Jump to content
IGNORED

Replacing stock motherboard IC's


 Share

Recommended Posts

While I understand the key chips inside all the versions of the A8 were custom, the majority of ICs on the motherboards seem to be totally stock. I wonder if it is worth spending £10 or so and swapping out the old 1980's vintage DIPS for modern ones? I ask because I had a truly infuriating bastard of an intermittent fault with the "SIDE2" that was only solved by replacing the "SN74LS08N" IC with one from a slightly newer "REV D/REV 4" motherboard. Looking up the product code online shows you can buy a brand new version of that chip made by Texas Instruments for 44p from 'Farnell' in the UK:

 

http://uk.farnell.com/texas-instruments/sn74ls08n/logic-quad-2-inputgate-14dip/dp/1740023

 

I dare say the same is true for many of the others - especially the RAM, decoders and other 'glue' logic.

Link to comment
Share on other sites

Well, I doubt there are very many "mission critical" Atari computers nowadays but you could probably decrease the chance of a failure with some refurbishing. Some things to keep in mind:

 

1. Some computers were totally socketed. I'd much rather track down an 800XL full of sockets than spend the hours desoldering everything.

 

2. Some chips won't be much newer than what you have. When did they last make 64Kx1 DRAMs?

 

3. Think about heat-sinking the 6502 and maybe Antic. Helping the big chips dissipate heat can't hurt.

 

4. Make sure your power supply isn't over 5V. If your power supply errs on the high side, you might get more life swapping it for another one.

  • Like 1
Link to comment
Share on other sites

Well, I doubt there are very many "mission critical" Atari computers nowadays but you could probably decrease the chance of a failure with some refurbishing. Some things to keep in mind:

 

1. Some computers were totally socketed. I'd much rather track down an 800XL full of sockets than spend the hours desoldering everything.

 

2. Some chips won't be much newer than what you have. When did they last make 64Kx1 DRAMs?

 

3. Think about heat-sinking the 6502 and maybe Antic. Helping the big chips dissipate heat can't hurt.

 

4. Make sure your power supply isn't over 5V. If your power supply errs on the high side, you might get more life swapping it for another one.

 

Excellent points Bryan and just the sort of advice I was looking for.

 

In regards the sockets, I am pleased to say I have two "REV C/3" motherboards which are i think 100% socketed - and one of which actually boots! - and a "REV D/4" with is perhaps 20% socketed if that. The latter also boots... for about fifteen minutes and then crashes to a lockup and corrupted screen. I suspect there is a broken track that parts when the heat rises too far. My 'working' machine is a bit of a Frankenstein's monster already with bits form the other two - specifically that <expletive of choice> 'SN74LS08N' from the REV 4 and a replaced 6502 from its sister. Sadly with just a little hubris one evening I replaced the CPU in back to front and was rewarded with a red screen... Nevermind - I have two more of those! Do you know if the 6502 used in the A8 is unique or if it can be bought away from dedicated Atari suppliers?

 

I also wondered about the TI part to be honest. I suspect it might not be physically that much younger than the ones on the boards, but hopefully has been kept in ideal conditions and also not subjected to use/heating.

 

In regards cooling are there any custom heatsinks available or is it a question of making one's own from strips of aluminium or whatever? I think I quite badly need to heatsink the ANTIC and CPLD chips that are attached to the VBXE. The other week I went back to the machine after accidentally leaving it on all evening in 'CON 80' mode and found the monitor full of rolling garbage. If I switched off and then back on I got a couple of minutes of service before it began to roll again and lock up. If I took the output from the standard monitor port I got a normal - if horrible composite video! - screen. I am pretty sure it is an overheating problem as that has not happened after I left it alone for a couple of days and tested it again. I was wondering about how to tackle it.

Link to comment
Share on other sites

There are adhesive thermal pads and heat sinks available that should do the trick nicely. I'd try something like the Aavid 507302B00000G (Digikey HS115-ND) and flip it upside down if needed for clearance reasons. The fins can be bent or cut back so it doesn't touch anything. Stick it down with something like the T-Global DC0011/05-L37-3F-0.25-2A (Digikey 1168-1794-ND). For this application, things like super-glue would probably work okay too.


These parts are meant for TO-220 packages, but should work nicely. If you visit the Aavid site, you might find something you like better.


The 6502 is (CO14806) is a custom Atari part.

Link to comment
Share on other sites

 

There are adhesive thermal pads and heat sinks available that should do the trick nicely. I'd try something like the Aavid 507302B00000G (Digikey HS115-ND) and flip it upside down if needed for clearance reasons. The fins can be bent or cut back so it doesn't touch anything. Stick it down with something like the T-Global DC0011/05-L37-3F-0.25-2A (Digikey 1168-1794-ND). For this application, things like super-glue would probably work okay too.
These parts are meant for TO-220 packages, but should work nicely. If you visit the Aavid site, you might find something you like better.
The 6502 is (CO14806) is a custom Atari part.

 

 

Many thanks Bryan!!!! I will give those parts a look over and see if I can get hold of them from a local supplier. I've had problems with digikey recently - utter nonsense where I had to sign some sort of agreement not build a hydrogen weapon out of three microswitches I bought to repair my Logitech trackball...!!!

  • Like 1
Link to comment
Share on other sites

heh. The issue you had with the SN74LS08N IC was not that the IC was bad. It was that the propogation timing differered from the IC that you replaced it with. That chip is directly involved in the buffering of the PHI2 signal from the CPU, which is the basis for read/write timing system wide.. As literally VOLUMES of pages have been written about on here before, plugging external devices onto the data bus (as is the case with SIDE2 and anything else that communicates directly in parallel with the CPU on an Atari) causes additional capacitance on the bus and can further increase discrepancies in PHI2-to-bus timing. The way this manifest itself is that many devices are succeptible to PHI2 timing variances in the form of instability issues when that device is connected.

 

So no, this has no bearing on whether or not you should replace the rest of the ICs in the machine. I am glad you solved your SIDE2 stability issue though.

  • Like 1
Link to comment
Share on other sites

While I understand the key chips inside all the versions of the A8 were custom, the majority of ICs on the motherboards seem to be totally stock. I wonder if it is worth spending £10 or so and swapping out the old 1980's vintage DIPS for modern ones? I ask because I had a truly infuriating bastard of an intermittent fault with the "SIDE2" that was only solved by replacing the "SN74LS08N" IC with one from a slightly newer "REV D/REV 4" motherboard. Looking up the product code online shows you can buy a brand new version of that chip made by Texas Instruments for 44p from 'Farnell' in the UK:

 

http://uk.farnell.com/texas-instruments/sn74ls08n/logic-quad-2-inputgate-14dip/dp/1740023

 

I dare say the same is true for many of the others - especially the RAM, decoders and other 'glue' logic.

Replacing TTL's only makes sense if they are of the 1st generation (i.e. 74xx) or if they're suspected to be the cause of a problem. The 1st generation TTL's used a lot of power and (because of that) produced a lot of heat. Usually you can swap these for later generation TTL's, like 74LSxx, 74HCxx or 74HCTxx. These use far less power, produce less heat and can be clocked at a higher speed. In some cases using non-LS TTL is required, however, especially if the IC has to drive a relative large load and sink a lot of current. That's not applicable in the A8, AFAIK all TTL's in the A8 are of the low power 74LSxx type.

 

You could revert to replacing TTL's with new ones to try to get your non-functional A8 back on its feet again, as buying them will cost you about 20 GBP or so at max. Swapping them over from the other A8 will cost you nothing, of course ;-)

I'd start by reseating every IC in its socket, chances are contact resistance between the IC's pins and the socket is causing the random lockups you're experiencing.

 

As Ken said, it was probably just a lucky shot that your troubles went away after swapping the LS08 with a another one.

 

re-atari

Edited by re-atari
Link to comment
Share on other sites

heh. The issue you had with the SN74LS08N IC was not that the IC was bad. It was that the propogation timing differered from the IC that you replaced it with. That chip is directly involved in the buffering of the PHI2 signal from the CPU, which is the basis for read/write timing system wide.. As literally VOLUMES of pages have been written about on here before, plugging external devices onto the data bus (as is the case with SIDE2 and anything else that communicates directly in parallel with the CPU on an Atari) causes additional capacitance on the bus and can further increase discrepancies in PHI2-to-bus timing. The way this manifest itself is that many devices are succeptible to PHI2 timing variances in the form of instability issues when that device is connected.

