Cumana CS400 switchable drive - noises and disk errors

for bbc micro/electron hardware, peripherals & programming issues (NOT emulators!)
iamaran
Posts: 288
Joined: Tue Mar 14, 2006 8:08 pm
Contact:

Cumana CS400 switchable drive - noises and disk errors

Post by iamaran » Wed Feb 27, 2019 11:20 pm

What a beast this is. Apart from a schoolmate in the early 80s that had 8-inch floppy drives rigged up to his BBC (down to his dad I assume), I don't think I've ever seen such a monstrosity. Even the locking button is massive (thought they all had levers). I can't remember what brand mine was back in the day (TEAC I think) but it was at least a third as long!

Anyway, it doesn't sound right in operation and nothing loads - disk error. Though I can get a *CAT of the few disks I tried. Not trying any more until I know what I'm doing. I don't want the drive to cause disk trouble or the disks to create drive problems.

There are only a few disks in my donated collection anyway, and from the directory listings so far I'm fairly confident they are all the usual stuff.

Attached is a recording of its operating noise when I do a *CAT.

Thanks, as always, to those way more knowledgeable than me.
Attachments
cumana_wav.zip
(430.26 KiB) Downloaded 6 times
cumana.mp3
(53.47 KiB) Downloaded 14 times
cumana_CS400.jpg
Last edited by iamaran on Thu Feb 28, 2019 11:42 pm, edited 4 times in total.

User avatar
AndyF
Posts: 1270
Joined: Sat Feb 23, 2008 10:16 pm
Location: Derby
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by AndyF » Thu Feb 28, 2019 7:32 am

Examine the visible media area of the disc that has been in the drive. Does it appear to have any damage ?

I do not have the specifics (others will be able to advise) and I can't listen to the audio at the moment.

I'm merely thinking of cases where either dust/dirt on the floppy disc itself gets stuck in the tiny head gap or there's some damage to the head anyway which then 'eats' the floppy disc.

Talking of eight inch drives, I have seen them and they were a huge affair although I've not seen one hooked up to a 'B , given the 8271 controller supported them (afaik it was designed for these perhaps) if the machine had arrived in say 79 or 80 then it may of been possible more of these would of been around and in use perhaps. Could be worse, could be the 3" discs***, to be fair these are very sturdy but a bit "lol price" to get now.

*** Yes 3" not 3.5" :) 3 as in the same as some of the Amstrad machines and later Speccy.
Last edited by AndyF on Thu Feb 28, 2019 7:34 am, edited 2 times in total.
Andy

* NEW * The Jetset Willy and Manic Miner community :)

Adventure games ported across to the BBC (in progress) as soon as I can find some time!

Kazzie
Posts: 778
Joined: Sun Oct 15, 2017 7:10 pm
Location: North Wales
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by Kazzie » Thu Feb 28, 2019 9:19 am

When you get a disk error, what error code does it report? (It may be listed in this document.)

The noise sounds to me like it may be the drive's stepper motor doing a merry dance. There should be a sensor that detects when the head is stepped all the way out (to track 0) and informs the computer of that fact. If the head isn't moving, or the sensor isn't working, the Beeb's floppy drive controller will give an error 14 after trying and failing to get to track 0.

(Other moving components in a floppy drive include the solenoid that loads (lifts/lowers) the heads onto the disk surface, and the motor that spins the disk. The motor shouldn't be able to make that kind of noise, and I can't think of a reason why the solenoid would be switched on and off so rapidly.)

If you can open the drive up and run it without the case, you should be able to see a bit more of what's going on. As well as the power supply at the back, you should see something vaguely similar to this:
floppy drive heads circled.png
The part I've circled is the floppy drive heads, which step in and out (left and right on the image) along a set of runners. With the power off, try pushing the heads along the runners, toward the right / front of the drive. They should slide along with a gentle effort. If they're stuck fast, don't force them.

After sliding the heads "in", turn on the drive and attempt to *CAT a disk. You should see the heads stepping back out toward the edge of the disk.

