OT: TI 36X Pro and Wishes for a solar RPN calculator


I am still holding out hope for a solar-powered RPN calculator from HP, even if it is similar to the 11C. However, a solar powered HP 15C or 35S would be sweet.

In the meantime, here is my review of TI's 36X Pro:

TI 36X Pro Review


This looks like a nice calculator, but I'm curious about the appeal of solar power. On a programmable calculator (like the 11C, 15C or 35s that you mention) you're absolutely going to need a battery anyway, so why bother with having a solar power in addition? I seriously doubt that a modern, high-speed programmable could run at a good speed on the tiny about of current you'd get form such a small solar cell array. I suppose that when running on solar power you could slow down the clock, but it would likely have to be really slow and what fun would that be?



I love solar in scientific calculators. It might not be entirely rational; I like the idea that if I was on a desert island, my calculator could last forever.

Most calculators now that use solar are dual powered, and I think that does help -- the batteries seem to last an extremely long time.

The last time one of my dual powered calculators had its battery die, I tried some experiments to see what it could do with the solar cell alone. In room lights, some things (possibly root finding, integration, and even statistics [because it didn't keep running totals as you entered data, but only after you call a stat function]) would cause the calculator to reset. In direct sunlight, though, I think it could do anything.


I love solar in scientific calculators....Most calculators now that use solar are dual powered...

I agree. Non-programmable scientific calculators (what few still remain) should without any question be dual solar/battery powered.

There's just something appealing to a small micro-power consuming device capable of scientific calculations that can power itself. The physics, electrical engineering, materials science, and mathematics that are incorporated into such a calculator are still an object of great fascination for me.

I also agree that a solar-power option for programmable calculators would be of questionable utility. I doubt any solar panel could be incorporated that by itself can power such a calculator during program execution.


Nice review, especially the comparison to the Casio fx-115ES and the Sharp EL-W516.

Based on my own personal comparison, the 115ES is an easier to use device than the W516. Both are available for about $17.50 in north Alabama. The only thing that the W516 does that I wish the 115ES did is preserve the state of the machine at shutdown so that it all appears when turned on again. The 115ES wipes the slate clean when it shuts off, manually or automatically.

That having been said, based on your description the 115ES would still be my calculator of choice, even ignoring the 30X Pro's 25 percent higher price. And, like all TI calculators produced over the past 20 years or so, the 30X Pro is god-awful ugly. (Who does TI's horrible industrial design?) I've developed such a liking for the 115ES that I've been able to get over it not being RPN. I carry one everywhere. It's an amazingly powerful device, and if it should get damaged, it's dirt cheap to replace.

As a matter of interest, the numerical integration scheme (Gauss-Kronrod) used by the 115ES greatly outperforms the Simpson's Rule method used by the W516 in terms of speed and accuracy. I wish HP would adopt that method for their machines. What does the 30X Pro use?


I haven't tried the TI36x pro. I'll admit I'm skeptical, because the only TI calculator I like (love, in fact) is the TI89.

But I prefer the 516w to the 115es.

I had the 506w before I got the 516w, and the upgrade went perfectly smooth. Not only does the 516w add abilities, it loses none.

What's particularly impressive to me (especially over the 115es) is that it added functionality, but it's still just as easy to use, and just about every function is still directly accessible from the keyboard (without needing menus). One function that is buried in a menu is ALG (allows you to repetitively substitute values into an expression), but the ability to assign any function to one of four soft keys fixed that.

I haven't directly compared the 516's integration to the 115es's, but I don't think the 516's is merely Simpson's rule. It can use Simpson's rule in the exact same way the 506 does, if you pass an integer argument after the integration limits, but if you don't give that extra argument I think it uses Romberg integration (since it determines on its own "when to stop")

I haven't used the 115es in a while, so I don't remember how it does constants, but I really like the 516's constant library. It's a big improvement over the 506's. The 506 required you to memorize a two digit code for each constant. The 516 has a menu that shows both the symbol AND the unit of the constant! (even some HP scientifics don't do that, though the HP50g does) You can browse though the menu to find your constant, but if you happen to remember its two digit code you can go to it instantly.

