mc68060rc50 internal heat sensor

[Nikolaj_sofus]

does anyone know how exactly the internal heatsensor works?

i tried to measure on the pins of my apollo acc, so i can conclude that it's not a pt100 resistor.

another thing, how much heat does the cpu tolerate?

Thought that this would be useful information if i should overclock

[Chain]

it should be 400 ohm at 25 celsius according to datasheet, but there isnt anything about PTC/NTC type so you need to calibrate it yourself

[yaqube]

RTFM:

2.12 THERMAL SENSING PINS (THERM1, THERM0)

THERM1 and THERM0 are connected to an internal thermal resistor and provide information about the average temperature of the die. The resistance across these two pins is proportional to the average temperature of the die. The temperature coefficient of the resistor is approximately 1.2 Ω/°C with a nominal resistance of 400 Ω at 25°C.

[Nikolaj_sofus]

thanks
I will try to solder a couple of wires to the pins and put up my multimeter at some point so i can see how hot it actually gets.

I also modified a copper heatsink to fit in the desktop wedge, so no i need to get some heat adhesive tape to stick it to the cpu and a gap pad to lay on top of it to see if it would lead some of the heat on to the keyboard.

I'm also on the lookout for a fan which blow the air out horisontaly, so i can actually fit it in the amiga and get the air in the right direction.
Does anyone know where to get this?

[rkauer]

Old notebooks can be a source for vertical fans (which blows air horizontally).

[buggs]

I know that this thread is a little stale. Nevertheless, I think that I do have some useful addition to the content in here.

Please don't rely on the resistance numbers posted in the 68060 user manual. There's a notice in the errata sheet regarding temperature reading (http://powerphenix.com/ct60/060_mask40.pdf, Spec Changes section, S3).

The actual resistance at 25°C is 780 Ohms, with a temperature coefficient of 2.8 Ohms/°C.

Example (A4000 in custom Tower, Matze's TK060, Rev.6@100 MHz, after 1/2h POVRay):

[Wrangler]

Whoa! - I want one of those ^^^^

[hooverphonique]

Originally Posted by buggs

Example (A4000 in custom Tower, Matze's TK060, Rev.6@100 MHz, after 1/2h POVRay):

Sorry for OT, but what is "Matze's TK060"? Google doesn't come up with anything useful for me..

[buggs]

Originally Posted by hooverphonique

Sorry for OT, but what is "Matze's TK060"? Google doesn't come up with anything useful for me..

This is a WIP 040/060 card by M. Heinrichs for A3000/A4000 which was recently announced at A1k.org. It'll most probably be open source once it works as intended. The card is not aimed to be feature-rich and (just) contains CPU socket, 128 MB onboard RAM and an I2C temperature/voltage sensor.
Up to now, most of Matze's projects were aimed at the DIY crowd. Some people do offer pre-built hardware based on these projects over at AmiBay.

[jkdsteve]

Originally Posted by buggs

This is a WIP 040/060 card by M. Heinrichs for A3000/A4000 which was recently announced at A1k.org. It'll most probably be open source once it works as intended. The card is not aimed to be feature-rich and (just) contains CPU socket, 128 MB onboard RAM and an I2C temperature/voltage sensor.
Up to now, most of Matze's projects were aimed at the DIY crowd. Some people do offer pre-built hardware based on these projects over at AmiBay.

This is a perfect product on that account, CPU/MEMORY..BAM!

[Daedalus]

Ha, bizarre that this cropped up, I'm literally measuring the response from a 68060's core temperature sensor as I type this.

[Daedalus]

I should have added though, thermistors aren't particularly linear, and especially positive temperature coefficient ones like the one in the 68060, are notorious for having a very non-linear response. Some even have a negative coefficient when they go outside a certain range, so the values are best taken with that in mind.

Anyway, with a little bit of testing here this evening, I've found that my rev. 4 LC060 responds differently to both the manual and the erratum. At room temperature, with the top surface of the chip at ~26 degrees, the therm resistance is 905 ohms. After some lha work for about 10 minutes at 75MHz, the top surface temperature was around 45 degrees, and the resistance was 998. That gives a response closer to 4 ohms/degree, although the real figure is likely less because of the difference between internal and external temperatures. Either way though, with a resistance in the 900 ohm range, it's a very different matter again to both the calculations mentioned here - enough for me to not assume the coefficient is the same either. Hmmm...

[alexh]

If you're interested someone in the Atari world did an investigation into the 060 temp sensor on the Atari Falcon CT60.

https://www.atari-forum.com/viewtopi...338498#p338498

The long and short of it was that each Revision of the 68060 the resistance varies wildly. He suggested you measure the resistance of your chip to calibrate the temp software.

(Might be worth reading the whole thread)

[Daedalus]

Ooh, thanks for that I'll check it out. I'm building up a calibration curve for this chip here as we speak, but there's always going to be a difference between the external surface and the core temperature so it'll only ever be a very rough approximation.

[Daedalus]

Okay, I have done some calibration on this rev. 4 68LC060 I have here. At 24 degrees, it has a resistance of 890 ohms, and at a surface temperature 46 degrees, it has a resistance of 984 ohms. That gives around 4.3 ohms per degree, though of course the core temperature was probably higher. So taking a bit of a guess at the temperature difference between the outer surface and the core (adding 33% to the difference from room temperature), I have what I think is a reasonable approximation of the actual temperature inside...

[Calabazam]

Hi.

Based upon your knowledge and discussion in this thread, could anyone give an opinion on what would be the typical temperature difference between a rev 5 68060 @ 50 MHz and a rev 6 @ 100 MHz if used in same airflow condition?

Thanks.