 

So no, this has no bearing on whether or not you should replace the rest of the ICs in the machine. I am glad you solved your SIDE2 stability issue though.

 

That is absolutely fascinating MEtalGuy66 and more of the information i was hoping to gain!!!

 

I think however there must have been at least some issue with the chip, even if that condition fell short of actually rendering it genuinely unserviceable. Especially as when I first replaced it with the same, rather beaten-up and cheap looking IC from the same revision motherboard the problem's persisted. Perhaps the 'new' chip has more narrow tolerances, or itself ran closer to specification therefore once installed the system as a whole had more leeway that allowed the disruption from the extra demands from the "SIDE2", "SimpleStereo" and "VBXE" to not push the PHI signal entirely out of its operating envelope? Is it significant that the 'SIDELoader' functionality was broken even before I had installed any mods at all - simply when attached as a cartridge to a totally stock system it didn't work? Might that not suggest a problem or at least an already fairly wonky timing situation existing inside the unmodified system?

 

This directly suggests a wider and rather dispiriting conclusion. If installing mods blindly is likely to cause these problems then what can we do deliberately to make sure they don't happen? How can we ensure the propagation for this and other signals remain within stable limits? From what you are saying it almost sounds like applying mods is a hit-and-miss affair with no real expectation of them actually working. Currently I do not know enough about the electronics of the system to be able to refute that stance, but it is not very encouraging for the further development and installation of mods. There seems to be a great deal of 'Well... It might work.' or at best 'well... It will probably work.'... If things are so delicately balanced then the mod vendors really should warn about the potential - or worse, likelihood - of their electronics not working.

Link to comment
Share on other sites

1. Some computers were totally socketed. I'd much rather track down an 800XL full of sockets than spend the hours desoldering everything.

 

well, keep in mind though that the sockets atari used in most of the XL line were single wipe only. They were using pretty much the cheapest sockets they could get their hands on so they could compete with prices. Some of these sockets may only be rated for a handful of insertions, and trust me, they do break. It's probably better to just replace any chip you replace with a precision socket, or at least a double wipe socket while you have it apart.

The 1st generation TTL's used a lot of power and (because of that) produced a lot of heat. Usually you can swap these for later generation TTL's, like 74LSxx, 74HCxx or 74HCTxx.

Well, 74HCxx might work, but it's NOT something you should regularly do. If the part was originally HC, then replace it with that, but HC is NOT a TTL level part, it's CMOS and the logic levels differ. In many cases, it'll work just fine, but there's no point in buying HC when you can get HCT and have it be TTL levels, and not have to deal with that potential issue.

 

 

That is absolutely fascinating MEtalGuy66 and more of the information i was hoping to gain!!!

 

I think however there must have been at least some issue with the chip, even if that condition fell short of actually rendering it genuinely unserviceable. Especially as when I first replaced it with the same, rather beaten-up and cheap looking IC from the same revision motherboard the problem's persisted. Perhaps the 'new' chip has more narrow tolerances, or itself ran closer to specification therefore once installed the system as a whole had more leeway that allowed the disruption from the extra demands from the "SIDE2", "SimpleStereo" and "VBXE" to not push the PHI signal entirely out of its operating envelope? Is it significant that the 'SIDELoader' functionality was broken even before I had installed any mods at all - simply when attached as a cartridge to a totally stock system it didn't work? Might that not suggest a problem or at least an already fairly wonky timing situation existing inside the unmodified system?

