4xEIDE99 trouble

[rogerxy]

Hi
I'm not a very advanced Amiga user so far, but I bought a 4xEIDE to use 2 CF-cards with my A1200. One CF card for the OS and one CF card in the back of the machine to move files.

When I've inserted the 4xEIDE and the OS-card on the first port of the flat cable it all works fine and the OS boots (photo 1 and photo 2).

However, when I add the CF-adapter and the empty CF-card to the end of the flat cable (photo 3) and start up it doesn't boot at all. I've tried to move jumpers back and forth on both the CF-adapter for the OS CF card and the CF-adapter for the empty CF card but it never boots. Sometimes the screen is black and sometimes it just reboots again and again. I've also tried to remove the empty CF card from the adapter but no changes.

Any idea what the problem might be here?

Thank you in advance

[Boing-ball]

The issue maybe down to the adapters not supporting Master and Slave settings.
Note that the A1200 has no settings for master and slave. Ideally the 44 pin built in IDE connector is only designed to support 1 x HDD.

The 4xIDE adapter is a Buffered 4 way IDE card. But again only 1 x Master can boot from this. The rest of the ports get activated later on by the use of Software drivers from the drivers disk.

As for file transfers, why not use the PCMCIA port with a PCMCIA adapter to CF card? Or are you wanting to use tha for something else?

[kolla]

Note that the A1200 has no settings for master and slave. Ideally the 44 pin built in IDE connector is only designed to support 1 x HDD.

It was designed to be compatible with the IDE standard, and that standard allow for master and slave.
Back in 1994 I had one 540MB Conner disk as master and a Seagate 1.2GB as slave, a few years later I had replaced the disks with larger ones and also added CD drive, going beyond 2 drives.

The 4xIDE adapter is a Buffered 4 way IDE card. But again only 1 x Master can boot from this. The rest of the ports get activated later on by the use of Software drivers from the drivers disk.

Whether a disk/device is master or slave is irrelevant for what the system boots from, it can just as easily boot from a slave device and out of the box A1200 IDE supports 2 drives, not just one!

IDE is an 8 bit bus, but the A1200 IDE port is 16bit wide, that’s why these 4x IDE cables/adapters work, they give you a second 8bit IDE for another master and slave, 4 devices in total. Buffering is a bonus, not strictly needed but noce to have. Activation of the second IDE bus can be done with IDEFix or, since OS 3.9, using a dummy kickstart module named “ATAPIismajik” which scsi.device looks for. If you burn your own kickstart with sufficiently new scsi.device (v43 something) and this module, the A1200 IDE will support 4 drives from coldboot.
https://aminet.net/package/driver/media/AtapiMagic

[kolla]

Any idea what the problem might be here?

Many IDE/CF-adapters do not have jumper for master/slave setting, and by default operate as master.

However, it appears that yours do have a MASTER/SLAVE jumper, so the one set as master goes to the end of the cable and the slave is attached on the middle connector of the cable. You can try this without the 4xIDE thing, just attach the cable directly to the A1200 IDE. If this doesn't work, there's something wrong with either the CF-adapters or the cable (neither very likely). If it works, try again with the 4xIDE, again with both CF-adapters on the same cable and make sure to use the primary port.

A second option is to use the other secondary port of the 4xIDE for one of the adapters, with both cf-adapters set as master.  But then you either need to load the "ATAPIismajik" rom module described above, or use IDEFIx, to activate the second adapter.

As mentioned, many CF-adapters don't have slave/master jumper, as was also the case for many IDE devices at the end of the IDE/PATA era. The trick is that by this time, one had figured something called Cable Select (CS) where devices set themselves to slave when located in the middle of the cable. Nothing had to be done on the IDE controller for this, just a special "cable select" cable and at least one CS-aware device that can set itself to slave, so his also works with A600/A1200.

[Boing-ball]

A second option is to use the other secondary port of the 4xIDE for one of the adapters, with both cf-adapters set as master.  But then you either need to load the "ATAPIismajik" rom module described above, or use IDEFIx, to activate the second adapter.

