WTF happened to this 810 (Tandon mech) drive ?

[cwilbar]

 

So I recently picked up another 810.  This one looks like it had very little use back in the day.  No signs of having been opened, factory stickers over the screw holes, no yellowing, great condition.

 

I booted a half a dozen disks fine with it.

 

Now it won't boot anything.  I removed the top, and if I push the head assembly back I can get it to read track 0.  And I'm not talking a little bit... I'd say a nearly a track's width or more off.

 

How can a drive go this far out of alignment ?

 

Can the coupler? coupling the stepper slip on the Tandon 810s ?  On the 6 (protected) disks, it was doing the ratta tat tat process on each disk.

 

I'm not happy that I had to break into this mint drive.  Not much dust or anything inside either.  Very clean unit.

 

Before I go messing with loosening the screws and attempting alignment I'd like to know more about that coupler and if that can cause an issue like this on these Tandon mechs ....

 

Thanks,

 

-- Curt

 

[_The Doctor__]

after a clean and re-seat of whatever you can with deoxit (side board connections/chips/jumpers/switches), including dusting/aeroduster blasting of switches sensor / detectors... gently check if the stepper motor can be moved... if it slipped as you suggested it should still slip around under slight pressure.

 

If all is tight, and the rails are clean... make sure the pressure pad on the top side of the head assembly hasn't fallen out of it's holder or the holder slide springs assembly that apply pressure haven't become stuck or bound... if it has... you can more than likely re-glue the pad (putting the glue remnant side back in the holder and the hair/soft clean side towards the disk), and the pressure assembly is pushing nicely against the disk as it should...

 

make sure the belts and bands are straight and where they belong...

 

plug it all back in and see if that hasn't fixed your issues... I've done this sort of stuff forever and it's normally just this simple.

 

If not, then it's meter time, starting with the voltage checks and track zero sensor check...

[cwilbar]

Before I figured out it went grossly out of alignment, I had tried swapping analog boards and side boards.  Behavior stayed the same.

 

I then noticed that when it went back to track 0 (hitting the stop), it seemed to settle away from the stop.  So, I pushed the head assembly back to the stop (the stepper motor resisted me), at which point it read track 0.

 

This is what is making me think that the stepper motor and head assembly have somehow 'drifted' from their working position.  It seems that the 'settle points' of the stepper are no longer aligned with the tracks.  I've confirmed the normal alignment system is tight and didn't drift.... also making me suspect stepper motor/head asssembly issues.  The head does track freely via the stepper, it just doesn't land in the right place.

 

I've never messed with the Tandon 810s much, so I'm not familiar with how the stepper/head assembly connect, and if that is prone to any issues.

 

And if I recall, no 810 has a track 0 sensor (a seriously nasty omission in my opninion).

 

[_The Doctor__]

@cwilbar

-=>Yes a generic response to tandon mechs, but you can still clean, de-oxit, blow out the drive and lube it(the rail and motor/stepper), also debris have gotten in between the track zero stop screw etc before and made drives pop back to the wrong track...

 

-=>if it ain't broke don't fix it... did you put the original boards back in? if so did that cure anything?

 

if not, mark where the stop screw is, and adjust slightly away from the head, a tiny bit. the mech may be very close to settling where it's supposed to be... but popping to the wrong location...

exercise the drive between all changes and adjustment.

you can move the head when no power is on all the way in both directions then put it about half way...

Make sure to try another single sided single density 810 disk as well... just in case something is up with your disk...

 

Once the drive reads a disk it's normal for it to settle to another point on the disk so it may not sit at track zero.

 

It would seem you are looking around the forum and have access to the FSM which is darn near everywhere these days... pretty much explains all your looking for.

 

I wouldn't negate the entirety of a response just because one takes issue with a single piece of said information. Same as if someone responds to 'my 1050 doesn't work' and then quibbles about minor differences between WST and Tandon mechs afterwards, though swapping boards, roms, or mechs is a really big deal between the two different types. The general guidelines still hold true... and you can still clean the track zero stop/sensor etc.

