New products for C64 and Amiga from Individual Computers - Page 15

kipper2k · 17 October 2012, 14:12

this may seem too simple, but is there a notch on the cpu socket so that the resistor close to the cpu does not prevent the board from sitting all the way down. If the notch was not there you would not be able to press the socket all the way down. Also if the resistor is out of alignment that could also prevent it sitting all the way down.

demolition · 17 October 2012, 14:38

Yes, I noticed that notch and it does not seem to touch the resistor, so I don't think that is the problem. I could try installing it in my other (but still not fixed) A600 simply to see if the fit is different.

alenppc · 17 October 2012, 16:03

Does anyone know when the preorders will start shipping from Amigakit?

Schoenfeld · 17 October 2012, 17:59

ACA620 units have already reached AmigaKit, they should already be shipping.

ACA1220 and ACA1232 will take another week or so. I have just picked up the first sample from the assembly place and will spend the rest of the day giving them a full test before the mass-production can continue.

BigRamPlus may also take a few more days. No good news there, it's a very low-margin product, so we have it on low priority after the accelerators. If you want Z3 Ram "now", ZorRAM is available "now". Production is running, so if you have pre-ordered BigRamPlus, you can be sure that you will get it.

Jens

alenppc · 17 October 2012, 18:56

Thanks Jens.
I was asking mainly about the A620, I can't wait to get mine. It would be nice if people who start getting shipping notifications from Amigakit post it here so we can know how long to wait approximately.

demolition · 17 October 2012, 18:57

Ok, I now have an update for my ACA620 problems and I think I know where the problem is.

I tried another PSU and it didn't change anything. I took out the ACA620, cleaned the CPU pins again and reinserted the ACA620 and now it seems to run a lot better although it is still not without problems.

I completed 8 full memory tests (2+11 MB) without a single error and this wasn't possible before. I can also maprom KS3.1 now without that weird chipmemory problem, so it must have been a socket connection problem.

I still think the socket connection seems a bit unstable, and it doesn't touch the motherboard as shown on the attached picture. It is very sensitive to vibrations which is probably why it crashed when I used the FDD.

Currently it seems to be running fine (in WB playing music right now), however if I press down on the PCMCIA connector only a few mm, it crashes.

kipper2k · 17 October 2012, 19:08

it seems like you have a cold solder joint somewhere that is causing intermittent failure. When pressing down on the pcmcia area you are probably causing the joint to open and of course when the cpu/board warms up the expansion is causing the intermittent breaks too.

Cold solder joints are very hard to trace. One thing to try is to use a pencil with an eraser and gently press down on the motherboard in different areas to see if you can home in on the area. To fix a cold solder joint you would have to use a bit of flux on all pads and touch up with an iron to remelt/reflow the solder. Don't just attempt to use an iron with no flux as you could dry up more joints.

demolition · 17 October 2012, 19:22

I agree that it is a bad connection somewhere, but the problem does not occur without the ACA020 connected, so I'm assuming it is not in the motherboard?
The bad connection might still be between the 68k and the CPU socket causing these failures?

kipper2k · 17 October 2012, 19:30

Quote:

Originally Posted by demolition

I agree that it is a bad connection somewhere, but the problem does not occur without the ACA020 connected, so I'm assuming it is not in the motherboard?
The bad connection might still be between the 68k and the CPU socket causing these failures?

get a big fan from somewhere, tabletop, stand up fan etc and point it at your amiga and see if that fixes it. I had one A600 that would crash on me when it got too toasty and as soon as i cooled it by pointing a household fan at it it would chirp along quite happily. Could never figure out why it got so hot, i think it was just the internals of the cpu that for whatever reason consumed too much power and got too hot.

Schoenfeld · 17 October 2012, 19:46

Quote:

Originally Posted by demolition

I agree that it is a bad connection somewhere, but the problem does not occur without the ACA020 connected, so I'm assuming it is not in the motherboard?
The bad connection might still be between the 68k and the CPU socket causing these failures?

The ACA620 pushes on the sides of the 68000 pins, which bends them in slightly. If there is a cold solderjoint (which I also believe is what you're observing), the effects will be stronger with the accelerator installed.

I'm a bit less gentle than kipper2k when looking for such a bad solderjoint: I connect PSU and monitor only, then take the board in both hands and twist/bend it fairly hard. If the computer continues to run, everything is fine. If it crashes, I need to re-solder the PLCC chips like kipper2k described: Apply a bit of flux all around, melt the solder for 2 seconds on each pin and usually the computer is immune against hard flexing/twisting after that.

