Individual Computers - New accelerator cards for A600 & A1200

[roy bates]

Originally Posted by Akira

Sorry if this question sonds stupid, but here it goes.
I just bought an ACA620 and m waiting for it.

In preparation, I thought I'd give a clean to the CPU legs just in case. The maual, however, says the following:
"Never use contact-spays for this task!"

What exactly is contact-spays? I swear I googled it and have no idea what it is.
Would regular rubbing alcohol work or do I need isopropyl alcohol?

the only contact spray(i think thats what you ment?) ive used is switch cleaner,and there right just use ipa instead.

[Amiga1992]

The manual says, literally, "contact-spay". If that's a typo, I have no clue. I pasted verbatim what it says.
Clarification would be welcome.

[Cymru]

Yes it is a Typo; a contact spray contains an electrolyte (such as a metallic salt) to increase the connection between two, well, contacts. Used on inert gold, that might be ok, although Hg might be better from a chemical stand point. They also contained a reducing agent (such as a sulfite) to de-oxidize corrosion. They were never in favor just as a can of flat-tyre "fixer" solves one problem but causes another. There is sold the de-oxidizer alone, but a stiff brush into slots, and a rubber eraser on contacts is more affective. That and Iso-PrOH is safer as the contact spray can jump a connection across the wrong contacts if too much gets in. I don't think they are sold any more.

For rusted parts stuck together (nuts & bolts, frozen pistons, etc.) the top product is called "Kroil" made by Kano Laboratories in the states.

[roy bates]

i dont know either akira,but if they did mean contact spray it might because its got lubricant and other chemicals that effects gold in it,ipa just evaporates with no resedue.

[Cymru]

Vide Supra.

[Amiga1992]

Originally Posted by Cymru

Yes it is a Typo; a contact spray contains an electrolyte (such as a metallic salt) to increase the connection between two, well, contacts. Used on inert gold, that might be ok, although Hg might be better from a chemical stand point. They also contained a reducing agent (such as a sulfite) to de-oxidize corrosion. They were never in favor just as a can of flat-tyre "fixer" solves one problem but causes another. There is sold the de-oxidizer alone, but a stiff brush into slots, and a rubber eraser on contacts is more affective. That and Iso-PrOH is safer as the contact spray can jump a connection across the wrong contacts if too much gets in. I don't think they are sold any more.

For rusted parts stuck together (nuts & bolts, frozen pistons, etc.) the top product is called "Kroil" made by Kano Laboratories in the states.

Thank you for your in depth explanation, Cymru!
So what should I use, generic rubbing alcohol or isopropyl alcohol? Orrather, can I safely ue regular alcohol? I don't have isopropyl.

[roy bates]

i would use isopropyl(ipa),you can get it cheap.

[Cymru]

What is "regular" alcohol? Ethanol? Denatured alcohol (Ethanol + Methanol)? Wood Alcohol (Methanol)? Rubbing Alcohol is an old name for 70% Isopropanol isn't it?

Because of its use in medicine, Isopropanol (IsoPrOH) is cheap; it comes in two versions, 70% and 90%. The other 30 or 10% is water which does cause oxidation (rust/corrosion) and does conduct electricity. I only use 90% and let it completely dry (evaporate) before applying current to anything.

[Schoenfeld]

Originally Posted by Akira

Thank you for your in depth explanation, Cymru!
So what should I use, generic rubbing alcohol or isopropyl alcohol? Orrather, can I safely ue regular alcohol? I don't have isopropyl.

Alcohol can be either ethanol or propanol - both are "alcohols" if you dig out your organic chemistry lessons: Ethanol = C2H5OH and Propanol = C3H7OH (the OH group makes it an alcohol). The cleaning property of Propanol and Isopropanol is pretty much the same, it's just that the OH group is at the middle of the three carbon atoms. That mainly changes the boiling temperature of the liquid, but not the cleaning property if I remember right (that was 11th grade.. 23 years ago, so memory may not serve right here).

It's just that ethanol is the stuff you can safely drink (within limits that everyone should know best for himself), therefore it's taxed. However, there's a way to get around the tax in Europe, and that's adding a salt that makes it undrinkable (the German word would be "vergällen" - dict.leo.org says that the English word would be "to denature"). Using that cheaper stuff would leave a residue, which is why many people think that ethanol should not be used for cleaning delicate parts at all. Fact is: If it's the pure stuff, it's as good as isopropanol.

The tool of choice is an old toothbrush. No need to dry with cotton or so, it dries on it's own after a few seconds, and you can safely operate the Amiga with a slightly wet CPU.

Long story short: Thanks for pointing out the typo, and don't use overpriced sprays.

Jens

[demolition]

Originally Posted by Cymru

Isopropanol (IsoPrOH) is cheap; it comes in two versions, 70% and 90%.

Don't tell that to the bottle I have on my table. It thinks it is 99%. Stuff like this will be different from country to country, and I have actually never seen alcohol below 96-97%ish here (well, except the drinkable stuff)..

[Amiga1992]

Originally Posted by Schoenfeld

Long story short: Thanks for pointing out the typo, and don't use overpriced sprays.

Dude, I cracked up when I googled spay...

Thanks a lot for your input, I will buy a bottle of isopropyl now and be ready for that beauty to come through the mail!

[Cymru]

Sorry, it probably does come in different strengths depending on the country. Pure Ethanol is not pure ethanol; the drinkable form is 95% EtOH and 5% water; the lab form is 99+ % and benzene. The alcohols form an azeotrope with water so that they act as one molecular group. One has to form another azeotrope with benzene to get a higher percent. The same applies to the other lower molecular weight alcohols (and to some extent with the higher molecular weight alcohols/"waxes"). Other solvents lacking the hydroxyl (OH) group would make for less oxidative solvents, but are not readily available. Others are such good solvents, such as Acetone, they may dissolve a wee bit too much of plastics, paints, and adhesives.

In general, Isopropanol is cheap and effective bearing in mind that the water molecule is electron "rich" and will conduct electricity (electrons); probably why allowing it to dry out over night is best. Further, the OH (oft shown as OH-) group loves metals (oft shown as Me+) and make, for example, FeOH3 or rust. This is called "oxidation" for the addition of Oxygen. In electronics it is not good. Throw in a "Basic" alkaline like Lye (KOH Potassium Hydroxide) or as in the case of our friend the barrel battery, and not only does it oxidize, it also dissolves the copper traces. No, "Acid" batteries use lead (Pb) and sulfuric acid and are found in cars. Well, not hybrids, but you should know the difference.

[Amiga1992]

OK so now that i received this, I want to install it tomorrow.
Is there any tool I can download from Aminet that lets me test the functioning of the machine? Something like an infinite archiving loop or something like that? How do you guys test yours to see if they are stable? Jens? Any advice? I wanna roll it for a bunch of hours after I install it so I am totally sure the unit is fine.

[demolition]

Originally Posted by Akira

Is there any tool I can download from Aminet that lets me test the functioning of the machine?

I used MBRtest for mine, which tests the memory using various patterns. I put an image of it in the zone for you.
In my experience, if the machine isn't stable, this will show it. Just set it to infinite loops and to pause on errors. Enable all tests except for float (very slow) and select all memory ranges. Leave it for a couple of hours if it is still running, it should all be good.

[Amiga1992]

Originally Posted by demolition

I used MBRtest for mine, which tests the memory using various patterns. I put an image of it in the zone for you.
In my experience, if the machine isn't stable, this will show it. Just set it to infinite loops and to pause on errors. Enable all tests except for float (very slow) and select all memory ranges. Leave it for a couple of hours if it is still running, it should all be good.

Thanks Demolition!! I'll do that )
That's the Microbotics XA1230 memory test, how come it works with the ACA620? Are results trustworthy?

[demolition]

Originally Posted by Akira

Thanks Demolition!! I'll do that )
That's the Microbotics XA1230 memory test, how come it works with the ACA620? Are results trustworthy?

It is just a memory test program and works with chip, slow or regular fast mem, no matter where it comes from. The only limitation I've found is that it doesn't recognize fast mem from PCMCIA SRAM cards (it doesn't appear on the list).

[Amiga1992]

[EDIT 2] I gave up on that CF-IDE adapter so I wil leitehr buy a new one or a flash-on-module unit. I am using it with teh A1200 one now and it works fine.

I ran MBR for a while now, and after the 7th cycle the machine locked up. I also tried to load Indy 4 Adventure WHDLoad and I got a black screen after a while and stuck there.

I am using ACATune with the options -maprom * -maxmem
Is the unit OK? Is it an ACAtune misconfiguration? Becuse before I used ACTune it got to 21 cycles without problem.
Also why is the memory separated in 4MB blocks (or smaller) on MBR-Test? There's like 4 different FAST RAM sections.

[EDIT3] Sorry I am talking to myself so much. I noticed the screen does weird things now when I have the acelerator on. Namely I can see color changes on bars on the screen that weren't there before.

Here's a video of the computer doing those bars. First you will notice some blueish hue bars while extracting things (not very noticeable on youube video unfortunately). Second you will see weird dark red bars on the WHDLoad startup screen while loading data to RAM (and freezing after a while). Tried with and without maprom, same results. Tried without ACATune, i don't see the bars then, but Beneath a Steel Sky still locks up:

[ Show youtube player ]

(I know I have an old WHDLoad, but the screen colour flashing occurs not just in WHDLoad, also noticed while I use MBR on the red text)

My machine has a rev 1.3 June Bug motherboard, caps replaced last year, standard A601 chip ram expansion, picopsu, cf ide and pcmcia card reader, nothing else.

Do I have to send it back?

[EDIT 4]: Kinda notice on the MBR test loop that it flashes colours more on the tests done on FAST RAM located at $600000. Look at this video, you can clearly see the flashing: [ Show youtube player ] (laugh at music )

[demolition]

Your ACAtune command line should work fine. Those color changes doesn't look like anything the ACA should be able to cause, but I could be wrong. Is the A600 100% fine without the ACA installed (or when disabled)?.

It is normal that you see many memory regions in MBRtest since the ACA620 fills out all the small unused memory spaces to optimize the amount of available memory, even though some of the blocks are rather small.
Also enable the testing of chipmem in MBRtest (the top one) since that communication is fully happening through the CPU socket where as fast mem access is handled locally on the ACA. You will see some flickering on the screen as the test passes through the visible screen area but this is normal and no persistent gfx corruption should happen (like it did on my 600 when it wasn't stable).

I'd try to clean the CPU thoroughly again, reinstall the ACA620 and see if it helps.

Is your board rev. 2? It should be, but if you got a rev. 1 by mistake, then you won't get it to work.

[jimbob]

I can't imagine why activating ACAtune would trigger it but that flashing looks a lot like the flickering some A600s get when accessing PCMCIA. Which video output are we watching? This probably doesn't apply if it's RGB but it might just be worth checking out the fix in this thread since you have a rev1.3 -

http://eab.abime.net/support-hardwar...ia-slot-2.html

This wouldn't directly explain the hangs but it could indicate a power issue, even though the picoPSU should be good. Have you got a brick you can try?

[Amiga1992]

Thanks for your replies guys.

Originally Posted by demolition

Your ACAtune command line should work fine. Those color changes doesn't look like anything the ACA should be able to cause, but I could be wrong. Is the A600 100% fine without the ACA installed (or when disabled)?.

Yes, this computer is 100% fine, been using it for a long time and was recapped last year. No problems at all.

It's not ACATune that's creating the problem, rather the extra memory ACATune adds that creates the problem.

You will see some flickering on the screen as the test passes through the visible screen area but this is normal

Does it look at all like what you can see in the video I posted?

I'd try to clean the CPU thoroughly again, reinstall the ACA620 and see if it helps.

Is your board rev. 2? It should be, but if you got a rev. 1 by mistake, then you won't get it to work.

Well the board's socket doesn't really sit flat on the motherboard, one of the corners, I think because of the socket's plastic, is lifted 1mm or less. I see no way I can fix this.

Also I ran the test for 1 hour 38 minutes without fails or lockups.

I assume the baord is rev.2... Sysinfo says "icomp.de 2" on description or something like that. Box had NO yellow sticker.

Originally Posted by jimbob

I can't imagine why activating ACAtune would trigger it but that flashing looks a lot like the flickering some A600s get when accessing PCMCIA. Which video output are we watching?

Hey Jim, you're onto something here. The software loaded on that video is indeed being pinched from teh SD card on my PCMCIA reader, and the display we're seeing is Composite. However, why does this happen only when ACATune adds all the RAM to the system? And this wasn't happening before without the ACA620.

EDIT: Just tested this again and it still happens even when I didn't use ACATune. So this is definitely one of the problems! However, how can I have power issues when I have a picoPSU? I am not sure now about this happening without the ACA620 installed. Gotta find a jumper to disable it and see what's up.