Swapping DF0: and DF1: on a500+

Hi all,

I got an A500+ and a gotek together the external drive interface from amigastore.eu. What I want to do is swap df0 to df1 and vice versa preserving the internal drive.

So I swapped the pins 13 and 14 on the EVEN CIA. It works for Gotek it responds to df0: but the internal drive goes crazy. I even tried changing it's internal jumper to DS1 (Chinon) but no avail. I checked the continuity and the signals are both reversed in both floppy connectors.

The internal drive still performs every track seek and read as it was a mirror from the Gotek. By now I just disconnected it but ultimately I would like to be able to use it. Looks like the drive makes no distiction between DS1 and DS0, placing the jumper on DS2 or DS3 disables it.

So any ideas of what could be wrong ?

Thanks,

RC

[RedskullDC]

Hi RC,

Internal and external drives get a different "Motor On" signal.

Pins 46 and 47 of GARY chip determine whether the _MTR0 signal (internal drive) or _MTRX (external) is activated.

You'll need to switch those as well.

Cheers,
Red

[hooverphonique]

obviously, the select signal going to the internal drive is doing something when it should not be.. do you have access to an oscilloscope?

@redskull: are you sure about that? as far as I remember, the motor latching is implemented in 74-style logic on the A500... Edit: I am wrong, I was thinking of the big box amigas ;-)

[Jope]

Swapping the selects is all you need. Both mechanisms should be jumpered to DS0.

Something else is wrong if the internal drive thinks its select is constantly low.

- broken drive mechanism
- broken drive cable with wires shorting to ground inside the insulation
- faulty gary pulling sel0 down
- faulty cia pulling sel1 (that has been rerouted to sel0) down
- your swap mod shorts sel0 to sel1 together by mistake
- some other short circuit somewhere, perhaps under the motherboard

I did more tests: before the mod the gotek worked fine as df1 connected to the external cable with the df0: internal. So no Gary or shorted cable/board.

Sel0 and Sel1 are not shorted. I checked the continuity, the sel0 reaches the pin 10 on external drive and sel1 reaches the pin10 on internal drive.

Then I swapped both drives: using the gotek as internal and the chinon as external both on ds0 jumpers and they do work, gotek as ds1 chinon as df0. But not the opposite.

Looks like the chinon refuses to work when connected to the internal cable. I even replaced the cable. Also I have a spare A600 drive and the same behaviour.

Could it be the motor signal ? Does gotek need the motor signal as well ? That would explain...

[Jope]

The gotek probably does need a motor signal.

Based on your description in the first post, the internal drive responds to the same commands as are sent to DS1, so it doesn't sound like a motor signal problem to me. Are you sure your internal drive is still jumpered to DS0?

On the subject of the two motor signals, gary generates the separate motor signals by taking in sel0 and mtr from the CIA and then activating mtrx if the motor signal wasn't supposed to go to the internal drive.

[Stedy]

RedskullDC is correct.

You need MTRx + logic to create a MTRx for the real floppy drive configured using DS1, otherwise the drive will only spin when DF0: is active but will not spin when DF1: is selected due to MTR0 (for DF0 only being present on the A500.

I have this in my Amiga drive notes file:
To make DF0:/DF1: switchable with a real drive has a slight complication. You need a Motor ON signal.

Swapping SEL0/SEL1 for a Gotek or HXC drive is easy, that's all they use. The A500/A600/A1200 have SEL0/SEL1 on the internal floppy cable but crucially, only MTR0 (MOTORON on A500). This enable the motor, on drive 0 at the same time as SEL0. I tried connecting two drives to MTR0, nothing worked as they caused bus contention. At least I know why. The A1500/A2000/A3000/A4000 have logic onboard to re-assign DF0:/DF1:

According to the timing diagrams, SELx and MTRx go active at the same time so you may be able to connect SELx to MTRx on the Amiga floppy as well as the SELx input. If not pick off SELx and MRES from the 23 pin external floppy connector, generate MTRx + SELx to create MTR1 for a real floppy and use a long Reset (CTRL-Amiga-Amiga) to switch between Real floppy and emulated. That's the basic idea, it may need some tri-state buffers and a delay circuit (or microcontroller) to control the drive swap.

Ian

I reversed Gary's 46/47 pins with no avail as well. So probably as Stedy said we need a glue logic for that. I'll pass that, I don't want to modify this much.

I got a A508IDE from kipper2k so I will probably reverse the mod then and use WHDLoad.

[Jope]

The gotek here was connected to the external port, wasn't it, hence you don't need to worry about MTRx, it is already going to the right place.

Jope, my problem is the internal chinon as DF1 it does not work when reversing the sel0-sel1.The gotek works fine as df0