DuoDisk Light Always On - Help Needed

Yeah, that flowchart is much too general to be useful. 

I've seen a Disk II drive exhibit this symptom when its Q1 power transistor failed with the E and C pins internally shorted together.  I'd imagine the same fault could occur at those big transistors Q1 and Q2 inside the UniDisk.

With the power switched off, attach a continuity meter to the E and C pins of transistor Q1.  There should be no continuity.  If the meter shows continuity in either direction then transistor Q1 has failed shorted.  My best-guess is that one of these transistors is okay and the other is shorted.

DuoDisk Q1.png

Skid-Wheel Playz wrote:

Ok, so I've checked that in both directions and the same thing happens on both of them; they either read no continuity at all (in one direction) or it reads 580 (in the other direction). I'm assuming that's not shorted then, right?

Your readings indicate "BAD".

No continuity in either polarity direction would indicate "GOOD".

Skid-Wheel Playz wrote:

Ok, so I've checked that in both directions and the same thing happens on both of them; they either read no continuity at all (in one direction) or it reads 580 (in the other direction). I'm assuming that's not shorted then, right?

Right, that's not shorted like the faulty one I was describing.

It could still be faulty in some other way, but it's not a dead-short from emitter to collector like the case I saw.  Mine was <1 ohm, and current flowed in both directions.

LaserMaster wrote:

 

Your readings indicate "BAD".

No continuity in either polarity direction would indicate "GOOD"

Don't forget, those transistors were being tested in circuit, so other components will contribute to its behavior.

On the Disk II analog card, the emitter of Q1 is cross-connected to the COM pin of the ULN2003.  The COM pin has internal diodes to the output pins of the ULN2003, each through its own diode.  And one of the ULN2003's outputs is connected to the base of Q1.  So the continuity meter is probably just driving a little current through the diodes inside the ULN2003, which is thus leaking back to the base of the transistor and allowing a small current -- in one direction.

So those test results are closer to "inconclusive" than "bad".  It just didn't exhibit the specific problem we tested for.

Steve blogposted about unwanted current paths in a UniDisk and posted a Disk II schematic, assuming their circuits are probably similar.

Depending on the version of the analogue card you have this might help.

https://bitsavers.org/pdf/apple/disk/5_inch/050-5014-D_Duodisk_Analog_Bo...

You mentioned checking ULN2003, the DuoDisk analogue card I think uses ULN2068 which are different and would affect stepper motor operations.

ggb wrote: Depending on the version of the analogue card you have this might help. https://bitsavers.org/pdf/apple/disk/5_inch/050-5014-D_Duodisk_Analog_Board.tif You mentioned checking ULN2003, the DuoDisk analogue card I think uses ULN2068 which are different and would affect stepper motor operations.

Right you are.

That is an excellent schematic, and it's doubly-different from my not-so-good best guesses.  That is to say, it contains details I never guessed it had and it omits details I guessed it did have.  And some of their design choices are just annoying, like the irritatingly-backward designations for Q1 and Q2: 

• Q1 controls drive 2.
• Q2 controls drive 1.

So, because Apple switched to the ULN2068 to control the steppers, they substituted the 74LS07 at location B5 to control the 12-volt gate transistors.  So there's an easy functional test that will help determine the location of the fault -- just remove B5 and switch on the power.  If the fault is at or before B5, then the motor and LEDs will stay off.  If the fault is somewhere after B5, then the motor and LEDs will still come on.

DuoDisk B5 highlights.png

(For more detail, right click the schematic and "Open image in new tab."  Or save it and open it in another program.)

Could you clarify how the Duodisk is connected.
Is the Duodisk connected to the disk controller card via Duodisk cable when you have provided the status of the motor on and state of the LED.

With the 74LS07 installed and when the computer is first turned on Does the motor stay on and the LED on and stay like that.
Normally when booting from a floppy if it can't find the correct data from the disk, it will keep trying to read endlessly.
eg. Motor on and LED on.

