Have you ever wondered how much current a coin-cell scientific actually draws? Well, I did, so I cracked (not literally) my TI-36X Pro open and took some measurements with my trusty digital multimeter. My coin cell had a voltage of ~2.95V throughout the test.

Disclaimer: I cannot guarantee every TI-36X Pro ever manufactured would yield the same numbers during such a test. My serial# is L-1123F, so if yours looks similar then it's most likely going to be about the same. Also, the solar panel didn't generate nearly enough power to keep the calc on in this test because I dimmed my lights as much as I could. It was probably contributing anywhere from 0 - 10 μA while doing the tests. More about the solar panel: it's sufficient to keep the calculator powered, even during heavy load, if you point a bright light directly at it. However, the calc will not power on without a CR2032 battery in the socket. The threshold voltage for it to think there's a battery present is unknown. If I run it from only the solar panel using dim-ish lighting then gradually the screen will lighten, and eventually look completely unpowered. If there's a very specific amount of light hitting the panel, it's enough to not reset the calc while corrupting much of its RAM.

Now, to the data!

* "MEMORY CLEARED" screen: 42 μA
* Idle at homescreen after a mem clear: 30 μA (spikes to 32 μA during cursor blinks)
* MODE screen: 67 μA (spikes to 70 μA during selection blinks)
* MODE screen (second page): 50 μA (spikes to 53 μA during selection blinks)
* COMPLEX menu: 65 μA
* MATH menu (MATH): 58 μA
* MATH menu (NUM): 62 μA
* MATH menu (DMS): 48 μA
* MATH menu (R<->P): 61 μA
* Data/List editor: 56 μA
* Entering a series of integers into the data/list editor: 56 μA - 119 μA [fluctuates]
* FUNCTION TABLE menu: 61 μA
* Exploring x^2 in the function table by spamming the down arrow key: 95 μA - 168 μA [fluctuates]
* Entering a MathPrint expression on home screen: 38 μA - 109 μA
* Calculating a numeric integral: 207 μA (209 μA when printing the answer just after calculation has finished)
* Summing a large series: 229 μA
* Solving for 1 solution of a random quintic: 240 μA
* "BREAK Error" screen: 46 μA

As we can see, the calculator overall stays below ~60 μA in its "idle" states and it's safe to assume it never goes above 300 μA in its "heavy calculation" states. Cursor blinks add about 2-3 μA due to LCD updates; the LCD is a moderately power-hungry component in general, it seems.

Assuming your CR2032 battery has a capacity of about 220 mAh, you should get a solid 3,000 hours of use out of this calculator.

Interesting how the cursor blinking takes up so much current...

Anyway, that amount of time is around 125 days, assuming constant 24/7 use of the calculator. However, most classes that require such a calculator usually last around an hour, and then maybe another hour would be used on homework. So let's say two hours per day. That amount of time would span several years, if not decades, of average use.

All from such a small battery. Crazy.