How Hot Do These 8-bit Computers Get?

[@tubeMonger]

Would be nice to see both, Fahrenheit and Celsius.

[@Psythik]

Yeah seriously; IDK why he used Fahrenheit when computer parts are *always* measured in Celsius, even in the US.

[@microcolonel]

You guys are ridiculous. 😂 Stop being ignorant, go learn Fahrenheit.

[@terosaarela4555]

Back in the 80s, a thermal camera would’ve required a trolley for the hydrogen tank that was needed, and it would’ve cost tens of thousands.

[@RetroHackShack]

Yeah. We've come a long way.

[@infinitecanadian]

CelGenStudios has one from the late 1970s that is small enough to be held.

[@Ramdileo_sys]

@@RetroHackShack at 15:01 .. A capacitor gets hot for only one reason Aron... bad ESR / internal resistance... is draining lots of power... almost in shortcut...

[@raycreveling1583]

In the mid 90's Was a Firefighter/EMT. Our volunteer department got a grant for one of these cameras. I believe it was US $30,000 for comparison a new AED Defibrillator was less then 10K.

[@SeanBZA]

Note about the ceramic packages is that you can see that they are actually better thermal conductors than the plastic, as even the hottest ceramic package is evenly warm, with only a barely visible heat area around the actual die inside. Thus you see the plastic DIP packages are rated for up to 1W in the small DIP40 package, but the corresponding ceramic packages are allowed up to 3W of power dissipation under the same circumstances, and thus run cooler. That was the issue with the first production runs of MOSTEK, the ceramic chips could barely handle the power they dissipated, but the plastic packaged versions of the chips ran much hotter, and thus they tended to fail fast, as both the thermal cycling, and the high temperature, aged both the bond of the die to the leadframe, and also tended to shear off the bondwires as the package expanded more than the die did. Took a good few revisions to get power dissipation down, and minimise leakage current in the die, plus also using a thin conformal coat over the die after bonding, but before epoxy encapsulation, to act as an interface, though it did little to help with the die separating from the leadframe itself, which is why so many fail when hot, but work when cold. Yes did add heatsinks to many DIP40 IC's, using even a small one made a big difference in reliability.

[@kse3]

If I remember right the heat issue on the Apple II was from one or more components in the power supply, which you wouldn't have seen show up here because of the big metal cover over it

[@RetroHackShack]

I'll have to remember to check that next time I fix an apple two power supply.

[@SeanBZA]

@@RetroHackShack Normally the power transistors, the diodes and strangely enough the output capacitors and inductors, all running at high power level, and run at the limit. You can do a lot with new capacitors that are lower ESR, and new diodes of higher current rating.

[@infinitecanadian]

My dad has a FLIR. We have a tradition of going out in winter and using it to view around his house in Vancouver.

[@elfenmagix8173]

Two things: 1) the 6502 on the Vic20 gets very hot, especially the older Petscii ones. as well as that huge voltage regulator. Other chips also gets hot but not as much as the 6502. 2) You are missing a lot of machines, as you are not including the European market. Sinclairs, BBC Micros, Acorn, and other machines also get very hot. You should talk with Nolan of Nolan's Retro Lab to discuss about those machines. Jean Beta is also another youtuber that deals with European machines.

[@RetroHackShack]

Yeah. I would love to do another round with European machines. I have a few Spectrums, but what others would you suggest. Beeb? Amstrad CPC?

[@jasmijndekkers]

Nice thermal camera and good to see how it works. Greetings from Steven from the Netherlands

[@v7bdv7bd34]

Thanks! The thermal imager is very useful and good weight.

[@RacerX-]

Nice experiment. Camera looks decent enough. I have that brand multimeter and it has been great. A note about the PLA. I have had a few 5-pin video early 64s and they have all had that 8411 REV3 PLA. I used to think that they were replaced because the 8411 seemed like a date code. Apparently it is not a date code and that is the original PLA. Those were programmable chips and are much better than the later MOS made PLAs. Thanks for another great video.

[@RetroHackShack]

Thanks for that info. I thought it was a date code too.

[@blackterminal]

On one of my c64 units i attached heatsinks to almost every chip. I would like to do more of my machines like that. Probably one of each type or maybe just the ones I actively use.

[@themax4677]

Instant read digital meat thermometers are another good low cost way of getting temps. I have a couple that I use for things ranging from general air temperature and fish tank temps to even oral human temps in a pinch. Just mind what you used them for last and remember they are rather pointy!

[@larryk731]

This seems like a great diagnostic tool.

[@RetroHackShack]

For sure

[@blackterminal]

I noticed when I had a Amiga 500 the keyboard IC gets very hot to the touch which surprised me. Its not a IC youd think would get hot.

[@MaggieKeizai]

Dude. I love your shirt. I totally remember that particular scratch n sniff sticker!

