External Roctec RF382C-A floppy drive - not working

[solarmon]

UPDATE: This is the current summary of the state of my Roctec/Citizen drives.

Drive #1 - WORKING:
* Replaced belt - initially with 71mm belt, then with 69mm belt
* read/write heads interfered with
* Recapped (here and here)
* Track alignment fine tuning
* Top head ribbon cable repair
* Top head spring missing, used coin as temporary solution.

Drive #2 - WORKING:
* Replaced belt - initially with 71mm belt, then with 69mm belt
* Recapped
* Track alignment fine tuning

Drive #3 and #4 - WORKING:
* Replaced belt - 69mm belt

Drive #5, #6 and #7 - WORKING:
* Replaced belt - 69mm belt

Please read the rest of the thread for further details.

In general:

* Avoid dismantling the read/write headers! You will get header alignment issues!
* Avoid dismantling the track motor! You will get track alignment issues!
* A 69mm replacement belt seems to be better fit than a 71mm one, from my experience.
* The belt can be replaced without dismantling the drive completely - use fine tweezers to help guide the belt through the drive to where it needs to be.
* Clean the read/write heads using isopropyl alcohol with a cotton bud.
* Clean up any old crusty lubricant on mechanisms using cotton bud, and replace with lithium grease, or similar.
* A re-cap could be attempted if all else fails, but I've never got a drive working just from a recap.

Importantly, about fine track alignment using the track motor at:

http://eab.abime.net/showthread.php?...92#post1335392

===============================================


Hi,

I had recently acquired an external Roctec RF382C-A floppy drive.

It was not working since drive belt had perished, (see perish drive belt in first picture), but I was able to replace it.

However, it does not seem to be able to read a known working floppy disk. The floppy icon appears but had "DF1:NDOS???" (or something like that) as the name.

EDIT: Just checked - it actually appears as "DF1:????"

The drive seems to be recognised/seen by my A1200 as DF1:, has the usual/regular seek clicking, but when a floppy is inserted, it seems to try to seek to track zero and but then stop/fails.

If I manually move the drive head to the furthest track along, and then put a floppy in, it does seek back to track zero - so I know the track motor works.

The replacement belt and drive motor also seem to work, as I can see (with the drive out of it enclosure and opened up) the floppy disc spinning.

It actually did once manage to read a floppy and the correct floppy name (it was an Amiga Format cover disk) was shown. But before and since then, nothing.

I have tried cleaning it up, but there must be some other electrical or alignment issue?

I'm aware of the DriveTest tool - but I still need need to figure out how to use it.

Any suggestions on how/what to troublehoot further?

Cheers!

[solarmon]

I managed to figure out how to use DriveTest (after getting some hints from [ Show youtube player ]).

It looks like the track 0 sensor is not triggering. I don't think it is because it is dirty, since if it was, then the sensor (if it was working) would be triggered all the time, since it is an photo interruptor switch and dirt/fluff would interrupt the optical (infra red?) signal going to the sensor part of it?

I must assume either the sensor is faulty, or a circuit/components around it is faulty.

I'll try to give it a another clean but I don't think it is going to resolve the issue.

Anybody had any experience with replacing/fixing the track zero photo interrupter sensor?

[solarmon]

Thinking that the photo-interrupter component for sensing Track 0 was faulty, I ordered a new one (or what I think is the equivalent - KINGBRIGHT - KTIR0611S - PHOTO INTERRUPTER) from ebay.

However, testing some more with DriveTest, I found (I think) that it might actually not be faulty.

I can get DriveTest to show the TRACK0 signal going high and low (by STEP'ing the head back and forth over Track 0 - i.e. interrupting the photo-interrupter sensor, and not), but only when I enable MOTOR when selecting SELECT1 (DF1) for the not working external drive.

This does not occur for when selecting SELECT0 (DF0) for the working internal drive. I can have MOTOR unselected and the TRACK0 signal changes status appropriately.

Anybody with experience troubleshooting and repairing floppy drives come across these symptons?

