This is the LAST time I will warn Kerm NOT to lie that you cant overclock a ti84+se with software. I took the overclocker program that overclocks a ti83 plus to 15mhz and i reverse-engineered the det celtic command with the hex string. I found that the first 4 digits (3E01), when converted into decimal are worth just above 15000. Bingo!!!! I incrementally upped the hex numbers up to 9C40 which is worth 40000 (khz), and voilla! My calc was running 2.6666666667 times faster. Some stuff wasn't able to be sped up, but most asm and all ti basic programs were running super fast. I was surprised that even the os functioned correctly. The calc refused frequencies above 40mhz and was crashing on every such attempt. Wolfenstein 83 ,minecraft,etc... started working *MUCH* better and faster than before. Same thing with basic and hybrid basic games like Ben Trettel's Wolf3D and my modded version of ASCII-PONG II and also my text-based port of california speed and so on...
The calculator seemed like it had dunked-down 5 cubic meters of energy drink! No overheating, but jus very mild warmth at the button cell battery hold(all cpus get a little warm). also, the video chip was getting a little warm with wolfenstein83 and minecraft due to the faster video input.
At 40mhz, some asm games experience EXTREMELY MINOR video anomalies, but this is not present at 35mhz (88B8 setting). Graphing was *MUCH* faster and the calc was generally more responsive than ever. I guarantee that overclocking an ti84+ or 84+se to 40mhz is 100% safe. I think the os of the ti83+/se cannot handle such speeds. Note : my calc didn't let me immediately go up to a very high frequency: to get it to 40 mhz i had to (in my case) do a very incremental increase. This probably has something to do with the fishy hardware revision that my calc is from(2009, stronger RSA keys, 48k ram chip instead of 128k) and the latter is not the only problem that i have with this revision.
Below I have supplied you the code that i used to overclock my calc(only one line). You must have DCS 6.0 or later and enabled celtic lib. I sincerely doubt that the calc will crash from direct oc-ing to 40mhz, but if anyone has problems, feel free to tell me and i will post the fail-safe 26.35 to 40mhz overclock procedure. The worst that can happen is a spontaneous shutdown or the need for a battery pull ( if you increase to abruptly on newer calcs or if some program goes insane). I seems that the frequency checker cant see the overclock, hence it only uses the 15 and 6 mhz modes and not the third(oc) mode. Enjoy *HUGE* (40mhz) performance improvements topped with a 99.99% no crash guarantee and a 100% guarantee that it wont render the calc useless or corrupt the OS. All guarantees are for the ti84+ and ti84+se with os 2.55MP.

PROGRAM NAME:
CODE:
:det(20,''9C40D320C9'')

P.S. I think Kerm was hiding this all from me on purpose and if he does something like that again, I will get VERY annoyed!!!!!!!!!!!!!!!!!!!!!!!!!!
P.P.S. No insults or negative comments are allowed on this thread.
P.P.P.S. Tari cannot put any replies on any of my threads!!!!! Exclamation Exclamation Exclamation Exclamation [/img]
Quote:
the video chip


What?
3E 01 is ld a,1 in assembly, not just over 15MHz. The TI-84+SE has only two speeds, 6MHz and 15MHz.
I used 9C40 (hex for 40000) instead of 3E01 in the program you can find in the miscelaneous progs for ti83+784+ here on cemetech.
https://www.cemetech.net/scripts/countdown.php?/83plus/asm/programs/OVRCLKR.zip&path=archives
For user assembly programs and apps on the monochrome calculators, all are clocked to run at 6MHz in order to maintain compatibility with the TI83+ and other calculators. The OS still runs at 15Mhz. The assembly snippet you posted does this:


Code:
   sbc a,h
   ld b,b
   out ($20),a
   ret

Which is really undefined behavior, which sets the processor to execute at 6, 15.1, 15.2, or 15.3 MHz.

The orginal was:


Code:
   ld a,$01
   out ($20),a
   ret

Which sets the processor, to 15.1MHz.

Therefore, you are noticing the speed increase because your assembly programs and other things are basically running 3x faster. The z80 processor inside the TI84+SE can only be clocked to 4 speeds, 6, 15.1, 15.2, or 15.3MHz.

EDIT:
Also, 3E01 ~ 15873.
15MHz ~ 15000000

You are a few orders of magnitude off Smile
CalcMax wrote:
P.S. I think Kerm was hiding this all from me on purpose and if he does something like that again, I will get VERY annoyed!!!!!!!!!!!!!!!!!!!!!!!!!!
P.P.S. No insults or negative comments are allowed on this thread.
P.P.P.S. Tari cannot put any replies on any of my threads!!!!!
QFT.

