PDA

View Full Version : How do I enter logarithms with bases on a TI-89?

pagemap
09-12-2002, 09:36 PM
Such as if I am trying to solve for C in this equation:

C = 2W log_2( 1 + S/N )^1/2

sleepminded
09-12-2002, 09:43 PM
Originally posted by pagemap
Such as if I am trying to solve for C in this equation:

C = 2W log_2( 1 + S/N )^1/2

hey...i was gonna ask that...eventually... :dodgy:

glagic
09-12-2002, 09:44 PM
convert it to natural log (ln). i forgot which is over which, but thats the only way. you CAN use logarithms on the ti-89, but it can only be base 10.

pagemap
09-12-2002, 09:48 PM
Ok here is the problem I was working on:

W = 3000 Hz
S/N = 250 (given)
So,

C = 2Wlog_2(1+S/N)1/2
C = 2(3000)log_2(1+250)1/2
C = 6000(1/2)[log_10(251)/log_10(2)]
C = 3000(7.97)
C = 23,900 bps

Guess I had to do it the long way.... oh well.

glagic
09-12-2002, 09:52 PM
did it work with converting the logrithm to natural log then?

pagemap
09-12-2002, 09:54 PM
That is Shannon's capacity theorem.. are you working on this type of stuff in class right now? And no, I didnt mess with natural logs.

Windsor
09-13-2002, 01:02 AM
I'm pretty sure the 89 can only do natural log, so you'd have to convert.

ChrisMG187
09-13-2002, 06:10 PM
actually the ti-89 can do regular base 10 log... there just isnt a button for it. Just use the log() function. I don't know how to do anything othere than base 10 on it, but you could just convert it to base 10.

i6s1
09-13-2002, 08:04 PM
Originally posted by pagemap
That is Shannon's capacity theorem.. are you working on this type of stuff in class right now? And no, I didnt mess with natural logs.

Hey, I did that in DataComm.

Log_10(1+S/N) / Log_10(2) = Log_2(1+S/N)

I had to do it the same way as you... I have an 83+.

spigidygak
09-15-2002, 01:46 AM
Originally posted by i6s1

Hey, I did that in DataComm.

Log_10(1+S/N) / Log_10(2) = Log_2(1+S/N)

I had to do it the same way as you... I have an 83+.

Yeah thats sorta all you need to do no need to convert just enter in the log base 10 then put it over the log of whatever you need it to be. No need to convert to natural logs. . . BTW this is where that 40" manual comes in VERY handy.

pagemap
09-29-2002, 01:48 PM
OK guys and gals, I have found a MUCH easier way to go about this.
Instead of this hairy beast: C = 2W log_2( 1 + S/N )^1/2

Why not use this instead?

C = 2W(.5)[log(S/N+1)/log(2)]

Where
C=max theoretical capacity
W=range of frequencies available
S/N= signal to noise ratio