If the drive fails to do these, either the head assembly is seized on its runners or the stepper motor is playing up. If the drive passes these tests, but continues making funny noises and throws an error, I'd think about checking the output of the track 0 sensor. (Do you have a multimeter available?)
BBC Model B 32k issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
Acorn System 1 home-made replica

User avatar
MartinB
Posts: 5229
Joined: Mon Mar 31, 2008 9:04 pm
Location: Obscurity
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by MartinB » Thu Feb 28, 2019 9:32 am

As often as not, these symptoms turn out to be an example of the recurrent theme of the head shield(s) falling off due to the mounting foam having rotted. There are lots of citations on the forum, clicky here for a typical example with piccies. Your drive may not be a variety that has such shields fitted but certainly worth a check.

iamaran
Posts: 288
Joined: Tue Mar 14, 2006 8:08 pm
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by iamaran » Thu Feb 28, 2019 11:24 pm

Thanks for the replies.

1. The disks are variable in quality, from a quick visual inspection, but nothing you wouldn't have at least tried in 1983!

2. It doesn't give an error code as such, just "Disk fault" when trying to execute a file. *CAT returns a list of files.

3. This drive is different! Link to video what happens when I *CAT
https://www.dropbox.com/s/rul9kce4o2zqw ... a.mp4?dl=0
Last edited by iamaran on Fri Mar 01, 2019 12:02 am, edited 1 time in total.

Kazzie
Posts: 778
Joined: Sun Oct 15, 2017 7:10 pm
Location: North Wales
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by Kazzie » Fri Mar 01, 2019 9:13 am

Well that's interesting. The part moving vertically is the head-loading mechanism, the very part I said would have no reason to be dancing about like that.

Under that moving metal arm, there'll be a solenoid that raises or lowers the arm. That arm it turn lifts the heads off the floppy disk, or lets it down onto the disk surface.

I can see the disk is spinning, so that motor is doing something. But the head assembly is hidden under that shiny piece of metal in the middle. Can you see if the head itself is actually moving down and up with the sprung arm? Is the shiny metal plate obviously removable (to let us see if the stepper motor can move the heads back and forth)?

The Shugart floppy drive standard (for 5 1/4" drives) doesn't have a dedicated wire to control the head load solenoid. The drive does it instead, by some combination of Drive Select, Motor, and related signals. There are usually some jumpers to configure this.

Could you take a nice, clear photo of the floppy drive from above, so we can identify what model of floppy drive Cumana used? It could be by Teac, Mitsubishi, Panasonic, or any number of other companies. Also, please look for any manufacturer model numbers on the circuit board or on a sticker fixed to the drive itself (not the outer box). If we can work out what inputs the drive is using to control its head load solenoid, we may be able to find the "faulty" signal it's responding to, or reconfigure the drive so it uses a different method to decide when to load the heads.
BBC Model B 32k issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
Acorn System 1 home-made replica

iamaran
Posts: 288
Joined: Tue Mar 14, 2006 8:08 pm
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by iamaran » Fri Mar 01, 2019 6:31 pm

Hello again.

Well, the shiny plate isn't obviously removable and I can't tell if the head is moving.

Photos are attached - one from above, and one of the model number. It's an Epson SD-540. I removed the sticker to see whether it was covering over a different model number, but it is blank underneath.
Attachments
IMG_6705.JPG
IMG_6702.JPG
Last edited by iamaran on Fri Mar 01, 2019 6:43 pm, edited 1 time in total.

Kazzie
Posts: 778
Joined: Sun Oct 15, 2017 7:10 pm
Location: North Wales
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by Kazzie » Sat Mar 02, 2019 8:27 am

Thanks for the photos. Now I can see a bit more of what we're working with. Don't worry too much about not being able to see if the heads are stepping: while the solenoid is still raising and lowering them, we won't be able to read a disk anyway.

It seems that documentation for the Epson SD540 is rather elusive. The best information I can find on the internet is of another person with the same problem as you. Not very helpful, but it does suggest that your problem isn't unique.