The one thing I like more about the 115es is that it does have the ability to make iterative calculations easier (I call this "weak programming" -- and it's not documented!)


I haven't directly compared the 516's integration to the 115es's, but I don't think the 516's is merely Simpson's rule.

The W516 manual says that it uses Simpson's rule. And for every example I've tried on both the W516 and the 115ES, the 115ES returned more accurate results much more quickly (half the time or less). The Gauss-Kronrod quadrature method is fairly sophisticated, and produces better results quicker, especially on classes of functions that are notoriously hard to numerically integrate.

I haven't used the 115es in a while, so I don't remember how it does constants, but I really like the 516's constant library.

I thought I would like that too, but I found that it takes longer (more keystrokes) to call a constant or conversion on the W516 than on the 115ES. But for the 115ES one has to refer to a chart of codes for such that's printed on the back of the sliding cover. I'll take that over the time the W516 requires.

The numerical accuracy of the 115ES is better as well. A simple show of that is Mike Sebastian's test arcsin(arccos(arctan(tan(cos(sin(9)))))) in DEG mode. The W516 produces 9.000000098906, while the 115ES produces 9.00000000733338. The error percentage is 13.48 times larger for the W516.

I also HATE the DRILL mode of the W516. It's absolutely idiotic. What a waste of firmware!

But I do like very much the preservation of machine state when the W516 shuts down. I wish the 115ES did that.

Edited: 28 Apr 2011, 12:55 a.m.


Thanks for the review - I'm looking forward to getting a TI-36X Pro for my collection when they become available near me.

Regarding the 115ES and the W516 I have both. I noticed a couple of weeks ago that there is a new version of the W516 available at my local Staples and bought one. When I got it home and compared it to my current unit, the only difference I could see is a REALLY significant re-styling. I didn't open the package and returned the newer version. Anyone else seen these? If so, IS there any functional difference?


Thanks, Mike.

I was not able to find the method in the 36X-Pro. Since the 36X-Pro has an 84+ interface, I am going to guess Guass-Kronod because the 84+ uses that method.

The 115ES is a favorite of mine overall - other than having to switch modes to use complex numbers, matrices, and vectors, it is an outstanding.

Casio also has a solar programming model, fx-50p. I had to get one online though. I wished that Casio did not bury the programming mode and commands though.

Edited: 28 Apr 2011, 10:05 a.m.


Nice review! I have noticed that in the fx-115ES (fx-991ES in Europe) there is no keyboard buffering so when entering a very long formula (with many factions, roots...) the calculator redraws after each key press, but as the formula becomes longer and slower redraw, it can miss keystrokes. Does it happen with the TI 36X-Pro and the Sharp's EL-W516? Can you test it please? Thanks

Edited: 28 Apr 2011, 10:10 a.m.


Are you noticing that in "pretty print" mode, line mode, or both? My guess would be that pretty print mode would be the worse for that, but I almost always using line mode.


Are you noticing that in "pretty print" mode, line mode, or both? My guess would be that pretty print mode would be the worse for that, but I almost always using line mode.

It happens only in pretty print mode.


Here are a few test calculations: (Sharp) EL-W516 vs. (Casio) fx115-ES vs. (TI) 36X-Pro. All three calculators are set to "pretty print" mode.

1200+3800+28700+2899+19000+2839+4657+2983-2930-10000 = 53148

I notice a little slowness in the 115ES towards the end, but not in the 36X-Pro. The EL-W516 works fine.


sqrt(7^2 + 7^3 + 8^2 + 8^3 + 9^2 + 9^3 + 10^2 + 10^3 + 11^2 + 11^3) = 65.80273551
[using the y^x key]

All three calculators work fine.

1 2
1 + --------- * --------------
1 2
1 + ---------- 2 + ------------
1 2
1 + ---------- 2 + -----------
1 2
1 + ------- 2 + ------
2 3

= 35/24

= 1 + 1 / (1 + 1 / (1 + 1 /( 1 + 1 / 2 ))) *
2 / (2 + 2 /( 2 + 2 /( 2 + 2 / 3 )))

The fx-115ES returns the answer and works fine, and so does the EL-W516.

