Atari 600XL measures C64 internal case temp with/without fan

[Britishcar]

I used a 600XL to measure the internal case air temperature of a C64 while it was searching for prime numbers during a two hour period. Here’s what I discovered.

 

I ran the experiment with one probe sitting outside of the case and one probe inside the case. The ambient room temperature was 74 at the beginning of the experiment and rose to 75 at twenty-two minutes in. It varied between 74 and 75 degrees until thirty-two minutes in and then remained at 75 for the duration of the experiment.

 

The internal case temperature also started at 74 degrees but quickly rose 10 degrees to 84 degrees in twelve minutes after the machine was switched on and began running a BASIC program that finds prime numbers. It rose more slowly an additional 10 degrees to 94 degrees at fifty-four minutes in to the experiment.

 

The internal case temperature was at 94 degrees at sixty minutes in when I turned on a small USB-powered fan I have installed in the case. The temperature began to immediately drop and was at 91 degrees in five minutes. At eighty-four minutes in — twenty-four minutes after turning on the fan — the internal temperature had dropped to 88 degrees where it remained steady until one-hundred and twenty minutes in at the end of the experiment. Over all, the fan created a 6 degree drop in the internal case temperature. The fan is located at the rear of the motherboard and faces upwards to blow towards the native vent slots in the rear part of the top cover.

 

Pictured (Atari 600XL; RF - Channel 3; Hitachi Ultravision) is the table showing the fifteen minutes between 60 and 74 minutes when I switch in the fan and the temperature begins to drop. Also pictured (Atari 600XL; RF - Channel 3; Hitachi Ultravision) is the chart showing the full time/temperature graph. The 3rd picture is the Commodore 64 (S-Video; Hitachi Ultravision) finding prime number 3041 in two hours running Commodore BASIC.

[TZJB]

9 hours ago, Britishcar said:

I used a 600XL to measure the internal case air temperature of a C64 while it was searching for prime numbers during a two hour period. Here’s what I discovered.

 

I ran the experiment with one probe sitting outside of the case and one probe inside the case. The ambient room temperature was 74 at the beginning of the experiment and rose to 75 at twenty-two minutes in. It varied between 74 and 75 degrees until thirty-two minutes in and then remained at 75 for the duration of the experiment.

 

The internal case temperature also started at 74 degrees but quickly rose 10 degrees to 84 degrees in twelve minutes after the machine was switched on and began running a BASIC program that finds prime numbers. It rose more slowly an additional 10 degrees to 94 degrees at fifty-four minutes in to the experiment.

 

The internal case temperature was at 94 degrees at sixty minutes in when I turned on a small USB-powered fan I have installed in the case. The temperature began to immediately drop and was at 91 degrees in five minutes. At eighty-four minutes in — twenty-four minutes after turning on the fan — the internal temperature had dropped to 88 degrees where it remained steady until one-hundred and twenty minutes in at the end of the experiment. Over all, the fan created a 6 degree drop in the internal case temperature. The fan is located at the rear of the motherboard and faces upwards to blow towards the native vent slots in the rear part of the top cover.

 

This is a good applied science experiment. Do you have an image of the 600XL and peripherals that went into this experiment please? How is it all hooked up to the 600XL? Additionally, could the Atari speed control the fan based on temperature?

[Britishcar]

8 hours ago, TZJB said:

This is a good applied science experiment. Do you have an image of the 600XL and peripherals that went into this experiment please? How is it all hooked up to the 600XL? Additionally, could the Atari speed control the fan based on temperature?

Hello TZJB! I don't have a photo of the 600XL "geared up" but I can take one and upload it at my next opportunity.

 

The hookup is using the AtariLab with Temperature module set up in a standard configuration with only one unusual addition. The standard setup for the Temperature module is the ROM cart it comes with, the breakout box, and one temperature probe plugged into joystick port #2. The only thing I've added -- and that I don't see any mention of in the manual -- is the addition of a 2nd probe that can be plugged into the breakout box and runs along side of the first one such that the program can measure two temperature probes (called blue and orange) at the same time. The retail package only comes with one probe so I assume users may have been able to purchase a 2nd one optionally but I'm not aware of any retail packaging on adding a 2nd probe that ever came out. The ROM cart and breakout box are all set up for a second probe, however, if you have one.

 

It would be very interesting to program the Atari to vary the fan speed based on temperature thresholds but the ROM cart it comes with has no such ability. A custom program would need to be written that could work with the probe(s) and joystick breakout box. I don't think there's any technical limitation to doing so, simply a bit of research and thinking about how to make it all flow.

 

As far as probe locations, the external one simply lays outside of the C64 bread bin about 8" away from it, not touching anything but air. The internal probe goes though the gap in case near the expansion slot with the tip of the probe suspended just a bit above the 6510.

 

I intend to run the experiment again with a different position and orientation of the fan to see if I can get a lower temperature by varying where the fan extracts the air.

 

[Rybags]

Can you do a comparitive control with the other computer just sitting there idle?

 

6502 does something on every cycle, even to a point when it's hung with a *2 opcode.  So over time, it should get warmer either way.

[Britishcar]

On 3/25/2023 at 11:35 PM, Rybags said:

Can you do a comparitive control with the other computer just sitting there idle?

 

6502 does something on every cycle, even to a point when it's hung with a *2 opcode.  So over time, it should get warmer either way.

Here is this your suggested run with the C64 sitting at the ready prompt only. As you predicted, it's very similar. I had to replace the probe so there's no knowing whether it was slightly closer to the 6510 or not.  The blue (C64 internal probe) climbed from 73 degrees to 96 degrees in an hour. After starting the fan at the 60 minute mark, it dropped to 89 degrees at 80 minutes in and stayed there until the full 120 minutes was up. It never went below 89 but the drop from 96 degrees was clear. Ambient (external orange probe) started at 73 and ended at 75 over the course of the two hours.

[Britishcar]

Slight correction. The C64 climbed to 95 degrees in 40 minutes and stayed there until it hit 96 degrees just before the 60 minute mark. The graph doesn't show 1 degree movements and seems to be in 2 degree resolution. You can see it "flatline" at 40 minutes.

[Rybags]

In theory if you had some SID music going that used filters it'd probably send the temp up more.

But really, the temps you are seeing barely represent what you'd see on a normally cool, idle PC on a winter morning.

 

My i7-4790 normally sits idle around 50C but will generally drop under 40 when the room is cold.  Though I normally run all SSDs which means the case is cooler than most.