Welcome, Guest. Please login or register.

Author Topic: A3640 rev. instabillity  (Read 2282 times)

Description:

0 Members and 1 Guest are viewing this topic.

Offline JohnFanteTopic starter

  • Full Member
  • ***
  • Join Date: Aug 2004
  • Posts: 162
    • Show only replies by JohnFante
A3640 rev. instabillity
« on: December 07, 2004, 07:38:52 AM »
I have an A4000 with an A3640 rev. 3 card. I recently bought a Voodoo 3 card for my Mediator and a KickFlash 4.0.

Before the new cards the machine was rock solid, but after i got the graphics card and the Kickflash installed it has begun to get a bit unstable.

At first it bagan to rest radomly, but that has disappeard after i got the screenmodes defined better.

Now one of the major problems is that when I change back and forth between PAL mode and my high res workbench screen the later corupts after a few switches and i have to reset.

Another problem is that when i reset after one of theese things the machine resets again and again. And the only solution is to take the Kickflash out.

I have the latest firmware in the Kickflash (1.6) and all the latest Mediator drivers. I use 3.9 BB2.

I have read alot about the A3640 rev 3 being unstable with graphics cards etc. and i was wondering if anybody else has had a similar problem and if there is a solution.

Thansk in advance.

Booooiiiing!!!!
 

Offline Castellen

Re: A3640 rev. instabillity
« Reply #1 on: December 07, 2004, 07:54:10 AM »
The info here is probably what you're looking for:
http://www.amiga-hardware.com/a3640.html
 

Offline Effy

  • Hero Member
  • *****
  • Join Date: Aug 2002
  • Posts: 2053
    • Show only replies by Effy
Re: A3640 rev. instabillity
« Reply #2 on: December 07, 2004, 08:23:39 AM »
Assuming he uses V3.0 ...

"V3.0

This revision will not work in A3000's at all (unknown if this also applies to the A3000T and A4000T) because of a conflict with the DMA handling of the onboard SCSI controller and some Zorro cards. Although it will work with A4000's. A bug in the design allows the board to grant DMA access to more than one device, causing the machine to hang. There is also a problem with the STERM* signal that can cause sporadic crashes with some cards such as the Cybervison 64. The PAL or GAL chip located at U209 should be -01.

V3.1

This card works fine in general however it does have a bug related to Zorro II DMA that causes some cards like the GVP Phonepak to stop working, when installed in an A3000. The PAL or GAL chip located at U209 should be -02.

V3.2

No known problems. The PAL or GAL chip located at U209 should be -03."


Offline Brian

  • Hero Member
  • *****
  • Join Date: Mar 2003
  • Posts: 1604
    • Show only replies by Brian
    • http://www.syntaxsociety.se
Re: A3640 rev. instabillity
« Reply #3 on: December 07, 2004, 08:35:25 AM »
I think it sais it all... though just to add to the horror check this link.

Offline Effy

  • Hero Member
  • *****
  • Join Date: Aug 2002
  • Posts: 2053
    • Show only replies by Effy
Re: A3640 rev. instabillity
« Reply #4 on: December 07, 2004, 08:46:42 AM »
"Apparently, C105, C106 & C107 were added as an afterthought. I can only theorize, but I'll bet that the engineer in charge of the A3640 passed the schematics by his senior engineer for approval. The senior engineer may have then suggested that a few more electrolytic capacitors be added for good measure."

Brain : that is indeed stuff for a horror tale  :-o

Offline JohnFanteTopic starter

  • Full Member
  • ***
  • Join Date: Aug 2004
  • Posts: 162
    • Show only replies by JohnFante
Re: A3640 rev. instabillity
« Reply #5 on: December 07, 2004, 10:45:19 AM »
Thanks for the replys

I had seen the info om amiga-hardware. That was what triggered my suspision.

Is it worth the effort to try to manually upgrade my rev 3 to a 3.2? I am not an electronics wiz at all.

If there is anybody out there that are able to do the upgrade I would be glad to make it worth their while.  :-)

Thanks in advance.
Booooiiiing!!!!
 

Offline patrik

Re: A3640 rev. instabillity
« Reply #6 on: December 07, 2004, 11:44:04 AM »
@JohnFante:

There is some more information on this page.


/Patrik
 

Offline JohnFanteTopic starter

  • Full Member
  • ***
  • Join Date: Aug 2004
  • Posts: 162
    • Show only replies by JohnFante
Re: A3640 rev. instabillity
« Reply #7 on: December 07, 2004, 12:20:43 PM »
@Patrik

Thanks for the link.

The problem is that I have neither the eqiupment nor soildering eksperience to do the hack.

If anybody else have done the hack with success I would be more than happy to pay that person to fix mine also. :-)




Booooiiiing!!!!
 

Offline Framiga

  • Hero Member
  • *****
  • Join Date: May 2003
  • Posts: 4096
    • Show only replies by Framiga
Re: A3640 rev. instabillity
« Reply #8 on: December 07, 2004, 01:56:08 PM »
@JohnFante

all those in the original A4000D case? and its original 145Watt PSU?

 

Offline JohnFanteTopic starter

  • Full Member
  • ***
  • Join Date: Aug 2004
  • Posts: 162
    • Show only replies by JohnFante
Re: A3640 rev. instabillity
« Reply #9 on: December 07, 2004, 04:20:22 PM »
My specs are:

A4000D/40
A3640 acc.
Maks RAM.
Creative CD/DVD drive
Internal and external floppydrive.
Mediator PCI 4000Di
Voodoo 3 3000
Internal ScandooblerFlickerfixer
SoundBlaster PCI 128
D-link DFE530TX Ethernet
KickFlash 4.0
Cocolino mouseadaptor

The case is closed at the moment, but as far as i remember it is an Lite-On powersupply.

The only thing i can spare is the external floppy. So i hope this is'nt a power problem. :-(

EDIT: My powersupply is an LiteOn PE-3151 that gives 151.6W
Booooiiiing!!!!
 

Offline yogisumo

  • Sr. Member
  • ****
  • Join Date: Apr 2004
  • Posts: 255
    • Show only replies by yogisumo
Re: A3640 rev. instabillity
« Reply #10 on: December 07, 2004, 06:33:38 PM »
I've had all 3 revs.  My original 3.0 would gradually slow down as system got warmer ( only with cards in machine ) and then lock up completely.  I now have a 3.1 and a 3.2.  I had Joe Archibald at www.archtech.ca change the backwards caps on my 3.1.  My 3.2 came out of an A4000T.

The cut and jumper part of the upgrade is not that difficult I think but I'm not sure about the pal upgrade.  You could try to contact Archtech or www.mrhardwarecomputers.com.  Unfortunately for you, they're in USA and Canada.
 

Offline JohnFanteTopic starter

  • Full Member
  • ***
  • Join Date: Aug 2004
  • Posts: 162
    • Show only replies by JohnFante
Re: A3640 rev. instabillity
« Reply #11 on: December 07, 2004, 07:30:42 PM »
I see what you meen about the rev 3.0 :-(

When the system is cold everything works fine. But when it gets warm the instabillity comes.

Will it help to put and a fan inside?

Booooiiiing!!!!
 

Offline K7HTH

  • Hero Member
  • *****
  • Join Date: Sep 2004
  • Posts: 573
    • Show only replies by K7HTH
    • http://NewPeopleString.com
Re: A3640 rev. instabillity
« Reply #12 on: December 08, 2004, 12:01:14 AM »
Quote

JohnFante wrote:
I see what you meen about the rev 3.0 :-(

When the system is cold everything works fine. But when it gets warm the instabillity comes.

Will it help to put and a fan inside?



Fans are ALWAYS good. The cooler your electronics stay the better they perform.
Own the Web... Go PeopleString! It\'s Huge!!
 

Offline yogisumo

  • Sr. Member
  • ****
  • Join Date: Apr 2004
  • Posts: 255
    • Show only replies by yogisumo
Re: A3640 rev. instabillity
« Reply #13 on: December 08, 2004, 12:09:17 AM »
Quote

JohnFante wrote:
I see what you meen about the rev 3.0 :-(

When the system is cold everything works fine. But when it gets warm the instabillity comes.

Will it help to put and a fan inside?



What I did is open the case and put a fan blowing in over top of the case for testing.  At least one of the problems with the 3.0 revision is the sampling of the STerm signal on the bus.  As the system gets warmer, electrical resistance goes up which causes the bus problems.  The 040 is not overheating, but the heat exposes a marginal signal.  It's probably easier to find an accelerator or another a3640 than to try to fix that one.  At least, all you would need is the board and not the cpu....

 

Offline JohnFanteTopic starter

  • Full Member
  • ***
  • Join Date: Aug 2004
  • Posts: 162
    • Show only replies by JohnFante
Re: A3640 rev. instabillity
« Reply #14 on: December 08, 2004, 06:49:01 AM »
Thanks for alle the advice. It seems thou that the problem was not my A3640 (at least not yet :-))but the order of my PCI cards. I changed the order and the problem disappeard.

From bottom up the order is:

Emty slot
Voodoo 3 3000
Soundblaster 128
D-link 530TX ethernet.

DoctorQ from DK (who helped me a lot :-) described the problem as DMA related. I think that i will eventually install a fan just in case.

No my only problem is that i can't get eternet card up an running. I will make a seperate post about that.

Once again thanks for all the help.
Booooiiiing!!!!