Reply
Posts: 3
Registered: ‎10-30-2015

SmaRTClock differences in F41x and F92x

Hello guys,

 

First time typing here.

Do someone can explain why after any type of RST, the F92x,F93x CPUs are clearing/reseting the 32-bits SmaRTClock counter/timer too?

F41x is not doing so with its 47-bits, which allows keeping the Real-Time-Counting.

Checked by one-and-the-same high level code, taking in account the differences in CPU implementations, ex. counter size, register addresses...

 

Thank you,

Ivan.

Posts: 1,955
Registered: ‎10-14-2014

Re: SmaRTClock differences in F41x and F92x

@Vanio

Do you mind to elaborate this in more detail, what page of the datasheet you see these difference on?

 

WeiguoLu
Posts: 3
Registered: ‎10-30-2015

Re: SmaRTClock differences in F41x and F92x

No, no. I am not describing difference in datasheets. In opposite, they are almost equal, but still not providing information about the fact that F412 is not reseting the RTC timer value during any kind of RST, while F930 is reseting it. Means, F412 is keeping counting based on value before the RST, but F930 starts from 0 after RST. Could be funny but this happens  even on software reset or on RST done by Flash programmer when code execution starts.

The problem, in fact a real and serious problem is that, a device using F412 needs to migrate to CPU with bigger RAM and better power characteristics, but F930 does not keep the same functionality regarding RTC. If the manufacturer is a different company, OK, but the reliability to follow the same company politic is obviously missed here. I have opened this discussion in order to have the Silabs' CPU designers opinion about how this issue is qualified: as an error that will be fixed or will be kept as it is. The answer of this question is important from our business point of view.

 

Posts: 442
Registered: ‎01-18-2004

Re: SmaRTClock differences in F41x and F92x


Vanio wrote:

Hello guys,

 

First time typing here.

Do someone can explain why after any type of RST, the F92x,F93x CPUs are clearing/reseting the 32-bits SmaRTClock counter/timer too?

F41x is not doing so with its 47-bits, which allows keeping the Real-Time-Counting.


I can't explain why the parts are different, but the data sheets for both devices are explicit.

 

For the '410, page 182 of data sheet rev 1.1 tells us about RTC0DAT: smaRTClock Data and its reset value is noted as "variable." 

 

For the '920/'930, page 205 of its rev 1.4 data sheet tells us that RTC0DAT is reset to 0.

 

Again, as for why the difference, I do not know, but the documentation is clear.

 

(This wouldn't be the first time that a peripheral that is in different SiLabs device families doesn't act the same in each.)

Posts: 3
Registered: ‎10-30-2015

Re: SmaRTClock differences in F41x and F92x

Hello Andy,

Thank you to taking this topic. But that is not the point. Datasheets difference you described here is not relevant to the issue. The topic is about RTC counter, not about RTC interface registers. Meanwhile I got a mail from Silabs sales (?) saying that they will check the issue.