/vr/ - Retro Games

[Anonymous No.8715547]

It has become a vogue amongst retrofags to heat sink C64s to the waist. Why? As it turns out, Commodore used an ancient grug caveman fabrication process from the 70s that resulted in their chips becoming little ovens that can fail in relatively short time if steps aren't taken to keep them cool. The VIC-II in particular used 1600 milliwatts of current, which is a lot for a 40 pin DIP and sinking it will reduce its heat output by a good 30%. Intel and Motorola had issues with this too in the Altair 8800 days, but quickly adopted newer processes that generate a lot less heat.

[Anonymous No.8715556]

Doesn't matter, heatsinks cost pennies and most ICs last longer when they run cooler and their off and on temperature difference doesn't get too high.
Might as well heatsink anything and everything you can't get new ones anymore if you want them to last, the cost vs longevity argument favors it.

[Anonymous No.8715558]

>>8715547
>Intel and Motorola had issues with this too in the Altair 8800 days, but quickly adopted newer processes that generate a lot less heat.
That explains why Intel went with 8086/8088, 80286, 80386 and even most 80486 having no heatsinks, to heatsinks becoming a requirement.

[Anonymous No.8715583]

>>8715556
I don't think you need to worry about sinking post-video game crash consoles anything like that since they don't have anything in them that gets particularly hot. For example the only thing in a Mega Drive that gets kinda warm is the Z80 in Model 1s and Model 2s switched to a CMOS Z80 that doesn't get above room temperature.

[Anonymous No.8715591]

>>8715583
Doesn't matter. Same logic applies.
Just because it's not "hot" for you, doesn't mean thermal physics aren't doing it's work to grinding down its life expectancy.

[Anonymous No.8715829]

>milliwatts of current
Triggered

[Anonymous No.8715838]

>>8715547
God I hate these cringy pseud blogs. Especially when the redditard is blatantly ignorant. I particularly like how you think a heat sink reduces heat output, when the entire purpose of a heat sink is to increase heat output. What a stupid googlestupid little faggot.

[Anonymous No.8715875]

>>8715591
>>8715556
>IP count did not go up between these posts

[Anonymous No.8715885]

>>8715875
Nice math faggot.

[Anonymous No.8715892]

>>8715583
>>8715591
i read on the Amibay forums about heat sinks and they were in universal agreement that Amigas don't need them at all as they use HMOS chips that don't get very hot. the problem affects primarily Commodore ICs using the 6xxx numbering system as those used the original 70s-vintage NMOS process.

[Anonymous No.8715915]

>>8715558
CPU dies were getting more and more dense so sinks became necessary. The 486 was the last generation of x86 to run on 5V power and the first to use passive cooling. The Pentium switched to 3.3V and added fan stacks.

[Anonymous No.8715946]

>>8715547
The C64C runs much cooler as it uses the newer HMOS 8xxx series chips and it has a better-ventilated case.

[Anonymous No.8715975]

to minimize thermal stress in the chip you want no more than a 20C difference between room and operating temperature. something like a 40C gradient is too much.

[Anonymous No.8716005]

>>8715547
The VIC-II in my experience seems more prone to overheating than the SID which is most likely to be killed from ESD or people doing retarded hardware mods. VIC-IIs get hot and run at 7Mhz internally, which is a pretty high speed.

[Anonymous No.8716059]

the ceramic VIC-IIs on the earliest C64s just don't work that well. they tend to overheat to the point where they won't run for more than an hour or two without glitching out and resulting in a screen full of garbage. but then again these were practically a prototype version of the VIC-II.

[Anonymous No.8716157]

>>8716005
FWIW a large portion of the die on VIC-IIs is used for the sprites, in fact about 70% of it. I would wager if you were just typing in BASIC the chip is under less load than it is playing games or demos.

[Anonymous No.8716329]

The PGA sockets which most modern ICs use have much worse thermal characteristics than the DIP and other such setups vintage electronics use. They don't heat evenly and can shoot up 30-50C in a matter of seconds. Then factor in lead-free solder and other such mishaps.

[Anonymous No.8716340]

>>8715975
Running stuff constantly is better than repeated power cycles, but even if you only power something on and off once a day that's 365 power cycles a year which is really fucking nothing.

[Anonymous No.8716551]

>>8715558
Cooling requirements for 486s varied but a passive sink was the minimum.

[Anonymous No.8716742]

>>8715875
I'm pretty sure he never pretended to be two different people.

[Anonymous No.8716751]

didn't someone mention that early Famicoms were heat sinked even though the PPU didn't really need it?

[Anonymous No.8716772]

>>8716751
They were. It's possible the first ones did actually need it since it was early and Ricoh may have improved them or used a different process later on.

>>8716059
The first version of something tends to have issues until the engineers get everything nailed down pat. Did you know that very early 6502s from 1975-76 did not have a working ROL instruction? In fact Microsoft BASIC specifically included a workaround for this, but it had been long fixed by the time the Commodore PET appeared.

[Anonymous No.8716779]

>>8716059
It's also hard to put a heat sink on these due to the raised metal center.

[Anonymous No.8717497]

>>8715547
the issue here was that as the chip heated up it would keep drawing more power and getting hotter and hotter until it finally blew up. the 8080s were known to self-destruct like this.

[Anonymous No.8717689]

>>8715892
"Need" is a vague statistic too, since nobody has tested them running for decades with vs without heatsink, having both sinked and control from the same batch.

[Anonymous No.8717963]

>>8715583
(ignore the assembly language LARPer in here)

You can use an infrared thermometer to see how hot a given chip gets and if you think it's too high (as in over 140F) then by all means put a heat sink on it. The data sheet for the chip (if available) should specify its maximum operating temperature. I've never heard of the chips in an Amiga getting particularly hot but you're welcome to measure them.

[Anonymous No.8717985]

>>8715583
>>8715892
Go take it up with him.

[Anonymous No.8718003]

>>8715547
The VIC-II gets hot, incidentally, because the chip is pushing that ancient 70s NMOS process to near its limits. The HMOS variant used in the C64C is much cooler because it's a more advanced process so the chip doesn't have to work as hard.

[Anonymous No.8718076]

https://www.youtube.com/watch?v=PYDn9idlWb4

this was pretty funny where he said he got flamed once on Usenet about the C64 power bricks.

>why didn't you have modern ATX PSUs in 1983 i kill you scum

[Anonymous No.8718367]

yeah I don't think any CMOS chips pre-486 get much above room temperature

[Anonymous No.8718379]

Fuck you OP.

[Anonymous No.8718419]

>>8716772
>Did you know that very early 6502s from 1975-76 did not have a working ROL instruction?
No, because hardly anyone needs to know that.

[Anonymous No.8718508]

>>8718419
>>8715556
>>8715591
>>8717689
All samefag. All assembly language LARPer.

[Anonymous No.8718634]

>>8718076
Look at the OP pic. The modern-style heat sinks on there weren't even available yet in the early 80s. They would have just had a flat heat spreader like the Colecovision used.

[Anonymous No.8718832]

>>8718634
of course. a lot of things we now take for granted like good power supplies and thermal shutoffs if something overheats literally did not exist in that time, or if they did they were only used by commercial or military equipment and would have been too pricey for consumer equipment.

[Anonymous No.8718930]

>>8718367
>yeah I don't think
Yeah, you don't

[Anonymous No.8718965]

>>8718076
I mean, Bill isn't really at fault for Commodore's Soviet bloc-tier manufacturing standards and management.

[Anonymous No.8719754]

>>8715547
>It has become a vogue amongst retrofags to heat sink C64s to the waist
this has been going on for decades
>Why? As it turns out, Commodore used an ancient grug caveman fabrication process from the 70s that resulted in their chips becoming little ovens that can fail in relatively short time if steps aren't taken to keep them cool
false. biggest enemy of early MOS chips is static electricity and wrong voltages being fed to chips. only chip known to suffer from heat issues was the TED chip designed for plus4 series.
>Intel and Motorola had issues with this too in
m68k chips required no cooling at all. never seen any cooling on a motorola that was slower than an '030. only time i've seen cooling on them were in very expensive accelerator cards that struggled to get air to begin with. your revisionist history is cringeworthy at best, and most wrong.

[Anonymous No.8719758]

>>8718634
those chips require no cooling at all. people are (misled) to believe that if they stick heat sinks onto chips, it will make them last longer, but this is false.

