Master Compact's custom PSU + double floppy drive unit

[DanCaster]

Hi guys, I'd like to introduce you my Master Compact's custom PSU + double floppy drive unit!

It's a nice little box made of plywood. I wanted to make kind of a mix between the Commodore's floppy drives plus an integrated PSU unit that can power the drives and the Compact at the same time... all of that in an Apple I's style wood box!
kind of a "monster" from that point of view, but I think that the final result is pretty nice

I made it for three main reasons:
1) I wanted to leave the Master Compact totally original touching the least I can speaking about circuitry or case modifications (I may just want to fit the DSF room and nothing else)
2) I wanted to stop using the original PSU so that I can preserve its original components that are still working today
3) I wanted something a lot smaller than the original unit, something that can consume much less space upon the desk

I also discovered it's pretty convenient having the second floppy in absence of the DFS ROM: I leave inserted in the second floppy drive a floppy with the DFS and a !BOOT that auto-loads the DFS ROM. In this way it's a lot faster to load the DFS at startup since I haven't to search for the image in the GOTEK and then swap to the image I really want to mount.

From the technical point of view this custom unit works pretty well, exception made when I power the Compact with the PSU unit inside my custom unit: if I run the welcome disk, I can see on the desktop background some disturbs caused, I think, by the high frequency noise generated by the switching PSU I'm using. It's a PSU unit bought from Aliexpress and normally used to power LED's strips (https://it.aliexpress.com/item/40001151 ... 4c4dT4EMwG). Putting a ferrite core on the power cable seems to do nothing. Please tell me If you have any suggestions on how to fix this or if you can suggest me a better PSU to fit inside my unit.

This unit uses a modified GOTEK with flash floppy and an high density (HD) floppy drive from an IBM PC compatible machine.

I had to make a shugart to IBM floppy drive interface adapter cable. I discovered that with this cable any IBM floppy drive can read an ADFS floppy BUT only really few drives can write ADFS floppies without corrupting the disks... only the SAMSUNG SFD-321B/EZ REV.S2 worked when writing. I REALLY would like to understand WHY... shorting the density select PIN to ground in the drive (PIN 1 with PIN 2) does nothing since PIN 2 seems to be N/C in all of my drives. I must put a piece of insulating tape on the right hole of my HD floppies to get them work with the HD drive connected to the Compact. I know that the WD1770 can be overclocked to work with HD floppies and respective drives but I will not touch it.
maybe these new floppy drives expect CMOS voltage levels and the WD1770 outputs TTL voltage levels instead? If you know the answare please tell me
edit - to let ANY drive work when writing, PIN 1 and 2 on its back should be shorted (the red wire and the immediately next one)

In a future update of this unit I would like to put inside it my "Master Compact expansion port development board" (viewtopic.php?f=3&t=19115) so that I can emulate "any" extra hardware through the expansion port of the compact like additional ROMs, RS-232 and MIDI interfaces. In this way every update to the machine will be external and easily removable!

[jms2]

That's very elegant, nicely done!

I have not tried using my new Sony drive to write an ADFS disc yet - I must try that, maybe I will find the same problem as you.

[DanCaster]

That's very elegant, nicely done!

I have not tried using my new Sony drive to write an ADFS disc yet - I must try that, maybe I will find the same problem as you.

Thank you jms2!

Anyway I've already tryied a SONY MPF920-E E/131 and a SAMSUNG SFD-321B/LEB REV.T4. They don't work at all when writing... just reading. I saw on youtube people who modified the SONY drive to work succesfully with a Commodore AMIGA... let me know when you'll try your drive

In my case the only working drive is the SAMSUNG SFD-321B/EZ REV.S2

[klintworth]

That's very elegant, nicely done!

Seconded - I think this is brilliant, and I'll try this with my PC128S keyboard unit! This has been added to my project list!! Paul

[jms2]

I found the same problem. Any attempt to write to the disc corrupts it. How annoying!

I found a few Amiga sites describing mods to the drive, but none of them explained what the changes actually do! I will have to watch one of the videos.

[DanCaster]

That's very elegant, nicely done!

