How to modify a Sony MPF 920 to provide a ready signal

[TFM]

Or handle the FDC error properly

See, that's the problem: The drive gives false information (Ready if not Ready!), that screws the FDC. 


In your case it seems to work. Have you ever had an error with SymbOS?

[khaz]

Bumping this thread, because I happen to have the same drive. I've modded it seemingly successfully, albeit not exactly as explained on the Internet. I do have a few questions about it and the various schematics posted in this thread.

I started by following this guide Sony MPF920-E E/131 modified to work with the Amiga, but it didn't work at all, I kept having a "disc missing" error. I have to say I didn't follow it completely: I didn't bridge SW3 to block the HD/DD verification, nor did I change the JC30/31 jumper.
- Can I just solder the switch and be done with covering HD disc holes with tape? Or do I still need tape for Windows and cpcdiscxp?
- Do I still have to change the JC30/31 jumpers if I don't want to use the drive as A? Does it even matter when I can use a switch on the cable?

What I did instead is that I used the disc-in switch to ground pin34. When a disc is in, the switch closes and pulls this part of the circuit low. By soldering a wire to the not-ground side of the switch, inserting a disc simply grounds pin34 as if there was a bridge on the cable between pins 33 and 34. This work fine, although AMSDOS and CP/M can get a bit confused between the two discs. Disckit3 has no problem pointing at the guilty drive when it asks for both to be empty after an operation though.

Earlier in this thread, this page was linked Lecteur 3"1/2 sur CPC [Quasar Net], and specifically these schematics were mentioned:

I'd have no problem implementing it if needed, but I have a lot of trouble understanding it, especially in relation to Drive 1. This is pin12 on the cable, right? Can someone explain what "Drive 1 select" does and what are the different outputs for the ready signal? One can use the disc-in switch, then the switch and 10kohms resistor are not needed, the gate is connected to the 5V side of the switch. This is what I would do.

If it's not already obvious, I have no background in electronics. I'm learning by reading stuff on the internet and tinkering with old circuits and asking stupid questions.

[pelrun]

Most of the signals are shared between both drives; the Drive 1/0 select pins tell the drives which one the CPC is trying to talk to. Deselected drives will ignore anything they see. The ABBA switch flips the select signal between the drives.


The switch in the circuit you posted *is* the disc-inserted switch; it's not meant to be an additional component. The resistor absolutely has to be there; you can't just go 5v->disc-in switch->gate, otherwise the input to the gate will be completely indeterminate when the switch is open. If there's already a resistor to ground connected to the switch on the drive pcb then you don't need another one, though.


The end result of the circuit is that the ready signal is pulled to ground only when the CPC is talking to the drive *and* the disc is inserted. Otherwise the signal is left alone so that it can be used by the other drive.

[khaz]

The switch in the circuit you posted *is* the disc-inserted switch; it's not meant to be an additional component. The resistor absolutely has to be there; you can't just go 5v->disc-in switch->gate, otherwise the input to the gate will be completely indeterminate when the switch is open. If there's already a resistor to ground connected to the switch on the drive pcb then you don't need another one, though.

But that's not what is described on the original page though?

(google translate below)

Deux solutions s'offrent à vous. La plus simple, mais pas forcément la plus évidente, consiste à localiser l'interrupteur de détection de présence de discs d'origine de votre lecteur 3"1/2 puis à détourner ses connexions avec le reste de l'électronique du lecteur pour l'insérer dans notre circuit "Ready". Il est important de ne pas isoler l'interrupteur du reste du lecteur car le signal qu'il envoie est parfois indispensable au bon fonctionnement de ce dernier : brancher vous en parallèle, ne coupez pas les pistes. Généralement l'interrupteur lui-même est facile à localiser, il est à l'avant du lecteur. En revanche, il doit être accessible au fer à souder et ça c'est une autre histoire... Dans le cas où vous repiquez les signaux de l'interrupteur d'origine, il ne faut bien sûr pas câbler le 5V et la résistance indiqué sur les schéma ci-dessus, mais relier directement le signal à l'entrée de la porte logique.

Two solutions Available to you. The easiest, but not necessarily the most obvious, is to locate the original discs presence detection switch your 3 "1/2 drive and divert its connections with the rest of the drive electronics to insert it in our "Ready" circuit. It is important not to isolate the switch from the rest of the drive because the signal it sends is sometimes essential to the proper functioning of the latter: you connect in parallel, do not cut the tracks. Generally the switch itself is easy to locate, it is at the front of the player. However, it must be accessible to the soldering iron and that's another story ... If you prick the origin of the switch signals, do not wire the 5V and the resistor shown the above scheme, but directly connect the signal at the input of the logic gate.

Most of the signals are shared between both drives; the Drive 1/0 select pins tell the drives which one the CPC is trying to talk to. Deselected drives will ignore anything they see. The ABBA switch flips the select signal between the drives.

Does it mean that if change the jumper to Drive 0, I would have to ground pin12 to have it back to B? This internal jumper, along with the pin11/12 cable switch, the internal ABBA switch, the cable twisting, etc. are a bit confusing to be honest.

The end result of the circuit is that the ready signal is pulled to ground only when the CPC is talking to the drive *and* the disc is inserted. Otherwise the signal is left alone so that it can be used by the other drive.

I think this is what I want. It would prevent AMSDOS to hang when I poll an empty A with a disc in B. Right?

[pelrun]

But that's not what is described on the original page though?

(google translate below)
Does it mean that if change the jumper to Drive 0, I would have to ground pin12 to have it back to B? This internal jumper, along with the pin11/12 cable switch, the internal ABBA switch, the cable twisting, etc. are a bit confusing to be honest.

Basically what the page is saying is that if you use the original disc-insertion switch, then the 5v and pulldown resistor are already present (because the original drive hardware uses it) so you don't have to add it yourself.

As for the drive select - each drive chooses which signal it listens to. Either it's hardwired, or there's a jumper that lets you select it. You need to ensure that no matter what is happening with an ABBA switch or a cable twist, that the two drives are configured to use opposite signals. If both drives are hardwired to the same select signal, then you need a cable twist (which swaps the DS0 and DS1 wires) to flip one drive. The ABBA switch then flips both drives.

I think this is what I want. It would prevent AMSDOS to hang when I poll an empty A with a disc in B. Right?

Yes, that's the intent. Hardwiring the ready signal to be always on means AMSDOS thinks there's a disc ready in A all the time, even when it's not true. Gating it with the drive select signal means that any other drives can correctly signal their own states to the CPC.

[Audronic]

I have been using the Modifications to a Sony MPF-920 which works . (These are Zaxon's Modifications).


Dropbox - Amstrad 3.5" Floppy Mod (Sony MPF-920).zip

EDIT :-  Changed Link
Good Luck    Ray

[khaz]

I have been using the Modifications to a Sony MPF-920 which works . (These are Zaxon's Modifications).


Dropbox - Amstrad 3.5" Floppy Mod (Sony MPF-920).zip

EDIT :-  Changed Link
Good Luck    Ray

This is basically the mod I've done, but I'm using the disc-in switch instead. It does its purpose, but it's not ideal and is buggy when polling an empty A with a non-empty B.

That was quick! Ok, here's the quick answer:

Getting a Ready signal:
Step 1 - Cut the track going to the headers pin 34 (marked with x in bottom left corner)
Step 2 - Connect pin 18 of IC2 to pin 34 of the header (shown as red line)

Making the Drive A: (If you want to use the drive as B: then this step isn't required)
Remove the bridge (black component with "0" on top) from JC31 and use it or a short piece of wire to short the JC30 contacts.

The other drive you posted seems to be exactly the same, just the layout orientation has been turned 90°, so this will work on both versions.

Bryce.

Edit: Just realised that the blurred first picture uses a different IC, can you possibly take a better picture of that drive or let me know the part number of IC2.

This is the first mod I tried. Pin 18 on the IC2 doesn't give a "disc ready" signal. Pin 17 sort of does, but has the same limitations as a disc-in switch mod.
[edit] I'm not trying random pins, but I saw a mod that used this pin instead of 18. Can't find it again though.[/edit]

The IC is D9848N 6943A. It's still sold everywhere but I just can't find a datasheet or a pinout for it.

[KaosOverride]

This should help with other drives http://freaknet.org/asbesto/roba/diskettes-jumpers.pdf


Sorry not useful  info about the ontopic Sony drive....

[khaz]

I'll bump this topic again, as I have another problem related to the Disc Ready signal.

The screen jumps, one frame is desync, every time a drive is accessed. At first I thought it happened only in my newly modified drive, but I have tried several other drives that behave the same. Only the HxC is perfectly stable. Simply doing |A at boot makes it happen, but it also happen at every track change, and when I insert (but not when I eject) a disc.

I've tried with several drives, modded, unmodded, with the pin 34 cut (no "disc ready", nor "disc change" signals), with a normal cable and a cable with a shorted pin 33-34. In all these situations, without a disc in, there is no interference. With a disc in, the screen jumps when doing |A. It also jumps when I insert a disc and the motor is still running, but not when I eject the disc. No interferences happen at all if the drive isn't powered.

Trying to isolate the problem, I thought the interference happened when both the pin 34 is grounded and the motor is accessed, but only for the first 1ms or so. But it also happen when the trace to pin 34 is cut and there is no forced grounding of pin 34. Also the interference at the disc insertion (with a spinning motor) baffles me too.

Based on all this, I don't think getting a proper "disc ready" signal with the previously linked schematics will modify this behaviour. I'm still going to do it though, the chip is on its way, but I would love your input on this problem.

[edit]You know what, haha, nevermind. I just realised that my Scart cable was drawing its +5V for RGB switching from the same source as the CPC and the external drive. Accessing the motor made for a very temporary drop in power which confused my TV. I just tried with my original CTM644 and everything works fine! Silly me

[Audronic]

@khaz


Can you adjust the Vertical hold on your Monitor ?? It may be that's the problem not the FDC components.


Good luck


Ray

[khaz]

It was totally the monitor. The cable I have plugs to the same power source for its RGB blanking signal, and when the drive motor draws some power, it confuses it. Nevermind!

[khaz]

Bumping the thread again, now that I have finally received my 74HC00[nb]All twenty of them. Does anyone want 19 SN74HC00N?[/nb]. I want to make sure I'm doing everything right.

So this schematics:

I want to use the Sony MPF 920 switch for Disc In, which has ground on one side and the other side connects to some pin on a chip. Inserting a disc thus closes the circuit and grounds that pin. How can I wire it so it can be the switch on the schematics? I have one spare NAND gate on the chip so I suppose I should use it, but I'm confused how. The rest should be straightforward: Drive 1 Select is pin 12 of the drive, ready is pin 34, and Ground is wherever.

[khaz]

Useless bump, finally done it. The real problem was to find some space for the chip and the bunch of components inside the metal casing. I'm sure someone could make a tiny board with smt components that would make any drive give a real ready signal. There must be a market, especially for Amiga users.