Strange behavior in A4000

Computolio · 03 September 2007, 07:54

I have here an Amiga 4000 that seems to work, mostly. The machine boots and runs stuff fine, but any software resets (or crashing) causes the expansion cards to disappear. They'll show up in the boot menu, but don't actually work unless I turn the machine off then on again. Manual resets using CTRL+LAMIGA+RAMIGA work just fine, which is really puzzling.
Another oddity is that the power light never dims. It's supposed to start out dim then quickly go bright on startup, and it's supposed to go from bright to dim whenever the audio filter is toggled.
I'm thinking that this could be caused by the battery - it made an amazing mess of the board when it leaked, and the RTC is dead. Maybe some other circuitry got messed up?
Another thing- the 040 in this machine does not have an FPU (it's an LC040). This results in crashes whenever anything tries to access the FPU. Maybe the lack of an FPU is causing some of the odd reset behavior, or perhaps something else is wrong with the CPU.

Computolio · 13 September 2007, 06:41

I replaced the 040 with one that has an FPU. I've also changed out the riser board, but it's no use. Boards are still disappearing after software resets.

A search on Usnet reveals that at least a couple of other people have had this problem, but nobody was able to solve it. Has anyone here run into this problem?

keropi · 13 September 2007, 07:48

do you have a v11 buster? maybe you have 9 and it causes this prob?

Computolio · 13 September 2007, 08:19

BUSTER is rev. 11. Even if it was a 7, it shouldn't be causing this as I only have Zorro-II cards installed. I've tried several combinations of installed cards and the behavior is the same.

keropi · 13 September 2007, 09:00

strange... I wonder what causes it!

alexh · 13 September 2007, 09:14

Perhaps dodgy kickstart chips. Did you prep the hard drive yourself? Check the startup sequence for anything out of the ordinary.

Toni Wilen · 13 September 2007, 09:28

Random guess: battery acid has damaged at least two traces, trace that goes to cia led/filter io-pin and reset line that connects to expansion connectors (and probably more)

EDIT (was wrong):

Better theory:

hard reset is long (100ms+)
soft reset is short (100us-200us or so)
-> there must be circuitry that keeps reset line active long enough (longer than soft reset) to reset all external devices. This reset circuitry is damaged and it fails to detect soft reset signal from the cpu or its time delay is now too short. (need to check schematics)

Toni Wilen · 14 September 2007, 17:24

A4000 reset circuitry:

U140 (74F32)
U141 (7407)
R120
R121
R122
R125

Computolio · 22 September 2007, 12:39

I managed to get a good look at the chips you mentioned. They're nowhere near the battery, but for some reason they have a noticeable amount of corrosion on the pins. That doesn't even make sense, as they aren't even near any capacitors.

I'd replace them, but I don't have any surface-mount soldering equipment of any kind. I was already thinking of sending the board to AmigaCenter after I send in the other 4000 board I have (which is totally dead), but until that time I guess I'm going to have to make do.

03 September 2007, 07:54	#1
Computolio Registered User Join Date: Apr 2004 Location: California Posts: 174	Strange behavior in A4000 I have here an Amiga 4000 that seems to work, mostly. The machine boots and runs stuff fine, but any software resets (or crashing) causes the expansion cards to disappear. They'll show up in the boot menu, but don't actually work unless I turn the machine off then on again. Manual resets using CTRL+LAMIGA+RAMIGA work just fine, which is really puzzling. Another oddity is that the power light never dims. It's supposed to start out dim then quickly go bright on startup, and it's supposed to go from bright to dim whenever the audio filter is toggled. I'm thinking that this could be caused by the battery - it made an amazing mess of the board when it leaked, and the RTC is dead. Maybe some other circuitry got messed up? Another thing- the 040 in this machine does not have an FPU (it's an LC040). This results in crashes whenever anything tries to access the FPU. Maybe the lack of an FPU is causing some of the odd reset behavior, or perhaps something else is wrong with the CPU. Last edited by Computolio; 03 September 2007 at 21:37.

13 September 2007, 09:28	#7
Toni Wilen WinUAE developer Join Date: Aug 2001 Location: Hämeenlinna/Finland Age: 49 Posts: 26,506	Random guess: battery acid has damaged at least two traces, trace that goes to cia led/filter io-pin and reset line that connects to expansion connectors (and probably more) EDIT (was wrong): Better theory: hard reset is long (100ms+) soft reset is short (100us-200us or so) -> there must be circuitry that keeps reset line active long enough (longer than soft reset) to reset all external devices. This reset circuitry is damaged and it fails to detect soft reset signal from the cpu or its time delay is now too short. (need to check schematics) Last edited by Toni Wilen; 13 September 2007 at 10:33.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
A4000 Motherboard strange behaviour	Marmes	support.Hardware	6	25 December 2014 10:38
Very strange SFS behavior.	Thorham	support.Apps	26	17 October 2009 15:04
Strange hdf behavior [not really a problem].	olesio	support.WinUAE	4	13 August 2009 08:34
discovered strange behavior	NfernalNfluence	support.WinUAE	7	26 May 2009 08:10
A4000: Strange reset	manicx	support.Hardware	5	22 August 2003 19:01

13 September 2007, 06:41	#2
Computolio Registered User Join Date: Apr 2004 Location: California Posts: 174	I replaced the 040 with one that has an FPU. I've also changed out the riser board, but it's no use. Boards are still disappearing after software resets. A search on Usnet reveals that at least a couple of other people have had this problem, but nobody was able to solve it. Has anyone here run into this problem?

13 September 2007, 07:48	#3
keropi . Join Date: Oct 2004 Location: Ioannina/Greece Posts: 5,040	do you have a v11 buster? maybe you have 9 and it causes this prob?

13 September 2007, 08:19	#4
Computolio Registered User Join Date: Apr 2004 Location: California Posts: 174	BUSTER is rev. 11. Even if it was a 7, it shouldn't be causing this as I only have Zorro-II cards installed. I've tried several combinations of installed cards and the behavior is the same.

13 September 2007, 09:00	#5
keropi . Join Date: Oct 2004 Location: Ioannina/Greece Posts: 5,040	strange... I wonder what causes it!

13 September 2007, 09:14	#6
alexh Thalion Webshrine Join Date: Jan 2004 Location: Oxford Posts: 14,339	Perhaps dodgy kickstart chips. Did you prep the hard drive yourself? Check the startup sequence for anything out of the ordinary.

14 September 2007, 17:24	#8
Toni Wilen WinUAE developer Join Date: Aug 2001 Location: Hämeenlinna/Finland Age: 49 Posts: 26,506	A4000 reset circuitry: U140 (74F32) U141 (7407) R120 R121 R122 R125

22 September 2007, 12:39	#9
Computolio Registered User Join Date: Apr 2004 Location: California Posts: 174	I managed to get a good look at the chips you mentioned. They're nowhere near the battery, but for some reason they have a noticeable amount of corrosion on the pins. That doesn't even make sense, as they aren't even near any capacitors. I'd replace them, but I don't have any surface-mount soldering equipment of any kind. I was already thinking of sending the board to AmigaCenter after I send in the other 4000 board I have (which is totally dead), but until that time I guess I'm going to have to make do.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)