In the photo you posted, there are two banks of jumper pins for configuring the drive. (There are also several solder-in wire links labelled jp1-20 on the board, but we'll ignore those, as they should be set at the factory.) We may be able to find another configuration that bypasses the fault to get your drive working again, or otherwise be able to work out what's causing the fault in the first place.
EpsonSD-540-Jumpers.png
Cropped and rotated from the image posted earlier
EpsonSD-540-Jumpers.png (200.49 KiB) Viewed 1059 times
The bottom bank, SS1, is used to control when the drive should be selected, as the Shugart standard supported up to four drives. (Hence options D0,1,2,3, and also MX representing "always selected".) This is configured as Drive 0, which is correct for a single-drive setup. These can be left alone.

The upper bank, SS2, also has five jumpers, which appear to be labelled IU ID OP MS HS, with jumpers currently on ID and HS. Without and documentation, it's a bit of a guessing game what each of these do.

Based on how other manufacturers label their jumpers, I suspect that the MS and HS jumpers control whether the motor and head solenoid are activated by the Drive Select lines or not.

At a guess, the IU and ID might be used to configure which control lines turn the front LED on (In Use vs Drive Select). Panasonic used the OP label on a jumper for "High Density" (1.2Mb) drives, but that doesn't make sense for a "Quad Density" (800kb) drive like this one.

:arrow: Iamaran, could you do a few tests to check what the HS and MS jumpers do?
  • Remove the HS jumper then power up the drive, and try to *CAT a floppy disk. Note what the head load solenoid is doing (still bouncing up and down, not moving, or another pattern). Check if the disk is spinning around when you *CAT.
  • Place the jumper on the MS pins instead, and *CAT a floppy disk. Is there any change in behaviour from the motor or the solenoid?
Just noting this information for later; the ICs on the circuit board include:
  • An MB8843 4-bit microcontroller, with built-in RAM/ROM
  • An MC3470 floppy disk read amplifier
  • A TD65203 for driving motors, solenoids, etc
  • A selection of 7400-series logic chips
Everything apart from the MB8843 should be replaceable is it turns out to be faulty
BBC Model B 32k issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
Acorn System 1 home-made replica

iamaran
Posts: 288
Joined: Tue Mar 14, 2006 8:08 pm
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by iamaran » Sat Mar 02, 2019 10:09 am

Thanks for the reply and information, Kazzie.

With the HS jumper removed the solenoid does not move. The disk spins and the beeb reports "disk error".

When the jumper on the MS pins it behaves the same as when it's on HS, i.e. bouncing up and down with the same pattern. The disk spins, "disk error".

Interestingly, I now get "disk error" when the jumper is back on HS whereas I was getting a directory listing yesterday. I'm probably wrecking the disk but that doesn't matter.

Kazzie
Posts: 778
Joined: Sun Oct 15, 2017 7:10 pm
Location: North Wales
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by Kazzie » Sat Mar 02, 2019 2:27 pm

iamaran wrote:
Sat Mar 02, 2019 10:09 am
Thanks for the reply and information, Kazzie.

With the HS jumper removed the solenoid does not move. The disk spins and the beeb reports "disk error".

When the jumper on the MS pins it behaves the same as when it's on HS, i.e. bouncing up and down with the same pattern. The disk spins, "disk error".

Interestingly, I now get "disk error" when the jumper is back on HS whereas I was getting a directory listing yesterday. I'm probably wrecking the disk but that doesn't matter.
Thanks for the measurements. It looks like both the HS and MS jumper locations control the solenoid (though not in a way that we'd like), and that the motor is configured elsewhere.

The next thing I'd like you to do is to take measurements on the pins of the HS and MS jumpers, to see what "instructions" are being given to the solenoid when the jumper is in position. For this step, you'll need either an oscilloscope or a multimeter. Do you have access to either of these?

