A week passed by and we’ve only got a bunch of measurements more and still no results from the impedance meter. Whatever setting I choose, whatever plug I change, it just doesn’t give me the results I want to get. Apart from the noise I already mentioned, there is also a 180° phase shift I don’t really understand. Capacitors are supposed to shift current to the left by only 90°, not 180°. The only possibility that I know of that could cause this additional 90° shift is the instrumentation amplifier, but that should only happen at higher frequencies, not at 1 kHz or lower. The picoscope also wasn’t there for me to use the whole time, so I had some time to do other measurements as well. As I’m slowly running out of time, I started doing the important things first: the question of temperature dependability. I went to San Agustin to perform measurements in a different, warmer environment with beans coming directly from the sun as well as beans that have been laying around for a while to get a bit cooler. For both cases, our measurements were pretty bad. That could be because of the fact that – so far – we didn’t use temperature in our calculations to link capacity to humidity; but it also could be the Gehaka industrial humidity measuring devices that we use as a reference point. These devices are not being recalibrated once in a while, they’re full of dust and they even differ in version as regards to both hardware and software. This could mean that there is a huge deviation between two of those reference devices. There’s only one way to find out: measure the same samples with both devices. No worries, it tops my to-do list.
Besides that, and on a more positive note, I also got my hands on one of these Gehaka reference devices from Sol y Café. It’s mine for the month to come, so now I only need a balance to get that perfect 142g needed to get correct measurements from the Gehaka. I’m not sure where I’ll find/buy/hire that, but it’s a pretty necessary tool to get good measurements. Tomorrow hopefully I’ll see Harry, to whom I’ve been trying to speak the whole week to get some documents and to see what’s possible for measuring on greater heights.
An idea for improvement that I had, was to alter our curve from linear to quadratic. I have made a pretty nice excel sheet so far with all my measurements, and it can be seen that quadratic approximations are the better fit for these data points. According to excel, the quadratic curve only fits just a little bit better, but at the edges of our range, I think this might mean significantly better results. The problem with quadratic regression so far was that our chip wasn’t powerful enough, but I think it’s possible.
Another thing that I am supposed to find out, is whether devices stay more or less calibrated after long periods of time (metal plates tend to oxidize and that might change capacity). I found two devices from Humasol students from one and two years ago, but neither had calibration coefficients in it, so I can’t tell whether these coefficients changed or not.
During the weekend, we rented a car with driver and drove to Gocta, famous for its giant waterfall (catarata). According to Wikipedia it’s the third highest in the world. I’ve asked some Peruvians and got some different answers, but it’s about 750m high, divided in two pieces. Let’s just assume it’s ridiculously high. It’s so high, and there flows so much water, that it pulls air downwards with it and creates a microclimate at the bottom. The weather during the (exhausting, but totally worth it) 4km walk towards it was normal, what you’d expect from a forest at 2200m height. When you get close enough, you start to hear it and feel some wind. If you get close enough, it’s as if it were raining. Of course, Jeroen and I went right beneath the waterfall and it’s pretty much a small storm down there. You might be thinking that we’d die from the tons of water falling on us, but fortunately that didn’t happen because most of the water just evaporates during the fall. Gocta is definitely one of the most beautiful places I’ve been to in my whole life.
After visiting Gocta and a lot of driving, we arrived in Chachapoyas to spend the night. From there, we went on to see Kuelap. Kuelap is a city of ruins at 3000m above sea level and is associated with the Chachapoyas culture. If you were tired of seeing so many stones, you could just look a bit further and get a beautiful view of the surrounding mountains. Nature in Peru is incredible, but I guess the pictures can tell that better than I do.
Found a 2 year old humidity sensor…
…And a one year old, unfortunately uncalibrated.
Calibration curve, looking pretty good. As you can see, in the endpoints, the quadratic fits better.
The mechanical team showing off their working generator at the yapangos.
My peruvian bike’s chain broke.
View from the roof of casa de voluntarios. Not bad, ey?
Random picture shot in the car to Gocta
It’s pretty easy to shoot nice pictures here 🙂
Freek and Michiel posing on a bridge, Laura in the background.
A mill for extracting sugar juice from sugar canes (caña).
Brecht posing under a gate on the road to gocta.
Volleybal game in Chachapoyas
Mathieu climbing an ancient wall in Kuelap
Posing around with the Humasol logo 🙂
We’ve got company!
Lamas admiring the view
Mechanical team, electrical team and humidity sensor team united in a chifa!
Rice fields on the road to San Agustin
*** footage from Gocta yet to come! ***