A590 issues

[timmc]

Hi there, I have a A500 rev 6a and a A500 rev 5, both US-NTSC. The A590 I have works with the rev 6a board but it prevents the Rev 5 board from powering up. I modified the A590 to pull power from the A500.

I can power on with the rev 6a (which also has a PiStorm installed) and I see the 2 MB of RAM on the A590. On the Rev 5, the power light on the A590 comes on, I hear the floppy drive make a noise, but no lights from the A500 keyboard. This is using the same power supply (I only have one working one that is currently using a MeanWell supply).

I tried the original A590 4.6 ROMs as well as 7.0 ones I normally use but no difference. The Rev 5 had a WB1.2 ROM but I replaced it with a burned 1.3 ROM. I haven't tried switching ROMs on the Rev 5 yet.

I ordered parts for the GottaGoFast RAM and IDE68k so I will probably just end up using those. Curious about this issue though.

Is there some incompatibility with the REv5 boards and the A590?

Thanks!

[Matt_H]

Can you run the 590 off its own power supply? The Rev 5 is more power hungry, so there's a possibility that's the issue.

Is your burned 1.3 ROM confirmed 100% good? If you're using 1.2 you need to set the disable switch on the 590 and load the driver from disk due to the Autoconfig bug in 1.2.

[thebajaguy]

The A590, particularly with a mechanical HD in it, was NEVER intended to be powered by the original A500 PSU. The load on the 12V rail is too much, and the 5V will likely be marginal on a stock A500 PSU. You would need an upgraded PSU to attempt it, and stick to replacement SCSI2SD or similar 5V-only drive replacement types. I would also measure the 5V line at the connector with the entire load on - it may be <4.8V and that can cause marginal operation issues. Additional DIP memory in the A590 can add to the marginal voltage levels.

One item to note: An old dried-out TH cap in the A500 reset circuit can make power-on fail if all of the gear is on the A500 PSU. The cap, now out of spec, releases the _RESET line to quickly, and the peripheral was not ready to be talked to by the 68000.

Internal peripheral expansions, and additional external modules off the ports are also power-thieves. Consider them in the load calculation.

[timmc]

Originally Posted by Matt_H

Can you run the 590 off its own power supply? The Rev 5 is more power hungry, so there's a possibility that's the issue.

I would like to but I don't have the power supply. I was thinking of just making one myself but wasn't 100% sure what to do with the power sense line. I have DIN4 cable and was just going to connect it to +12v and +5v. From the other post, it sounds like I may need to try it on its own supply.

Originally Posted by Matt_H

Is your burned 1.3 ROM confirmed 100% good? If you're using 1.2 you need to set the disable switch on the 590 and load the driver from disk due to the Autoconfig bug in 1.2.

I am wondering if the ROM is an issue. AmigaTestKit can't detect the ROM version but it boots it as well as Workbench without issue.

One point I forgot to mention was that I am only testing it with 2MB of RAM, no HD so far. I have a ZuluSCSI RP2040 I want to use with it.

[timmc]

Originally Posted by thebajaguy

The A590, particularly with a mechanical HD in it, was NEVER intended to be powered by the original A500 PSU. The load on the 12V rail is too much, and the 5V will likely be marginal on a stock A500 PSU. You would need an upgraded PSU to attempt it, and stick to replacement SCSI2SD or similar 5V-only drive replacement types. I would also measure the 5V line at the connector with the entire load on - it may be <4.8V and that can cause marginal operation issues. Additional DIP memory in the A590 can add to the marginal voltage levels.

One item to note: An old dried-out TH cap in the A500 reset circuit can make power-on fail if all of the gear is on the A500 PSU. The cap, now out of spec, releases the _RESET line to quickly, and the peripheral was not ready to be talked to by the 68000.

Internal peripheral expansions, and additional external modules off the ports are also power-thieves. Consider them in the load calculation.

Thanks for the reply. I forgot to mention I am not currently using it with a HD, just 2MB of RAM. I would like to use a ZuluSCSI in it however. Think I may do some more research and cobble together a separate power supply.

[pipper]

I can confirm that the A590 works with the ZuluSCSI, 2MB fast and Kick 1.3 without external PSU. I measured quite low voltage getting to the ZuluSCSI (barely 4.7v), but it is functional. Sysspeed claims to read at 1.4mb/s with a stock 68000 in this system (Rev5 mobo)

[LIV2]

Originally Posted by timmc

I am wondering if the ROM is an issue. AmigaTestKit can't detect the ROM version but it boots it as well as Workbench without issue.

Does the insert disk screen say "1.3" ? if not it's 1.2 which as stated above does not support autoboot.

[thebajaguy]

Originally Posted by LIV2

Does the insert disk screen say "1.3" ? if not it's 1.2 which as stated above does not support autoboot.

Actually, Kickstart 1.2 will cause any AutoConfig peripheral with a ROM to crash (LED flash) at boot time. The reason is C= goofed:

"Kickstart 1.2 (v33.180) did not have autoboot capability. Although it contained AutoConfigTM code, it was broken due to a simple bug: the autoconfig init code jsr had a wrong a6 register (the base register at the jsr time should be expansionbase but is execbase instead)."

[patrik]

Would patching this error in a 1.2 ROM make it work with autoconfig, or are there more errors later in the chain?

[Jope]

If you look at the blizkick sourcecode, above the label .not_33180, that should be all that is needed.

[Dale423]

Originally Posted by timmc

I would like to but I don't have the power supply. I was thinking of just making one myself but wasn't 100% sure what to do with the power sense line. I have DIN4 cable and was just going to connect it to +12v and +5v. From the other post, it sounds like I may need to try it on its own supply.

I have the PSU below. It powers the 500 and 590 fine without mods.

Check it out here.

[Mikeywikey]

The two kickstart issues with rev5 (and rev3) are the “oops” bug which prevents 512k kickstarts working (obviously not an issue for 1.3) and lack of pulldown for EPROMs, this can make EPROMs unstable, mask ROMs are fine - the east fix is to use a kickstart adaptor board which fixes the 512k issue and adds pulldown resistors.
No idea if the pull down issue is causing you a problem, but it can be the source of unknown instability - just another thing to think about on top of out of spec power!

[levo-ace]

I have what must be one of the rarest things on the planet right now , A FULLY WORKING A590 PSU!!
I have opened it up and plan to make a circuit diagram for everybody to reference.

The 5v 'sense' line actually drives a 5v relay that can switch 8A to act as a power switch on the supply line in the psu .

i.e. the 5v sense switches the mains to the A590 power supply .

easy enough to do with a 5v relay , a fuse and a meanwell psu!!

mains---fuse----relay----psu

[levo-ace]

The specified rating is 5v @ 3A and 12v @ 0.85A

The 12v comes from a 78S12 regulator and the 5v is switch mode controlled by a SG3524BN ic.

And a nice chunky transformer as well!!

[timmc]

Ah ok, that power supply sounds interesting. I would be curious to see the schematic.

In terms of the Rev 5 A500 and the a590, it was the Kickstart ROM that I burned. It powered on with a 1.3 message and seemed to work OK but AmigaTestKit couldn't determine the version. I swapped in a 1.3 mask ROM from my 6a A500 and the Rev5 and a590 worked great.

Guess I am still having issues burning a Kickstart ROM correctly...

Thanks to all for the help.

[levo-ace]

It is on my website
https://www.levosretrocomputerprojec...amiga/a590-psu