Welcome, Guest. Please login or register.

Author Topic: A3000 D Schematics Rev 9 Board  (Read 1745 times)

Description:

0 Members and 1 Guest are viewing this topic.

Offline orange

  • Hero Member
  • *****
  • Join Date: Dec 2003
  • Posts: 2794
    • Show only replies by orange
Re: A3000 D Schematics Rev 9 Board
« Reply #14 from previous page: July 04, 2005, 12:05:24 PM »
@Castellen

I have a lot of green sockets in one dead a2000. But it worked fine. That green stuff is VERY hard to remove. I've tried replacing sockets (68K and ROM) with no luck.

BTW, I'd say that 'reseating' chips might not be a good idea when there are green pins as it might actually loose contact.

Do you have any solution for those sockets other than replacing them with new ones?
Better sorry than worry.
 

Offline melangeTopic starter

  • Full Member
  • ***
  • Join Date: Jun 2005
  • Posts: 235
    • Show only replies by melange
Re: A3000 D Schematics Rev 9 Board
« Reply #15 on: July 05, 2005, 04:54:15 AM »
Thanks again...

@Castellen

I have begun carfully checking all traces from chips located in the acid damaged area.  So far I have checked every trace from paula and some from denise.  I found one trace out from U480-P6 to U470, U471, U472 - P2.  Didn't change a thing.

I should point out that this whole problem started with a machine that still booted up for a while before it crashed, but the display output was messy(maybe sync problem).

I tried swapping out chips as a first attempt to see if that made any difference and it sure did, I broke two legs of the Agnus trying to get it out.  This infuriated me because I have removed agnus chips many times and never done anything like that.  I thought (bugger) and decided to yank the one from the A500 (also 2 meg chip agnus) and promptly broke 2 legs off that as well

I then painfully resoldered the legs onto the A500 chip and reinserted it into the A500. It still works

Also did the same to the A3000 chip and I have no Idea if that works because all I get is a really unstable screen with green wavy lines and a flashing power led.  (which is exactly what I had b4 I removed it BTW)

I'm going to stick with this for quite a while yet.  Thanks for your help, back to the schematics

 :-(  :-(  :-(
2 x A1200 Blizzard IV 030@50Mhz
1 x A3000D 030@25Mhz
1 x A2000 030@25Mhz
1 x A2000, 2 x A600
2 x A500
 

Offline melangeTopic starter

  • Full Member
  • ***
  • Join Date: Jun 2005
  • Posts: 235
    • Show only replies by melange
Re: A3000 D Schematics Rev 9 Board
« Reply #16 on: July 05, 2005, 05:07:13 AM »
Oh, I forgot one thing.  I was having a bad day I should have stopped.  I also inserted yet another Agnus got interupted by my 4 year old son, and inserted it the wrong way and that one is now Dead.  Having done that I have to wonder if I have killed other components on the board.  DAMN!!!  i could be wasting my time trying to find dead traces when I may have killed something else.  I guess i can only keep at it for a while and hopefully figure out what I have done wrong!

 :pissed:  :pissed:
2 x A1200 Blizzard IV 030@50Mhz
1 x A3000D 030@25Mhz
1 x A2000 030@25Mhz
1 x A2000, 2 x A600
2 x A500
 

Offline melangeTopic starter

  • Full Member
  • ***
  • Join Date: Jun 2005
  • Posts: 235
    • Show only replies by melange
Re: A3000 D Schematics Rev 9 Board
« Reply #17 on: July 05, 2005, 07:58:06 AM »
Nope got that wrong as well.  After looking at the datasheet for the U480 chip I figured out that it's wasn't pin 6 but pin 5.  there is no problem with that connection so I have found nothing

The schematic is a little hard to read, some numbers are too small to read properly.

Doom and despair hangs over me :-(
2 x A1200 Blizzard IV 030@50Mhz
1 x A3000D 030@25Mhz
1 x A2000 030@25Mhz
1 x A2000, 2 x A600
2 x A500
 

Offline Castellen

Re: A3000 D Schematics Rev 9 Board
« Reply #18 on: July 05, 2005, 12:40:19 PM »
@orange
Yes, the corrosion is basically impossible to remove from inside the socket contacts.  By attempting to clean it, you only stretch the socket contacts, making things even worse.

