Rosemary Ainslie Quantum Magazine Circuit COP > 17 Claims

[MarkE]

Low power DC load temperature calibration is complete. I've plotted one cool-down cycle on the graph as well as the 10 runs at various DC power settings.

The final plot is nice and linear. Given a measured temperature at the 60 minute mark, one may confidently read the equivalent DC power level from the plot.


Real data has a certain beauty to it. I have always thought that was one of Nature's great truths.

Nice job!  One can certainly resolve down to around a tenth of a Watt or better with that data.

[MileHigh]

TK:

If time is not a factor may I suggest 75 minutes.

My logic is as follows:  At about 20 minutes you are at about 90% of the final temperature.  So I mentally shift your cool-down curve  to start at 20 minutes, and I observe how how it "eats" into the temperature rise.  When the shifted cool-down curve is close to zero that indicates when thermal equilibrium would set in with a "head start" of already being at 90% of the final temperature.

So I massage that all in my head and come up with 75 minutes as being "very very close" to thermal equilibrium.  That's in contrast with 60 minutes being within a "few percent" of thermal equilibrium.

Also, since the goal is near-thermal equilibrium, the difference between the ambient temperature and your oil is critical.  Note that if the ambient drops 5 degrees, then the oil at near thermal equilibrium also drops 5 degrees.

Sorry for sticking my nose in!

MileHigh

[MarkE]

TK:

If time is not a factor may I suggest 75 minutes.

My logic is as follows:  At about 20 minutes you are at about 90% of the final temperature.  So I mentally shift your cool-down curve  to start at 20 minutes, and I observe how how it "eats" into the temperature rise.  When the shifted cool-down cure is close to zero that indicates when thermal equilibrium would set it with a "head start" of already being at 90% of the final temperature.

So I massage that all in my head and come up with 75 minutes as being "very very close" to thermal equilibrium.  That's in contrast with 60 minutes being within a "few percent" of thermal equilibrium.

Also, since the goal is near-thermal equilibrium, the difference between the ambient temperature and your oil is critical.  Note that if the ambient drops 5 degrees, then the oil at near thermal equilibrium also drops 5 degrees.

Sorry for sticking my nose in!

MileHigh

I think it is fine the way that it is. 

The longer thermal time constant eyeballs to about 10min.  Even at 80C rise, the rise from 50min to 60min is about 1C.  That is getting into the resolution of a thermocouple based sensor.   I don't think it is worth a day or two of runs to flatten out 1C.

[TinselKoala]

I agree. This data is entirely usable for the purpose: read a temperature and find the equivalent DC power dissipation. I can even use the 20-minute temperature for that, since everything is so well behaved (once convection starts, anyway.)

Later on when I'm doing experimental trials and "Joule tests --- because Joules doesn't lie --- " (tm Donny), it may turn out to be important to get all the way to complete equilibrium. We shall see.

[MileHigh]

That's fine.  For what it's worth the time constant is about 15 minutes, eyeball + calculator.