For illustration, here's how I expect the two jumpers are wired up:

Code: Select all

         _________
        |         |
>-------|---H S---|---\
        |         |   |------> toward the solenoid
>-------|---M S---|---/
        ¦         ¦
Though without a circuit diagram, we're effectively reverse-engineering the layout. :wink:

Having removed the jumper, two of the four pins will still be connected together electrically: either the top pair of the bottom pair. These will be the "outputs" from the jumper bank that connect to the solenoid. You can do a continuity test with your multimeter to check which two they are. This test can safely be done with the power off. We'll be interested in the readings we get on the other two pins though, the inputs.

Turn the drive on, then use either an oscilloscope or a multimeter in voltage mode to measure the voltage of each of the "input" pins before and during a *CAT operation. (For a multimeter, place the red probe on the pin, and the black on the grounded screw in the corner of the board, near the floppy data cable. It might be useful to have an extra hand to press buttons on the keyboard!) You should find a constant value before running *CAT, and a different value while the drive is trying to read the disk.

We don't know what kind of signal is actually being sent to the solenoid, so if you're using a multimeter take both a DC and AC voltage measurement, if you can.

Depending on the kid of signal you see, we should be able to work out if the fault is before or after the jumper bank.
BBC Model B 32k issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
Acorn System 1 home-made replica

iamaran
Posts: 288
Joined: Tue Mar 14, 2006 8:08 pm
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by iamaran » Sat Mar 02, 2019 9:54 pm

Hi. I own neither, but can borrow a multimeter on Monday. Thanks for your patience.

Kazzie
Posts: 778
Joined: Sun Oct 15, 2017 7:10 pm
Location: North Wales
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by Kazzie » Sun Mar 03, 2019 6:05 am

iamaran wrote:
Sat Mar 02, 2019 9:54 pm
Hi. I own neither, but can borrow a multimeter on Monday. Thanks for your patience.
Great. I look forward to hearing back from you.
BBC Model B 32k issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
Acorn System 1 home-made replica

iamaran
Posts: 288
Joined: Tue Mar 14, 2006 8:08 pm
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by iamaran » Tue Mar 05, 2019 10:00 pm

Thanks for your time and patience.
This multimeter has crocodile clips and even though they're quite small it was impossible to touch the pins in isolation. Once I'd put pins in the clips it was slightly easier but still not a doddle!

However, here are the (hopefully accurate and meaningful) results:
The bottom pair gave a reading (this multimeter doesn't beep) so the top two must be the inputs.
In AC voltage mode with the dial set to 20V then 2, then 200m, nothing gave a reading regardless of whether trying to read a disk.
On DC with the dial set to 20V, the top left (MS) gave 4.85 before the *CAT and 0.2 during. The top right (HS) gave 4.5 before the *CAT and 0.22 during.

Kazzie
Posts: 778
Joined: Sun Oct 15, 2017 7:10 pm
Location: North Wales
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by Kazzie » Wed Mar 06, 2019 8:02 am

iamaran wrote:
Tue Mar 05, 2019 10:00 pm
Thanks for your time and patience.
This multimeter has crocodile clips and even though they're quite small it was impossible to touch the pins in isolation. Once I'd put pins in the clips it was slightly easier but still not a doddle!

However, here are the (hopefully accurate and meaningful) results:
The bottom pair gave a reading (this multimeter doesn't beep) so the top two must be the inputs.
In AC voltage mode with the dial set to 20V then 2, then 200m, nothing gave a reading regardless of whether trying to read a disk.
On DC with the dial set to 20V, the top left (MS) gave 4.85 before the *CAT and 0.2 during. The top right (HS) gave 4.5 before the *CAT and 0.22 during.
Thanks for the results.

The lack of readings on AC mode isn't a concern, as there should be a clear DC voltage when the drive is operating correctly. (An AC reading would have suggested that the solenoid was dancing on and off because the head-load signal was alternating on and off.)

The head-load signal, like most control signals in a floppy drive, are active-low signals. That means that the voltage is high (~5V) when they're off, and low (~0V) when they're on. 4.5~4.85V and 0.2~0.22V are well within the range of acceptable values.

So it seems that the head-load signal is correct at the jumpers, but either it is getting "corrupted" between the jumpers and the solenoid, or else there is a fault with the solenoid itself.

