Longer than expected WDT timeout on 10F206...?

[spero]

PIC: 10F206
Vdd = 2.2V
WDT: Active

I have set the WDT period to 72mS (1:4 prescalar) and putting the PIC to sleep expecting it to wake up 72mS later. However it wakes up approximately 95mS later...trying to figure out why it's taking so much longer.

Initially I thought that it was the Device Reset Timer (Tdrt) adding 18mS which brings the wake-up time roughly inline with the 95mS that I'm seeing. However on closer inspection I see that the DRT timer is 18mS only for a power-on reset and 10uS for subsequent resets, that's if I understand the datasheet correctly; see Table 9-3 on page 46.

Can the long WDT reset time be due to the fact that I'm running the PIC at the low end of its Vdd range?

Cheers,
Spero.

[PCM programmer]

The 10F206 data sheet doesn't have graphs of WDT timeout period
with respect to Vdd voltage. However, the 18F452 data sheet does:
http://ww1.microchip.com/downloads/en/DeviceDoc/39564b.pdf
Go to page 300 in the Acrobat reader and look at Figure 23-19.
Looking at the curve for 25 degrees C, the timeout period is about
17 ms at 5.0v. At 2.2v, it's about 25 ms. So definitely, with lower
Vdd voltage you will get a longer WDT timeout period.

[Ttelmah]

It is also worth looking at the data sheet figures, and reading the 'comments'. At 5v, the standard timout period, is 9mSec minimum to 30mSec max, with the comment that 'these figures are characterised but not tested'. So even at 5v, the figure can be expected to range btween these values, and may even range slightly more!. With voltage change as well, the range may even be larger, and for the extended temperature range, yet larger still!...
The centre column giving the 18mSecc value,has the comment 'Data in the Typ column, is at 5v, 25c, unless otherwise stated. These parameters are for design guidance only'.
The watchdog timeout, is an _inaccurate_ time. It must not be relied on for anything that requires reasonable accuracy.

Best Wishes