Normally resetting Apple II will force the disk controller card to turn off the Motor.
Have you tried pressing <CTRL> Reset?
If you do, does the motor stop and LED go out?

Could you provide a photo of the Duodisk analogue card showing the part numbers for C6 & C7.

Can you measure the voltage on B4 Pin 4, Pin 5 and B5 Pin 8 and Q2 Collector when first turned on after the head seek.
Then repeat the measurements again after a Reset.

B4 Pin 5 and B5 Pin 8 do seem odd after reset.

Can you remove the connectors for drive 1 from J1 and J3 on the analogue card. Power on, then Reset and measure the voltage on B4 Pin 4, Pin 5 and B5 Pin 8 and Q2 Collector again.

Seeing B5 Pin 8 at 4.25v doesn't make much sense unless the Open Collector output of the 74LS07 isn't behaving.

With both drives disconnected from Duodisk Analogue card headers J1 and J3.
Power on the computer and Reset.
Could you please measure voltage on B4 Pins 4, 5, 12, 13 and B5 Pins 8, 9, 10, 11.

The photo you provided earlier showing the version of the analogue card, the surface of chip at B5 (74LS07) looks unusual like it's been lasered and the labeling of the part isn't visible compared to the majority of the other chips.
Is this just a trick of the photo?

Skid-Wheel Playz wrote:

The 12.06 on P10 of B5 kind of took me by surprise, is there anything wrong with this chip?

Nothing is wrong there.  The 7407 can tolerate up to 30 volts at its "open collector" outputs.  When the drive is not enabled, B5 simply disconnects from that output and allows resistor B13 to pull it up to 12 volts.  When the drive is enabled, B5 pulls pin 10 down to 0 volts.  (That's what "open collector" means in the 7407 IC description.)

In fact, that's an illustration of how it ought to behave.  By comparison, pin 8 of B5 is showing unexpected voltages.  I suspect one of the drives was enabled when you took that last set of measurements.  (eg: maybe it hadn't been RESET, so it was trying to read a diskette?)

I've notated that last batch of measurements in the circuit diagram as follows...

DuoDisk voltages.png

1) Check for shorts in the cable. Using the DB25 connector, check for shorts from Pin 11 to any of the three Pins 3, 10, 12

2) With the drives still disconnected from the Duodisk analogue card, can you remove chip A1 and B2.
Can you power on and after Reset check the computer is at the prompt.
Then measure voltage at B4 Pin 5.

Skid-Wheel Playz wrote:

I can assure you that the results were with drive 1 reset. However, this could be our problem. The led is always on, even when reset. Maybe an outside source is putting it partially into use when it's not supposed to be, enough to activate the led and making pin 8 weird.

Good observations -- yes indeed, it's probably just another indication of the same problem.  If there's "an outside source" that's influencing the voltages, the disconnections suggested by @ggb in comment 16 and comment 21 can really help isolate the source.

 

Skid-Wheel Playz wrote:

I really hope it's not the MC3469P that's the issue because that was the first chip I tried replacing after recieved the broken DuoDisk, and wow is it a real pain to source, but since it gave the same results, I don't think it is.

That's logical.

Let's tally some components that can probably be ruled out -- things that probably aren't causing the issue, based on your tests:

• MC3469P (A1) is probably not at fault.  Only pin 9 connects to the circuits that affect the LED & motor, and that's an input pin so it would require an improbably-catastrophic failure for the MC3469P to affect the motor circuit.  The fault persisted when the MC3469P was removed, so it was not a likely cause.
• 74LS33 (B2) is probably not at fault.  Like the previous component, it only connects to the motor circuits through input pins, so it wouldn't likely affect the LED & motor even if this IC failed completely.  The fault persisted when the 74LS33 was removed, so it was not a likely cause.
• 2SB596-0 (Q4) power transistor on motor speed control board.  This component hasn't been mentioned in any previous discussion because it's hidden away inside the drive chassis -- and, candidly, I don't know whether the DuoDisk even uses the same type of transistor as the Disk II.  Because it connects to both "MTR ON" and "+12V ENBL" wires, it's feasible that this transistor could feed a bad signal back into the "ENABLE 1" signal.  But if Q4 was at fault, then the fault would have resolved when you disconnected drive 1 (J3) for the test suggested in comment 16.  But it didn't resolve; in comment 17 and comment 19, you indicated that the LED stayed on even when J3 was unplugged -- and your voltage measurements suggested the fault was still occurring.  So the power transistor on the motor speed control board is not a likely cause.

Here's another circuit diagram, with the various drive-enable signals highlighted various colors.

DuoDisk enable signals.png

Maybe the "ENABLE 1" signal is faulty before it reaches the DuoDisk?  If the pinout of the cable is correct, then  B4 pin 5 inside the DuoDisk should connect through the cable to the disk controller's A2 pin 8.  (Pin 8 of the 74LS132)  If the cable has continuity, then perhaps A2 itself is generating a faulty signal from the controller card?

As @S.Elliott suggested, next step is to test continuity from disk controller to analogue card.
With the computer off, measure resistance from Apple 5.25 disk controller or IO controller from chip A2 Pin 8 to Duodisk analogue Card chip B4 Pin 5.
If continuity then with some luck if you have a controller with socketed chips and you have a spare 74LS132 or another disk controller with a socketed 74LS132, you could try swapping the 74LS132 and then test again.
Power on, then Reset and measure the voltage on B4 Pin 5 of the Duodisk analogue card and check the LED behavior.

Skid-Wheel Playz wrote:

Alright so there is continuity between A2 Pin 8 and B4 Pin 5. Unfortunately, I do not have a spare chip nor a spare disk controller. I will try my best to get my hands on one though. It is in a socket though which is a plus. Anything I could do in the meanwhile to make sure it really is a faulty signal from the 74LS132?

Thanks again for your help, both of you. I would never have mana

You can replace the 74LS132 with 74(LS)00.

Skid-Wheel Playz wrote:

Alright so there is continuity between A2 Pin 8 and B4 Pin 5. Unfortunately, I do not have a spare chip nor a spare disk controller. I will try my best to get my hands on one though. It is in a socket though which is a plus. Anything I could do in the meanwhile to make sure it really is a faulty signal from the 74LS132?

Even though the errant "enable 1" signal is coming out of the 74LS132, it doesn't necessarily mean that the 74LS132 itself is to blame.  It might be responding to a bad signal from a faulty 556 at D2, or 74LS259 at C2, or 74LS05 at B2.  (FYI: C2 is silkscreened "9334", Fairchild's old number for a 74LS259 equivalent)

While troubleshooting the controller you can keep the drive connected or unconnected, whichever is more convenient.  Or try taking measurements both ways to determine whether the drive or cable might be affecting the controller.

With the controller installed and powered on, and after RESET to stop all drives, try measuring voltages at the following pins:

• A2 pin 8 (74LS132) should be 5 volts when drive 1 is off.  If this signal is less than 4 volts when the cable is unconnected, then there's a fault somewhere in the controller -- possibly in one of those ICs at A2, B2, C2, or D2.
• A2 pin 9 (74LS132) should be 5 volts when drive 2 is off.  This measurement will probably be okay.
• A2 pin 10 (74LS132) should be 0 volts when both drives are off.  If pin 10 is correct while pin 8 is not, then A2 is probably faulty.  If this pin measures over 0.25 volts then D2 (556) is probably faulty.

If all three of those measurements are correct, then the controller is probably not at fault.  Especially if the measurements are correct when the controller is unplugged.

If it looks like the faulty enable signal is coming from the controller, you can further isolate the cause by measuring at these locations:

• C2 pin 9 (74LS259 or 9334) should be 0 volts when both drives are off.  If this signal is not right, then C2 is probably faulty.
• B2 pin 6 (74LS05) should be 5 volts when both drives are off.  If this signal is not right, then either B2 or D2 (556) is probably faulty.

To help understand the results of the measurements, have a read of this link.

https://learn.sparkfun.com/tutorials/logic-levels/all

C2 Pin 9 would be logic Low and feeds B2 Pin 5, B2 Pin 6 is the output of inverter (input from Pin 5) and would be logic High and is as expected.

A2 Pin 10 is Logic low which should force A2 P8 to Logic High (NAND gate needs all inputs High to get output Low, otherwise output is High).
A2 Pin 8 output is at a voltage that is not considered a valid Low or High state.
Based on this I would be replacing A2.

retro_devices wrote:

You can replace the 74LS132 with 74(LS)00.

That sounded plausible.  Just to test it, I found an old Disk II controller and replaced its 74LS132 with a 7400...and it works just fine.

So, why did Apple choose LS132 instead of the more common LS00?

My best guess is that Apple chose the LS132 because three of its inputs would be driven by the NE556, a linear component that only outputs a "high" signal of 3.3 volts.  Perhaps someone wasn't sure 3.3 volts would be high enough to reliably drive the inputs on an LS00, and therefore chose the LS132 for its Schmidt-trigger inputs?  (3.3 volts was later accepted as Low-Voltage TTL, aka LVTTL)

I was thinking the Schimdt trigger inputs for the 74LS132 might be used to minimise risk of multiple clock pulses going to A3 and C3 during Q3 transitions and/or due other injected noise via the drive cables.

Who knows if that disk is good. Could have been bad from before you got it, or could have been damaged during your troubleshooting. Unless you have a good, known working disk, it's hard to say if the drive is functional.

Apple did put out a tech note that duos can potentially wipe a disk in error. To prevent it they recommend removing c29 and c30 from the analog board:

"NOTE: There is a chance for data on the diskette to be lost in operating the DuoDisk. This can occur in attempting the "Open Apple CTL-Reset" technique for rebooting, or when using software with certain copyright protection schemes. If a unit exhibits this problem and passes the Drive Acceptance Program, check the analog board. Analog boards with part numbers 676- [] 101 and 676- [] 102 may have this problem. The fix is to carefully identify and cut two capacitors off the board. The capacitors are labeled C29 and C30 in zones Bl and Al (refer to the DuoDisk Analog Card Chip Swapping Chart and Figure 1, #4 and #5). Use small wire clippers or simply jiggle the capacitors to snap the connections."

No idea how common this issue actually is. I removed them from my duo just to be safe. 

You mentioned you have a IIe.
Without knowing what else you have available and assuming you only have the one diskette that came with the Duodrive, I would have a look at ADTpro and using the Apple cassette ports. You should be able to transfer a disk image to the diskette assuming the Duodrive is working correctly and there is no damage to the diskette.

https://adtpro.com/
https://adtpro.com/connectionsaudio.html

Skid-Wheel Playz wrote:

Alright so good news! I just replaced two of the three chips, the NE556 and the 74LS132 with a 74LS00. It works better than before, after CTRL Reset the light actually turns off. However, the random disc that came with the DuoDisk on eBay is not reading. It's just spinning and spinning with no data ever showing up on the screen. Is this normal? Maybe the disc's data got damaged when testing with malfunctioning chips? What should I do next? I don't have any other disks, however maybe I could try writing to this disc? There's also a faint clicking/ticking noise when trying to read or write. Is there yet another problem? I can't tell if the head ever moves after the initial reset to track zero.

The drive might be working now.

