What steps will reproduce the problem?
1. Use r1.1 code with default voltage reference (1.1V)
2. Use a Witespy KV modded minimOSD board (22k/1k dividers)
3. Increase GUI calibration to max (255)
4. Voltage displayed on screen is still too low (example, shows 10.2V when it
should read 12.4V).
I fixed the issue on my project by simply changing the hardcoded voltage
scalers (mw_osd.ino, lines 588 & 598) from 0.0001 to 0.0003. My voltage display
then calibrated properly using a GUI value of 104. Center of 255 calibration
range is 122. So 104 is very close, and should allow proper calibration with
all currently available Witespy OSD boards.
Using the 5V reference is not a very good option, because resolution is lost.
Also, users may still need 1.1V for RSSI & current.
The current system checks out OK with basic theories and maths, but in practice
we may have a problem low input impedance. The mega328 sheet does suggest
output impedances <10kohm. I haven't measured linearity with the current
hardware. I impedance is a problem, linearity should be verified. If
significant non-linearity is found, then maybe the system should be changed to
use lower value resistors in the future (like 10k & 450)?
Original issue reported on code.google.com by imel...@surewest.net on 3 Jun 2014 at 6:51
Original issue reported on code.google.com by
imel...@surewest.net
on 3 Jun 2014 at 6:51