william_teter

Fair enough. Thank you. Any ideas on how to get rid of the offset issue? I have tried multiple solutions, but this has worked the best sofar

The number of bits refers to the size of the data pipe. C will probably hide this from you this, but if you are programming with assembly, you will find if you need to add a 32 bit number to another 32 bit number, you will need far fewer instructions to do so on the 32 bit uC vs the 8 unless it has a special instruction to accomplish this. The address space will be smaller in an 8-bit uC.. but once again hardware can allow for bank switching and allow the 8bit uC to access just as much memory as a 32bit uC could. For just a simple hello world code written in c, you may not notice a difference... it's when you get into more complex applications that this can become a problem.

I have the Rigol. It is quit nice and modifiable to act as a 100Mhz DSO. You can see a good review of it on Dave's EEV Blog. I believe he said Agilent designed it. I don't know anything about the other scope.. the wider display looks nice

This circuit takes the output of a dac (min of 0v and max of ~4.096v.. [depending on % off]) that comes in through the Y port and creates a signal used to drive a galvo amplifier. The galvo amplifier requires a balanced bipolar differential signal between -10v and 10v. In this circuit the ILDA Y + line outputs between -5v to 5v and the ILDA Y - line outputs between +5v to -5v. I think this circuit suffers from the following: -Dead zone caused by op-amp's dc offset(?) -Independent voltage references (both appear to be sourcing about 1.2mA on the vRef lines) can be off by 0.5% (and that's ignoring temperature differences!). That means the center could be shifted a full 1% off! -Off-shifted center will mean a dac step above the center value will not be the same size as a dac step below the center value of the dac on the ILDA Y+ and ILDA Y- outputs. Edit: It's kinda hard to read, but that top left voltage says 2.048V

I have a circuit that needs two voltage references, lets say 5v and 2.5v What would be the best way to ensure the lower value was exactly 1/2 the value of the higher voltage (e.g. accounting for % tolerances on the 5v voltage reference)? Temperature variances won't be too much of a concern, but if a solution accounted for it, that would be nice. I've thought of: 1. Using an individual 5v and 2.5v vref IC 2. Using a resistor divider using high precision resistors I think the latter idea is a better one since two resistors are far cheaper AND will account for the 5v reference tolerances. The downside would be no current which I shouldn't need much at all of (it is the vref for a dac with microamps current specs on the pin) Just a theory question!