The 36X-Pro returns "MEMORY LIMIT" Error when using the fraction key, but expression works when using the division key and parenthesis.

# 4:
pi^2.4 - pi^pi + e^(2^3.1)
pi * ----------------------------- - 1.4 ln(5)
e^2.4 - e^e^1 + pi^(2^3.1)

= -1.348316376

fx-115ES: Calculator returns Syntax error (at pi^(2^3.1)). The calculator doesn't take every key stroke once I get to the last term

36X-Pro: calculation has to really "refresh" itself as I am finishing the denominator

EL-W516: This calculator wins with this example. Entry was responsive without slow down the entire calculation

# 5:

Integrate 3x^3 - x^2 + 1/x - 1/(x^2) + 1 from 1 to 10

= 7176.652585

The fx-115ES works fine. I can enter the integral on the 36X-Pro fine, but took several seconds for the integral to calculate. The EL-W516 operates similarly to the 36X-Pro.

# 6:

Sum of sin(tan(x)) + tan(sin(x)) - cos(x) - cos(x)^2 from 1 to 10 - Radians Mode

= -0.2223937769

fx-115ES: entry is fine, but the calculator took a few seconds to find the sum

36X-Pro: entry is fine and the sum was instantly found

EL-W516: entry is fine but the calculator took a few seconds to find the sum

# 7:

sqrt(3^6 - 2^3 + 1) 2.7^2.5 - 1.7^2.5
-------------------- + sqrt( --------------------- )
2 * cuberoot(4) 1.7^2.5 - cuberoot(6)

= 10.51498832

fx-115ES: the entry really slows down when I am entering the second term

36X-Pro: The calculator flashes with each keystroke at the second term, but very little speed slow down

EL-W516: This calculator handled the expression well - no slow down


I think the EL-W516 wins with these super long expressions.


I would hope that _IF_ HP made a solar RPN calc, it would have the ability to display actual voltage/current/wattage from the solar cell(s) and could also be used as a photographic light meter, IR programming input and a data logger for the light readings. You know, none of the humdrum everyday just solar powered stuff...



dona nobis pacem


That sounds really cool!


Do any of these machines provide a modulus (or remainder) function? Its lack on the fx-115ES is driving me crazy!

Possibly Related Threads...
Thread Author Replies Views Last Post
  OT: Graphing calculator for Android ? Ivan Rancati 4 809 11-18-2013, 12:33 PM
Last Post: Alberto Candel
  [PRIME] RPN: another attempt at returning more than one value to the RPN stack Marcus von Cube, Germany 5 1,023 11-05-2013, 02:44 AM
Last Post: Marcus von Cube, Germany
  RPN-67 Pro 2.0 is here! Willy R. Kunz 12 1,259 09-28-2013, 10:50 AM
Last Post: Michael de Estrada
  OT--'The Big Bang Theory' and RPN/HP--Any connections? Matt Agajanian 19 1,957 09-05-2013, 08:27 AM
Last Post: Adrien Bertrand
  OT TI-59 for the iPhone Matt Agajanian 10 1,250 09-04-2013, 08:57 AM
Last Post: Eddie W. Shore
  OT-The ideal calculator Matt Agajanian 0 376 09-04-2013, 01:11 AM
Last Post: Matt Agajanian
  OT: Reversing Sinclair's amazing 1974 calculator hack - half the ROM of the HP-35 Egan Ford 6 1,151 09-02-2013, 10:54 PM
Last Post: Thomas Klemm
  OT--TI-36X Algorithms Matt Agajanian 48 4,425 09-01-2013, 08:13 PM
Last Post: robert rozee
  TI-36X Final Answer Matt Agajanian 0 340 08-31-2013, 08:40 PM
Last Post: Matt Agajanian
  OT: Simulating a TI calculator with crazy 11-bit opcodes Egan Ford 8 1,159 08-13-2013, 12:06 AM
Last Post: Paul Dale

Forum Jump: