From: "Alexander W. Chin" Subject: Re: 32kHz vs. 48kHz tracking, 0nce used tapes Date: Tue, 10 Jun 1997 07:00:22 +1000 Heiko, >From: Heiko Purnhagen >Subject: Re: 32kHz vs. 48kHz tracking >Date: Sun, 08 Jun 1997 20:30:27 +0200 > >Unfortunately my both DATs (57ES and D7) seem to be more prone to >tracking error in LP than in SP. From monitoring some capstan servo >signals on an oscilloscope, I have the impression that these tracking >errors are partly caused by (random/load-depended) irregularities in >capstan speed - since the capstan is spinning at less than 1 Hz (60rpm) >for LP. I observed the tracking problems mostly in the last part (last 5 >to 30 min) of a tape. Anybody with similar experiences ??? Yes! I have this exact same problem with both my D7 and D8. You are never going to believe what causes it, and even better, you will never believe the solution! The D7 and D8 never have a problem when brand new. Thing is when they get old, the get his problem, and many other people have told me individually about it too with the same machines. After a lot of time and research, get this... the problem is polishing! It works like this: When you buy a brand new D7 (or replace the heads - I have one spare one on my desk), the heads have a rough surface. You need a very powerful magnifying glass or microscope to see this. The surface of the drums (the spinning part) and the lower drum (stationary) are roughened up in the factory, very very find roughening though. For me, after about 6 months use @ 4 hours per day say around 600-700 hours, then the problem begins at the end of the tape. In the last 30 minutes or so, mistracking every 6 seconds or so. When the heads go under the magnifying glass again, they are nice and smooth and polished. From what I gather, you need a certain amount of friction on the heads if not they become to smooth and the tape sticks to them, literally. There is a layer of air between the tape and head, and if they are baby-bottom-skin smooth, then the tape sticks tighter to the heads. Sometimes the head rotation even seizes, and stops spinning! The solution? Sandpaper the heads a little, and all is well. You will all start laughing, but really this fixes the problem. The mistracking is caused not so much by the head, but since there is more resistance to the capstan which pulls the tape, it slips. On my machines, you see the capstan turning, but the tape movement is erratic. I guess this is less of a problem on big machines with huge capstans and pinch rollers, but on the tiny D7, the tape really slips. So, a little sandpaper works wonders. I have another fix prior to this before discovering the sandpaper: paint the capstan with black paint so it is more sticky. That way the tape doesn't slip. Go figure! Cheers Alex ============================================================================== Alexander W. Chin Unisearch Multimedia Suite 232 Systems Administrator National Innovation Centre E-mail : alexc@newt.phys.unsw.edu.au Australian Technology Park Phone : (612) 9209 4065 Garden Street, Eveleigh Fax : (612) 9209 4061 NSW 1430 http://gerulf.acsu.unsw.edu.au/~alex Australia From: "Alexander W. Chin" Subject: Re: 32kHz vs. 48kHz tracking Date: Sun, 15 Jun 1997 00:02:05 +1000 >From: pkmin@scimail.risc.rockwell.com (Peter Min) >Date: Tue, 10 Jun 1997 19:05:43 -0700 > >Ted-very good analyses here. Heiko monitored the capstan speed with a >scope somehow and it appears the capstan speed can vary. Alex's theory as >to why it varies (smoothing out of the heads thus causing higher friction) >makes great sense. You summed up a bunch of thoughts beautifully! Actually, I have more to add here. This experiment I went through had even deeper roots. I narrowed the problem down finally to the heads, as said before, but at first I noticed the speed variation that Heiko reported. I found out that during the mistracking, the capstan speed varied a bit, around 5 % if I remember. So, naturally, I thought that the problem was the capstan not being powerful enough to pull the tape. I suspected that the capstan motor was so small and weak and wimpy it could not deal with additional resistance in the tape path. I also noticed one more fact. The tape tension is not as constant as we would all like to believe it to be. It is less near the TOP of the tape, and more towards the END of the tape. This comes about from the mechanical tension feedback loop in the D7/D8 and also a lot of VHS VCR machines. I shall not go into details of how it works, but put simply, the reel diameter is less near END, and torque/friction at the hub provided by a pad gives more tension at the END compared to the huge reel at TOP. Also, since the hubs are not perfectly round, there is a cyclic tension variation each revolution, approximately 6-8 seconds. Measured this with a tension meter (Thanks Mike Scott - my very good buddy). Anyways, back to the capstan. Since I thought the motor was not strong enough, I hooked up a power amplifier I designed to provide more power to the driving capstan. Had to modify the D7 drive circuits a little. I powered the amplifier with at 15 volt supply, which would give ample power to pull the tape. After all this, I still saw the speed variation, and same mistracking. Didn't help a little bit. I also changed the capstan motor. Didn't help a bit either. New pinch rollers went it. Nope. New spring (Thanks Klay) for pinch roller tension. Nope. New reels. Nope. I even cut the pinch roller tension spring shorter to give higher tension/pressure. Helped a bit. Except the pressure was now too high, and could cause the capstan to stall totally (especially in LP). When this happened, very funky sounds came out... sounded like blah-daaa-daaa-daaa-daaa-daaa-daaa-daaa-daaa...