A typical logic IC isn't able to drive a solenoid itself, it doesn't provide enough current to do so. One would usually use a transistor to amplify the current, and use that amplified current to operate the solenoid. The TD62503P chip has seven transistors on it, for this kind of purpose. When the input to these NPN transistors go low, their output goes high, so the jumpers could be connected directly to one of these transistors. Alternatively, the signal may go through some of the 74xx logic chips first before going to the TD62503P.

:arrow: With the power off, could you do a continuity test from one of the "bottom" (output) jumper pins to each of the pins of the TD62503P chip? If there is a connection, I'd expect it to be to one of pins 1-7:
TD62503P.png
TD62503P.png (11.96 KiB) Viewed 908 times
If you do find a continuity connection, could you power up the drive, connect the meter to the corresponding output pin, and repeat the same AC and DC tests as earlier? That'll tell us whether the transistor chip is operating correctly.

Thanks.
BBC Model B 32k issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
Acorn System 1 home-made replica

iamaran
Posts: 288
Joined: Tue Mar 14, 2006 8:08 pm
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by iamaran » Thu Mar 07, 2019 8:29 am

Thanks for the reply. I seem to be getting a reading on every pin of the TD63503P from the bottom output pin of HS with the multimeter in resistance mode. If this makes absolutely no sense, then the explanation is my poor eyesight and unsteady hand and I'll get my son to look at it with me tonight after school and work.

Kazzie
Posts: 778
Joined: Sun Oct 15, 2017 7:10 pm
Location: North Wales
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by Kazzie » Thu Mar 07, 2019 11:21 am

iamaran wrote:
Thu Mar 07, 2019 8:29 am
Thanks for the reply. I seem to be getting a reading on every pin of the TD63503P from the bottom output pin of HS with the multimeter in resistance mode. If this makes absolutely no sense, then the explanation is my poor eyesight and unsteady hand and I'll get my son to look at it with me tonight after school and work.
Is that a zero (or near-zero) resistance between HS and every pin on the TD62503P? That would be very strange.

Readings of more than an ohm or so indicate a resistance greater than you'd get with a simple stretch of wire (i.e. some other component in-between, or a break in the wire). For a continuity test, we're only talking about really low values (e.g. 0.5 ohms), which is when a more fancy multimeter (with a built-in continuity function) would be beeping at you.

Lets see what a fresh pair of eyes/hands brings to the situation. :)
BBC Model B 32k issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
Acorn System 1 home-made replica

iamaran
Posts: 288
Joined: Tue Mar 14, 2006 8:08 pm
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by iamaran » Thu Mar 07, 2019 11:06 pm

Hi again. So... I am a #-o

This multimeter doesn't have a dedicated continuity mode so I've been using the resistance mode. This has a variable selector dial and I've had no idea what to set it to. The results vary - at the higher levels everything seemed to have a connection and at the lower, none did. Now I'm fairly certain (ha ha), having tested in each pin in each dial position, there is no connection to any pin on the TD62503P - because at every position of 20K and below on the dial there is no resistance reading to any pin on the TD62503P.

Thank you for your time and patience with this beginner!

Kazzie
Posts: 778
Joined: Sun Oct 15, 2017 7:10 pm
Location: North Wales
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by Kazzie » Fri Mar 08, 2019 11:47 am

Okay, having no connection to any of the TD62503's pins makes a lot more sense than a connection to all of them. (I'll just check you're happy you were testing against the correct "side" of the MS/HS jumpers: if you were on the "far" side of the removed jumper, then the multimeter would give an open circuit measurement all the time1)

If the MS/HS jumpers don't connect directly to one of the transistors in the TD62503, they may well lead to one of the 74** general purpose logic chips on the board first, or perhaps the big MB8843 microcontroller. (They won't be connected to the MC3470 near the front of the drive, so don't worry about that one. The blue chip is a set of terminating resistors for the floppy data cable; again it makes no sense to use those to control the solenoid.)

:arrow: Could you repeat your continuity test from the MS/HS jumpers to each of the pins on the 74** chips, and the MB8843? For any chips you get a continuity reading for, please note which chip and pin number you found continuity on.