The attached picture was taken with quite a bit of trickery with light&macro lens, so the crack is clearly seen as a shadow. Believe it or not, this A600 was working without an accelerator - exactly your case.

Jens

demolition · 17 October 2012, 19:59

Quote:

Originally Posted by Schoenfeld

The ACA620 pushes on the sides of the 68000 pins, which bends them in slightly. If there is a cold solderjoint (which I also believe is what you're observing), the effects will be stronger with the accelerator installed.
The attached picture was taken with quite a bit of trickery with light&macro lens, so the crack is clearly seen as a shadow. Believe it or not, this A600 was working without an accelerator - exactly your case.

That does seem like a very probable cause, so I guess I know what to do next weekend.

Edit: BTW, Vesalia changed the description of the ACA620 from originally '9+1 MB' to '12+3 MB', how come? I guess it is still not possible to get more than 11.3MB available, 1MB for maprom and a bit for a freezer?

prowler · 17 October 2012, 20:01

Quote:

Originally Posted by Schoenfeld

The attached picture was taken with quite a bit of trickery with light&macro lens, so the crack is clearly seen as a shadow. Believe it or not, this A600 was working without an accelerator - exactly your case.

That photo is a masterpiece which deserves to be included in a gallery of pictures illustrating what to look out for when troubleshooting an intermittent fault.

It really needs to be seen full size

, so here we go:

Thanks for including it, Jens!

demolition · 17 October 2012, 21:40

I just tried installing the ACA620 in my other (non-working) A600, and mechanically it feels a lot more solid when installed in it. It seems to touch the motherboard all the way around and there is more of a 'locking' feeling when pressing it down (and no squeaking noises). The markings of the CPU are the same so I'm not sure why this could be.

I also tried booting the A600 without the ACA620 installed but with Acatune in startup-sequence and the program crashes. If I boot the Amiga without startup-sequence and simply type 'Acatune' it says 'Card detected: ACA630-25MHz' so obviously it will crash if it tries change something..

Edit: Jens, I tried your twist-test and it revealed a perfectly stable stock A600 capable of running memory tests while I was twisting the motherboard quite a bit (with no bottom shield). Combined with the mechanical feeling described above I'm still thinking the bad connection is between the CPU and the socket as it doesn't feel completely right. I am not sure why that is or what to do about it though. That being said, I will probably still resolder all the CPU pins for safety's sake if I haven't found another solution by this weekend. As you said, the socket pushes on the pins which could make a gap in a bad solder.

Smakar · 17 October 2012, 21:49

Hello Jens, is there any update on the status of the ACA500? I have an A500 and have pre-ordered the ACA1220, but will not be able to use the ACA1220 until the ACA500 is available. I apologize in advance if the status has been posted elsewhere.

moijk · 17 October 2012, 22:14

I'm ready to preorder a couple of aca500's as well.

demolition · 18 October 2012, 09:45

I think I have managed to get the ACA620 99% stable now. I did a third, and extra thorough, cleaning of the CPU pins, and now I am able to twist the motherboard with the ACA620 installed, without crashes. Pushing down on the PCMCIA also doesn't cause any bad things to happen.

I assembled the A600, and I played several WHDload games for at least 30 minutes (previously I could hardly play for one minute without it crashing), after that I did 30 minutes of memtests without any errors, but shortly after that it froze as I opened a folder in WB. I'm sure the ACA620 was as hot as it would get at this point, so maybe it still needs some extra cooling?

Now it also feels a little better mechanically and there is less squeaking when installing it. I just recapped this machine and it had some acid damage in various places and I had to repair several traces and vias, so maybe some of the fumes from those caps also caused the CPU pins to oxidize, resulting in the bad connections that I had.

I will now try to sand off the CPU pins a bit with some grain 1000 to see if I can get the last 1% stability.

melbourneben! · 19 October 2012, 15:13

Jens, will the aca 1220 have any compatibility issues with the later A1200 revisions due to chipset timing? or do the timing issues only become evident on cards with faster cpu's?

Schoenfeld · 19 October 2012, 20:29

Quote:

Originally Posted by melbourneben!

Jens, will the aca 1220 have any compatibility issues with the later A1200 revisions due to chipset timing? or do the timing issues only become evident on cards with faster cpu's?

It will most probably not work even with the slowest ACA1220. The way that the slower CPU cards access the A1200 is exactly the same as the faster CPU cards, so I'd expect that timing fixes are still required. I did a tiny change in the timing, which improved a lot on my test-machines, but I don't have all-positive feedback from external testers.

The ACA1220/ACA1232 is in mass-production now, and I have the first few units here. They will be shown tomorrow by Alinea at Amiga-Meeting Nord in Bad Bramstedt, Germany (if DHL express doesn't mess it up again), and I have just eMailed Matthew from AmigaKit (who is currently at Amiwest) that he's free to take orders, because I can ship "right now".

Jens

alenppc · 19 October 2012, 20:38

Quote:

Originally Posted by Schoenfeld

I have just eMailed Matthew from AmigaKit (who is currently at Amiwest)

That explains the wait-time for the ACA620s from Amigakit...

amigakit.com · 20 October 2012, 00:41

@allenppc

Thanks for your preorder.

I will arrange for your ACA620 to go out on Monday.

17 October 2012, 18:57	#286
demolition Unregistered User Join Date: Sep 2012 Location: Copenhagen / DK Age: 43 Posts: 4,190	Ok, I now have an update for my ACA620 problems and I think I know where the problem is. I tried another PSU and it didn't change anything. I took out the ACA620, cleaned the CPU pins again and reinserted the ACA620 and now it seems to run a lot better although it is still not without problems. I completed 8 full memory tests (2+11 MB) without a single error and this wasn't possible before. I can also maprom KS3.1 now without that weird chipmemory problem, so it must have been a socket connection problem. I still think the socket connection seems a bit unstable, and it doesn't touch the motherboard as shown on the attached picture. It is very sensitive to vibrations which is probably why it crashed when I used the FDD. Currently it seems to be running fine (in WB playing music right now), however if I press down on the PCMCIA connector only a few mm, it crashes. Last edited by demolition; 02 May 2013 at 16:00.

17 October 2012, 21:40	#293
demolition Unregistered User Join Date: Sep 2012 Location: Copenhagen / DK Age: 43 Posts: 4,190	I just tried installing the ACA620 in my other (non-working) A600, and mechanically it feels a lot more solid when installed in it. It seems to touch the motherboard all the way around and there is more of a 'locking' feeling when pressing it down (and no squeaking noises). The markings of the CPU are the same so I'm not sure why this could be. I also tried booting the A600 without the ACA620 installed but with Acatune in startup-sequence and the program crashes. If I boot the Amiga without startup-sequence and simply type 'Acatune' it says 'Card detected: ACA630-25MHz' so obviously it will crash if it tries change something.. Edit: Jens, I tried your twist-test and it revealed a perfectly stable stock A600 capable of running memory tests while I was twisting the motherboard quite a bit (with no bottom shield). Combined with the mechanical feeling described above I'm still thinking the bad connection is between the CPU and the socket as it doesn't feel completely right. I am not sure why that is or what to do about it though. That being said, I will probably still resolder all the CPU pins for safety's sake if I haven't found another solution by this weekend. As you said, the socket pushes on the pins which could make a gap in a bad solder. Last edited by demolition; 17 October 2012 at 22:09.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Santa's come early... A new scandoubler from Individual Computers	NovaCoder	News	708	18 October 2016 22:43
Individual Computers Announces Clone-A Project	Paul	News	83	14 December 2013 16:05
Individual Computers Silver Sponsor of the Revision	gibs	Amiga scene	1	22 April 2011 16:43
Individual Computers: New products, RoHS comliance, Vacation until july 26th	Paul	News	31	21 July 2007 19:22
New products by individual Computers	Paul	News	0	30 November 2004 15:58

17 October 2012, 14:12	#281
kipper2k Registered User Join Date: Sep 2006 Location: Thunder Bay, Canada Posts: 4,323	this may seem too simple, but is there a notch on the cpu socket so that the resistor close to the cpu does not prevent the board from sitting all the way down. If the notch was not there you would not be able to press the socket all the way down. Also if the resistor is out of alignment that could also prevent it sitting all the way down.

17 October 2012, 14:38	#282
demolition Unregistered User Join Date: Sep 2012 Location: Copenhagen / DK Age: 43 Posts: 4,190	Yes, I noticed that notch and it does not seem to touch the resistor, so I don't think that is the problem. I could try installing it in my other (but still not fixed) A600 simply to see if the fit is different.