(Mateo beat me to the disassembly so I have nothing else to add here. Have the port 0x20 reference I guess.)
I've managed over 40. Try 75!!!
CalcMax wrote:
I've managed over 40. Try 75!!!


How did you overclock it that high?
caleb1997 wrote:
CalcMax wrote:
I've managed over 40. Try 75!!!


How did you overclock it that high?


He didn't. It's impossible.
It is possible. I cant tell you the limit so that you don't accuse me of blowing your calculator!!!!
At the 75 setting, it is noticeably faster than at 40.

Note: some software will max out after a certain speed and not get any faster even with more OC!
CalcMax wrote:
It is possible. I cant tell you the limit so that you don't accuse me of blowing your calculator!!!!
At the 75 setting, it is noticeably faster than at 40.

Note: some software will max out after a certain speed and not get any faster even with more OC!


If you tell me, and my calc blows, you have my word that I will not blame you. It's my decision to use this, and my fault if something goes wrong.
I am not sure, after 100mhz, the speed doesnt increase in a linear path. You will have better speed at 150, 200 or even 250 settings, but i dont think path is linear. I currently have 500 set. After 100, the speed isnt same as setting. NO overheating, just minor warmth!
CalcMax wrote:
I am not sure, after 100mhz, the speed doesnt increase in a linear path. ... NO overheating, just minor warmth!


Even if the number there somehow does represent CPU clock in kHz (the disassembly shows otherwise), how did you get past 65.535 MHz?
"the speed doesnt increase in a linear path", obviously Evil or Very Mad
Now now, gentlemen, let's not mock the newbie just because he's not listening to the experts and is instead angrily standing his ground. CalcMax, since you seem to be ignoring (1) Tari and I and MateoConLechuga about what your assembly actually does, and (2) points about how you could even be getting the numbers you claim, I suggest you try an empirical test. Write a TI-BASIC program that counts from 1 to 1000, displaying each number. Use a stopwatch (or the stopwatch mode of your phone, if you have one) to time how long that program takes at the default 15MHz, and then with the calculator "overclocked" and compare the result. Because the TI-OS sets the CPU into its fastest possible mode for TI-BASIC anyway, another option would be to try a time-consuming command like rand(500). You'll see that your "overclocked" mode is not overclocked at all.
Sto. I was using sin(XX)XX and a few other functions. Minecraft worked to some point.
KermMartian wrote:
The TI-OS sets the CPU into its fastest possible mode for TI-BASIC anyway

I'm actually pretty disappointed in 84+ performance, even with 2.43 instead of a MP OS. I'll do empirical speed comparisons between calculators in another thread, now that time is no longer in a shortage for me.
Have code that lets you get 20x faster speed than 15mhz, but only if aaa cells are VERY FRESH!!!

This special code is top secret and i wont give it out. It has 5 or6 digits instead of 4.
So we can try and verify this ourselves, how fresh? Like "let them charge for 8 hours and immediately place them in the calc" fresh? How long will the batteries work in the calc until the over clock stops working?
The OC doesnt really stop. It just slows down a few tens of mhz, but sometimes recovers when it can pull more juice. I think my calc is definitely running at about 300mhz. NO MATTER HOW BIZERCK THE COMPLEXITY OF THE FUNCTION IS, THE CALC GRAPHS IT VERY QUICKLY(1.5-2 MINS). THESE FUNCS WOULD PROBABLY TAKE 40 MINS WITHOUT OC. I have had some on 15 mhz take 15min and they were much simpler. Alot of stuff has maxed out now. I have tried icreasing further, but HAD to revert due to the sierpinski triangle prog loosing presision on the extreme. Apps have reached limit and havent been really improving. I was impressed with the performance on the graphong and in wolfenstein83.

I MAY SHARETHE 300MHZ CODE WHEN I FEEL LIKE IT, BUT NOT NOW!!!!!!!

P.S. Did you try 75 and 40? For 75 just convert 75000 to hex and use instead of 9C40, just leaving the rest of the code intact.
  
Register to Join the Conversation
Have your own thoughts to add to this or any other topic? Want to ask a question, offer a suggestion, share your own programs and projects, upload a file to the file archives, get help with calculator and computer programming, or simply chat with like-minded coders and tech and calculator enthusiasts via the site-wide AJAX SAX widget? Registration for a free Cemetech account only takes a minute.

» Go to Registration page
» Goto page 1, 2, 3, 4, 5  Next
» View previous topic :: View next topic  
Page 1 of 5
» All times are UTC - 5 Hours
 
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum

 

Advertisement