Quote of the Day
I am only an average man, but by George, I work harder at it than the average man.
— Theodore Roosevelt. I think about this quote often. I have worked with some very bright people, but the folks who really accomplished things were all hard workers.
While I am in management, I still get an occasional circuit design question – I really like these questions because the get me away from some of the monotonous aspects of management (i.e. budgeting). This morning a power supply designer stopped by and wanted to discuss an issue he was having with fixing an issue with an old design. I immediately dropped my budgeting activity and leaped into action.
Basically, the old design was working, but the output voltage needed to be made variable to compensate for variations we have observed. Here are his constraints:
- No major changes to the circuit because we need to use the same test fixture
- No added parts
- Minor trace changes are allowed
- Test points cannot be moved
- The required output voltage changes required are small
- We do have a spare Digital-to-Analog (DAC) converter available
While the circuit was quite complex overall, the part involved with setting the output voltage is simple (Figure 1). The solution ended up summing the spare DAC's output voltage with the power supply feedback control signal. The focus of this post is how I determined the correct value of the summing resistors. It was a nice application of Mathcad that I am going to use for a training class that I will be conducting early in 2016.
The requirements are straightforward:
- DAC can put out 0.25 V to 2.5 V
- We need the output voltage to vary from 0.7 V to something higher than 1.1 V.
- The DAC is tri-state (i.e. high-impedance) on powerup. The circuit must be well-behaved with the DAC output tri-stated.
Figure 2 shows my circuit analysis, which determines the resistor ratio (RT/RS) as a function of the minimum output voltage level.
Figure 3 shows a graphical presentation on how the output voltage varies with different minimum output voltages.
Figure 4 shows how I selected my components. In this selection, I used a program that I wrote that computes the optimal resistor values for a given ratio.
While this was a routine calculation, it does illustrate the process of optimizing a circuit subject to many constraints.
The circuit test was just completed today (4-March-2016) and the circuit passed all tests.