[solarmon]

I still have not got this external floppy drive working, but I thought I'd post to provide an update so as to document my struggle with it.

I found out that some floppy drives only set the TRACK 0 signal along with the MOTOR signal. So this explains the behaviour that I was seeing.

I also noticed that the READY state was flapping. One of the conditions for the READY state is the rotational speed of the spindle needs to reach 90% of the specification.

In my case, it looks like the new drive belt I had put in was slipping.

When I had replaced the belt I noticed that there was was 'stuff' on the drive belt motor shaft - I thought at the time that this might have been the remnants of the perished belt, so removed it by scraping it off.

Now I think it might have been there to provide some friction. I decided to add some B-7000 Glue Adhesive to the motor shaft where it drives the belt to both provide some friction and to increase circumference.

This seemed to have done the job and the belt was no longer slipping and the READY state was not flapping.

However, that still did not get it working. It seems that there could be some alignment problems - so I tried playing about with the top read head alignment with no luck.

Then I thought that usually/typically with alignment problems, the drive is able to only read floppy disks that it had formatted/created. So I tried to format a floppy disk but that failed too.

So my current conclusion is that there might be some other component that has probably failed - maybe a capacitor or chip.

This is now getting beyond my skillset, but I've learnt a lot about how floppy drives work during this troubleshooting!

[mogouk]

@solarmon i feel your frustration, i have 2 of these drives with extaclty the same symptoms, i have repaired a few of these by just replacing the drive belt but the last 2 i have will just not have it, only thing i can think of now as a last resort is to replace the caps, has anyone had similar issues and got a working drive with replacing the caps on these models, i have spent hours of my time on these drives. and reluctant to give them any more. but if there was a good chance it was the caps maybe i would have another go. the caps look good, no leakage very shiny but they are 30 years old

[mech]

Have you considered just converting a pc drive to amiga and replacing the entire mechanism? this would be a cheap and easy way. Maybe adopt the sony mpf-920 to it?
This would do away with belt drive problems forever.

[mogouk]

Originally Posted by mech

Have you considered just converting a pc drive to amiga and replacing the entire mechanism? this would be a cheap and easy way. Maybe adopt the sony mpf-920 to it?
This would do away with belt drive problems forever.

yes i've converted mpf920 sony drives for standard cases, but these are slim type drives with slimmer cases so standard drives don't fit.

[solarmon]

So, I'm still trying to fix this drive....

I done a few things since.

I have proven that the 23-pin External floppy connector on my A1200 works - using an external drive adapter for Gotek and PC drives (https://www.ebay.co.uk/itm/Amiga-ext...y/272437676307). I can get an external Gotek and PC drive to work, so I'm confident that there is no issue with the external floppy drive connector.

I even connected this external drive adapater to the 23-pin port of the Roctec drive itself and it works - so this shows, at least, the pass-through/daisy-chain part of the Roctec is working.

The Roctec seems to work in every aspect apart from reading of data from the floppy drive/disk. Tests using Amiga Test Kit and DriveTest shows the functions, other than reading, seems to be working.

On the Roctec controller board, there are two chips. An unsocketed HD7407P chip which seems to be some kind of buffer chip. And a socketed AMI 9139MZH 18CV8PC-25 chip which seems to be some kind of CMOS programmable chip.

I decided to replace the HD7407P chip, just in case this was faulty, and at the same time I put in a suitable DIP socket for it. This didn't seem to resolve the reading issue.

I didn't replace the AMI 9139MZH 18CV8PC-25 as I don't have any way of programming a new one, or even have the source/image to program it with.

I'm confident that the controller board itself is OK, so it must be the Citizen drive itself. I manage to acquire the another, same, Citizen drive. When I connected it (after replacing the drive belt and cleaning the heads) it was initially reading/showing a floppy disk (well, it was showing the name of it) but then it disappeared from Workbench after a few seconds. After a few attempts with this behaviour it eventually stopped working altogether and seems to behave exactly like the first Citizen drive.

So I can only assume that now both Citizen drive have the same issue!