Here's a simple exercise that tests a whole bunch of components in the controller and disk drive.  With a fully-functioning Disk II controller and drive, these commands cause the controller to print threee bytes in inverse text: 00 then FF then 7F

• Press CTRL+RESET to get a BASIC command prompt.
• Type the command  CALL-151  and press RETURN to go to a Monitor prompt.
• To exercise drive 1 type the command: C0E9 C0ED I C0EA N C0E3 I C0E8 C0E2 N
• To exercise drive 2 type the command:  C0E9 C0ED C0EB IC0EE NC0E3 IC0E8 C0E2 N

With a properly-functioning Disk II controller and disk drive, those commands will cause it to print several lines of cryptic numbers, and the lines printed in inverse text print the bytes  00, then  FF, then  7F.  Does your DuoDisk do that?  Do both commands cause it to print the same bytes or different ones?

capture804.png

Looking ahead...

Are you willing to sacrifice the unknown contents of that diskette in order to build a bootable disk?  Or do you want to preserve the diskette so that you can inspect its contents later?  Either way, you'll probably want to buy additional disks so you can put both drives to use.  (Apple can use "low density" aka SD or DD, but not "high density" or HD)

Visit the AsciiExpress web site to learn how to use the tape port to download disks or software onto the Apple through your phone's headphone port.  The "GAMES" section can download popular games directly onto your Apple, without using the disk drive at all.  The "DISKS" section can download an entire disk at a time, but requires you to blindly overwrie the contents of the target diskette.

Skid-Wheel Playz wrote:

When I tested it, it didn't give those characters. Here's what it gave with both drives empty:

Ok, that might be a good thing!

Since the DuoDisk responded with "00" every time, it's probably just indicating the status of the write-protect sensor every time.  Unlike the Disk II, a DuoDisk drive probably won't generate "FF" bytes unless there's actually a diskette physically present at the "write protected" sensor.

A simple way to test this theory is to repeat the test while there's a write-protected diskette in the drive -- if all three bytes change to "FF" then it confirms that the three printed bytes are all indicating the same thing: write-protect status for that drive.  That test would also confirm that the DuoDisk's LS125 and LS32 are functioning properly.

Skid-Wheel Playz wrote:

Should I cut off the capacitors like nick3092 said?

I can't answer definitively.  But if it was my DuoDisk, then I wouldn't snip off the capacitors unless I could find more corroborating evidence for it.  Especially since the technote specifies capacitors C29 and C30, but the schematic doesn't show capacitor C30.  (In the schematic, capacitor C29 is paired with C28.)

I ran the monitor commands on my working Duo for reference. With no disk in the drives, I had all AA's for the regular lines and all 20's for the inverted lines on both drives.

When I repeated with a write protected disk in drive 1, the non inverted lines were still AA. But the inverted lines changed to A0 for EA and E8, while E2 changed to 20.  

Edit: if it matters, I have a Yellowstone and it was in Disk II mode. 

nick3092 wrote:

I ran the monitor commands on my working Duo for reference. With no disk in the drives, I had all AA's for the regular lines and all 20's for the inverted lines on both drives.

When I repeated with a write protected disk in drive 1, the non inverted lines were still AA. But the inverted lines changed to A0 for EA and E8, while E2 changed to 20. 

Those Monitor commands were tuned for an Apple Disk II Controller or Apple IO Controller.  With other types of controllers (eg: IWM, Franklin, A2, or Yellowstone) those commands will activate the drive but won't print the same bytes.  In particular, IWM and Yellowstone controllers cannot be commanded to generate the inverse "7F" byte, even if you attached a 20-pin Disk II drive.

On your Yellowstone controller, the inverse bytes will contain 3 bits from the controller's status register, probably followed by five 0-bits from its mode register.  In Disk II mode, all the mode bits will be zeroes.  So I would guess your Yellowstone was telling you:

• A0 = 1010 0000   Bit 7 indicates write-protected diskette, bit 5 indicates the motor is on
• 20 = 0010 0000   Bit 5 indicates the motor is on

Are you troubleshooting a DuoDisk, too?  Or just tinkering to compare how your hardware behaves?

Just tinkering as a second set of data to compare the results knowing my duodisk works, since it seemed like maybe the values for a duodisk weren't known for sure.