Timing internal to the OM4(T) comes from a 32,768Hz crystal, which are
common clock crystals. There is a 2**n divider which generates various
timing signals as well as the CPU counter to determine shutter speed in
manual, spot, and memory mode. The external divider has 27 stages,
enough to give the longest timeout of 68 minutes for the memory hold
function, ie. 4096 seconds. The other major timeout is the 120 second
shutter limit in (center-weighted) auto mode. In auto exposure, the
shutter timing (Tv) is determined by the voltage integrated onto a
capacitor from the photo diode compared to another voltage generated
from the aperture and ASA settings (Av-Sv). Spot, memo, and the Av-Sv
signals can be digitized with an 8 bit resolution of a logrithmic
values, with the LSB being 1/12 EV. The full scale range is 256*1/12 =
21.33 EV. The complete details of how everything is calculated can be
found in:
http://www.patents.ibm.com/details?pn=US04618238__&language=en
Timing for manual, spot or memory modes is done by presetting a counter
in the CPU. The count may not be exact binary counts as you suggested.
The accuracy of the stored Tv value is 1/12 EV, and according to the
patent is approximated in the counter as:
Tv=12(12*X + Y + 1/12*Z)
"where X, Y, Z are integers Accordingly, an exposure period T can be
expressed as follows:
T = (1/f)*2^(Tv/12) = (1/f)*2^(12*X + Y + 1/12*Z)
where f represents the frequendy of the clock pulse CK. This can be
approximated by the following equation:
T = 1/f)(1+Z/12)^(2 12*X +Y) <-- can't quite read
Accordingly, when presetting the Tv value (M8) in the timer counter,
the Tv value is initially divided by 12, and then a duodecimal fraction
(which is assumed to comprise four bits) is determined. <and a whole
lot of more stuff, shifting bits, etc.> The timer counter is then
decremented with a period of 1/f, thus counting the exposure period.
..."
So, it seems there is plenty of range to do the 4 minute max spot
exposures. The auto exposure is limited by a hardwired counter circuit.
It's getting late and there may be some typos....
Wayne
Hopkinton, MA
< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >
|