[@stevethepocket]

The really interesting thing for me was seeing all those little chips on the Apple II that weren't getting hot at all on top but were heating up the board around them a good deal, producing a nice little ricer underglow effect. I'm guessing they all had thick ceramic lids, no doubt at considerable expense to the manufacturers (and the customers, but that's a given; it's a freaking Apple).

[@SeanBZA]

Ceramic packages in the early RAM was because of radioactive contamination in the epoxy fill used on those early chips. The plastic contained small amounts of radioactive material from both natural sources and atmospheric bomb testing, and this would, during part of the decay series, emit alpha particles into the chip, causing random bit flips. The solution was to use the ceramic packages, as those were made with higher purity fillers, and not contaminated by the tiny amounts of natural radioactives, plus the gap between the lid and the die would allow most of them to naturally absorb 2 electrons, and form neutral helium, before interacting with the silicon. Plastic packaging took off with better refining, and not using fillers with contamination, allowing the price to rapidly fall, as the ceramic package is, in most cases, much more expensive than the die inside. So going to a plastic case, where the highest value part was the die, followed by the copper in the leadframe, and then the plastic fill, came in much lower in cost. Plus also volume manufacturing also made the cost per die cheaper, and process shrink increased that, along with the newer processing using larger wafers with higher yields.

[@user-marco-S]

When you did measure the temperature, you forgot to add an overlay with the temperature in Celsius. Outside the USA, only Celsius is in use.

[@Fred_Raimer]

Wow! They lowered the price to $289 and still have the $60 coupon! While I certainly don't *need* it, your great presentation and the fact that I can't resist a bargain is testing me! I will have to blame you if my wife gets on my case for buying yet another "toy" LOL. Thanks for all of your effort and great attitude!

[@michaelstoliker971]

C64 is a hot mess! LOL

[@michaelstoliker971]

A cheap way of finding a hot chip quickly is to use a spray bottle of Isopropyl alcohol. Spray it on the chips and it will evaporate quickly on hot chips.

[@horusfalcon]

This looks like a fun way to spend some time... 😄

[@Ramdileo_sys]

A capacitor gets hot for only one reason Aron... bad ESR / internal resistance... is draining lots of power... almost in shortcut...

[@PrimeRibb69]

The c64 power bricks could get hot enough to fry an egg.

[@mzwandula]

Fahrenheit Fahrenheit Fahrenheit... The rest of the world does not give a shit on Fahrenheit! Neither do I. I'm out... Will not watch this to the end

[@KLund1100]

Atari had to remove the plastic shells around the OS and RAM cards in the later 800's because of repeated heat failures. SO you are missing a major player in the heat issue. Taking a 'short cut' by using a 800XL instead of getting, using a original 800, is doing dis-service to your topic. Please consider pulling this video down and redoing it properly. ( you will have to run the 800 with the cover on for a while to heat up, then remove it and quickly check the temps)

[@RetroHackShack]

I'll need to get an original 800 first. 😀

[@michaelstoliker971]

@@RetroHackShack I've got an original 800 with the plastic shells. Nobody ever calls for removing cartridge shells, what's that about? You can send me your thermal imager and I'll report back.

[@michaelstoliker971]

The plastic shells were removed for cost savings. I've rarely heard of an Atari failing from overheating. Not saying it doesn't happen but the story about the shells being removed for overheating sounds good but isn't supported by any facts I've seen.

[@KLund1100]

One of the most definitive books on Atari, Business Is Fun by the late Atari expert himself Curt Vendel, page 465: "Not too long after the release of the Atari 800 computers, Service Bulletins were issued to service centers informing them that later releases of the Atari 800 would eliminate the expansion module enclosures for the Operating System and Memory modules (wrapping the PC boards within these cases caused overheating problems, computer lockups and component failures). Atari would also later offer the Atari 800 computers fully maxed out with 48K of memory, and the top cover latches of these were replaced with two screws and washers locking down the top cover from access." (this last sentence is how you can tell early version 800's from the later, but not CTIA or GTIA versions. Also the top covers between these two versions are no interchangeable.) It would interesting to have a heat test of the 800 added to this video

[@rager1969]

Why did you apologize for using Fahrenheit? You don't apologize for you accent or your gender, do you? If a viewer uses metric, they should be able to deal with it.

[@RetroHackShack]

Just to be nice 🙂

[@stevethepocket]

I'm American and I still had a poor frame of reference for how hot these chips are because thermal monitoring for modern computers is always done in centigrade-with typical results ranging from about 70-90.

[@MaggieKeizai]

I love how people get bent out of shape and call Americans stupid because they don't understand something. Meanwhile, we're all fluent in metric, so it's like complaining about people in Hungary speaking Hungarian from where I sit.