(This is an example of where a proper continuity function is handy, as you could sweep the probe along the pins, and just listen out for a beep. Without this function you'll have to poke each pin and look at the multimeter in turn. It's slower, but it will still give correct results.)

We'll be able to look up the details of those chips and work out what they're doing with the signal, and where to measure next as we search for the broken bit on the circuit board. It may take a bit of back-and-forth to get there, but there's nothing inherently complicated in the measurements that need to be made.

If you're happy to remove the floppy drive from the outer casing, it could be useful to look for the (two?) wires coming from the solenoid to the circuit board. That could help us work backwards as well as forwards in working out how the jumpers are connected to the solenoid.
BBC Model B 32k issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
Acorn System 1 home-made replica

User avatar
1024MAK
Posts: 9035
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by 1024MAK » Fri Mar 08, 2019 12:11 pm

Can you post up the model name and number of the meter you are using, or a photo of it please.

The resistance range(s) (Ohms ranges) on multimeters are designed to measure the resistance of resistors, or the resistance of circuits.

A manual meter has multiple ranges in order to cope with the wide range of resistances that a user is likely to want to measure. This is done because the electronics in the meter and the display can only work over a limited range.

The following assumes a 1.999 count (or 3½ digit) digital meter is being used. The same principles also apply to other meters, just the numbers are different.

When measuring resistance, if the value is unknown, you start on a high range (say 200K) and see what the result is. If the displayed value is the overrange indication (see later) switch to the next range up (say 2000K or 2M). If the displayed value is less than 19K, switch down to the next range (say 20K). Again, if the displayed value is less than 1.9K (or 1900 ohms) switch down to the next range.

There are various indications for overrange. They all mean the same thing. The meter cannot display the value as it is either too high or too low for the currently selected range. Or it may be beyond the meters capability.

Some examples of overrange are the left most digit showing “1” with all the rest of the digits blank. Or the left most digit showing a flashing “1”. Some show “OL” instead.

On a resistance range, an indication of zero (0.0) may not mean zero unless you are on the lowest range. This is due to the limitations of the meter circuitry.

Analogue meters with a moving coil meter, that is a needle pointer work a bit differently.

One other important point about using the resistance ranges. To measure the resistance, the meter has to apply a voltage and a current to the circuit or component under test. If the voltage applied by the meter is 0.3V (some types of diode) or 0.6V (silicon diodes and transistors) or above, some semiconductors may start to conduct as the voltage causes them to start to switch on. This can give some misleading results.

Other passive components can also confuse matters. For example resistors or inductors in parallel will produce lower than expected values. Capacitors in circuit will cause the displayed value to keep changing as the capacitor charges up.

Mark

User avatar
1024MAK
Posts: 9035
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by 1024MAK » Fri Mar 08, 2019 1:04 pm

Here are some photos showing what I talked about earlier:
6575C5FF-C1E5-4118-8055-A05B5DBB37DF.jpeg
DT830B on 200 Ohm range with leads open showing overrange
.
5143699F-30FC-4FF2-8E6E-1A41832A3D61.jpeg
DT830B on 200 Ohm range with leads shorted
.
1268069B-7658-4B0F-834F-732E7338FD2C.jpeg
DT830B on 200K range with leads shorted
.
F9133C5B-5266-4FD2-AF0C-94E181426214.jpeg
DT-4000ZC with leads open showing overrange
.
E12D6EB8-AF62-4ED6-955D-7F221A77DF89.jpeg
DT-4000ZC with leads shorted
.
478642BE-A532-46EE-A341-F10307A815FC.jpeg
XL830L on 200K range with leads open showing overrange
.
49D91A78-5C26-4D61-A743-6DA13ACCDD90.jpeg
XL830L on 200 Ohm range with leads open showing overrange
.
1C35B249-3ADC-4B64-AFA8-D3104E0FEA25.jpeg
XL830L on 200 Ohm range with leads shorted
.
9A4BD5F4-692F-4417-96BE-958435983B97.jpeg
XL830L on 200K range with leads shorted
Mark

iamaran
Posts: 288
Joined: Tue Mar 14, 2006 8:08 pm
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by iamaran » Fri Mar 08, 2019 9:15 pm

Kazzie, thanks again. I'll double check everything so far and reply again when I'm 100% certain I know what I'm doing!

Hi Mark, and thanks.
The multimeter is a generic type, model G-44 but I think that's pretty meaningless. I've attached two photos.
Attachments
closed.jpg
open.jpg
Last edited by iamaran on Fri Mar 08, 2019 9:21 pm, edited 2 times in total.

iamaran
Posts: 288
Joined: Tue Mar 14, 2006 8:08 pm
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by iamaran » Mon Mar 11, 2019 10:29 pm

Hi, and sorry for the delay - I think I know what I'm doing now! I've redone everything reported so far and am fairly confident I've done it correctly. [-o<
The only recorded continuity is on pin 28 of MB8843. I did test each pin of every 74**.
The attached (not too great) photo hopefully shows the correct side of the MS/HS jumpers and the aforementioned pin of the MB8843.
Thank you :)
Attachments
P1120141.jpg
Last edited by iamaran on Mon Mar 11, 2019 10:33 pm, edited 3 times in total.

Kazzie
Posts: 778
Joined: Sun Oct 15, 2017 7:10 pm
Location: North Wales
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by Kazzie » Tue Mar 12, 2019 2:07 pm

Hello, again, welcome back. (Don't worry about the delay!)

Good work tracing all those various pins. Unfortunately, the fact that the jumpers are connected to the MB8843 is a bit unfortunate for our purposes. While all the 74xx chips are simple logic chips, with well-defined functionality, the MB8843 is a programmable microcontroller, which means that it has a custom program "baked-in" which decided what to input and output on its pins, and when. As a result, while we know where our head-load signal goes in, we can't know where that signal comes out again.

That doesn't mean that we're defeated, though. While we can't carry on tracing the signal forward, we may be able to trace the signal backward from the solenoid. There should be a pair of wires connecting the solenoid to the circuit board at some point. If you can locate these, you could use the same "continuity test" method to work out where they're connected to. (I'd expect one to be connected to ground (e.g. the mounting screws) the 5V or 12V power rail (look to the white power plug at the back of the floppy drive), and the other to be connected to one of pins 10-16 of the TD62503 transistor chip.)

Another possibility is to observe the voltage on each of pins 1-7 of the TD62503 when you try to *CAT a disk, both when the Head Select jumper is present, and absent. I'd expect to find that most readings will be the same, but that one (or more?) pins will have a different measurement when the jumper is removed. This would also help us work out what's going on on the far side of the microcontroller.
Last edited by Kazzie on Tue Mar 12, 2019 4:38 pm, edited 2 times in total.
BBC Model B 32k issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
Acorn System 1 home-made replica

User avatar
1024MAK
Posts: 9035
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by 1024MAK » Tue Mar 12, 2019 3:00 pm

As TD62503 chips are open collector transistor arrays, I would expect any load, such as a solenoid to have one wire/terminal connected to a positive power rail (+5 V or +12V) and the other wire/terminal to be connected to one (or two) of the chips output pins.

Datasheet here :wink:

Oh, and be careful because the coil windings of the solenoid are likely to have a relatively low resistance.

Mark
Last edited by 1024MAK on Tue Mar 12, 2019 3:01 pm, edited 1 time in total.

Kazzie
Posts: 778
Joined: Sun Oct 15, 2017 7:10 pm
Location: North Wales
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by Kazzie » Tue Mar 12, 2019 4:38 pm

1024MAK wrote:
Tue Mar 12, 2019 3:00 pm
As TD62503 chips are open collector transistor arrays, I would expect any load, such as a solenoid to have one wire/terminal connected to a positive power rail (+5 V or +12V) and the other wire/terminal to be connected to one (or two) of the chips output pins.

Datasheet here :wink:

Oh, and be careful because the coil windings of the solenoid are likely to have a relatively low resistance.