Certain mechs had stop rings on their rails in a 1050 and dust bunny build up around the ring stopped those mechs from achieving full travel as well... so keep these things in mind if you've got the tinker bug and want to mess with drives for whatever reason- Be it a 1050 or an 810.

Edited June 24, 2020 by _The Doctor__

[cwilbar]

wasn't trying to quibble... just stating that to the best of my knowledge that all 810s had no track zero sensor, and was willing to be corrected if such an unusual beast did exist in some variant.

 

And I did all the usual.... check for obstructions to the mech, I lubed the rails, cleaned the heads, all the usual stuff.  The last thing I want to have to touch (until I'm sure it is off) is the alignment.

 

I've been through different disks, disks that load fine on other drives.  I posted as I'm at a loss to explain why it so drastically went out of alignment when nothing that I've found so far is loose.  It is off by anywhere from 0.75 to 1.25 track width (estimated).  Which seems way out (more than I've experienced to date).  And to do that so suddenly, I wanted to put it out there for input from anyone who may have seen that in case there might be something specific to Tandon 810s (as most of my 810 experience is with MPI 810s).

 

I did revert back to the original parts once I verified via swap that the issue did not seem to be coming from the electronics.

 

Just remembered that I have a Tandon mech with a dead head.  I'll investigate on that mech how the parts all go together... that might help me understand better what to look at as well.

 

Appreciate all the input.  I do have an orignal of the FSM (which does cover the Tandon mech too).  But as I've found, there is a body of knowledge out there that was never captured in FSMs ?

 

Thanks,

 

-- Curt

 

Edited June 24, 2020 by cwilbar
omitted info

[_The Doctor__]

Give the stop a tweak as noted... and check the rabbits fur pressure pad etc.

 

I wouldn't go about alignment until seeing if stop adjust doesn't change anything.

 

as for alignment, same deal mark precisely where things are before adjusting...

 

since you say you moved the head manually to track zero... did it start reading and working? (of course it would fail on re power or next zero seek) but this lets us know the head is still ok..

 

stepper motors rarely fail but when they do... sometimes re-orienting them helps you get by until you replace them. (or forget since it works and get another 5-10 years out of it, stranger things happen)

Edited June 24, 2020 by _The Doctor__

[cwilbar]

When I manually moved the head (I had to fight the stepper), it did start reading track 0, but then when it tracked from there it stopped reading.

 

I believe that the shaft has slipped on the rest of the mech, and now the steps inside the stepper are out of alignment with the tracks.

 

I'll do a bit of investigation tonight.

 

I always loved the MPI 810s.  Never thought much of the Tandon ones :-).  But this one is as close to new as I've seen, and for an 810 it had pretty quiet operation (well except for copy protection 'snark' behavior.... but that is to be expected).

 

[_The Doctor__]

okay, that makes sense

 

with this in mind (and I hate to say this).... with power on, you can very carefully, quickly, loosen the stepper and turn it till it just touches the track zero stop... and tighten it again.... please make sure nothing touches stresses or short while doing so (I use plastic or wood to support and insulate).

you should be pretty close to where you need to be...

booting should work once more...

after that to do a perfect job requires and o-scope and alignment disk.

a damn near perfect job can be had with a factory disk of almost any kind and marking the stepper adjustment and dialing the beast in with the marks and the oscilloscope... it won't be cat's eye, it will be discernible as to where the strongest output is...

 

finishing up make sure everything it tight when done... make sure that you do so without anything at full travel so you don't stretch or bump anything... probably will run forever.

Edited June 24, 2020 by _The Doctor__

[+bob1200xl]

My first 810 was brand-new out of the box and worked just fine... until it read a bad sector. This knocked the head out of alignment when it drove against the crash stop. I can only suggest that some of the Tandon mechs were not properly torqued from the factory. I just aligned the head and tightened the screw on the coupling. Worked perfect, forever.

 

Bob

 

[cwilbar]

Well, the coupling I believe was the source of the problem.  I was able to hold the shaft of the stepper, and w/o great difficulty move the coupler on the stepper shaft.

 

Sadly, while working on it, I screwed up big time.

 

While checking the screw that holds the band to the shaft, the shaft rotated, the screwdriver slipped off, crashed into the bottom of the head carriage, popping the rails out of the retainer on one end and breaking the band.  Now to go BHH somewhere..... sigh.

 

Thankfully I have a parts Tandon mech with what appears to be a dead head.  I've already got the parts I need to resurrect this one again.  I'm just probably going to have to do a little extra fiddling to get the alignment roughly dialed in, and then I need to figure out how to align with the alignment diskette without blowing the head.  Wonder if I should be using 10X probes.... maybe just using the coax to j hook cheapo leads exposes the heads to too much load ???

 

I seriously hate mistakes like this.....  I'm only thankful I had a parts drive (although maybe the bands are available via B&C or Best.... but I'll only need one should I measure the head of my parts drive compared to a good one and find it is not blown).

 

This rapidly got more involved than planned.....

 

[Kyle22]

Good luck re-aligning it. The best way I can think of is run RPM under SpartaDOS-X. Make a batch file to run it over and over. Anything to keep the drive reading. A BASIC pgm. would work. Anything that does a constant read of a sector.

Turn the stepper motor one way [slowly] until you begin to hear read errors. Mark that spot as to the position of the stepper motor.

Turn it the other way until you get read errors, and mark that spot.

Place it in the EXACT mid-point between the two extremes.

 

That's a start, anyway. It is best if you have an original Atari formatted disk.

 

Hope this helps.

 

[cwilbar]

Well, that band is fun to change ?

 

I didn't want to mess with the stepper motor, or the head carriage subassembly (that the cam screw moves for fine alignment).  Nor did I want to mess with the track 0 stop.

 

So I did manage to transplant that band, and reattach it to the head.  It still seemed a bit off, so I loosened the coupler and rotated the opposite end of the stepper motor a bit and retightened....

 

And I booted M.U.L.E. !

 

I'm sure it is not dead on....  I'm going to check what probes I have (genuine ones vs the cheap 'Chinese?' things I was using) and get my other scope (I have 2 of the same, and the one I was trying with the 810 has noisy noisy controls).  And try to get an idea with my alignment disk how close I am.

 

So, went from bad to worse, and back to pretty good.

 

Only thing else I discovered is that on the left rail there are two tabs below the rail and one wider one above, now there is only one and one :-(.  That was another casualty of the accident I didn't discover right away.

 

So far so good.... now I have to prefect doing fine alignment.

 

Don't like the prices but I was wondering how some of the Fluke Scopemeters would work for doing alignments.  Nice and portable, and doesn't take up much space. 

 

-- Curt

 

[_The Doctor__]

the sample time isn't the best on most of the scopemeters for the money involved, I still use an old scope for these sorts of things... otherwise I'd go for some of the very nice fast (but awfully expensive) digital scopes with very high sampling rates. even for our old Atari speed is king... sometimes due to how the timing is, slower sampling meters and scopes don't quite catch what you want to see.

[cwilbar]

yeah, I do like my analog scope.... though I'm thinking of adding a Rigol or similar to the collection sometime for the advantages a digital scope has for some things.

 

I'll pull my scope over from my 'arcade laboratory' and see if I can find two genuine Tektronix 10X probes and try that route.

 

I should get some newer probes for my old scope (465B or 475B, can't recall now) as the few I have are old stiff and crusty ?

 

 

[+bob1200xl]

The non-OEM probes seem to work fine on my 2465A. The heads should have a fairly low impedance so I wouldn't worry about x1 probes. Most inputs are 1 meg, unless you select 50 ohms. Regardless, you won't likely hurt the heads.

 

Bob

 

 

[cwilbar]

@bob1200xl These leads are just coax terminated in two j hook leads.  Something I got off eBay IIRC.

 

I tried to align one 810 with Tandon mech several weeks ago, and had the scope hooked up, and at a point I stopped getting signal.  It never loaded a single sector after that.  I had been playing around with scope settings/etc.  I still don't know what happened, but I don't want it to happen again :-).

 

If I have 2 Tektronix probes, I'll start with those, and use the scope that doesn't need a serious amount of switch contact cleaning (I have two of the same scope).