(record skipping)..pop....buzzz...(the machine reset itself). So finally, new head unit. Uhm... where did all this mistracking suddenly go?? The tape worked fine, movement was very smooth (even the mechanism noise was much lower) and all problems gone. I even put back all the old reels, capstan motor, pinch roller, spring.. and restored the circuit board back to original. Still fine. So it had to be the head. The old head was still good, the CRO showed an almost perfect RF pattern. The heads themselves were not worn that much (or else the signal would be distorted). So it had to be physical, and the microscope pointed that out. I also took out the old head and dropped it into my then brand new D8 (Thanks Mike again), and got the mistracking shit again. So even a *brand new* D8 can show this problem with a very friction-ful head. To sum it up, after all this research, it is definitely slippage. It is not a question of capstan strength. The speed variation that I measured (and Heiko too) turns out to be the servo system compensating. When the tape slips, its movement slows down relative to what it should be. The servos sense this and speed up the capstan to pull it back to speed. This would work fine in theory, but if the tape has slipped, it can't move fast, no matter how fast the capstan turns. And if the capstan/tape interface were perfect, then you wouldn't see the speed change in capstan, it can't speed up without speeding the tape up at the same time, and then by doing that, your tape is now too fast! >Alex do you have a problem during the last 30 minutes of a 2 hour tape? Yes, I had. Now I don't. But my then brand new D8 is now beginning to show these symptoms. So looks like sandpaper will be happening soon. >Perhaps this is *one* reason why a tape gets *eaten* on occasion. It gets >stuck then you try to open the D7 and......(I have had 1 eaten tape >also)(another could be when capstan and head speeds become too discrepant). > If too discrepent one way --> snap. If too discrepant the other---> >slack. Yes, this problem was prevalent when the heads got too smooth. To be safe now, everytime I load a tape into my D7/D8, I hit the FF and let the deck fast wind first. After it drops out of fast wind, and goes full speed, then I drop into play mode. Oh one more thing. I should correct you since I have observed this with the cover off. If the head seizes the machine will reset. When the machine resets, it will unload the tape back into the cassette shell. But if the tape seized the head drum, it is stuck to the drum. So it cannot go back into the shell. It is still stuck there, even though the guides have retracted back. If you try to open the cover now, you have a very very very serious problem, since the tape is still inside the machine. If you ever get a reset, do what I do. Press FF to get the tape reloaded (which it could be still loaded anyway, being stuck to the drum). In most cases, dropping into fast wind can free up the adhesion. Then stop it, and unload it normally. What this means is that if ever your D7/D8 power cycles, always reload it, never open the door. I'll be getting a D3 soon, and I'll confirm if the D3 has this problem too (I am made to believe the D7/D8 is based on the D3 chassis). >"On my machines,you see the capstan turning, but the tape movement is >erratic. I guess thisis less of a problem on big machines with huge >capstans and pinch rollers,but on the tiny D7, the tape really slips." > >You haven't figured a way to observe the D7 in action, have you? Oh yes I have. ;) Just open it up and take the top cover off. Then open up the tape cage, and also take out the top plate that covers the head. You need to remove 6 screws in total. Then get a DAT tape, slide the bottom off, and expose the top flap. Load that into the machine, and put something heavy on it (I have a glass bottle of Parker Quink Ink, filled with coins for this). Then click the catch (door catch) and the machine will load the tape as usual, and you can watch it do its stuff. One thing though. Do not observe this under bright light, because the tape beginning and end sensors are optical. If the ambient light is too bright, then the machine thinks the tape is at END or TOP and refuses to work. Then you need to eject it and reload. >"paint the capstan with black paint so it is more sticky. That way the tape >doesn't slip." > >someone crazier than me? I thought the permanent marker to my mic grills >was a feat!!! > >I would say that if people are observing the same probs in home decks with >stronger motors than the D7, then Alexs, Ted Summed it up great, and Heiko >who did some research, theory best explains it(over my wimpy motor theory). > And you could also say that when the heads "freeze", sony has a shut off >feature. Yes, if the heads freeze, the machine resets. And I apologise for the long post. Most of it will not be relevant to non D7/D8 users, but I thought you would all like to hear some technical details. I would imagine any DAT deck will suffer from 'sticktion', but the small D7/D8 design makes it very apparant. Cheers Alex ============================================================================== Alexander W. Chin Unisearch Multimedia Suite 232 Systems Administrator National Innovation Centre E-mail : alexc@newt.phys.unsw.edu.au Australian Technology Park Phone : (612) 9209 4065 Garden Street, Eveleigh Fax : (612) 9209 4061 NSW 1430 http://gerulf.acsu.unsw.edu.au/~alex Australia