swaggy said:
Curious - what improvements will the new board have?
Haven't nailed it down yet, but I plan to address all of the limitations I encountered with the first version.
For example, the first version was intentionally made using fixed resistors. Less board space, cheaper, lower noise, less drift. However the big limitation with that has been trying to tweek gains to exactly align full scale input with webcontrol A/D full scale. The new board will use 20-turn trimpots for the gain.
While using one of the boards with a 3-axis accelerometer, I would have liked a power supply +V available at the analog input connector, so that's going in too. (5 pins instead of 4).
Installing resistors and a jumper in either of two places to set inverting or non-inverting operation was ok, but if I replace the resistors with trimpot, setting the mode would require jumpers either way. I'm putting pads to set the mode, but also putting tracks between the pins for the most common configuration (non-inverting).
Things like that will make little difference to most people, but tidies it up a bit in other regards, and makes assembly quicker. The total cost will be more though, because the 20-turn trimpots are more expensive than fixed resistors, etc.
Oh, and to keep the board the same size I've had to go to a SOIC (surface mount) amplifier. Makes it harder for the average DIY to assemble, but I can always solder the chip here before shipping.