Amiga 4000 A3640 CPU Card 3.1 Revision Capacitor Issue

crazyegg · 16 October 2009, 19:48

Hi All,

I have a A3640 card for my A4000. It is running the LC version of the 040 chip and everything is currently working fine.

The revision number on the board is 3.1. However the U209 GAL chip is -3 and not -2.

Question:
Do I have a 3.1 board or a 3.2 board? More importantly, do I have to replace the 3 caps due to incorrect polarity issue?

There is no evidence of cap leakage and the solder pads are very shiny and not dull.

FYI, I've also measured the ohms of the caps as specified in another thread. The negative side of the cap was around 0.8 Ohms and the positive side was 0.1 Ohms. Is that good?

Regards...ed

alexh · 16 October 2009, 20:53

Quote:

Originally Posted by crazyegg

Do I have a 3.1 board or a 3.2 board?

It doesn't matter. With the A3640 the only question should be "Is it a 3.0 or not?" The difference between a 3.1 and 3.2 is almost negligible. A few issues with the now defunct GVP PhonePak card?

Quote:

Originally Posted by DaveHaynie

The A3640 had two problems in its Rev 3.0 form. The first was a marginality -- its sampling of the local bus STERM* signal was marginal. This is fixed on Rev 3.1 with a cut and jumper. But beware, some boards marked 3.1 didn't get this fix, though apparently they're a small number.

The second problem on the A3640 Rev 3.0 is a real live bug. This was a flaw in the bus arbiter that could allow the '040 and any local bus master on the local bus at the same time. Rev 3.1 incorporates -02 of the U209 PAL to fix this problem. It's not a perfect solution, though, in that it creates a potential for the local bus master to be locked out of the local bus for 10's of microseconds, even if the '040 isn't using the bus. This was corrected in -03 of the U209 PAL, which makes your Rev 3.1 A3640 into a Rev 3.2 A3640. Clearly, if you're not using cards with a DMA problem, this is not an issue.

Quote:

Originally Posted by crazyegg

More importantly, do I have to replace the 3 caps due to incorrect polarity issue?

I haven't.

crazyegg · 19 October 2009, 02:43

Hi,

Thanks for the reply!

I had no previous notion that leaky caps were such a big problem for us Amiga folk. Based on other threads at EAB, it appears that cap replacement is recomended for A600/A1200/CD32/A4000....

Is this just due to age and did Commodore just cheap out and use capacitors that are prone to leaks? Are non-leaking caps that much more expensive?

Regards....ed

kipper2k · 19 October 2009, 02:57

Quote:

Originally Posted by crazyegg

Hi,

Thanks for the reply!

I had no previous notion that leaky caps were such a big problem for us Amiga folk. Based on other threads at EAB, it appears that cap replacement is recomended for A600/A1200/CD32/A4000....

Is this just due to age and did Commodore just cheap out and use capacitors that are prone to leaks? Are non-leaking caps that much more expensive?

Regards....ed

Hi Ed,

Most electrolytic caps are rated by hours service, 3000 normally being the max (guaranteed) life. Even if the usage is not that high, time will wear out the case due to the acid and leak causing damage to the board. Capacitors are really cheap so its not that expensive to replace them, just need a steady hand and the right tools (Digikey.ca has all the caps you could possibly need

alexh · 19 October 2009, 09:19

Quote:

Originally Posted by crazyegg

I had no previous notion that leaky caps were such a big problem for us Amiga folk. Based on other threads at EAB, it appears that cap replacement is recomended for A600/A1200/CD32/A4000....

I've not had any trouble with any of mine and have no intention of replacing any "in case". (or ever probably).

Jope · 19 October 2009, 11:19

I have also never replaced caps in my old gear. It's always been failed ICs so far.

MagerValp · 19 October 2009, 16:23

My A4000 suffered a fair bit of damage from leaking 22 µF caps:

http://picasaweb.google.com/MagerValp/A4000

crazyegg · 20 October 2009, 16:22

Hi All,

How can I test on my A3640 if the caps are on backwords or not? My caps legs are very shiny and do not show signs of leakage

I located some confusing (well, confusing for me

) instructions on another thread:

"You can verify this by locating position U103, just below the oscillator. Vcc (5V) and G(nd) should be printed on the device. Check using an ohm meter for a connection between VCC and the black side of the capactors, similarly check for a connection from G to the other side of the capacitors."

Can someone explain exactly what I should be doing with my ohm meter (i.e. where do I put the probes and what am I looking for)?

Regards...ed

Chain · 20 October 2009, 17:39

just measuring voltage on each cap will give you hint which are correct and which are reversed.

desoldering all those b*stards and replace them with new ones with correct polarity will be the best solution

alexh · 20 October 2009, 18:30

Visual inspection should be good enough.

There is an A3640 on Ebay at the moment, says all capacitors replaced and correctly fitted.

As you can see C105, C106 & C107 are the opposite way round to the silk screen.

crazyegg · 20 October 2009, 20:12

Hi All,

Thanks again for all the great help.

I think I will just leave the caps alone on my version 3.1 board and not replace them for now

There is no evidence of leakage anywhere and that is why I thought that perhaps the caps are indeed in correct alignment.

FYI, the caps polarities match the silk screening outline on the board and I was thinking that MAYBE the silk screening outline may have been "corrected" on my board. Wishful thinking????

Regards....ed

rkauer · 21 October 2009, 02:39

And even if the capacitors didn't leaked, they will not work as they should be.

Replace at least the three offended units.

MagerValp · 21 October 2009, 15:04

The problem with the A3640 boards is that the silk-screen is wrong. On the boards that were fixed, C105-C107 are mounted "backwards". The C105-C107 should have the black stripe towards the + (which is "backwards"), all others should have the black stripe away from the +.

Commodore apparently got a batch of bad 22 µF caps, as can be seen in the common problems with A4000, A3640, etc. Using them with reversed polarity (such as the A3640 caps mounted backwards, or the A4000 audio circuit with negative A/C voltage) shortens their life significantly. When they start to leak, the corrosion will eat through solder joints, board traces, and wreak havok. Replacing them is easy, and you can replace them with regular through hole components, so there's no need for SMT soldering equipment.

Paul_s · 21 October 2009, 15:32

use this guide

http://amiga.serveftp.net/A3640_capacitor.html

alexh · 21 October 2009, 16:38

Nice, I'll kill the link above to the bad site.

16 October 2009, 19:48	#1
crazyegg Registered User Join Date: Mar 2008 Location: Canada Posts: 122	Amiga 4000 A3640 CPU Card 3.1 Revision Capacitor Issue Hi All, I have a A3640 card for my A4000. It is running the LC version of the 040 chip and everything is currently working fine. The revision number on the board is 3.1. However the U209 GAL chip is -3 and not -2. Question: Do I have a 3.1 board or a 3.2 board? More importantly, do I have to replace the 3 caps due to incorrect polarity issue? There is no evidence of cap leakage and the solder pads are very shiny and not dull. FYI, I've also measured the ohms of the caps as specified in another thread. The negative side of the cap was around 0.8 Ohms and the positive side was 0.1 Ohms. Is that good? Regards...ed

20 October 2009, 18:30	#10
alexh Thalion Webshrine Join Date: Jan 2004 Location: Oxford Posts: 14,337	Visual inspection should be good enough. There is an A3640 on Ebay at the moment, says all capacitors replaced and correctly fitted. As you can see C105, C106 & C107 are the opposite way round to the silk screen. Last edited by alexh; 21 October 2009 at 16:40.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Amiga 4000 audio capacitor help	r.cade	support.Hardware	2	27 January 2012 21:32
WTB: A3640 CPU card for A4000	Akiko	MarketPlace	1	11 June 2011 11:42
FS: A3640 for Amiga 4000	coze	MarketPlace	5	22 June 2007 12:46
Ebay at 1 USD : A3640 Rev 3.1 for Amiga 4000 running at 32 Mhz !!!	Effy	MarketPlace	21	01 August 2005 15:05
[FA] A3640 040@25MHz [AMIGA 4000]	scan_x	MarketPlace	0	28 December 2003 13:03

19 October 2009, 02:43	#3
crazyegg Registered User Join Date: Mar 2008 Location: Canada Posts: 122	Hi, Thanks for the reply! I had no previous notion that leaky caps were such a big problem for us Amiga folk. Based on other threads at EAB, it appears that cap replacement is recomended for A600/A1200/CD32/A4000.... Is this just due to age and did Commodore just cheap out and use capacitors that are prone to leaks? Are non-leaking caps that much more expensive? Regards....ed

19 October 2009, 11:19	#6
Jope - Join Date: Jul 2003 Location: Helsinki / Finland Age: 43 Posts: 9,861	I have also never replaced caps in my old gear. It's always been failed ICs so far.

19 October 2009, 16:23	#7
MagerValp Registered User Join Date: Aug 2008 Location: Göteborg / Sweden Posts: 237	My A4000 suffered a fair bit of damage from leaking 22 µF caps: http://picasaweb.google.com/MagerValp/A4000

20 October 2009, 16:22	#8
crazyegg Registered User Join Date: Mar 2008 Location: Canada Posts: 122	Hi All, How can I test on my A3640 if the caps are on backwords or not? My caps legs are very shiny and do not show signs of leakage I located some confusing (well, confusing for me ) instructions on another thread: "You can verify this by locating position U103, just below the oscillator. Vcc (5V) and G(nd) should be printed on the device. Check using an ohm meter for a connection between VCC and the black side of the capactors, similarly check for a connection from G to the other side of the capacitors." Can someone explain exactly what I should be doing with my ohm meter (i.e. where do I put the probes and what am I looking for)? Regards...ed

20 October 2009, 17:39	#9
Chain A-Collector, repairments Join Date: Jul 2001 Location: Czech Rep. Age: 50 Posts: 1,536	just measuring voltage on each cap will give you hint which are correct and which are reversed. desoldering all those b*stards and replace them with new ones with correct polarity will be the best solution

20 October 2009, 20:12	#11
crazyegg Registered User Join Date: Mar 2008 Location: Canada Posts: 122	Hi All, Thanks again for all the great help. I think I will just leave the caps alone on my version 3.1 board and not replace them for now There is no evidence of leakage anywhere and that is why I thought that perhaps the caps are indeed in correct alignment. FYI, the caps polarities match the silk screening outline on the board and I was thinking that MAYBE the silk screening outline may have been "corrected" on my board. Wishful thinking???? Regards....ed

21 October 2009, 02:39	#12
rkauer I hate potatos and shirts Join Date: Oct 2007 Location: Sao Leopoldo / Brazil Age: 58 Posts: 3,482	And even if the capacitors didn't leaked, they will not work as they should be. Replace at least the three offended units.

21 October 2009, 15:04	#13
MagerValp Registered User Join Date: Aug 2008 Location: Göteborg / Sweden Posts: 237	The problem with the A3640 boards is that the silk-screen is wrong. On the boards that were fixed, C105-C107 are mounted "backwards". The C105-C107 should have the black stripe towards the + (which is "backwards"), all others should have the black stripe away from the +. Commodore apparently got a batch of bad 22 µF caps, as can be seen in the common problems with A4000, A3640, etc. Using them with reversed polarity (such as the A3640 caps mounted backwards, or the A4000 audio circuit with negative A/C voltage) shortens their life significantly. When they start to leak, the corrosion will eat through solder joints, board traces, and wreak havok. Replacing them is easy, and you can replace them with regular through hole components, so there's no need for SMT soldering equipment.

21 October 2009, 15:32	#14
Paul_s Registered User Join Date: Nov 2006 Location: Amigaville Age: 46 Posts: 3,334	use this guide http://amiga.serveftp.net/A3640_capacitor.html

21 October 2009, 16:38	#15
alexh Thalion Webshrine Join Date: Jan 2004 Location: Oxford Posts: 14,337	Nice, I'll kill the link above to the bad site.

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