atd stuck bits

[jasl8r]

I am using the atd in 10-bit mode on a 68hc912b32, and the 2 LSbs never change from '10'. This appears to be a somewhat similar issue to this one. Despite the +-2LSb accuracy, shouldn't I still be seeing a variable value in these two bits?

I'm using this code to begin the conversion.

LDAA #$C2
STAA atd_ctl2
;Delay 100us
LDAA #$E1
STAA atd_ctl3
LDAA #$44
STAA atd_ctl5

~Jeremy

[bigmac]

Hello,

According to the datasheet, the HC912B32 device does not have a 10-bit mode - the ATD is 8 bits only.  Or am I looking at the wrong datasheet?

Regards,

Mac

[jasl8r]

I am reading the M68HC12B family datasheet which covers the MC68HC912B32. It describes 8 and 10-bit atd operating modes and does not mention that any processor only has an 8-bit mode. Can you post a link to the datasheet that you see that in?

Thanks,
Jeremy

[bigmac]

Hello Jeremy,

This is very curious!

I was reading from a quite old "Technical Summary" document (MC68HC912B32TS/D) that specifically listed 8 channels of 8-bit resolution.  However, I could not find any errata document that covered more recent upgrading of the ATD from 8-bit to 10-bit resolution.  It is not unknown for Freescale to do this, but an A-suffix is usually attached to the part number when changes of this magnitude occur.

If the new parts do actually handle 10-bit resolution, I wonder if the part you have been testing is an old one.

According to the Freescale website, it would seem this part is no longer being manufactured.

Regards,

Mac

[jasl8r]

On the freescale processor page it does describe it as having the 10-bit atd. I also do in fact have the word sized memory locations for storing the results, with the lower byte just stuck at '0x80'. I did try my program on two different chips as well, so I don't think that there is any hardware error.

Message Edited by jasl8r on 2006-12-0709:39 AM

[jasl8r]

Okay, so I was writing to the incorrect register. Figures.
0xE1 => atd_ctl4, and I was writing to atd_ctl3. Sorry for the trouble.

Jeremy