Proxmark3 community

iceman

Administrator

Registered: 2013-04-25

Posts: 9,538

Website

LF divisor inconsistancy

LF part uses a 12 000 000 (12Mhz) clock to generate the carrier freq on antenna.

To get 125Khz (125000) we need a divisor to adjust the carrier freq.
Hence  12 000 000 / 125000 = 96.
However in the helptext for "hw setlfdivisor" it states

> (19-255) Drive Antenna at 12Mhz / divisor+1

Its here we get the 125Khz divisor from "hw tune" output,  95,  since 96-1 == 95.
So far so good.

134Khz,  gives 12 000 000 / 134000 = 89.55 

Testing:
 divisor | (divisor * 134000) | diff  (12 000 000-nnnn)
-------------------------------------------------------------
  89     | (11926000)         | 74000  
  90     | (12060000)         | 60000

Smallest marginal is a divisor of 90,  which would translate into PM3 settings like 90-1= 89.

Now this raises more questions.  What is the definition of optimum divisor?  The one which hits the sought after Khz? Or the one which diviates less from the target Khz?

In PM3 code:
all places in code uses "88" as optimum divisor for 134Khz.

This is also an issue on github
REF:  https://github.com/iceman1001/proxmark3/issues/74

iceman

Administrator

Registered: 2013-04-25

Posts: 9,538

Website

Re: LF divisor inconsistancy

And the real reason for bringing it up is Cotag,   which has 132Khz from Reader,  while the card answers in 66Khz.

Which raises the question of optimal divisor for 66Khz.

12000000 / 66000 = 181.8181  
       181 (11946000) 12000000-11946000 == 54000
       182 (12012000) 12000000-12012000 == 12000

PM3:
 182-1=181   or
 181-1=180 

marshmellow

Contributor

From: US

Registered: 2013-06-10

Posts: 2,302

Re: LF divisor inconsistancy

my take on the 134khz:  the differences are marginal.  it likely won't matter which is used. 

since 134 is so close to 125 they may have rounded away from 125 to have less overlap. ( you can "see" 134khz tags on a 125khz read and the opposite too, just not very well )

but i believe typically the goal would be to get as close to the target khz as possible.

marshmellow

Contributor

From: US

Registered: 2013-06-10

Posts: 2,302

Re: LF divisor inconsistancy

as to your cotag card, i seem to recall all lf tags say they respond at 1/2 the carrier frequency.  i don't think you need to adjust the pm3 frequency to 66khz to listen to the tag, 134khz should work.

iceman

Administrator

Registered: 2013-04-25

Posts: 9,538

Website

Re: LF divisor inconsistancy

Its different I guess,  I've seen in documents 134 Kh and 134.2 Khz.     PM3 is using 134.8 Khz
Haven't seen too many 134 tags in the wild.