Help with Programmable Current Source-

nicsky

Joined Aug 1, 2009
11
@Ta Ba
Responding to your post#15:
You say "DAC" but you describe a "PWM" signal. How much voltage amplitude ripple is on the "DAC" signal at its output (that will drive the current source) = 0V? =0.1mV? =1mV? That is, a PWM signal is a series of discrete pulses that must be averaged to convert the duty cycle to a voltage. I am asking, if the duty cycle remains constant (e.g. at 10%), how much ripple (not removed by the averaging circuitry) will be present in the voltage presented to the input of your current source?

You say duty cycle ranges from 1/(10us +40us) to 1/(10 us+1s); that is, from 20000 to 0.999990. That is not possible; duty cycle must be in the range 0 - 100% (or 0.000 to 1.000 as a ratio). What are the actual duty cycle limits?

You say "DAC" period, actually PWM period, is 10us, fixed. This PWM signal has a resolution of 10us/(1000/2)=10us/500=20ns. That's a pretty short interval for a PWM, meaning that each "step" increase in duty cycle corresponds to an interval of 20ns. Can the MCU you are using actually produce such a fast PWM signal?

To design a current source it will be necessary to know how rapidly the "DAC" signal can change. Theoretically, from the info you have provided thus far, that signal could change from 0V to full-scale in 10us (one PWM period). That's pretty fast. Realistically, what is the fastest change expected in the real signal? For example, it might be 1mV/sec, or 400mv/sec, or ???
The design of this type of current source is well defone ny a lt app note lt6082 data sheet and i designed a similar cct but with instrument amp with gain change plus selectable sense resistor this gives four ranges 10u to 10n
This cct operates via a load to ground. The frequency response of this is likely to be an issue with such larger impedances. Shielding and guarded tracks are essential. My biggest question is. Why use a dac to gemerate a pwm signal? The input signal could be much bigger and less sensitive
 
Top