3000D SCSI issues

[dreamcast270mhz]

Okay, I got this old Mitsumi SCSI cdrom I tried to use with my 3000D. I plugged the CD rom in, put the centronics to 25-pin cable on the rear SCSI port, then booted it up. The HDD, the only internal device refused to boot. So, I removed the CD-rom and tried to boot the machine again, nothing... My SCSI cable has a terminator soldered onto the end of it. Now, I have reason to suspect it screwed something up. I have the prototype WD chip, buster rev7, dmac-02, I think those have everything to do with the SCSI. i even read the manual of the cd-rom, it apparently shipped with an 8bit ISA card and was for MS-DOS. The manual says nothing about SCSI or termination, merely talks about DMA and IRQ and stuff like that. That has led me to believe this is a parallel CD-rom, and it screwed something up. The computer now turns on and just sits there, I can't get to the early bootup screen (maybe because I rely on the AMBER?)

[shaf]

Hello,

I think there are Terminators at the External SCSI Connector that need to be renoved if connecting an external device. Anso ensure that the SCSI ID is different for each SCSI Device, I believe the A3000 Internal ScSI Device defaulted to SCSI ID 6.

I don't have my CBM Service Tech notes handy but will try to verify when I find them.

Shaf

[Zac67]

You need to remove the motherboard terminators when you connect devices to both internal and external connectors. The last external device has to be terminated.

[Castellen]

Entirely possible it is a parallel port CDROM drive, there were a few of those kind of things once upon a time.

If you plug a parallel port device into the A3000 SCSI connector (it's the same connector type for both ports) it usually shorts pin 25 to ground (5V termination power) which exceeds the series diode's foward current, causing it to go short circuit, causing a track on the bottom side of the PCB to fuse.

Afterwards there will be no termination power supplied by the Amiga to the SCSI bus and it'll refuse to boot in the same way as you've described.

Check this by measuring for 5V on pin 25 of the A3000 ext SCSI connector.

The diode is D800 and lives next to the SCSI port.  The original is a 1N4148 or something, you can replace it with a standard 1N400? or any other (non-zener) diode you come across.

You'll need to take out the main board and repair the PCB track underneath, should be obvious where it's fused.  You'll need to put a small wire from D800 to pin 25 of the connector.

Oh, and on some A3000 boards the silk screen detail for D800 was reversed.  Obviously the cathode (stripe end) connects to pin 25 and the anode to +5V.

Double check the details at your own leisure, I could look it up in the service manual, but I'm too lazy to walk to the shelf to grab it.  If you can't work it out, let me know and I'll get off my arse...

[chiark]

I do wish people would put a link to all the threads they post on each Amiga board

Brilliant, thanks for the info Anthony: I dug out the schematics and posted the following on amibay:
Pin 25 is connected via a diode (d800) to +5v - that could have gone.
Pins 8, 10, 11, 12, 21,22 and 23 are connectedt through pull up resistors in RP802 (220 or 330 ohm) to +5v
Each pin on the connector is connected through a ferrite bead (EMI1800-1817)

The schematics say the diode is a 1N914, for what it's worth...

[tone007]

The computer now turns on and just sits there, I can't get to the early bootup screen (maybe because I rely on the AMBER?)

You'd be able to get the early startup menu using the 3000's scandoubler.

[Castellen]

@tone007
As I recall, you still can't get to the early startup menu with the A3000 in this state.  Essentially the SCSI controller enters an endless loop or something which causes the entire machine to hang during boot.

@chiar
Yep, close enough.  Most of the schematics seem to be permanently imprinted in my brain, suppose that's what happens when you work on Amigas for a living.

The resistor network you mention is one section of the passive termination network, there should be three of the same resistor network modules.  Each line in the SCSI bus is connected to the junction of 220 and 330 Ohm series resistors between 5V and GND; the aim being to bias the line at 3V and provide an AC termination of approx 130 Ohms per line.

[tone007]

@tone007
As I recall, you still can't get to the early startup menu with the A3000 in this state.  Essentially the SCSI controller enters an endless loop or something which causes the entire machine to hang during boot.

..right, I should've added if the 3000 were working properly.

This reminds me of when I was 12 and tried to stick a 64K Apple II memory board into an IBM XT.  The slot was the same size!  Seriously, though, I can't imagine confusing a parallel CDROM for SCSI one, the lack of SCSI ID selection on the drive would be a dead giveaway.

[dreamcast270mhz]

Well i'm a moron for forgetting the SCSI ID and termination pack, Cause it's requiring all my strength not to bust the POS Mitsumi drive.

[dreamcast270mhz]

I repaired the track with some 22 gauge wire, i also took the opportunity to fix my 3640. So, now all i need is that diode, time to run up to RS. Anthony, you were right, the track was burned right up, so the PSU is okay. I'll see if I can't get her working again.

[chiark]

@Castellan: thanks for the info.  I didn't see/realise that it was used as a voltage divider: makes sense now I look at it again with that frame of mind

@dc270: excellent, if it's just that then that's really not too bad, especially as you've got the skills to fix it yourself.