Friday, August 4, 2017

No one ever says "It's just like using an oscilloscope."

It has been quite a week.

We have spent two full days trying to figure out why our payload kept tripping the fuse. Startup current for our pump was popping the HASP fuse, so we've been trying a bunch of mitigation strategies: reservoir capacitors, a "slow start" to ease into our full current draw, installing our own fuse, and at one point we even considered flying a battery to even things out. None of them worked.

At this point in the project, the necessity of looking at our electrical profile under an oscilloscope became increasingly apparent to us. It seems obvious in retrospect, but given that none of us have ever done this before we can't be too hard on ourselves. To add a bit of insult to injury, the physics student and electrical engineering majors both needed to re-learn how to use one. ("There's a reason that the expression is 'just like riding a bicycle' and not 'just like using an oscilloscope'." -Julie) The LSU faculty were very accommodating. Actually, everyone's been amazingly accommodating. We've been swapping spare parts with the other teams in post-midnight hotel-room liaisons. The faculty advisor for UB spent hours going over our circuit diagrams for us and gave freely (thanks, Larry!) from his impressive stock of circuit components and gubbins.

Anyway, we discovered, once we re-learned how to use basic lab equipment, that our little pump is just noisy. as. hell.

This waveform isn't even from the initial surge required to get the pump going; this is what it looks like during regular operation. That little spike in the middle (it's kinda hard to see, but the darker line above the 'M' at the bottom) represents a few nanoseconds' worth of ~1.75 A draw. You may recall that we are fused to never exceed our current allotment of 0.5A. Our pump is blowing that out of the water, for very small time intervals, multiple times per second. We never noticed it through all our testing because we've been using standard multimeters, which don't register data at anything close to that level of fine resolution. We tried evening this out with some more circuitry magic, but weren't having any luck.

The upshot of all of this is that we decided to just leave our pump off permanently. We're not going to use it. Our payload is now chugging along happily without its centerpiece. The Bernoulli Air Bath Intensifier (BABI) will be as quiet and low-pressure as the rest of the payload. This is sad and disappointing, but vastly better than flying a brick of non-functional electronics, which in turn is worlds better than not flying at all. We passed pre-integration and integration again, GOAT is hanging out in the vacuum chamber now, and we're about to start chilling for the first part of the thermal test.

It's going to be a wonderfully slow day compared to the shrieking half-blind anxiety of the last few. We're updating our documentation, reading, and chatting with the other teams. If this goes well, then we will truly have reached the end of the work for this project. That's quite a feeling.

In celebration, permit me to share with you a picture of our electrical/software lead airing his feet out after a long day in the high-bay: