Issue with Track 0 sensor on 1050 Tandon

[Retrobecanes]

I purchased a replacement for the track 0 sensor for the Tandon in my 1050.  I bought it from DigiKey based on a recommendation that I found in the forum (OPB933W51Z, DigiKey part number 365-1782-ND).  It looks identical to the original at the exception that it has  (5) wires instead of 4.  

Here are the wires on the sensor that I purchased.  Is there a way for me to use this on the Tandon or do I need to buy a new part?

I have also attached the datasheet of the part I bought to this post.

 

 

OPB930-940-L-W.pdf

[TGB1718]

Pins 5 and 2 can be connected together (both need to be ground) so still only 4 connections.

 

https://uk.farnell.com/tt-electronics-optek-technology/opb840w51z/opto-switch-slotted/dp/1497932?

 When I replaced my Track Zero Sensor, I used one of these, the only difference is pins 1 and 2 wires

have to be swapped If I remember correctly, it's in the writeup in the below link.

 

Here's a link to the thread I started for it.

 

Track Zero Sensor.pdf

Edited May 5, 2023 by TGB1718

[Retrobecanes]

Thank you.  I think that I have wired it correctly now since the head parks itself when the drive is turned on.  The head does bang against the stop a few times but I assume that some of that will be mitigated once I adjust the position of the sensor.  What is weird however is that after the head parks itself, the drive stops, then a few seconds later spins a few times and then stops again, and so on  and so forth.  Also the computer doesn't seem to recognize that the drive is connected.  What could this be?

[spookt]

A pulsing motor like you describe usually means that the drive has failed it’s power on self test. This is often a failure in the logic on the board. I can’t remember if a failure to detect track 0 would cause it to fail POST but it may be related. Do you have another drive mechanism to test the board with?

Edited May 6, 2023 by spookt

[TGB1718]

Try moving the track 0 sensor out a bit so it doesn't go all the way back until you get the drive going,

alignment doesn't matter too much until you do an alignment test. (unless you don't have another way to load

the test software disk) in which case, you will have to make little adjustments until a disk will load.

 

What I did was mark the drive with the position of the old sensor and when I fitted the new one, set it to

that position, luckily mine loaded the disk and I could do an alignment.

 

[xrbrevin]

ive found that the pinout can vary despite the unit looking identical to the original

[Retrobecanes]

So I have a working 1050 drive (let's call it system X) and this non-working 1050 (let's call it system Y).  I have been testing with both and I conclude that the logic board is not the issue:

- assembly X with logic board X: as this is the working system, all works well and it boots from the disk

- assembly Y with logic board Y: this is the original non working 1050, parks head, spins intermittently, no boot.

- assembly X with logic board Y: behaves normally (parks head, no intermittent spin) but provides a boot error

- assembly Y with logic board X: parks head but produces intermittent spin, doesn't boot from disk, same behavior as system Y

Based on this I believe that the logic board Y is ok but that the drive assembly Y is the problem.  I checked all the connections and they seem ok.

What else should I try?

 

[Retrobecanes]

I have done some additional since the previous post (see above):

I migrated the FDC, ROM, MPU, RAM and RIOT from system X to system Y.  As a result system Y would not work at all, I would get power to the board but absolutely no start up activity on the drive.  It turns out that the ROM (U10) from system X does not work at all in system Y.  The FDC, ROM, RAM and RIOT from system X worked fine in system Y but did not solve the issue.

Based on what I am reading in this forum and other sources, this intermittent spin of the drive is a sign that the drive is failing a self-test.  But I have no idea where to get from there.

I have attached the ROM from system X and from system Y to this post.

 

[_The Doctor__]

@Retrobecanes

clean the living daylights out of Y assembly head and align the Y assembly... you may end up with a good mech.

Y PCB prolly needs a Rom or FDC fixed. Pay close attention to Connector orientation, or you may damage sensors.

That eprom has a pin issue, you may have found your fault

Edited May 6, 2023 by _The Doctor__

[Retrobecanes]

Thanks, I indeed didn't put the ROM back in properly after my testing, it is now corrected, it wasn't the source of the issue.

How would you recommend that I align the Y assembly?  Is there a procedure for doing so?

I also notice the the JP1 to JP4 settings next to the ROM are different between my two boards.  I am not finding any reference about how to set JP1 to JP4.

I move the FDC from system X to system Y but it didn't make a difference.

Could a ROM issue explain the problem that I am experiencing?  What version of the ROM should I replaced it with?

 

[_The Doctor__]

I say do a google site search or elastasearch on AA and you will find a multitude of threads going from using a commercial disk or two and dialing alignment in for the track sensors and then if still needed track alignment. They run the gambit from using next to nothing all the way to oscilloscope and special alignment disks made for the task.

[Retrobecanes]

Also to continue with the troubleshooting, I programmed a new ROM K EPROM and replaced the original in system Y, no change.  The ROM is not the culprit.

 

[_The Doctor__]

depends on the mech, WST mechs need a different ROM than Tandon mechs.

Also the eproms may require proper jumper selection on the PCB since it is not a mask ROM.. You never know what someone else has done or what you may have done.

Edited May 6, 2023 by _The Doctor__

[Retrobecanes]

The original ROM in system Y was a K revision, as per the label, so I reprogrammed a K revision.  No effect.  