The only thing better than replacing the sockets with new ones, is removing the sockets and soldering the ICs directly into the board.
Those sockets are truley hideous, just a general cause of problems all round, even without leaking batteries to make things worse.

My advice is that if any socket contacts have any hint of green in them, the socket needs to be replaced.
It may work fine today, but start causing intermittent problems tomorrow.

There's an easy technique for removing those sockets and minimising risk of damage to the PCB.
I think I've explained it in another thread somewhere, but can re-iterate if you like.


@melange
Fault finding by measuring continuity like you're trying to do can work, but as you'd have already noticed, is very time consuming.
If you want to do it that way, you need to measure from the leg on the IC body, to another known good contact elsewhere.
That can give you incorrect readings, as when you press down on the IC leg to measure it, it presses the leg into the socket, perhaps making temporary contact just while you're measuring it.

A much faster way to fault find is to run the machine and check for signal activity on appropriate IC legs using an oscilloscope.  This is the method I usually use for finding such problems.
The problems there is you need an oscilloscope, and you need a good idea of what you're looking for to begin with on each signal line.

But for your problem, the best approach would be to replace sockets which are corroded and in the process, carefully check continuity on tracks which are suspect.  As in tracks which might be corroded, ones which run through the RTC area and especially tracks which run to vias near the RTC area.

I drew a crappy wee diagram which explains via corrosion and how to repair them:
http://amiga.serveftp.net/images/ViaRepair.jpg

This is more of a problem in A4000, though it's worth being aware of it.  The A3000's PCB uses a slightly different method to track between layers which is more robust to corrosion.

And just how did you manage to damage so many IC legs? :-P
There are tools for removing DIL devices such as those, which are cheap and work well.  Shaped like a large pair of tweesers with 90° hooks at the ends.
 

Offline melangeTopic starter

  • Full Member
  • ***
  • Join Date: Jun 2005
  • Posts: 235
    • Show only replies by melange
Re: A3000 D Schematics Rev 9 Board
« Reply #19 on: July 05, 2005, 04:22:41 PM »
@Castellen

measuring continuity is the only way I know how to approach the problem.  Yes, it is real time consuming, however I think this machine is worth the effort.  As you can tell my electronics knowledge is a little "simplistic" to say the least, however over the years I have managed to repair a multitude of devices with my simplistic knowledge.

I have done things like replace flyback transformers in TV's and Monitors for instance.  Recently I had a cap blow in a 17" Monitor which took a string of components out with it when it went, including a chip, transistors and resistors.  I managed to repair this, but not by my own fault finding.  I found technicians accounts of what had been replaced on the internet, because it was a common problem with that monitor type and I simply ordered all the parts listed and replaced them all and it works.  I have built a ton of kits over the years including a SCRAM (SCSI/RAM) interface for the A500 which worked perfectly first time.

So I'm not a complete loon, but I can get out of my depth sometimes, and I think I may be getting there now :lol:

I have also replaced sockets on Amiga Mobo's with success in the past, but I would like to avoid that ATM if I can.  There is some quite nasty corrosion in the Denise socket, but I have checked for continuity on the most suspect pin to the underside of the board and to the chips the traces branch off to and it reads fine.

As for removal of the Agnus, well that was just sheer stupidity on my behalf and a small element of impatience.  Yes I should have a tool to remove it, but I remembered trying to use one of those in the past (long time ago) and I couldn't seem to get the chip out. Maybe it was a crap tool.  My A500 Mobo had some holes beneath the socket which made it quite easy to get it out with a small push from beneath.

  I am going to have to invest in some better tools.
The lack of quality gear is also causing me some problems ATM.  My good Weller soldering station blew it's heater element and I am currently using a real cheap and nasty iron which has hampered my ability to solder anything.  And taken me a lot longer to do things I could do easily with the other iron.  For instance putting the legs back on the Agnus took me about an hour to get it right for each of the two chips. I wanted to be sure they were on well enough to not come off when I inserted it into the socket.