Seconded - I think this is brilliant, and I'll try this with my PC128S keyboard unit! This has been added to my project list!! Paul

I'm glad to read this

Wish to see soon your custom unit!

[DanCaster]

I found the same problem. Any attempt to write to the disc corrupts it. How annoying!

I found a few Amiga sites describing mods to the drive, but none of them explained what the changes actually do! I will have to watch one of the videos.

I saw the videos... they don't explain what these modifications do... in one of them a person said that the modifications were explained in some forums not in english and probably not in Italian

Let me know if you do some researches. I'm very curious about this topic

[BeebMaster]

Wow...wood! And an Olivetti - very nice!

The Sony MPF920Z will definitely work with ADFS (and DFS) as I used to use these in the long-forgotten BeebMaster dual floppy drive. I think it was jut about the last floppy disc drive in production. That model has automatic density select, so using a DD disc or covering up the HD hole would put it in DD mode. Possibly the 920E either has a jumper for density select, or doesn't support DD at all - I can't find the datasheet for it.

[DanCaster]

Wow...wood! And an Olivetti - very nice!

The Sony MPF920Z will definitely work with ADFS (and DFS) as I used to use these in the long-forgotten BeebMaster dual floppy drive. I think it was jut about the last floppy disc drive in production. That model has automatic density select, so using a DD disc or covering up the HD hole would put it in DD mode. Possibly the 920E either has a jumper for density select, or doesn't support DD at all - I can't find the datasheet for it.

Thanks BeebMaster.

it took many time to me before discover that I had to put a piece of tape in the hole of a floppy to let it work as DD instead of HD... probably because before that discovering I knew there were defferences between floppyes but not in detail... I did not really know what is the difference between SD DD HD and so on...

[tricky]

Beautiful box, do you think it needs any ventilation, maybe only on the back?
The meanwell PSU is the only other PSU that I know of that is used in a beeb, but mostly because it fits in a model B.

[DanCaster]

I found the same problem. Any attempt to write to the disc corrupts it. How annoying!

I found a few Amiga sites describing mods to the drive, but none of them explained what the changes actually do! I will have to watch one of the videos.

I saw the videos... they don't explain what these modifications do... in one of them a person said that the modifications were explained in some forums not in english and probably not in Italian

Let me know if you do some researches. I'm very curious about this topic

Maybe I discovered what these modifications do:

1) They let the drive be seen as 0 without a special cable (like the one I made) that swaps wires between the connectors
2) They bypass the density selector switch so that every floppy is treated as a DD one
3) They change the "Disk change" signal into the older "ready" signal

Anyway the modification 3 for the compact is useless since that signal is not brought into the Compact's WD1770,
the modification 2 can be done with a simple tape to the right hole of any floppy (like I do) and modification 1 can be done simply twisting 3 wires like in the picture of my first post.

So why certain drives do not work with the Compact remains a mystery to me... I should plug a logic analyzer or an oscilloscope to try to figure out something more...

[DanCaster]

Beautiful box, do you think it needs any ventilation, maybe only on the back?
The meanwell PSU is the only other PSU that I know of that is used in a beeb, but mostly because it fits in a model B.

Thank you tricky,
my PSU does not overheat... I'll anyway drill some holes on the bottom and on the back.
The meanwell PSU could be too big for my unit and probably overheats a bit but I'll think about that. Thanks

[jms2]

Maybe I discovered what these modifications do:

1) They let the drive be seen as 0 without a special cable (like the one I made) that swaps wires between the connectors
2) They bypass the density selector switch so that every floppy is treated as a DD one
3) They change the "Disk change" signal into the older "ready" signal

Anyway the modification 3 for the compact is useless since that signal is not brought into the Compact's WD1770,
the modification 2 can be done with a simple tape to the right hole of any floppy (like I do) and modification 1 can be done simply twisting 3 wires like in the picture of my first post.

I think you are right - those are the three modifications that I have been able to find. As you say, these should not be relevant to this problem. Its a bit of a mystery!

I will see what I can find out about the Sony MPF-920E (which is the drive that I am using ... I think).

[jms2]

Correction - mine has a sticker which says:

SONY MPF920
Z/121

It dates from 2005.
I don't know if that means it is an MPF920Z, but it appears to. Beebmaster says that these should work, so I'm puzzled.

I managed to find a reference to using these drives on an Atari ST (https://www.exxoshost.co.uk/atari/last/ ... d/main.htm) which says that drives containing a SONY controller chip don't work, whereas later drives with a "BH955" controller do work. But when the say "don't work", what they are talking about is swapping the drive position jumper from DS1 to DS0, which seems to be essential for the ST. In our cases, we have not moved this jumper.

I might try to format a disc using my drive to see what happens. Maybe I was just a bit unlucky before...

[jms2]

Well that was "interesting"!

In ADFS, the drive formats without problems, but then when it comes to Verifying it reports "Disc error 50" immediately.
In DFS (2.45), the formatting doesn't do anything - it just hangs the machine.

Reading and writing to the Gotek drive work fine.

[BeebMaster]

That's very odd.. the MPF920 Z/131 should work fine. Are you definitely using a DD disc or covering up the HD hole?

Maybe there's something special about the Master Compact disc interface? Could you try the drive in a Master 128?

[CMcDougall]

yeah, very weird not writing discs, I had (now got MMC) two drives on mine working perfect with DFS as ADFS is pants.

[jms2]

I don't know about Dan, but I'm using the "proper" DD (blue) discs. One thing I have not tried is a different disc - maybe this one is just dead?

The Compact application note does mention that the floppy disc routines in ROM have been sped up compared to the Master 128. Also, it says that write pre-compensation (whatever that is) has been applied to all settings of *CONFIGURE FDRIVE. This gave me an idea, so I tried different values of *CONFIGURE FDRIVE to alter the timings. However, all this did was make the drive fail more quickly...

Next step is to try another disc- but I will have to find a blank one, any attempts to write are likely to end in disaster!

Whilst the Compact's floppy code may be a bit faster, I find it hard to believe that it's any faster than the comparatively modern PCs that the Sony drive was intended for.

The only strange thing about the Compact's interface is the way that the lower-numbered wires in the cable are not connected, and there is a permanent 0v feed to pin 4 which causes the disc LED to light when the drive is in use. This feature was needed for the original Sony drives that the machine shipped with, but isn't really needed for more modern drives. The MPF-920Z datasheet says that pin 4 is not connected, so I don't think it can be that.

[CMcDougall]

always use one hole blue discs, rest are point less

some pics I had , that may or may not help....

[DanCaster]

I found the fix to the problem! now all the drives work perfectly!

Who did this picture wanted to short to ground the "density select" PIN but instead grounded a N/C PIN:
In the picture PIN 3 and 4 are are shorted, not PIN 1 and 2 that are the right ones...
I already saw this image in the past and understood what it wanted to do, but the PIN disposition confused me. That's why I originally said that grounding these PINs seemd to do nothing and that they were N/C: because I was grounding the wrong PINs like in the original picture

Now, as you can see, everything works fine also with the Sony floppy drive

So at the and of the story I can tell you that if you want to use an IBM PC's floppy drive with a Master Compact, you should place a piece of tape in the right hole of the floppy (only if it's an HD one... otherwise there are no holes on the right...) AND short PIN 1 and 2 like in the picture. Anyway some floppy drives like my SAMSUNG SFD-321B/EZ REV.S2 don't necessarely need to have PIN 1 and 2 shorted to work... maybe they have some protection circuitry that auto-grounds floating PINs

I'm going to edit my main post with the right informations

[jms2]

That is great news Dan. But I want to make sure that I understand what is going on...

Col's photo is correct, pins 3 and 4 should be shorted in the original Compact floppy cable. This is because the original Sony MPF63W drives needed a 0V signal to be provided on pin 4 (IN USE). This isn't necessary on the MPF920Z.

The MPF63W pin assignment shows the IN USE signal on pin 4:

However, the MPF920Z datasheet shows that all four pins 1-4 are not connected:

I did manage to find some references to pin 2 being Density Select 1=Low/0=High, but:
- The MPF920 automatically selects density so doesn't use this pin
- On drives which do use this pin, shorting it to 0v would select High density, which is not what we want.

So I don't understand why shorting pins 1 and 2 helps...? Unless the Sony datasheet is wrong?

[CMcDougall]

yes that IS my picture, taken from my original untouched PSU box, but with my PC drive connector shoved on so could copy stuff to drive 1/3

[jms2]

Actually, this might explain it - you have the MPF920E, and on that model pin 2 selects the data rate (although the datasheet doesn't explain which setting relates to 5v and which one to 0V):

So this could work for your drive, but I'm still not clear on why it would help for mine. But maybe in the later days of floppy drives the manufacturers got careless and made mistakes on their datasheets?

[BeebMaster]

According to this on the MPF920Z, Key means:

Maybe shorting something to it when it doesn't exist is causing the problem?

[jms2]

Maybe shorting something to it when it doesn't exist is causing the problem?

Good point, but if none of these pins do anything then it shouldn't matter.

Another question is why these drives apparently work on a normal BBC. In that situation nothing is shorted, but there is a cable going from each pin back to the floppy interface. I'm still trying to figure out the effect of this.

[DanCaster]

jms2, you did a really interesting research and all you explanations are very convincing!

Anyway I have to say: "eppur si muove" (And yet it moves)

I tried the SONY MPF920-E E/131 and a SAMSUNG SFD-321B/LEB REV.T4 that before my "mod" were not working at all when writing. After I shorted PIN 1 and 2, they magically started working perfectly, even better than the original floppy drive: not a single floppy wrongly formatted.

Even stranger: when I measure the resistance between PIN 1 and 2 my tester measures zero like if these PINs were N/C (as for PIN 3 with 4 and 5 with 6 that are N/C by standard in IBM PC drives). Anyway shorting PIN 1 and 2 has an immediate effect.

last but not least, the SAMSUNG SFD-321B/EZ REV.S2 keeps working even with my "mod".

I suggest you to try this "mod" with you floppy drive too. three different working floppy drives I think can hardly be just an isolated case...
In the future I'll may have two more floppy drives to test. I'll update you in case

my tests consisted in:

*FORMAT 0 L
*.
*BACKUP 1 0
*.
*!BOOT

[danielj]

Just a note of caution. Covering holes in floppies will result in unreliable data. High density disks use a different coating which requires a stronger magnetic field to write. If you use those disks as double density it'll probably work initially but HD disks written as dd often "fade" over time.

[Kazzie]

Pin 1 (index) is an input to the 1772 floppy drive controller. There's a pull-up resistor in the Master Compact, which ensures the input is high if left disconnected. Because it's an active low signal, the 1772 will think it's getting no response from the drive if the wire is disconnected. Having shorted pin 1 and 2, it's permanently low, so there's a input detected.

Put another way: this fix isn't for the drive's benefit, it's a fix for the FDC because the newer drive can't give it something it wants.

[jms2]

Pin 1 (index) is an input to the 1772 floppy drive controller. There's a pull-up resistor in the Master Compact, which ensures the input is high if left disconnected. Because it's an active low signal, the 1772 will think it's getting no response from the drive if the wire is disconnected. Having shorted pin 1 and 2, it's permanently low, so there's a input detected.

Put another way: this fix isn't for the drive's benefit, it's a fix for the FDC because the newer drive can't give it something it wants.

I think we are using different pin numbering systems here. The numbering system that Dan and I have been using up to this point is 1 = pin 1 on the drive. This is the same as pin 1 on a normal BBC as well. However as Col's table above shows, the Compact has a bizarre arrangement in which the drive's pins 1-7 are not connected to the host machine at all. The only thing you can do with these is locally (ie at the back of the drive) connect the even numbèred signal-carrying pins to adjacent grounded pins in order to tie them to 0v.

At the Compact itself, pin 1 is connected to the drive's pin 8 = index, so it's not this 'pin 1' which is giving us a problem.

I notice that people using these discs on Amigas and STs have to short out a connection internally to force the drive into DD mode. I wonder if this is the same as shorting pins 1 and 2. I also wonder why it is necessary - unless maybe they are using HD discs and don't want to tape over the hole?

[daveejhitchins]

I've got to ask . . . with my tongue stuck hard against my inner cheek . . .

Does this unit meet the relevant EMI requirements?

But, seriously . . . That's a good looking piece of wood-work

Dave H