Problem

Solution

Details

Q47 Drive 1 and 3 fails backup. Customer has two Q47's each failing drive 1 and 3 when running a backup.

Found that each Q47 is setup in a tri buss configuration, they have DLT 4000 SE 50 pin drives. The controller / Host cable goes to a Black Box that is labeled "S S C L" from there the cable goes to the HP system. Drive A is daisy chained to drive B and the cable from B goes to a Black Box labeled "S S C L" from there the cable goes to the HP system. Drive C is daisy chained to D and a cable from D goes to a Black Box labeled "S S C L" from there the cable goes to the HP system. The controller, Drive A, and Drive C have a terminator. The Q47's pass confidence tests. The Cable to the "S S C L" box from drive D was rerouted in place of drive A's terminator. The cable to the "S S C L" from Drive B was rerouted in place of the terminator of the controller to daisy chain all the drives. With a laptop connected to the controller in remote diagnostic mode we ran drive load unload tests with no errors, also ran read write tests to drive A & C with no errors.

The problem is not in the Q47, it must be application related or something with these interface boxes they are using.

Drives C & D are down in a Q47.

Check the drives for status lights on their face plates. If either drive has any illuminated LEDs, the drives could be the problem. This can be verified by swapping the drives in question to another position. If the problem continues with the slot, the problem is related to the power supply or the connection associated with that slot. If the problem follows the drive, the drive is the issue. If neither drive has illuminated LEDs, chances are, the "Big Dog" power supply is out. Check the power supply and replace if needed.

On a BHT Q47, drives C & D are down. This is indicated by the front panel LCD displaying "drive not present" status messages for these drives. The drives or their shared power supply are the suspects.

On a Q47, a tape is left stranded. In this case, it was in one of the drives half in and half out and the tape is physically blocking the path of the picker so that any subsequent operations are not possible.

Remove the tape from its position and attempt to put the library back online. If this can be done, the drive load unload test should be run to check proper operation. If this passes, the incident is probably a fluke but if the diag fails or if the same problem occurs in the future, check the picker pads and clean with tap water. In the case of a Q2, clean the normal force wheels. When pick place errors occur, do the following:

1. Check picker pads and clean with tap water and a rag.
2. Clean the laser assembly lense.
3. Observe initialization sequence and confirm the laser targets the calibration label just below center.
4. Check to see if pick and place errors occur with the drives, the magazine cells, or the fixed cells. Each area has its own offsets.
5. Try resetting the datums.
6. Confirm the failure is not due to specific DLT cartridges.
7. Confirm that the magazines properly eject the tapes

The library mishandled the tape and must be checked.

Error: 150

1. Checked carousel for free movement.
2. Cleaned intensity label with an eraser and laser with alcohol wipe.
3. Tried different barcoded tapes and checked barcodes for contaminants.
4. Checked power supply voltages and adjusted.
5. Tried swapping the theta motor, then motor driver, then power supply.
6. Checked scanner output by connecting a laptop and performing the "calc sensor load" test. The output was 11 and typically should be between 6 to 30. (lower values indicate a bad laser.

Q47 library was getting barcode scanning errors. The library code was 4.43 and was updated to 4.51 due to better handling of barcode reading in updated code. Since the 4.51 code was installed, the Q47 now getserror 150 (theta drift)

Front Panel LCD is dim or not illuminated. Drive lights flashing, tape will not eject.

Measured and adjusted the 2326 power supplies to 5.25 volts dc as needed The problem remained the same so the positions of the two 2326 power supplies (left-most and middle positions) were swapped. This made the LCD illuminate normally. Rethreaded drive leader and tested.

The cause was the 2326 power supply in the number two (middle) position.

Unable to config changer on a Q7. Drive load unload passes, setup menu works normally and unit displays no sign of any problems yet the robotic CPU does not show up when the host bus is scanned. Drives show up when bus is scanned.

Defective CPU

Unable to config changer on a Q7. Drive load unload passes, setup menu works normally and unit displays no sign of any problems yet the robotic CPU does not show up when the host bus is scanned. Drives show up when bus is scanned.

Q47 getting errors while running and locks up. Drives going offline. Tape stuck in drive. Tape stuck in picker. Init error 141.

Drive A gets cleaning required light with multiple tapes and cleaning carts. Swapped drives D and A between their positions so that the A drive is in the D slot and vice versa. The problem moved with the drive. The error 141 is symptomatic of the drive not being ready to accept the insertion of the tape being moved by the picker.

Q47 getting errors while running and locks up. Drives going offline. Tape stuck in drive. Tape stuck in picker. Init error 141.

Error 18 on a Q47 due to binding in Y axis

Check all axis to be sure none are binding and move freely. Remove all shipping locks and any other source of motion impedance from the interior of the unit before turning on the power.

Shipping locks were not removed during install. When power was applied, the unit could not move the robotic picker assy thus the unit failed to initialize and the error 18 was displayed.

Get I/O errors from Q47

This problem occurs on any drive that has the i/o cable from the computer. The big dog p/s was swapped because the problem initially looked like it stayed with drive D. The scsi buss length is 14ft on this system, which includes the hot to library cable and the 18 in daisy chain cables and internal cables. The SE version of this library only supports up to 9.8 ft. Will swap daisy chain cables for 12in or shorter if possible. Also will swap library controller. After swapping cables, the problem seems to be improved, but not resolved...still getting i/o errors, but not a frequently. Conferenced Breece Hill Boulder in...they found that the original info we had been given was not correct for DLT4000 SE drives. The SCSI buss length is the standard SCSI spec not 9.8 ft. So we have about 19ft to play with. The buss connects directly to one of the tape drives then daisy chains from drive to drive finally terminating at the library controller. Swapping the library controller had no effect. The errors always seem to occur if more than three drives are in the chain and occur on the drive the buss from the computer is connected to. Had ae reverse buss and connect to the drive controller first then daisy chain to the drives. The i/o errors stopped. No reason it should make any difference to the scsi buss, but this seems to have resolved the problem.

Get I/O errors from Q47

Error 8 PICKER JAWS FAILED TO OPEN

One or both of the picker solenoids did not open after 20 tries. Per the QSx70 Library Service Guide available here, this could be due to the picker solenoid flag being out of adjustment, the Z-Flex cable, the motor driver cca, the z carriage assembly, or the y flex cable.

Error 8 PICKER JAWS FAILED TO OPEN

Error: 14, Error 22 on a Q7 or Q47 Movement of the robotics is slow.

Error 14 T_Home_TOUT - Home of Theta axis took too long.Error 22 Mag_Prs_Pos_Err - Theta Axis difference between desired and actual position exceeded tolerance during magazine present detection phase.

Error: 14, Error 22 on a Q7 or Q47 Movement of the robotics is slow.

Q7 or Q47 with pick or place retries.

Reset the datums via the maintenance menu. Make sure that the cartridge presence senser switch is working properly and that the cartridge presence LED illuminates when the switch is activated. Check and adjust offsets as necessary.

Q7 or Q47 with pick or place retries.

Unable to config changer on a Q7. Drive load unload passes, setup menu works normally and unit displays no sign of any problems yet therobotic CPU does not show up when the host bus is scanned. Drives show up when bus is scanned.

Defective CPU

Run drive load unload test. Check and adjust power supplies as needed. Cable to the CPU only and do the bus scan. Replace the CPU.

+5 Volt Power supply voltage drifting on a Q7/Q47 2326 Power supply.

Try Cleaning the contacts on the Molex Connector. Adjust voltage to +5.20 and check again.

The Contacts on the Molex Plug on the power supply sometimes get dirty causing the voltage to be unstable. The 2326 power supply is a phenolic card with a power module bolted to it. The power module has a cable which is routed to a molex connector on the phenolic board. The molex connector is routed to the card edge connections via internal traces which is how the power supply is connected to the tape library mother board/backplane. Also worth noting is that the 2326 is an uprev to the 2325 power supply. Both cards are identical in fit, form, and function with one exception. The 2326 produces more amperage in order to support the power demands of the DLT7000 drives. If there are DLT7000 drives in a Q7 or Q47, the 2326 power supply is required in order for the drives to work properly. If the Q7/Q47 does not have DLT7000 drives, the 2325 will work fine, however if the 2325 power supply should ever be replaced, the 2326 will likely be the replacement as BHT no longer produces the 2325 power supply.

Q47 has intermittent barcode read errors.

Appears that barcode reade errors occur on different tapes and may actually misread some labels. Laser appears to start reading at bottom of first label and steadily gets off more as laser scans labels moving towards top of unit. AE to check laser alignments and to reset datums. AE will also check p/s voltages, run confidence tests, and swap labels around and with other q47 if possible as they have another that is offline. Cust will change barcode labels....Conferenced Don S. with BHT Boulder in to check for other issues with q47 code and any other suggestions. The cust is running 4.33 code the new 4.51 code will help barcode read issues. Don will send a couple of pages of BHT labels to verify if the labels are the problem. Also suggested cleaning white laser calibration sticker on the theta assy with an eraser. Cust will get about 20 tapes that are currently not in use and put new labels on them to verify that the labels fix the problem.

Q47 unable to read bar code labels. Some read ok, others have errors. Failures are random and not with particular labels.

1) Replaced labels with new labels purchased from Stevens Labels (Peaks Storages' vendor for bar code labels, 4875 Moline Street. Denver, Colorado (303) 373-1034). Previously confirmed failures are random and not specific to any particular label or storage position. Sometimes, a weak theta motor turning a fully loaded carousel causes this problem. This can also be caused by a worn theta bearing. The motor is field replaceable but in the case of the bearing, the chassis must be replaced as the bearing replacement can only be accomplished at the Breece Hill factory due to alignment procedures involved when re installing the carousel assembly.

2) Reset datums.

3) Verified offsets entered in the CPU match the values on the offsets listed on the back of the machine on the label next to the power receptacle.

4) Cleaned laser lens.

5) Checked laser to be sure it fires against the calibration label properly. This is done by verifying that the laser fires against the calibration label at a point approximately in the lower half of the calibration label but above the lowest third. If it is found that the laser is firing against the calibration label too high or too low, the Y flag position can be adjusted. The Y flag is located below the Y bridge and is attached with two screws. In some extreme cases, the mounting holes in the Y flag can be elongated. In this case, the offsets were correct and the laser did fire against the calibration label within the normal range.

6) If all the above is attempted and the problem continues, the firmware should be up revved. In this scenario. 4.41 is the code rev on the box. This rev of firmware is known to have bar code reading issues per Breece Hill Boulder.

7) Adjusting the offsets is the last thing to try to correct this problem.

Pick and place failures on a Q7 or Q47 a banging sound may be heard as well.

Y bridge not perpendicular and or the counterweight hits the bottom of the cabinet.