Radiophilic

Our set of carbon fiber test materials for the bike project is in from Soller Composites!

As a little plug, I have to say that their customer service is superb. When NoScript defeated their order system, the owner of the company actually got back to me about the error and helped me sort it out. They were also very good with handling the order, and shipping it. So far, I couldn’t be more impressed.

We’ll be testing the material out CIV102 style soon to establish whether it is indeed as strong as claimed, and working towards the first set of test structures for the frame.

Speaking of the frame, we’ve decided that we’ll pursue a frame construction method using aluminum tubing (1″-3″ in diameter) for the structural sections, joined together by carbon fiber joints. This will allow us the flexibility and rapid prototyping capability associated with hand layup for joints, and keep costs low with aluminum structural sections.

Here are a few pictures:

Paul and I have been putting some time in to the trike project, and we’re happy to report some early successes.

For example, we’ve got a motor running on Microchip’s MCLV dev board! Check out the video below.

I’ll be speaking at InPlay 2011, so if you’re into games, check it out! You can find specific information about our talk here.

So, a friend and I have  been talking about and working sporadically on this trike project for a few long months now, and we’ve decided that this is the summer we’re going to get some real work done on it.

So, to kick off a productive summer, we put together the dev environment for the drive side of things. We put together a motor control dev board, (rewound) motor, and Hall Effect sensors together. And mounted most of it on a tub, as per below:

More progress to come, of course, so stay tuned.

Alright, here’s part two.

Conclusions first, for the impatient.

The motor has been successfully reconfigured! The Kv value is roughly 105 rpm/V. I used 21 half-turns instead of 15. This is at the upper limit of optimal values for my proposed drivetrain (another post for another time) and will do perfectly, since this means maximum power output!

I suspect that the Kv/Number-of-Winding relationship becomes somewhat non-linear when you push relatively small motors to low Kv values. In this winding configuration, I predicted a Kv of 65-85, based on information gathered from Scorpion’s website about the other motors with the same geometry but different winding configurations.

Qualitatively, motor winding is conceptually easy but hard in terms of time taken (it took me almost 6 hours!) and concentration required (ticking off windings as you do them helps.)

I’ll be doing more detailed current and torque measurements eventually. For now, this is good enough: at 12V, I can’t trigger the power supply current limiter (presumably somewhat above 24A) and I can’t stop the motor with my hands (by grabbing the bell) — the friction heats the motor/my hand enough that I have to let go or risk burning myself.

Here’s a timelapse of the assembly process.

It’s been a day of conventionally earned lessons (this being a potential subject for another post, of course.)

In short (heh, pardon the pun.)

I’ve been playing around with a Scorpion S-4035-250Kv motor for the trike drive system. Yesterday, I did two things:

1) Characterization: It turns out that the motor really does run at pretty close to 250Kv. The testing setup is pretty much detailed (less a tachometer improvised from an IR-based opto-interrupt sensor hooked up to my trusty ‘scope) in my last video, so I won’t waste too much time on that here.

2) Re-termination from Delta to Wye. This was interesting because the resultant Kv was about 190 rpm/V instead of 250 * sqrt(3) = ~150 rpm/V as expected. I’m not really sure what happened here. The motor ran smoothly enough, though.

The conclusion? Re-winding required! I’ve attached a time-lapse video of the de-winding process that might serve as a good qualitative guide to what’s involved. The re-winding part is coming, soon.

Some interesting lessons about rewinding this Scorpion motor:
- Magnet wire sheathing rated at 200C is TOUGH. Don’t expect to be able to melt it off. My soldering iron goes up to 450C, not nearly hot enough to make even a scratch in the wire coating.
- Scorpion motors are well built. If they’re all made like my unit, the stator is press-fit and epoxied onto the bearing holder with high-temp glue that won’t submit to heat gun torture (though the stator laminations did sort of get sticky, presumably melting lamination material between stator layers.)
- I’ve managed to unwind the stator with the bearing holder still attached, and I’m pretty sure I’ll be able to rewind it, too, so the last point isn’t the end of the world. On the other hand, time will tell.

I extracted the factory winding pattern from the motor as I unwound it. It’s attached here, for your perusal. The view is the ‘top’ view, looking down at the stator from the side to which the bearing holder is attached.

Just put together one of the motors I plan on using for my e-trike project. Videos arguably tell about 30,000 words per minute (1000 words/picture * 30 pictures/second) so without further ado:

I tested the motor using an old controller left over from a teeny tiny R/C car. Aside from having some problems with a $15 Canadian Tire soldering iron the assembly was straightforward (ie. plugging things into other things.)

The only thing of note not shown in the video is the ATX power supply used to power the beast. Rated at 24A on the 12V rail, it was almost laughably easy to trigger the short protection of the supply.

More to come on this topic.

Astronomy always puts things in perspective for me. It’s probably the scale of things, both in time and space.

On the 21st, it was possible to see the first lunar eclipse on the winter solstice in over 400 years. The last time this phenomenon was seen dates back to the dawn (pardon the pun) of analytical heliocentrism.

Shot with a Canon 7D and 70-200mm f4L IS lens, tripod mounted. ISO800, 0.5s exposure @f4.5.

I’ve recently been doing a fair amount of video and audio work. Among this work is a trip to Welland, Ontario with Darrett Zusko, to record his concert there.

Without further ado, here are the videos! I’ll be posting a semi-brief howto guide along with these, as soon as I get a chance to write it.

Those of you who know my photo habit know I’ve had a Canon 7D for about a year now. And while I’ve played around with video, this is definitely the first time I’ve produced something even remotely serious.

Adrian Anantawan is a good friend of mine, and a very talented violinist. It was my pleasure to work with him on this project, and I’m sure we’ll have many more coming down the pipeline.

For those of you who are interersted, the audio was done using a stereo pair of Neumann TLM103s, in pure A-B stereo, with a mic separation of approximately 50cm. Because of the recording geometry, the stereo image derives almost entirely from phase shifts due to difference in acoustic path length. This has a neat, if subtle effect.