17 October 2012, 16:03	#283
alenppc Registered User Join Date: Apr 2012 Location: Canada Age: 44 Posts: 910	Does anyone know when the preorders will start shipping from Amigakit?

17 October 2012, 17:59	#284
Schoenfeld electricky. Join Date: Jun 2010 Location: out in the wild Posts: 1,256	ACA620 units have already reached AmigaKit, they should already be shipping. ACA1220 and ACA1232 will take another week or so. I have just picked up the first sample from the assembly place and will spend the rest of the day giving them a full test before the mass-production can continue. BigRamPlus may also take a few more days. No good news there, it's a very low-margin product, so we have it on low priority after the accelerators. If you want Z3 Ram "now", ZorRAM is available "now". Production is running, so if you have pre-ordered BigRamPlus, you can be sure that you will get it. Jens

17 October 2012, 18:56	#285
alenppc Registered User Join Date: Apr 2012 Location: Canada Age: 44 Posts: 910	Thanks Jens. I was asking mainly about the A620, I can't wait to get mine. It would be nice if people who start getting shipping notifications from Amigakit post it here so we can know how long to wait approximately.

17 October 2012, 19:08	#287
kipper2k Registered User Join Date: Sep 2006 Location: Thunder Bay, Canada Posts: 4,323	it seems like you have a cold solder joint somewhere that is causing intermittent failure. When pressing down on the pcmcia area you are probably causing the joint to open and of course when the cpu/board warms up the expansion is causing the intermittent breaks too. Cold solder joints are very hard to trace. One thing to try is to use a pencil with an eraser and gently press down on the motherboard in different areas to see if you can home in on the area. To fix a cold solder joint you would have to use a bit of flux on all pads and touch up with an iron to remelt/reflow the solder. Don't just attempt to use an iron with no flux as you could dry up more joints.

17 October 2012, 19:22	#288
demolition Unregistered User Join Date: Sep 2012 Location: Copenhagen / DK Age: 43 Posts: 4,190	I agree that it is a bad connection somewhere, but the problem does not occur without the ACA020 connected, so I'm assuming it is not in the motherboard? The bad connection might still be between the 68k and the CPU socket causing these failures?

17 October 2012, 21:49	#294
Smakar Amiga Fan in Phoenix Join Date: Aug 2012 Location: Phoenix, AZ USA Age: 66 Posts: 60	Hello Jens, is there any update on the status of the ACA500? I have an A500 and have pre-ordered the ACA1220, but will not be able to use the ACA1220 until the ACA500 is available. I apologize in advance if the status has been posted elsewhere.

17 October 2012, 22:14	#295
moijk License to Bubble Join Date: May 2010 Location: Grimstad, Norway Age: 46 Posts: 293	I'm ready to preorder a couple of aca500's as well.

18 October 2012, 09:45	#296
demolition Unregistered User Join Date: Sep 2012 Location: Copenhagen / DK Age: 43 Posts: 4,190	I think I have managed to get the ACA620 99% stable now. I did a third, and extra thorough, cleaning of the CPU pins, and now I am able to twist the motherboard with the ACA620 installed, without crashes. Pushing down on the PCMCIA also doesn't cause any bad things to happen. I assembled the A600, and I played several WHDload games for at least 30 minutes (previously I could hardly play for one minute without it crashing), after that I did 30 minutes of memtests without any errors, but shortly after that it froze as I opened a folder in WB. I'm sure the ACA620 was as hot as it would get at this point, so maybe it still needs some extra cooling? Now it also feels a little better mechanically and there is less squeaking when installing it. I just recapped this machine and it had some acid damage in various places and I had to repair several traces and vias, so maybe some of the fumes from those caps also caused the CPU pins to oxidize, resulting in the bad connections that I had. I will now try to sand off the CPU pins a bit with some grain 1000 to see if I can get the last 1% stability.

19 October 2012, 15:13	#297
melbourneben! Registered User Join Date: Feb 2012 Location: adelaide Posts: 44	Jens, will the aca 1220 have any compatibility issues with the later A1200 revisions due to chipset timing? or do the timing issues only become evident on cards with faster cpu's?

20 October 2012, 00:41	#300
amigakit.com Registered User Join Date: Aug 2004 Location: www.amigakit.com Posts: 2,015	@allenppc Thanks for your preorder. I will arrange for your ACA620 to go out on Monday.

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