My next step is to replace the SMD caps on them. There are the following 4 smd caps values, if I have read them right:

C35 = 2.2uF 35V
C39 = 10uF 10V
C101 = 2.2uF 35V
C102 = 2.2uF 35V

I have removed these caps (snipping off the cans first and then de-soldering the legs). There was only one cap that still seem to have fishy smelling liquid in it (can't remember which one) but the others all seem dry and the solder pads were quite dry and crusty. But after they were cleaned up the pads looked OK.

I'm now just waiting for the new caps to be delivered, so I'll report back once they have come in and see if I'm successful in soldering them back on. This will actually be good practice before I re-cap my A1200!

[solarmon]

Some photos of the caps removal:

(Sorry, I still don't know how to re-size the linked images using the the markup language in this forum)

Board with original caps:



Caps snipped off:


Close up of crusty pads:





Closeup of cleaned pads:

[solarmon]

I've been trying for ages to find the pinouts of the 26-pin connector on this Citizen U0DC-68A drive.

Looking at the logic/controller board it was easy enough with a multimeter to find that the VCC (+5V) and GND pins are:

VCC (+5V):
1
3
5
7

GND:
13
15
17
19
21
23
25

None of the 26-pin floppy drive pinouts I had found exactly matched this, so I was a bit cautious of using them for troubleshooting.

However, I have now found some references to such a 26-pin pinouts that matches my VCC and GND pinouts:

http://www.vcfed.org/forum/showthrea...X-floppy-drive

https://groups.google.com/forum/#!to...lt/Xyu1lw9I_yc

https://www.vogons.org/viewtopic.php...57883&start=60

These all had the pinouts as:

TABLE B-9 FDD Connector Pin Assignment 

Pin Signal I/O        Pin Signal I/O 
  1    VCC               2  INDEX 
  3    VCC               4  FDSELA 
  5    VCC               6  DSKCNG 
  7    VCC               8  READY 
  9    NOTCH0           10  MONA 
11    LOWDNS           12  FDCDCR 
13    GND              14  STEP 
15    GND              16  WDATA 
17    GND              18  WGATE 
19    GND              20  TRACK0 
21    GND              22  WPROTC 
23    GND              24  RDDA 
25    GND              26  SIDE

The last link references this 26-pin floppy drive pinouts to the Toshiba T3200SX laptop and I was able to find the Service Manual for it:

http://www.minuszerodegrees.net/manu...e%20Manual.pdf

An this has the same pinouts as stated in the other references:



So, I think now I'm confident enough to use these pinouts to troubleshoot this Citizen U0DC-68A floppy drive, should I need to.

(Or maybe use them to convert the 26-pin to 34-pin so that a standard floppy drive could be used)

[jurassicman]

Interesting thread, thanks for sharing your experience...
I have the same drive and I have replaced the belt without fixing it...

[solarmon]

So...

The caps arrived and in my eagerness, I quickly solder them on to the first Citizen drive.

The best soldering work, but it was my first time doing SMD caps and the contacts seem to buzz out OK with a multimeter:

(Notice I accidently melted parts of the brown connector - oops!)



Now, bear in mind that this drive had been complete dissasembled - including the read/write heads - so there will be misalingment problems anyways.

Long story short: It still cannot read floppy disks.

After messing about with Amiga Test Kits calibration test to try to align the heads, I can't get the upper and lower heads to align. X-Copy checkdisk shows the following:



Really disappointed and annoyed!

Not sure if the re-cap is a re-crap, or there is some other fault, or it is just a case of misaligned heads.

I'll park this one for now (or forever) and move on the the second Citizen drive, which I have not taken completely apart, include the top read head.

[solarmon]

So...

Second Citizen drive, the one that has not been totally disassembled - just the belt replaced.

This time, I did the recap without taking it completely apart. It was a bit tricky with the one near the metal housing, but I don't think I melted anything and, again, it buzzed out OK with a multimeter:



Now this one is a bit more interesting.

ATK calibration test shows that it all looks OK:



But it still can't read:



Even though it can write:



But ultimately, it still cannot read a floppy disk correctly. Strangely, the first and last sectors of both upper and lower sides are read OK.:



So, still the common issue is that either the logic board and/or the drive itself still cannot read.

Next, I think I will try to map out the read lines and see if I can spot anything.

And I'll probably try a format (since it can seem to write data) and see if it can read its own formatted floppy.

Any other suggestions? As I'm running out of ideas.

[solarmon]

I tried a Workbench format in X-copy of a floppy disk in the 'faulty' drive, and it went through OK:



It was able to read the disk using the 'Disk Info' option:



However, the Check Disk option, showed that it was not able to read the first sector of each side:



So, this meant that Workbench was still seeing it as "DF1:????".

I might try another floppy disk, just to rule out an issue with this one. But there is still the issue of head misalignment.

[solarmon]

Another update.

I tried replacing the capacity 22uF capacitor on the logic board. This did not resolve anything.

I then acquired two more faulty slim external drives which had similar (one same: U0DC-68A ; and one similar: U0DC-09A) floppy drives/ Both of these had a different logic board design - see https://imgur.com/a/3s346oQ

For both of these drives, again the drive belt had perished.

On the topic of drive belts. I originally used a 71mm belt that had been recommended on a few articles/videos, but found that this was just too loose and it was slipping too much:

https://www.ebay.co.uk/itm/FLAT-SECT...t/301376271629

I found a slightly smaller 69mm one (but slightly wider by 0.2mm) which seems to be much better and works without any problems.

https://www.ebay.co.uk/itm/FLAT-SECT...t/281080851960

Once I had to replaced the belt and cleaned the read/write heads (without taking it fully apart - so didn't mess with the alignment, assuming that it hasn't already been messed with), they both worked OK with their respective logic boards, and with my original one. Both were able to format a floppy and read it back, and the the same floppy could be read and checked OK by my internal DF0 floppy drive.

So now at least:

1. I have proven the logic board, even after the chip and capacitor replacement, is definitely working.
2. I have two working Citizen floppy drives as reference to be able to check/compare against.

[jurassicman]

@solarmon
I just want to thank you for this thread.

I bought the belt you suggested and now my Roctec drive finally works!
As you did, I previously had bought the other belt but that was too "loose" while with this one the motor mechanism works perfectly.

This is the quietest drive I ever had!

[r.cade]

I don't think I've even ever seen 3.5" drives with belts. I thought by that time all the motors were direct drive...

[Jope]

Originally Posted by solarmon

I found a slightly smaller 69mm one (but slightly wider by 0.2mm) which seems to be much better and works without any problems.

https://www.ebay.co.uk/itm/FLAT-SECT...t/281080851960

Thanks for the report, I'll add it to the belt thread.

[solarmon]

Originally Posted by jurassicman

@solarmon
I just want to thank you for this thread.

I bought the belt you suggested and now my Roctec drive finally works!
As you did, I previously had bought the other belt but that was too "loose" while with this one the motor mechanism works perfectly.

This is the quietest drive I ever had!

I'm glad this thread helped somebody else!

I just wish I could get my faulty Roctec Citizen drives working.

I have actually made a little progress that I had updated this thread with yet. I need to go back and review what I did and will update this thread.

Originally Posted by r.cade

I don't think I've even ever seen 3.5" drives with belts. I thought by that time all the motors were direct drive...

There seems to be a lot of these Roctec slim drives around - well, looking on eBay there are quite often these for sale (usually as parts/spares because of the perished drive belt issue)

Originally Posted by Jope

Thanks for the report, I'll add it to the belt thread.

Sorry, I wasn't aware of a 'belt thread', otherwise I would have reported to that too.

[Jope]

Originally Posted by solarmon

Sorry, I wasn't aware of a 'belt thread', otherwise I would have reported to that too.

It only talks about the RF-332, so it's easy to miss. I'll edit the title to show this other model too.

http://eab.abime.net/showthread.php?t=50077