 

This directly suggests a wider and rather dispiriting conclusion. If installing mods blindly is likely to cause these problems then what can we do deliberately to make sure they don't happen? How can we ensure the propagation for this and other signals remain within stable limits? From what you are saying it almost sounds like applying mods is a hit-and-miss affair with no real expectation of them actually working. Currently I do not know enough about the electronics of the system to be able to refute that stance, but it is not very encouraging for the further development and installation of mods. There seems to be a great deal of 'Well... It might work.' or at best 'well... It will probably work.'... If things are so delicately balanced then the mod vendors really should warn about the potential - or worse, likelihood - of their electronics not working.

no, there's really not anything you can do to guarantee you won't have timing issues. Atari's are known to sometimes have marginal phi2 signals, and those that do barely work on their own. Adding mods is usually a hit or miss thing, although it usually is not miss, if the mod is designed well. There's really nothing you can do to help the situation, it's the nature of NMOS and capacitive loading that cause these issues. You'd likely have to redesign the whole motherboard, and the atari custom chips as well, to use something other than NMOS, and for the board to be better laid out to reduce some of the capacitive loading caused by the traces themselves. NMOS does not 'drive' a 1 very strongly, it's internally a resistor and with enough capacitive loading, signals have reduced rise time and they can become problematic if it gets bad enough. Speeding up the parts in the path of that signal to have lower latency can help the issue. If you had replaced that part with a 74F08, it probably would have worked better as well. NMOS actually drives the output lines low, so usually fall time is not as much of a problem, but in the case of a signal like phi2, where it goes literally everywhere, and has trace inductance, as well as capacitance, you can still get issues.

Link to comment
Share on other sites

 

well, keep in mind though that the sockets atari used in most of the XL line were single wipe only. They were using pretty much the cheapest sockets they could get their hands on so they could compete with prices. Some of these sockets may only be rated for a handful of insertions, and trust me, they do break. It's probably better to just replace any chip you replace with a precision socket, or at least a double wipe socket while you have it apart.

Well, 74HCxx might work, but it's NOT something you should regularly do. If the part was originally HC, then replace it with that, but HC is NOT a TTL level part, it's CMOS and the logic levels differ. In many cases, it'll work just fine, but there's no point in buying HC when you can get HCT and have it be TTL levels, and not have to deal with that potential issue.

 

no, there's really not anything you can do to guarantee you won't have timing issues. Atari's are known to sometimes have marginal phi2 signals, and those that do barely work on their own. Adding mods is usually a hit or miss thing, although it usually is not miss, if the mod is designed well. There's really nothing you can do to help the situation, it's the nature of NMOS and capacitive loading that cause these issues. You'd likely have to redesign the whole motherboard, and the atari custom chips as well, to use something other than NMOS, and for the board to be better laid out to reduce some of the capacitive loading caused by the traces themselves. NMOS does not 'drive' a 1 very strongly, it's internally a resistor and with enough capacitive loading, signals have reduced rise time and they can become problematic if it gets bad enough. Speeding up the parts in the path of that signal to have lower latency can help the issue. If you had replaced that part with a 74F08, it probably would have worked better as well. NMOS actually drives the output lines low, so usually fall time is not as much of a problem, but in the case of a signal like phi2, where it goes literally everywhere, and has trace inductance, as well as capacitance, you can still get issues.

 

I'm getting repetitive, but once again many, many thanks Joey Z - more fascinating material I wanted to know!!! What really shames me is I used to know this stuff - or at least had exposure to it and a comfort level. Bloody hell - age has nothing to recommend itself.

 

Would you suggest I should get hold of a '74F08' unit? Could you point me in the direction of one?

 

Something occurs to me; swapping out that IC solved the problem... for my current "Ultimate1MB"/"VBXE" setup. I still have a replacement "SimpleStereo" coming in the post. I am somewhat concerned that is going to screw the timing further and I am going to face a dodgy 'SIDELoader' all over again once I install it!

Link to comment
Share on other sites

There is a solution for buffering PHI2 without delay, regardless of capacitance load (within a reasonable range).

http://www.atari.org.pl/forum/viewtopic.php?id=12727

 

That at least sounds like something I would like to have a try at. I just wish I spoke Polish!!! Is there any chance you could translate that post for me Simius? The forum won't let me download the *.PNG either.

Link to comment
Share on other sites

 

I'm getting repetitive, but once again many, many thanks Joey Z - more fascinating material I wanted to know!!! What really shames me is I used to know this stuff - or at least had exposure to it and a comfort level. Bloody hell - age has nothing to recommend itself.

