Expected life expectancy of Atari 8-bit hardware?

[Gunstar]

On 11/6/2019 at 1:08 AM, Sugarland said:

IIRC the 1200XL is known for having cold solder joints. Also the video quality is known to be bad. Though the one I just got stock video looks good so I don't know what are the statistics. The 1200 has many components on the motherboard, increasing failure risk. 600/800XL were not merely cheapening of the platform, but a very needed optimization.  I agree the 1200 does feel more like a premium quality product. Its external aesthetic design is awesome.

I've had 4 1200XL's come through my hands, I still own three. I've never had a cold solder joint issue either and they all worked fine when I got them, except for the known keyboard trace issue at the connector, once fixed, all the (best 8-bit home computer keyboard ever) keyboards worked great and in a decade have not had another keyboard issue. I have modded and upgraded my 1200XL's more than any other Atari and I never had a problem. The very first time I tried to remove a bad IC from an 800XL. soldered to the mobo, I lifted some traces in the attempt had had to repair them, the issue continued and after 3 IC removals with pulled traces or pads, it actually caused me to decide to upgrade my de-soldering tools for the first time in 20 years.

 

The ONLY thing keeping us (1200XL owners)from referring to the 1200XL as a TANK like we (800 owners-yep that one too) is the heavy metal shielding on the 800 compared to the tin shielding in the 1200XL (which is far ticker and sturdier than other XL/XE machines still-by a large margin).

 

But the aesthetics and keyboard of the 1200XL makes it my favorite and I quickly sold off the crappy-build-quality  130XE (I owned from 1985-2006, which I had to replace the keyboard and/or mylar 3 times, though the XE never quit working, but the case was so cheap that over time it became warped and the function keys and case were no longer flush and it looked terrible) the minute I upgraded my first 1200XL to 256K.

Edited November 7, 2019 by Gunstar

[Faicuai]

Interesting subject, indeed... 

 

Generally speaking, the simplest and most straight-forward indicator could be how frequent (and how many) issues owners report regarding a specific model... If you see little issues reported, you may consider it overall reliable. This, of course, depends on the model's popularity and intrinsic HW design / quality, which are the two key factors that will make these issues surface or not.

 

As a past and present owner of 400, 800, 800XL and 130XE, this is what I've found, so far (my apologies if lengthy):

 

1. Decay when exposed to friendly environment (e.g. low humidity, no sun-light, controlled-temp, etc.): most HW here seems to fare well on easy / forgiving environment, in general. The typical materials-decay (e.g. items glue such as 400/800 door-shields, plastics yellowing, mylar-traces on 800/XL/XE keyboards, corrosion of metal components, corrosion on IC's/ chips' pins, etc.) they all seem to advance pretty slowly. Would be up to external power-supplies reaction to longer periods of inactivity.
2. Decay when exposed to un-friendly environment (e.g. high humidity, indirect sun-light, hot environments, etc.): here's were things become more interesting (and delicate). On a scenario of higher-humidity, you will see cart-door RF-shields de-gluing on 400/800, mylar traces decaying rapidly on ALL models when present, plastics will begin changing colors (being the ENTIRE XL-line the worst offenders). But where things are going to get really bad (and uneven across models) is on the CORROSION department:
   • On the 400/800 you will see the tin-can metal shell covering the RF modulator going gray/black, and the heavy-duty aluminum shield / cage developing some whitish spots along the most critically exposed surfaces. However, everything else contained INSIDE the aluminum-cage will fare relatively well, especially the MoBo where there are no wide-tracks / traces making direct contact with the outer-edges of the RF-cage. External yellowing will be moderate-to-high, depending on exposure, and original case color could be restored to pretty close its original form, and re-yellowing will hardly re-occur, if stored on friendly conditions.
   • The 600XL/800XL/1200XL/130XE, however, will not fare that well, and will sustain significant decay, especially under humidity. The outer screws will mostly oxidize, the internal shields will turn mostly BLACK, dark-gray, the Mobo wide-tracks connecting to the shield will go BLACK, and humidity will travel down to the core of HW, all the way to all large VLSI chips, RAM, ROM and supporting chips (pins / legs will turn dark-gray & black). Mylar components will decay with 100% certainty, even Type-1 keyboards (the best ones) will make false contacts on keys, and external plastics will turn markedly yellow, and will RE-YELLOW again after being treated and stored properly.
3. External materials: besides external coloring / yellowing, I have not been able to detect any inherent differences on the plastics strength / composition between the older 400/800 and newer XL/XE:
   • The reason for this is that I have drilled both, and have seen the reaction and by-products, as well as the surface reaction when exposing to drill-bit heat. No significant difference in strength / hardness, that I could see.
   • The differences, again, seem on the finish (e.g. the 400/800 have a consistently texturized exterior finish, whereas the XLs seem to vary, with a more pronounced yellowing on the latter, as well).
   • HOWEVER, there are differences in overall shell / chassis rigidity. In particular, the 800 has a TON o space inside, especially on the right-side, and top-and-plastic convers have longer surfaces and joints on the back and sides, which makes the shelf-rigidity dependent on where the actual edges (top-and-bottom shells) join. There are very few screw-studs or joint posts/guides along, except on the very outer edges where the screws tie-up everything. This will make the 800 crackle / pop along those lines clearly more pronounced than the 800XL/XE, where the joints are much tighter with a lot less shell-surface on those joints.
4. Decay during operation:
   • They 800 keyboards will surely developed cracked-plungers (and non-functional space-bars) especially of you suffer from heavy-and-fat fingers-typing. You would be better off with the 400 keyboard or an 800 XL ALPS, heavy-duty keyboard which is probably the best and most durable keyboard ever made.The 400/800 9V power supply are notorious for blowing fuses easily, although running on a beefy and substantial AC/DC stage hosted on the system, which is easy to access and overall reliable (with a possible rare exception on the power-transistors attached to the large heat-sink). This may also apply to the 1200XL, which is in reality half-800 and half-XL.
   • On the 800 & 130 XL/XE, you will be out-of-luck if you get an ingot-type power-supply, which once it goes bad (due to excessive heat) will blow a slew of passive components on your computer, as it injects a generous and lethal flow of A/C current into a electronics designed solely for DC. If you have a rebuildable external P/S, then you will be in good-shape.
   • On the 400/800, RAM-failure (and to some extent ROM-failure) is rather common among the issues reported. However, and in the case of the 800, serviceability is bar-none, because a) ALL I/C packages socketed, and b) replacing RAM or ROM is as easy as removing the expansion-bay top-cover, and swapping in / out the offending module. About a 2-min operation, assuming parts are available. The 400 is more closed, it will take more work.
   • On the XL/XE (including 1200XL), serviceability is not at the level of the 800, as I've had to FULLY disassemble the cases, remove all shields (which come attached in different and stupid ways), in order to be able to trouble-shoot and repair them (that is if and ONLY if the system is fully-socketed, which is not the case for the XL/XE line, beyond Rev.C / Hong-Kong MoBo, except the 1200XL where I have always seen sockets everywhere). 
   • I have had to replace more discrete and VLSI components from the 400 and XL/XE than any other models I have had. PIAs, POKEYs and ANTICS among the most common ones.
   • Now, in case of a passive (e.g. resistor or capacitor) failure, and possibly affecting a key sub-system, such as Video, I/O, etc., it could potentially be laborious and harder to trace down on the 400/800/1200XL rather than the simplified 600/800/130 XL/XE systems. That is my experience when repairing blown passives on 800XL and 130XE.
5. Expandability and post-install integrity: well, let's face it:
   • As soon as you see yourself removing shields, cutting MoBo traces, de-soldering chips to socket-them, drilling holes on the MoBo, soldering spaghetti-looking jumpers all over, packing sandwiches of ribbon-cables, chopping chunks of the external case, etc., you already know that such system was not really designed to be expanded. 
   • Neither the 400 (outside of its internal slots) nor the 130/600/800/1200 XL/XE systems were really designed for placing internal expansions, except for a PBI slot where present.
   • And as for the 800, which is being the most expandable of all, it also suffers from not being able to bring stuff OUT of it (due to the thickness and placement of its heavy-duty shields).
   • All in all, I would say that the 800 would be the platform better geared for INTERNAL expansions and overall post-expansion reliability (while keeping virtually all of its frame and components intact). As for the XL/XE, that could be the case as well but for a much more limited # of upgrades because, for the most part, you will be forced to shed internal components and introduce significant external changes to make these upgrades happen.
   • System stability-wise, it is all about what people report after the upgrades, and plenty of that available on this forum.

 

Edited November 7, 2019 by Faicuai
Typos, additional detail...

[ACML]

On 11/7/2019 at 10:26 AM, Gunstar said:

I've had 4 1200XL's come through my hands, I still own three. I've never had a cold solder joint issue either and they all worked fine when I got them, except for the known keyboard trace issue at the connector, once fixed, all the (best 8-bit home computer keyboard ever) keyboards worked great and in a decade have not had another keyboard issue. I have modded and upgraded my 1200XL's more than any other Atari and I never had a problem. The very first time I tried to remove a bad IC from an 800XL. soldered to the mobo, I lifted some traces in the attempt had had to repair them, the issue continued and after 3 IC removals with pulled traces or pads, it actually caused me to decide to upgrade my de-soldering tools for the first time in 20 years.

 

The ONLY thing keeping us (1200XL owners)from referring to the 1200XL as a TANK like we (800 owners-yep that one too) is the heavy metal shielding on the 800 compared to the tin shielding in the 1200XL (which is far ticker and sturdier than other XL/XE machines still-by a large margin).

 

But the aesthetics and keyboard of the 1200XL makes it my favorite and I quickly sold off the crappy-build-quality  130XE (I owned from 1985-2006, which I had to replace the keyboard and/or mylar 3 times, though the XE never quit working, but the case was so cheap that over time it became warped and the function keys and case were no longer flush and it looked terrible) the minute I upgraded my first 1200XL to 256K.

After reading through this thread again, I pulled out my five fully assembled 1200XLs and the three spare main boards (total of 8 examples).  I wanted to, one exercise the caps and two, verify they were all still "Code 1" (USAF slang for no write ups - ready for next mission).  All eight have the ClearPic 2002 modification and duel OS (800XL & Omniview) using the RF channel 3-4 switch re-purposed as the OS selector.  Every one of them fires right up, passes memory test and were indeed Code 1.  37+ years and not an issue.  That just blows my mind.  What other tech item can you point to that has a reliability like that?  Like I mentioned earlier, I've had dozens of 1200XLs come in and out and every one was just as reliable.  Can't say that for the 400s and 800s I've owned.  If you are an Atari aficionado, and if you're reading this, you probably are, you have to get one of these machines.      

Edited November 10, 2019 by ACML

[ACML]

On 11/7/2019 at 10:33 AM, Faicuai said:

The 400/800 9V power supply are notorious for blowing fuses easily, although running on a beefy and substantial AC/DC stage hosted on the system, which is easy to access and overall reliable (with a possible rare exception on the power-transistors attached to the large heat-sink). This may also apply to the 1200XL, which is in reality half-800 and half-XL.

The fuse on the 9VAC power supply blows because the 400/800 had a short and blew the fuse.  The issue is not with the 9VAC power supply.  I've only lost the 9VAC brick to 400 PWR boards that went bad.  Are you saying the 9VAC brick goes bad on its own?  Maybe I'm misinterpreting what you are implying. 

[BillC]

1 hour ago, ACML said:

The fuse on the 9VAC power supply blows because the 400/800 had a short and blew the fuse.  The issue is not with the 9VAC power supply.  I've only lost the 9VAC brick to 400 PWR boards that went bad.  Are you saying the 9VAC brick goes bad on its own?  Maybe I'm misinterpreting what you are implying. 

The original 400 power supply was only 15VA, this may be the 9VAC power supply known for commonly blowing fuses, Atari later shipped the 400 with the same 31VA power supply used with the 800/810.

[spacecadet]

I don't think I've ever had an actual failure of an Atari computer out of the five or six that I've owned. I did have an XL power supply fail (the ingot) but not the computer.

 

Like Mclaneinc said, I think the main issue is wear and tear on the physical components. In that sense, I think the XL's are actually the best built because they have fewer wear and tear items than the 400/800, they're lighter and their cases seem a little less brittle at this point in time. The one 800 I owned had several cracked and broken "ribs" on the top vent area, the foam on the cartridge door had completely disintegrated, the shield had detached, and the door hinge itself was broken in some way (I never really bothered trying to figure it out, but it just didn't open or close properly). The XL line just doesn't even have those items to wear out.

 

I have less experience with the XE line but I do have some, and they seem to be just generally cheapened. Thinner casing, mushier but stiffer keyboard, etc. They don't *feel* like computers built to last, at least, although I'm sure most of them are still working.

[ACML]

 

2 hours ago, BillC said:

The original 400 power supply was only 15VA, this may be the 9VAC power supply known for commonly blowing fuses, Atari later shipped the 400 with the same 31VA power supply used with the 800/810.

Ah yes, I forgot there was the low wattage 9VAC supply BITD.  I'm was thinking exclusively of the 31VA version of the C017945.  That now makes sense.  Thanks.

 

[BillC]

4 hours ago, ACML said:

 

Ah yes, I forgot there was the low wattage 9VAC supply BITD.  I'm was thinking exclusively of the 31VA version of the C017945.  That now makes sense.  Thanks.

 

I blew the fuse in my 15VA model soon after I upgraded my original 400 with a Mosaic 64K RAM board, it uses 32-4116 chips so 4 times the power consumption for the RAM compared to 16K, my 400 only came with 8K RAM. The later 4164 based 64K RAM board from Atari actually reduced power consumption.

 

Typical power consumption 465mW/4116: 64K upgrade increases consumption by 11.16W.

Even if mostly +12V that's about 1A. Even a 4116 based 32K board would increase it by 3.72W.

Typical power consumption 135mW/4164: 64K upgrade reduces consumption by 2.64W, that's a difference of 13.8W!

 

This shows why the 15VA power supply would blow fuses, depending on upgrades.

[Chilly Willy]

11 hours ago, BillC said:

I blew the fuse in my 15VA model soon after I upgraded my original 400 with a Mosaic 64K RAM board, it uses 32-4116 chips

Weird - my Mosaic 64K board uses eight 4164 chips. The 32K board (from Atari that I got originally with my A400) used more chips. I'd guess yours is an earlier model before the price on the 4164 dropped enough to make it usable.

 

[CommodoreDecker]

On 11/5/2019 at 3:43 AM, xrbrevin said:

old tech was not built to expire, thats why theyre still going strong now! ?

even the devices that have issues are repairable. diagnosis is the trickiest part, the actual component replacement is relatively easy due to through-hole PCB construction.

however, the XE range were built to a lower price so extra care must be taken with the PCB - they are prone to lifting traces. i suppose this could be an indication of 'the older the better'?

 

To my current understanding -- older technology, in general, was created under a vastly different set of tools and fabrication processes - at varying stages of standardization for various materials, though by the late-1970s they were probably built a bit cheaper compared to a couple years prior when they went overkill (not unlike the A800's excessive RF shielding compare to the 1200XL, hehe.)  Notice how the microchips circa 1980 were typically made in a 6 or 3 micrometer circuit path thickness whereas today's chips are made with a much smaller 10nanometer, hence in part the radically higher transistor count in today's newer chips.  (1.5um was reached (but not a ubiquitously used process) in 1981, 100nm around 2000, 5nm in 2019...  )

 

So not only are the newer chips' trace paths much smaller and using less material, as well as being faster in part (but not exclusively) because the electrons have a much smaller distance to travel between transistors, they're going to be more susceptible to electromigratrion - or where electricity jumps across the traces, leading to system instability. Hence lower voltage and amperage requirements and in part why a modern PC USB port, if enough amperage is provided by the port (!!), could even power an 800XL.

 

Noting your reference of the XE line: There's a ton of chatter on how numerous units have poor quality RAM.  (And if a person is selling an XE, even with RAM Expansion kit, ask them to do a self-test and send a photo of the results.)  It has also been said the XE has comparatively brittle plastic.

 

Heat is a big killer of electronic components, whether it be ambient to components surrounding the chip or generated by the chip itself. The more heat will wear out the chip faster. The hotter it is the faster it will be. System instability like lockups is an early indicator of a heat-related problem within a chip, or if enough ambient heat melts and thus weakens a soldered conductivity joint (which is more the case for smartphones these days, older computers rarely if ever had that problem). For now I'll focus on the CPU, but it's true for any IC in general. Anyone who says "it's designed to still work at 100C so don't worry" when the manual shows the TDP being 100 or even 105C is somewhat unaware and I'd love to namedrop a forum where people actually believe such malarkey. TDP is when it fails but in real life it's still best to never ever run it anywhere that hot to begin with. And any generic Intel PC overclocking site (where everyone spazzes if a person reports a temperature of 50C or higher) will corroborate that. But all modern computers use Intel chips so no brand has any special magic to defy the laws of physics and they're all made in the same plant with the same processes and materials too. But I digress - as a chip is used, over time, it will get warmer to the touch. A hot chip is pretty much defective or is about to be and there's little that can be done. Adding in little makeshift heatsinks only goes so far, noting the Atari 7800 game console has a massive heatsink on a voltage regulator module (VRM), but nothing on the individual microchips. But from what I can tell so far, they don't get warm enough to the touch and if they did then they should have had one mounted at the factory. Never ignore cooling, but these older systems never needed it - which adds to their net lifespan from the get-go because they don't get warm to the touch.  Now if one wanted a 7800 for centuries, making a specially designed case subjected to liquid nitrogen cooling might help... It's also cost-prohibitive...

 

And, of course, there are the effects of good old fashioned entropy. What is the optimum half-life of each material on a system board? The one material with the shortest half-life is your best case scenario. Then compare to how much is being stressed the moment the power is turned on?  We do know the chassis is made out of plastic, though depending on model and build date there could be impurities in the plastic or a different combination of chemicals that comprise it. Some forms of plastic will shatter after a couple decades' worth of regular use and especially if it gets warm enough (like a CRT television set.)

 

But, as a crude estimation, for - say - an Atari 2600 video game console purchased in 1980. For up to five years it may have been used an hour per day on weeknights followed by perhaps three hours each weekend day. So that's 11 hours of usage. Assuming the power brick delivers clean energy output, as noisier energy will degrade components faster as well as showing weird waves across the monitor/TV, etc... so that's 14,300 hours. Then the unit is sold or thrown out or put into storage (and how - like how much dust, stored in the dark or in broad daylight, temperature and humidity, was their an earthquake or hopefully not bird with gastrointestinal problems, etc, etc) and forgotten about and that's where an even wider array of variables takes over.

 

Never mind how many power surges have existed, quality of the power supply unit, and so on... video game consoles like these, the NES, etc, aren't like that top tier video card that sold new for $1000 but two years later sells for $800 and the seller is not going to say if overclocking was done on it or not, and chances are it was used for far more than an hour per day...

 

Sadly, the 7800 appears to be a soft-based power switch - I'd rather see a hard switch at the AC line so power isn't constantly flowing through the device even when switched "off"... and, of course, as with power surges, anything sitting idle but receiving power initially will get a little extra stress too. Some systems incorporate methods to soften this blow, others do not... but I digress again.

But all in all, any 8-bit system under ideal conditions will outlast any modern piece of kit (that hasn't been overclocked) under the identical environmental considerations.

 

I've an 800XL that powers on to a faceless green screen - but whoever sold it either got it secondhand or swapped good RAM chips for generic timer chips that didn't even have the same pin count. The chassis is still of nominal use as is the keyboard, and other chips if they're good. All I know is, some of the RAM or RAM controller are toast. Makes you wonder what the computer's history was, what snuffed it to prompt a chip swap and then to resell the thing to someone who didn't know who then sold it to someone else citing "untested", which is possible since not everybody has a RF modulator adapter or is going to care - they just want to clean up their garage.

 

Edited November 11, 2019 by CommodoreDecker

[Pig]

I have what might as well be a brand new 600XL mint in box, still it's natural white as if being opened for the first time 30 years ago. Still wrapped in the plastic, too. I haven't tested it out yet because I actually don't have a TV it can hook into, so I may do a video when I finally do of what that'll be like.

 

That said, I'm more curious about how long the 5.25 disks last. I picked up a collectors edition of H2G2 with all the nifty swag included, and again, until I get everything hooked up, I have no way of knowing. But if it (the disk) doesn't work, I still have my precious relic anyhow.

[pjedavison]

I've had two power supplies just stop working, seemingly after being plugged in for too long -- one for an 800XL, one for a 130XE -- and a 1050 just started giving "BOOT ERROR" messages without even spinning up when I turned the connected system on, but the computers themselves are absolutely fine. They're the same computers I grew up with, but there was probably a 20+ year gap where they'd been dormant in my parents' loft before they were kind enough to give them to me.

[Gunstar]

59 minutes ago, pjedavison said:

I've had two power supplies just stop working, seemingly after being plugged in for too long -- one for an 800XL, one for a 130XE -- and a 1050 just started giving "BOOT ERROR" messages without even spinning up when I turned the connected system on, but the computers themselves are absolutely fine. They're the same computers I grew up with, but there was probably a 20+ year gap where they'd been dormant in my parents' loft before they were kind enough to give them to me.

What do you consider plugged in for too long? PSU's won't last forever, of course, but besides a few years in storage over the last 35+ years, my PSU are ALWAYS plugged in, 24-7-365 and I've have yet to have one die on me for that reason. I've had fuses blow, but only because the device (I'm pretty sure only disk drives) developed issues which blew the fuses, a quick 5 min. repair for me. Of course my (legacy) equipment, except for 2 devices, all use the 31VA "universal" PSU.

 

My 130XE's PSU, which I sold after using it for about 20 years of similar use still worked when I sold it. I am repairing a 800XL right now that had the infamous "ingot" PSU that fried it, but I am now using it with a PSU that isn't an "ingot" from a salvage 600XL and it still works, but I don't know it's history as it was given to me. But I can't say if my 130XE's PSU would still be working today if I'd kept it and continued to leave it plugged in for the last 14 years since I sold it. Maybe these later 5V PSU's don't last as long?

 

Longevity of the PSU's is one reason why I haven't done the modification to my 1200XL's for direct 5V PSU's; I highly doubt that on of the modern, small, wall-wort PSU's used by those that have done such modifications will last as long, and they may even cause issues with the computer when they do go out. As the old saying goes..."they don't make them like they used too." Maybe they are much more power efficient, but my electric bill has certainly never been noticeably larger due to old PSU's always being plugged in. 

 

As a general rule, from my experience, electronics last longer with at least occasional use, rather than sitting in storage for years. Also, solid-state electronics without moving parts will outlast any that has mechanical moving parts, so the computers should work fine much longer than disk drives or tape drives and printers with moving mechanical parts. I'd be willing to bet that any cartridge based consoles (like the 2600), even being 40+ years old, will still out last any more modern consoles with disc drives, HD drives and fans. In fact I have proof of it with a 2600, 7800 and Jaguar which are all still working perfectly and then there are the dozen non-working Xbox 360's and PS2's I have been given and am salvaging to make one or two working units out of them all.

Edited November 12, 2019 by Gunstar

[pjedavison]

Hard to say really, but both of them seemed to go after I'd left them plugged in (but the computer not on) all day. Could well have been a coincidence, but it's strange that it happened twice. As you say, though, I'd also expect these bricks to be able to stand up to a day of being plugged in without any difficulty! I haven't done any scientific testing or anything, just to be clear; it's just my best guess as to what fried them.

 

I remember when I was a kid my Dad always insisted we turned all the computer plugs and stuff off when we went to bed at night. I'm wondering if he encountered a similar issue at some point.

[qrysdonnell]

 

1 minute ago, Gunstar said:

What do you consider plugged in for too long? PSU's won't last forever, of course, but besides a few years in storage over the last 35+ years, my PSU are ALWAYS plugged in, 24-7-365 and I've have yet to have one die on me for that reason. I've had fuses blow, but only because the device (I'm pretty sure only disk drives) developed issues which blew the fuses, a quick 5 min. repair for me. Of course my (legacy) equipment, except for 2 devices, all use the 31VA "universal" PSU.

 

My 130XE's PSU, which I sold after using it for about 20 years of similar use still worked when I sold it. I am repairing a 800XL right now that had the infamous "ingot" PSU that fried it, but I am now using it with a PSU that isn't an "ingot" from a salvage 600XL and it still works, but I don't know it's history as it was given to me. But I can't say if my 130XE's PSU would still be working today if I'd kept it and continued to leave it plugged in for the last 14 years since I sold it. Maybe these later 5V PSU's don't last as long?

 

Longevity of the PSU's is one reason why I haven't done the modification to my 1200XL's for direct 5V PSU's; I highly doubt that on of the modern, small, wall-wort PSU's used by those that have done such modifications will last as long, and they may even cause issues with the computer when they do go out. As the old saying goes..."they don't make them like they used too." Maybe they are much more power efficient, but my electric bill has certainly never been noticeably larger due to old PSU's always being plugged in. 

 

As a general rule, from my experience, electronics last longer with at least occasional use, rather than sitting in storage for years. 

 

My take on all of this is that problems are rare enough that it's not worth 'doing' or 'not doing' anything about it. I keep my stuff plugged in. If it breaks I replace it. For the most part nothing breaks, so I'm not replacing things. And let's face it, I have 6 Atari 2600s in storage, and I could likely be dead before they all break on their own.

 

 

 

[Gunstar]

@pjedavison

Well, your father was right, in as much as doing it to save power I suppose, and can't hurt, unless there is a power surge when they are plugged back in. That's the reason my father always wanted everything off at night. But if you don't have them plugged into surge protectors (one thing I ALWAYS do) it could have just been a slight power-surge on that particular day that ended their lives. It might not have even been noticeable and you didn't realize it. This would be my educated guess, since both died on the same day, probably a one-in-a-million chance if there was no power surge.

Edited November 12, 2019 by Gunstar

[BillC]

On 11/11/2019 at 6:29 AM, Chilly Willy said:

Weird - my Mosaic 64K board uses eight 4164 chips. The 32K board (from Atari that I got originally with my A400) used more chips. I'd guess yours is an earlier model before the price on the 4164 dropped enough to make it usable.

 

I haven't had the board out in years to verify the chips, but what I do remember is that the board even extended down alongside the edge connector for more surface area and was packed.

I also purchased it late 1982/early 1983, so around when the 1200XL was introduced.

I would think that the 4164 would have dropped enough in price by then, it's possible the card was manufactured before that happened or that Mosaic was reducing existing inventory of 4116 chips before releasing a new 4164 version.

Edited November 12, 2019 by BillC

[Chilly Willy]

Yeah, sounds like you got old stock. Bummer. I later bought some 256Kx1 DRAM that I planned to use in modifying my Mosaic card with, but then got an Amiga and moved away from the various Atari plans I had at the time. Still got those DRAM.

 

[ClausB]

On 11/10/2019 at 9:31 PM, BillC said:

I blew the fuse in my 15VA model soon after I upgraded my original 400 with a Mosaic 64K RAM board, it uses 32-4116 chips so 4 times the power consumption for the RAM compared to 16K, my 400 only came with 8K RAM. The later 4164 based 64K RAM board from Atari actually reduced power consumption.

 

Typical power consumption 465mW/4116: 64K upgrade increases consumption by 11.16W.

Nope, that's max power, not typical power, and it's at the chip's max cycle rate. At the Atari's rate of 1.8 MHz, power is considerably less. Also, depending on whether the Mosaic design keeps 24 chips in standby when 8 are active, consumption can be even lower. See analysis on 48K here: