Tags

In a previous post, I decided to remove the buffer (voltage follower) op amp within the EC’s rectifier circuit.  I did this because I was too lazy to observe what caused a surge in voltage readings when measuring the lower ECVin+ (the “shrunken waveform” that is the voltage input to the EC Gain loop – the protagonist of the EC Circuit). After reviewing with Chris, he did the right thing and “suggested” I stop being lazy.  From a circuit design perspective, keeping the buffer op amp will keep surges coming from the capacitor to mix in with the DC voltage going into the ADC.  This makes sense because the waveform going into an op amp is met with a high impedance while the waveform going out has a low impedance.

# The Goal

The goal of this post is to characterize (even better understand!) the voltage surge that occurs when the multiplexer switches what is being measured by the EC rectifier circuit from ECVin+ to ECVout.

# Discussion

When I ran an Arduino sketch that measured the value for ECVin+, the majority of the time the readings would indicate a surge in voltage that eventually settled down.  This was most obvious when measuring ECVout (a waveform 6 x that of ECVin+) first.

## Observations

### Scope Probe on Vout of Voltage Follower Op Amp

The video below shows the results on my scope on the ADC AIN0 pin (relative to GND):

• (6 seconds into the video) take 100 samples of ECVout first.  Here is the results the Arduino read:
| Count: 100** ECVout ** Average: 1585.37mV| Std deviation: 0.58
• (15 seconds into the video) take 5000 samples of ECVin+ after taking the ECVout samples.  The readings started at 1353mV then gradually went down to 249mV.  249mV is the expected value.

• When starting to sample the ADC from my Arduino sketch, the voltage readings for ECVin+ (the shrunken waveform in the EC Circuit) are higher than the DC value should be.  To settle down the DCVin+ DC voltage value, it is best to sample 5,000 times before taking a sample to measure.
• ECVout measurements do not have an initial voltage surge.

### Scope Probe on Vin From Multiplexer

If the multiplexer causes the voltage surge, I expected to observe changes in the waveforms coming into the rectifying op amps.  To observe this, I turn on AC Coupling on the Scope channel.  By doing so, DC biasing will be eliminated.  Since I am still learning – I’ll pause and give my interpretation of how the Scope does DC biasing and what that means.  The scope puts a capacitor at the beginning of the probe circuit. By doing this – since i = C(dv/dt) and DC’s frequency is close to 0, i = 0 for the DC current.  Only the AC current will be measured.  (ok, I’ll look back at this explanation in a year and cringe….always practicing…always learning…).

In this video, the scope’s probe is on the EC rectifier’s input pin (i.e.: the output of the multiplexer) relative to GND.

# Next Steps Based On Observations

Given my current understanding of circuit analysis, I’m concluding that the above observations point out to a slight voltage surge that occurs when the circuit for the multiplexer is chosen and the waveform starts flowing through the EC rectifier’s op amps.

The residual of this surge in voltage finds its way into the ADC readings.  To not include this artifact in ADC readings, at least 5000 samples need to be taken prior to taking samples that will be used in measurements.  By throwing out the initial 5,000 samples, artifacts of the voltage surge caused by switching inputs into the EC rectifier are eliminated.

As far as the hardware part of the circuit, I’m putting the voltage follower op amp back into the EC circuit.  Here is the current design (Dare I dream that I can move forward to another Osh Park layout?):