[Anonymous No.8719764]

>>8718508
your inferiority complex is so extreme that it's just embarrassing to witness. aren't you the same pedophile that's always here that just makes up absolute bullshit about computer history as you go, gets it wrong, and then reports people? i've seen you around. you're the only pedo on this board that posts like this. when will you be livestreaming your suicide?

[Anonymous No.8719781]

>>8719754
>biggest enemy of early MOS chips is static electricity and
No really the heat issues with Commodore's ICs are well known. VIC-IIs tend to glitch out and display garbage if they get too hot.

>m68k chips required no cooling at all
Not the 68000 but the 6800 which was mid-1970s tech and used triple rail voltage.

[Anonymous No.8719786]

>>8719781
>>8719758
Every 10C of temperature cooler a chip is the longer it will last.

[Anonymous No.8719808]

68000s that use the original HMOS process get a little warm, later CMOS variants never get above room temperature. The thing about NMOS/HMOS is that the transistors draw power whether they're under active load or not, while in CMOS they don't use any power when idling thus no heat generation.

[Anonymous No.8720253]

>>8719786
>every goo goo ga ga words i read
How embarrassing

[Anonymous No.8720489]

>>8719781

[Anonymous No.8720693]

>>8715547
>a vogue
>the waist
>superfluous meme words to describe that it's old
I hated reading this. D-, apply yourself.

[Anonymous No.8720707]

>the waist
Anyone gonna tell me WHAT HE FUCKING MEANT by this?

[Anonymous No.8720801]

>>8720707
tiktok will. You should go back there.

[Anonymous No.8720876]

>>8715547
This is bait, right?
Current is measured in amps, not watts, and cooling =/= reduction of the amount of heat produced.

[Anonymous No.8721091]

>>8720876
>This is bait, right?
OP's not the guy who posts frog images.

[Anonymous No.8721131]

My C64 never need additional heat sinks

[Anonymous No.8721143]

>>8715547
Holy shit, I’ll need to run my emulator in less cores.

[Anonymous No.8721217]

>>8720876
>and cooling =/= reduction of the amount of heat produced
how long do you think your PC's CPU would last without the heat sink and fan stack?

[Anonymous No.8721379]

>>8715547
>It has become a vogue
The term is either "en vogue" or "in vogue" little fella.

[Anonymous No.8721429]

>>8719808
I read on Atari Age where a guy said he found that Synertek 6502s in Atari 8-bits seemed to have a high failure rate, and also they ran about 10C hotter than NEC and Rockwell ones. So yeah, different manufacturers could vary in the process and manufacturing standards used and some were better than others.

[Anonymous No.8721450]

>>8719808
I read on Atari Age where a guy said he found that Synertek 6502s in Atari 8-bits seemed to have a high failure rate, and also they ran about 10C hotter than NEC and Rockwell ones (but still not as hot as MOS's 6502s). So yeah, different manufacturers could vary in the process and manufacturing standards used and some were better than others.

[Anonymous No.8721738]

ITT: The assembly language LARPer schizo strikes again

[Anonymous No.8722121]

Heat sinks work. Want proof? The TMS9918 is known to be a very sturdy chip that rarely goes bad when compared to many of its contemporaries such as the VIC-II because it was heat sinked, the datasheet for it explicitly specified that a sink was required (there is also a modern FPGA version that outputs RGB and eliminates the four sprites per line limit if you prefer that). Also TI in general had higher manufacturing standards than the likes of Commodore.

[Anonymous No.8722398]

>>8722121
I don't mind the RGB output, the original chip had that anyway but the sprite flicker is part of the whole retro experience. Imagine for example a NES game without massive sprite flicker. Unconscionable.

[Anonymous No.8722475]

I saw a photo of a batch of dead chips removed from C64s and the VIC-II and CPU had uneven, pitted surfaces suggesting thermal damage had warped the plastic shell. The other chips did not have visible damage and may have been killed by ESD or something.

[Anonymous No.8722578]

>>8722121
too bad the rest of the Colecovision isn't so sturdy from the shitty controllers and PSU

[Anonymous No.8722980]

what happens from excessive temperature fluctuations is the plastic shell of the chip develops stress cracks and the die can get loose and moisture leaks inside and corrodes it. the aim of passive cooling is to minimize thermal expansion and contraction when the chip is powered on or off.

[Anonymous No.8723934]

>>8716059
you'd think the ceramic shells would have better thermal capacity than the plastic ones