Re: Cygnal/Silabs F12x MAC0, does it work?

Do you have a question? Post it now! No Registration Necessary

Translate This Thread From English to

Threaded View
Hi Gary,

Quoted text here. Click to load it

I supposed that the IDE might have a problem reading out the
MAC0 registers. That's why I did move the result from the
MAC0ACCx registers to variables in the DATA space, and watch
these variables. But these variables are always zero too :-(

Quoted text here. Click to load it

It's the original IDE from Silicon Labs, latest version 1.9
downloaded from their website.

Of course I asked them before I posted the problem here.
No answer from them since 3 days :-(

Michael

Re: Cygnal/Silabs F12x MAC0, does it work?
have you set SFRPAGE register?
Quoted text here. Click to load it



Re: Cygnal/Silabs F12x MAC0, does it work?
Hi,

Quoted text here. Click to load it

yes, SFRPAGE = 3

have you already tested the MAC0?

Michael

Re: Cygnal/Silabs F12x MAC0, does it work?
Quoted text here. Click to load it

MAC0ACC3 should be 0 and MAC0ACC2 should be 03 with the values you are using.

Sometimes there can be bad chip batches, but you tried a 122 and a 123 and the
odds that those two would exhibit the same problem would be small, as far as I
know. Of course, for all I know the 123 is just a 122 in a different package
and they might be from the same batch.

It's probably not something to do with external hardware as the multiplier is
completely internal.

It's not an SFR watch window problem since you're reading them into low RAM as
well and they confirm that it's 0. The MAC status seems to read fine and it
also says it's 0.

Does the IDE have a simulator mode? If so, does that work?

I wouldn't think it has anything to do with any JTAG weirdness as other stuff
seems to be coming through just fine.

I'm stumped. Hopefully Silicon Labs will be able to shine some light on what is
going on.


Re: Cygnal/Silabs F12x MAC0, does it work?
Hi Gary,

Quoted text here. Click to load it

that's correct.


No, there is no simulator mode.
 
Quoted text here. Click to load it
seems to be coming through just fine.

agreed. Everything else is working fine in these chips.
There have never been any problems with the JTAG interface.
 
Michael

Re: Cygnal/Silabs F12x MAC0, does it work?
About the only other thing I can think of to try is to use a faster clock. On
reset, the internal 24MHz oscillator is used and that is divided by 8 to
generate the SYSCLK. You might try changing it to divide by 1. People tend to
operate at faster speeds so maybe the MAC wasn't tested at the default 3MHz
SYSCLK.


Re: Cygnal/Silabs F12x MAC0, does it work?
Hi Gary,

Quoted text here. Click to load it

I have already tested at the default clock and at 99MHz.
Same negative results.

Michael

Re: Cygnal/Silabs F12x MAC0, does it work?

Quoted text here. Click to load it

I have tested it on Keil's uV2 with C8051F120 and everyting works as
supposed...

regards

Dejan



Re: Cygnal/Silabs F12x MAC0, does it work?
Hi Dejan,

Quoted text here. Click to load it

it's good to hear that it works. Seems that I got some early
prototypes with a bug in the silicon.
Hopefully SiLabs will shine some light on this problem
soon...

Thanks,
Michael

Re: Cygnal/Silabs F12x MAC0, does it work?
Hi all,

the problem is solved. Silicon Labs answered that it's a bug
in the silicon in F122/F123 controllers that were shipped
before October 31, 2003.

Thanks,
Michael

Re: Cygnal/Silabs F12x MAC0, does it work?
Quoted text here. Click to load it

So much for their quality control.


Site Timeline