 

Would you suggest I should get hold of a '74F08' unit? Could you point me in the direction of one?

 

Something occurs to me; swapping out that IC solved the problem... for my current "Ultimate1MB"/"VBXE" setup. I still have a replacement "SimpleStereo" coming in the post. I am somewhat concerned that is going to screw the timing further and I am going to face a dodgy 'SIDELoader' all over again once I install it!

well, I don't know that you'll need the 'F08, but if you have issues with SimpleStereo, then it's worth a try. Also worth a try would be a 74ACT08 (not 74AC08). You can find these from many chip vendors, I don't live in GB like you, so I can't really point you to any suppliers that ship there. You're best off looking on your own, or asking someone who lives there.

Link to comment
Share on other sites

At least the pictures.

I'd be a bit worried about using a PLL to do phi02. PLL's have startup times before they lock to the input frequency, they're not instant. Maybe it has more than enough time to lock though. You have to hope that the atari stays in reset long enough for the PLL to lock and stabilize before the system tries to start, or you'll basically have the same problem but a lot worse until the PLL locks. as it is, the atari probably has to stay in reset until the normal crystal oscillator starts up. I'm not sure how close this timing is already, but on some machines the PLL might take too long and cause a new issue that wasn't already there.

 

Nevertheless, if a machine is having issues which may be phi2 related, then I suppose it's worth a try, but I wouldn't just do it to a machine for no reason. If it weren't for the PLL lock/stabilization times, it seems like it'd have no downsides though.

Link to comment
Share on other sites

well, I don't know that you'll need the 'F08, but if you have issues with SimpleStereo, then it's worth a try. Also worth a try would be a 74ACT08 (not 74AC08). You can find these from many chip vendors, I don't live in GB like you, so I can't really point you to any suppliers that ship there. You're best off looking on your own, or asking someone who lives there.

 

No worries Joey Z! I thought they might be a bit specialized, but it seems Farnell have them all.

 

The "SN74ACT08N" goes for 75p:

 

http://uk.farnell.com/texas-instruments/sn74act08n/logic-quad-2-in-pos-and-gate-14dip/dp/1749910

 

However it seems the "SN74F08" has been replaced with the "SN74F08N", which they claim is a direct replacement and isn't going to break the bank at 66p:

 

http://uk.farnell.com/texas-instruments/sn74f08n/logic-quad-2-inputgate-14dip/dp/1740005

 

In both cases it looks like the 'N' variation is the more modern unit, unless that brings with it some incompatibility I think I will order a couple of both just to cover all eventualities.

Link to comment
Share on other sites

 

and

 

At least the pictures.

 

Very cool guys!!! Are those little devices still available Simius? If so could you give me an idea how much one would cost?

 

Presumably they are a direct slot-in replacement for the "SN74LS08N"?

Link to comment
Share on other sites

well, I don't know that you'll need the 'F08, but if you have issues with SimpleStereo, then it's worth a try. Also worth a try would be a 74ACT08 (not 74AC08). You can find these from many chip vendors, I don't live in GB like you, so I can't really point you to any suppliers that ship there. You're best off looking on your own, or asking someone who lives there.

 

Looking at the data sheet for the "SN74F08N" and "SN74ACT08N" it appears the former is the more responsive IC. The switching time for the latter is actually almost double on average, however it does have better survival characteristics in case I want to take my Atari to the North Pole or the Gobi!!!

Link to comment
Share on other sites

 

and

 

 

Very cool guys!!! Are those little devices still available Simius? If so could you give me an idea how much one would cost?

 

Presumably they are a direct slot-in replacement for the "SN74LS08N"?

 

Direct replacement. Cost about US$5 + shipping. Unfortunately, at this time there are only bare, unassembled PCBs.

Link to comment
Share on other sites

 

Direct replacement. Cost about US$5 + shipping. Unfortunately, at this time there are only bare, unassembled PCBs.

No worries Simius! If/when you do get some completed modules would you mind giving me a PM? This sounds like it might be the 'magic wand' to cure a lot of my marginal - and apparently timing/capacitance-related - problems.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
 Share

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...