[TGB1718]

Have you tried moving the track 0 sensor out a bit, it may be that the head is moving all the way back against the end stops

and never getting to the sensor, so it's registering that as a fault, then tries again.

I would put it in a position approx. 1/2 way along the slot in the sensor plastic as a basic start position and go from there.

[_The Doctor__]

double check 5 through 7 the other look correct

Edited May 6, 2023 by _The Doctor__

[_The Doctor__]

I see someone already soldered and moved the jumpers around... so something is up, other hands have been in there.

They look correct, just make sure they are not shorted etc

 

[_The Doctor__]

@Retrobecanes

oh I just remembered we had a batch of bad U8 sockets at on point, so you might want to check or replace that. Since this drive has had mods installed and removed as evidenced by the solder work and label fun. I'd make certain the sockets are good, you won't believe the number of times the sockets don't make perfect contact especially when modding has taken place. U8's sockets were a problem from the factory though.

Edited May 6, 2023 by _The Doctor__

[Retrobecanes]

@_The Doctor__ I replaced the U8 socket but no effect.  I am now thinking that the problem may originate from the track 0 sensor.  The original one was toast so I installed a new one but it was a five-wire model and i don't think that it is fully compatible with the drive, although it does park the head at start up.  I have an original track 0 sensor on the way, I will see if it passes the self test.

[TGB1718]

5 hours ago, Retrobecanes said:

i don't think that it is fully compatible with the drive

Just had a look at the datasheet and although it's not for the model you have, I'm assuming it's

probably compatible as everyone who sells these uses the same datasheet.

 

Saying that, you have the OPB933W51Z 

if that's the case, then aside from the 5 wires ( which is easily solved) the datasheets says

it's an Open Collector Inverting Output model.

 

Then that would mean it's going low when you want a high and vice versa, from the datasheet, the 3rd number

is the type which in this case is a 3 (933)

 

Electrical Specification Variations:

0 = Buffered Totem-Pole Output

1 = Buffered Open-Collector Output

2 = Inverted Totem-Pole Output

3 = Inverted Open-Collector Output

 

I think you need a passive one like the original, these are active and have buffer circuits in them.

Edited May 7, 2023 by TGB1718

[+kheller2]

First, you can't swap the EPROM and ROM between the two boards. Those jumpers next to the chip are setup to provide proper signals and power to either a ROM or an EPROM.  They have to be setup properly for whatever chip you use.

 

assembly X with logic board X: as this is the working system, all works well and it boots from the disk

  === Good

 

assembly Y with logic board Y: this is the original non working 1050, parks head, spins intermittently, no boot.

  === POST Error, failing startup checks

 

assembly X with logic board Y: behaves normally (parks head, no intermittent spin) but provides a boot error

=== Interesting, that a Good mech on Y causes a boot error.   Does the drive spin when the computer is powered on or is that constant boot error message with no spin of the disk?  The distinction is important.

 

assembly Y with logic board X: parks head but produces intermittent spin, doesn't boot from disk, same behavior as system Y

== This doesn't make sense unless Y Mech has failed components, like track 0 not aligned etc.. its failing POST.

 

It smells like Y has two issues, one on the board and one on the Mech.  Perhaps first step would be to align Y Mech Track 0 on working X board.  Assuming you plugged the connectors all in correct (remember the Tandon mech, one of the plugs is backwards from the rest).  Also make sure you have the head plugged in the right direction (red wire towards front).

 

 

FSM:  https://archive.org/details/1050FSMRebuiltAKH

Jumper settings are on PDF page 72.

 

[Havok69]

Saw a post that lists another part number. Mouser will have it back in stock in August, part # opb842w51z. Or get it from Best; part # CB101148, $16.

[Retrobecanes]

@kheller2 and @_The Doctor__, here is an update of the current situation, I am hoping that we can pick up where we left off:

1. I have replaced the track 0 sensor in assembly Y!  It took a while to get it, hence the delay in this thread.

2. Now if I combine the fixed assembly Y with board X, the system works perfectly well and passes all tests from the 1050 diagnostic test disquette.

3. However if I place assembly X or assembly Y with board Y, the system fails with constant "boot errors" (and drive spinning).

4. I have also noticed that if I place the drive cable in the connector closest to the drive id switch, system Y doesn't react at all when the computer is started.  System Y will only react and attempt to boot if the drive cable is placed in the connector furthest from the drive ID switch.

5. I swapped all the socketed chips between system X and Y and this had no effect (at the exception of the ROM chip which cannot be swapped as the jumper settings are different - see discussion above).

I am now convinced that the remaining problem is on board Y.

What would you recommend that I do to test board Y?

[_The Doctor__]

Time to check the traces and connections on the SIO sockets on the back of the board, reflow and if needed run jumpers pin for pin (also make sure no pins are broken/bent within SIO sockets)... check 3086 chip if still an issue after that.

Edited May 16, 2023 by _The Doctor__

[+kheller2]

Show a picture of the solder side of the SIO (Drive cable connector) ports on the Y board.  Those lines should be wired through between both jacks, and if one is not working then you probably have broken solder joints.

Let's talk about your #3.   Both floppy mechs are TANDON, correct?  What happens when you do this:  Power on Y, insert a bootable disk, wait for the drive to stop spinning.  Now turn on the computer.  Do you get boot errors AND the drive is spinning?

 

DOC beat me to it.  But yeah it could be your 3086 chip in the back of the drive malfunctioning.