Mark
Good catch, thanks. Post now edited.
BBC Model B 32k issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
Acorn System 1 home-made replica

iamaran
Posts: 288
Joined: Tue Mar 14, 2006 8:08 pm
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by iamaran » Wed Mar 13, 2019 7:43 am

Thanks guys. I'll see about getting the drive out of the casing soon.

iamaran
Posts: 288
Joined: Tue Mar 14, 2006 8:08 pm
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by iamaran » Sun Mar 17, 2019 7:12 pm

Here it is removed from the casing.
Attachments
P1120151.JPG
P1120150.JPG

User avatar
MartinB
Posts: 5229
Joined: Mon Mar 31, 2008 9:04 pm
Location: Obscurity
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by MartinB » Sun Mar 17, 2019 7:46 pm

On a quick scan, I can't see the keyboard links mentioned in the thread - are they all open circuit as in the picture? (Two of them, 3 & 4 from the left, set the head-step settling time and it can be set faster than some drives can cope with....)

Keyboard DIP links.jpg

.
Last edited by MartinB on Mon Mar 18, 2019 8:51 am, edited 1 time in total.

Kazzie
Posts: 778
Joined: Sun Oct 15, 2017 7:10 pm
Location: North Wales
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by Kazzie » Mon Mar 18, 2019 2:51 pm

Thanks again for the photos. Before I go any further, I'd like to echo MartinB's advice to check the keyboard links. They're all absent by default, and that'd be fine, as that represents the slowest, most conservative timing settings for a floppy drive. (There are more details on what the links do here.)

The picture of the underside of the floppy drive surprised me: I wasn't expecting a second circuit board under there! On the plus side, I can see the connector for the solenoid: it's the one with two wires (red and white) on the bottom circuit board.

We can't see what's on the other side of this circuit board, but I'd suspect the bottom left (as you photographed it) will be involved with controlling the head stepper motor, some of the upper right will be involved with detecting the "write-protect" notch in the floppy disk, and much of the rest is probably related to the motor (which spins the disk). It seems that the two boards are connected together on the edge of the drive, via fourteen pins labelled J3 on the upper board. (Do you see something like this on the side of the drive?)

I have a hunch that very little of the circuitry on the bottom is related to the head-load solenoid (other than a flyback diode), so it may still be possible to use a continuity check to find a connection between the connector for the solenoid and the J3 connections to the upper board.

:arrow: To do this tracing back, I'd suggest:
  • disconnecting the red/white connector, replacing the HS/MS jumper (if it isn't on the board already) and testing the drive with your Beeb. If you don't hear the solenoid moving when you try to *CAT a disk, then we've got the correct connector.
  • Disconnect the floppy drive, leave the connector off, and measure the resistance between the connector pins (on the board, not the lead: we don't want to measure the solenoid). Then reverse your meter's probes and measure the resistance the other way around. There should be a significant difference in the readings, due to the diode that's in parallel with the solenoid.
  • Test for continuity from one of the connector's pins to each of the "J3" pins that lead to the upper board. Then test for continuity from the other pin to each of the J3 pins. You should hopefully find a match for both of them. Note which J3 pins they are.
  • One of the connector pins should be connected to a power source (5V or 12V, but perhaps not directly), and the other should hopefully be connected to one of pins 10-16 of the TD62503. Turn the floppy drive the right way up again, and see if you find continuity between either of the J3 pins and the TD62503 (or the 5V / 12V power cable connectors).
If we can back-track from the solenoid to the TD62503, then we can start measuring what voltages we see around there when the drive is running, and hopefully find the fault.

There's also a PM heading your way shortly.
BBC Model B 32k issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
Acorn System 1 home-made replica

iamaran
Posts: 288
Joined: Tue Mar 14, 2006 8:08 pm
Contact:

Re: Cumana CS400 switchable drive - noises and disk errors

Post by iamaran » Mon Mar 18, 2019 9:44 pm

Thanks MartinB and Kazzie. The keyboard links are all absent, and again I've learnt something new so thanks for that.

Post Reply