 |
 |
| View previous topic :: View next topic |
| Author |
Message |
treitmey
Joined: 23 Jan 2004
Posts: 1094
Location: Appleton,WI USA
|
| Math help, I2C 1MHz sspadd calc |
 Posted: Tue Nov 13, 2007 9:37 am |
 |
|
I am trying to verify that the compiler is working right.
I am trying to get a 1MHz I2C setup. I'm using
| Code: | #use I2C(MASTER,SDA=FRAM_SDA,SCL=FRAM_SCL,FORCE_HW,FAST,RESTART_WDT)
|
The list file shows
| Code: | .................... #use I2C(MASTER,SDA=FRAM_SDA,SCL=FRAM_SCL,FORCE_HW,FAST,RESTART_WDT)
56E4: BSF F94.3
56E6: BSF F94.4
56E8: MOVLW 18
56EA: MOVWF FC8
|
| Code: | #use I2C(MASTER,SDA=FRAM_SDA,SCL=FRAM_SCL,FORCE_HW,FAST=1000000,RESTART_WDT)
|
The list file shows
| Code: | .................... #use I2C(MASTER,SDA=FRAM_SDA,SCL=FRAM_SCL,FORCE_HW,FAST=1000000,RESTART_WDT)
56E0: MOVLW 90
56E2: MOVWF FAB
56E4: BSF F94.3
56E6: BSF F94.4
56E8: MOVLW 09
56EA: MOVWF FC8
|
The lower 7 bits SSPADD @ 0xFC8 controls the speed. for fast it is set to 18, for fast=1000000 it is set to 9
Is this calculated correctly?
I use a
#use delay(clock=40000000)
#fuses h4,wdt16384,noprotect,nolvp,put
10MHz clock and h4 pll.
Last edited by treitmey on Wed Nov 14, 2007 4:55 pm; edited 3 times in total |
|
 |
treitmey
Joined: 23 Jan 2004
Posts: 1094
Location: Appleton,WI USA
|
|
| Posted: Tue Nov 13, 2007 9:45 am |
|
|
| Code: |
TABLE 17-3: I2C™ CLOCK RATE W/BRG
Fosc FCY FCY*2 BRGValue FSCL
(2 Rollovers of BRG)
40 MHz 10 MHz 20 MHz 18h 400 kHz(1)
40 MHz 10 MHz 20 MHz 1Fh 312.5 kHz
40 MHz 10 MHz 20 MHz 63h 100 kHz
16 MHz 4 MHz 8 MHz 09h 400 kHz(1)
16 MHz 4 MHz 8 MHz 0Ch 308 kHz
16 MHz 4 MHz 8 MHz 27h 100 kHz
4 MHz 1 MHz 2 MHz 02h 333 kHz(1)
4 MHz 1 MHz 2 MHz 09h 100 kHz
4 MHz 1 MHz 2 MHz 00h 1 MHz(1) |
|
|
|
Ttelmah
Guest
|
|
| Posted: Tue Nov 13, 2007 9:58 am |
|
|
Two things. Remember the values are hex. The division is by val+1.
So for 400K, is is dividing the clock by twenty five (0x18+1). The clock is selected as master clock/4. So:
40000000/(4*25) = 400000
Spot on.
For the 100K version, it divides by ten. So:
40000000/(4*10) = 1000000
Again spot on.
The compiler is doing it right.
Best Wishes |
|
|
treitmey
Joined: 23 Jan 2004
Posts: 1094
Location: Appleton,WI USA
|
|
| Posted: Tue Nov 13, 2007 10:07 am |
|
|
Cool
Thanks a bunch.
((Ttelmah:100k should read 1M in your post))
Belt and suspenders theory.
I'll also write a tight loop and check with scope.
With scope I get 1uS for half(5v part) cycle of the SCL pin.
I guess thats correct. But not sure. I expected 1us for a FULL cycle. |
|
|
Ttelmah
Guest
|
|
| Posted: Tue Nov 13, 2007 4:00 pm |
|
|
Er. I don't have '100K'in my post anywhere. The value I have, is 1000000, which is 1M, and is correct.
I'm easily capable of missing a zero, but I haven't...
Best Wishes |
|
|
treitmey
Joined: 23 Jan 2004
Posts: 1094
Location: Appleton,WI USA
|
|
| Posted: Wed Nov 14, 2007 4:44 pm |
|
|
no worries..
| Quote: | | Posted: Tue Nov 13, 2007 9:58 am |
| Quote: | Spot on.
For the 100K version, it divides by ten. So:
40000000/(4*10) = 1000000
|
Thank you for your help.
 |
|
|
|
|
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
|
Powered by phpBB © 2001, 2005 phpBB Group
|
FAQ 
Search
Register
Profile 
Log in to check your private messages 
Log in