Just to add here, be careful when using IDEFIX’97 as this does not play nice with AmigaOS 3.2. You can use it to say for adding in CD ROMs, but do not allow the change to the Startup-sequence.

Atapimagic is the preferred tool for AmigaOS3.2.

[Boing-ball]

It was designed to be compatible with the IDE standard, and that standard allow for master and slave.
Back in 1994 I had one 540MB Conner disk as master and a Seagate 1.2GB as slave, a few years later I had replaced the disks with larger ones and also added CD drive, going beyond 2 drives.Whether a disk/device is master or slave is irrelevant for what the system boots from, it can just as easily boot from a slave device and out of the box A1200 IDE supports 2 drives, not just one!

IDE is an 8 bit bus, but the A1200 IDE port is 16bit wide, that’s why these 4x IDE cables/adapters work, they give you a second 8bit IDE for another master and slave, 4 devices in total. Buffering is a bonus, not strictly needed but noce to have. Activation of the second IDE bus can be done with IDEFix or, since OS 3.9, using a dummy kickstart module named “ATAPIismajik” which scsi.device looks for. If you burn your own kickstart with sufficiently new scsi.device (v43 something) and this module, the A1200 IDE will support 4 drives from coldboot.
https://aminet.net/package/driver/media/AtapiMagic

Thanks Shelly.

Don’t forget that some people’s experiences can differ from yours. Have owned a few of the Elbox 4-way adapters and never did I get a device to boot from the slave setting. It was always on the primary master. This was using both IDEFIX’97 and AtapiMagic.

On the A4000 I have managed to get ATAPI/IDE Zip drives to boot in the slave setting. But never on a A1200. Perhaps this maybe down to revision of the board.

[rogerxy]

Thank you all for your replies, really appreciate it!

[kolla]

But never on a A1200. Perhaps this maybe down to revision of the board.

There's one thing about this.

The "original" A1200 kickstarts (3.0 and 3.1) tend to not wait long enough for devices to spin up and also spin them down on every reset. The drives are spun up one after the other starting with unit 0, then unit 1 etc, so in case of IDE, master first, then slave (and I'm sure a zip drive needs a bit more time than a CF card). The solution is to either cut pin 1 so that they don't spin down on reset and just accept that you need to ctrl-a-a on power up, or use a later kickstart that waits longer (3.2 waits very long).

A600 kickstarts typically don't have this problem, as that kickstart is also for A500 where slow SCSI drives were commonplace.

[kolla]

Just want to add that many years ago, I put together this IDE riser for two devices, with cable select, which I have used in various (A600, A1200, SX32) to easily copy between SD cards, CF cards, SSDs, old hard drives etc
(The name is homage to the original KA 47)

https://oshpark.com/shared_projects/M6rM2c0D

[F0LLETT]

Thanks Shelly.

Don’t forget that some people’s experiences can differ from yours. Have owned a few of the Elbox 4-way adapters and never did I get a device to boot from the slave setting. It was always on the primary master. This was using both IDEFIX’97 and AtapiMagic.

On the A4000 I have managed to get ATAPI/IDE Zip drives to boot in the slave setting. But never on a A1200. Perhaps this maybe down to revision of the board.

IDEFIX sucks and besides newer OS doesn't need it, as you have updated scsi and filesystem, so atapi is majik is enough to get extra channels seen.

On side note, please chill out on the Shelly business. Me and Kolla might have disagreements, but I would not go around calling him things. Its not the way things are done in civilized society.

There's one thing about this.

The "original" A1200 kickstarts (3.0 and 3.1) tend to not wait long enough for devices to spin up and also spin them down on every reset. The drives are spun up one after the other starting with unit 0, then unit 1 etc, so in case of IDE, master first, then slave (and I'm sure a zip drive needs a bit more time than a CF card). The solution is to either cut pin 1 so that they don't spin down on reset and just accept that you need to ctrl-a-a on power up, or use a later kickstart that waits longer (3.2 waits very long).

A600 kickstarts typically don't have this problem, as that kickstart is also for A500 where slow SCSI drives were commonplace.

3.0 Maybe, but 3.1 waits about 25 to 30 seconds for drive.