that the new limit has been reached.

Contact me (Adam Newsome) for more detailed information.

The load cell which is located on the crane itself to display the load to operators was reported not working during the week of November 15. It is planned to investigate on Nov. 25 and this e-log will be updated with the status.

Edit (2022-02-02): the load cell is functioning again. No changes were made.

The bridge motion was slowed due to encoder homing issue. It is likely that the problem which occurred yesterday caused an velocity fault (mismatch between encoder speed and commanded speed), which unlatches the homing status for the bridge motion axis.

When the bridge was in the southmost position (bumper pressed against wall), the position readback was: 0.353 m, encoder count = 5164003.

The bridge was moved to the "zero position" as indicated on the remote display - the tick was aligned with the zero.

The encoder was homed at this position. The home position readback was: 0.356 m, encoder count = 5164447. Note that this position is in alignment with the measuring tape zero mark, not pressed up against the wall.

The bridge was run to its full extreme positions in either direction to confirm proper slowdown and stopping behaviour. It appears to be functioning normally again.

A fault occurred which presented as "114. Main Hoist Cable Broken (3200PX)". This occurred when a SEG block was suspended approximately 1 m off the ground.

Upon inspection, it was determined that the cable was in fact not broken, but wrapped around itself. There are no guides, and it is possible for the cable to double wrap if loaded in an usual fashion. This condition triggered the "Hoist Rocker Switch 1" sensor, 3200PX. The sensor is not visible or accessible from ground level, and could not be inspected, but it was confirmed that there was no 24 VDC PLC input at 30000EA:I.Data[7].0, as would be expected under normal operating conditions.

To rectify the issue:

• The sensor was bypassed in the control panel by connecting wire 32001 directly to 24 VDC after confirming that there would be no adverse affects from doing this.
• With careful supervision, the load was lowered, while watching the cabling.
• The load was removed from the crane.
• The hoist was then lowered to its extreme lower limit, and the cable unwrapped itself naturally as it was being lowered.
• The sensor bypass was removed.
• The sensor began functioning normally again when the wrapping issue was resolved.
• The hoist was raised and lowered multiple times to confirm the cable wrapping issue was no longer present.

The crane is returned to its normally functioning state. Operators will keep a lookout for any more issues relating to the hoist cable. The crane will be inspected on Sept. 22, 2021.