I wish I had and knew how to utilise an oscilloscope.  I have always been of the understanding that you needed a service manual that outlined what type of signal should be seen on the oscilloscope at certain test points.  Guess I should get some rudimentary electronics books and try and up my skills a little.  :-D

Checking for suspect traces is hard when the corrosion is masking the area so badly that I'm not sure if there is/was a trace there at all.  That is why I have resorted to the schematic to study the RTC area closer and try and make sure I'm not missing anything.

Thanks again for your help and I'll keep you informed on my progress or lack thereof  :lol:
2 x A1200 Blizzard IV 030@50Mhz
1 x A3000D 030@25Mhz
1 x A2000 030@25Mhz
1 x A2000, 2 x A600
2 x A500
 

Offline melangeTopic starter

  • Full Member
  • ***
  • Join Date: Jun 2005
  • Posts: 235
    • Show only replies by melange
Re: A3000 D Schematics Rev 9 Board
« Reply #20 on: July 05, 2005, 04:43:14 PM »
@Castellen

Is it likely/possible in your opinion that my stupid mistake of inserting the Agnus the wrong way could have killed a nearby component of some sort?
2 x A1200 Blizzard IV 030@50Mhz
1 x A3000D 030@25Mhz
1 x A2000 030@25Mhz
1 x A2000, 2 x A600
2 x A500
 

Offline Castellen

Re: A3000 D Schematics Rev 9 Board
« Reply #21 on: July 06, 2005, 11:39:25 AM »
Running the device while in backwards would have damaged the the device, but it's fairly unlikely to have damaged anything else.  No guarantees though!

You mentioned the Denise socket has signs of corrosion in it.  The very first thing I would do is replace it, regardless if the connections seem fine or not.

If the Denise socket is corroded, the Paula socket probably is as well.  Replace them both.

The sockets themselves are cheap enough, just a little bit time consuming to replace them.  But it would be a much quicker fix than your current approach. :-)
 

Offline melangeTopic starter

  • Full Member
  • ***
  • Join Date: Jun 2005
  • Posts: 235
    • Show only replies by melange
Re: A3000 D Schematics Rev 9 Board
« Reply #22 on: July 06, 2005, 05:10:14 PM »
Thanks yet again for your reply.  I agree with the replacement of sockets.  i have replaced sockets before and it's not the easiest job but I can do it.  However first I have to fix my Iron, which is going to cost me about $65.00 AU that I can't afford ATM.  I'm still hoping to get it going for now and then replace the sockets whan I have a more reliable iron.  However as you have kindly pointed out I may get nowhere until I replace them.  I located a dead trace within the affected area today, fixed it but still no change, just flashing power led and rolling horizontal green lines on the screen.  I'm fairly sure the green line is going to be a green screen if the video problem comes good before the boot problem is located and repaired.  Could still have some Agnus dramas possibly, if the green screen is indicating Chip Ram as the fault.

Thanks for your help hope i have some good news soon... :-)
2 x A1200 Blizzard IV 030@50Mhz
1 x A3000D 030@25Mhz
1 x A2000 030@25Mhz
1 x A2000, 2 x A600
2 x A500
 

Offline melangeTopic starter

  • Full Member
  • ***
  • Join Date: Jun 2005
  • Posts: 235
    • Show only replies by melange
Re: A3000 D Schematics Rev 9 Board
« Reply #23 on: July 12, 2005, 10:43:39 AM »
I gave up.  At least for now.  I found a work around and now have my Tower working and in use :-)

It would appear there was more wrong with that board than i realised.  The (other) board I now have works better than the last.  I think there is more wrong with the RTC Circuit area than I had thought initially.  This board actually maintains the time, where as the last board used to lose it everytime it took a trip to the Guru.  I'm still using the same (replacement) battery which seems to be holding a charge, it wasn't holding a charge on the last board.

It is likely that I will get that dead board in my posession again oneday and I'll have another go at fixing it then.  

Now all I need is a Network Card.

2 x A1200 Blizzard IV 030@50Mhz
1 x A3000D 030@25Mhz
1 x A2000 030@25Mhz
1 x A2000, 2 x A600
2 x A500