Friday, November 1, 2019

Ongoing diagnosis of load function for Telex 8020 drive A


Sleazebag spammers promoting services and goods have been relentlessly posting fake comments on my blog using profile names such as "American Football". I have had to moderate comments now, so that I can keep them from benefiting in any way.

There has always been a low level of such spammery, with generic comments that include links to other sites in the body of the text. When it was infrequent I didn't mind them sitting there for a day or two, but now that I can get 4-6 comments a day cropping up, it became intolerable.

To legitimate viewers of the blog, it means that your comments will be delayed as much as a day until I can spot the legitimate comment and publish it. I wish I didn't have to moderate, but it is the only defense I have. Reporting dozens of spam posts has done nothing to stop the relentless waves of unwanted comments.


We are zooming in on the interval from when the Hub Vacuum sensor detects that the tape has wrapped around the take-up reel hub, until the 'dump' stage causes the tape to pull off the take-up hub. More specifically it is the timing of the transition to 'dump' that is the likely problem.

It seems that the correct behavior after detecting the hub vacuum would be to continue threading for a few more seconds, to achieve a few more turns on the take-up reel. Once sufficient tape is on that reel, dump state can begin. Its first action is to switch the vacuum and blower into run mode, where the vacuum is in the columns inside the tape path and the air pressure flows into the air bearings.

In dump state, the supply reel rotates clockwise to feed tape into the tape path and the take-up reel rotates counterclockwise to also feed back into the tape path. The vacuum in the columns will pull the tape down or up into the column as the reels rotate. When the tape in each column is in its proper position, the rotation of that reel stops. When both are in the proper position, the drive transitions to moving tape forward seeking the BOT reflective spot.

The symptom we see is that the tape has pulled entirely off the take-up reel and thus falls into the lower vacuum column, losing vacuum and causing a Load Check. This can occur for a number of reasons:
  • Not enough tape on the take-up reel so that dump can feed tape to the proper position in the lower vacuum column without the end of the tape falling off the reel.
  • Sensors in the lower vacuum column are not detecting the tape position correctly leading to continual feeding of that column
  • Servo loop for the take-up reel not working correctly, thus continuing to feed tape even though the tape reached the proper position in the column.

A properly equipped Telex customer engineer would have a very useful tool to help diagnose this - a plastic tape path cover. After swinging the metal cover open, the clear plastic plate fits atop the path permitting the CE to watch the tape movement. I don't have this so I have to work blind, instead recording sensor signals and watching external actions.

From detection of the Hub Vacuum condition, the logic should set a 5.5 second timer during which the tape continues to thread around the take-up reel, before it switches to dump state. In dump state, the diverter solenoid is dropped and vacuum switches to the tape columns. The rotation of the reels change to feed tape into the vacuum columns. Do I see the 5.5 second pause and is the take-up reel still rotating?

I set up my iPhone to record the loading process, with the VOM in view to see when the Hub Vacuum is detected. I recorded it in slow motion mode so that I could better observe timing, rotations of reels and other behavior. The VOM will slightly lag the signal change, but a missing 5.5s delay will be really visible.

As the old saying goes, a watched pot never boils. In this case, there are two possible ways that could be true. It could have successfully loaded (Not today). It could fail to thread down to the take-up reel even after dozens of tries. Sadly, that is what happened.

By the end of the day, I ran somewhere between 70 and 100 load attempts and exactly once the tape wound around the take-up reel. I know it wasn't 8.5 seconds of turning after the tape touched the hub vacuum ports, so my attention will turn to the timers on the control board.

This board uses a single 555 timer chip but has four transistor circuits to switch in different resistor values. These determine the different time delays used in the control logic. I thought I needed to use micrograbbers to route signals out from the control board where I can hook them to my scope. I want to see the various time durations of the 555 and relate that to the intended intervals. Fortunately, there is a test point (TP4) on the edge of the board for exactly this purpose.

Selectable resistance for 555 timer delays
When I figure out whether the duration is odd for all or just the 8.5s wrapping interval, it helps set the strategy for which components to check next. There may be multiple transistors conducting, thus changing the resistance in the circuit to the parallel equivalent, or one not conducting, or components that drifted off value.

No comments:

Post a Comment