The View Up Here

Random scribblings about kites, photography, machining, and anything else

Archive for July, 2015

Auditory Epiphany

Posted by Tom Benedict on 24/07/2015

Earlier in the week I went to pick my daughter up from class, only to find I’d arrived an hour too early. I didn’t have anywhere else to be, so I grabbed my recording gear and proceeded to play. I tested my hot glue filled mono mic against the newer designs with the screw-on backplates (no difference). I tested them against the built-ins to see if the lack of bass response I was getting in my prototype pseudo-SASS array was inherent in the mic body design, or if it came from the SASS (it came from the SASS). Then I just clipped my lavalier mics up in a tree and kicked back to record various bugs, frogs, and birds until my daughter’s class ended.

When I got home I loaded the last track into Audacity to see what I could see. Or hear, rather. I switched it over to spectral view – what would be considered a sideways waterfall view from a radio standpoint – and looked and listened. That’s when the epiphany took hold.

While I was making the recording, I could tell the audio was dominated by a bunch of crickets. What I saw in Audacity was that there were three species of cricket, all going on slightly different frequencies. I could hear them start and stop their call in the audio, identify which one it was from where it appeared in stereo space, and see which frequency it was calling on in the display. WOW!

Then I went back and listened to the birds. Then the frogs. That’s when I saw the signature of the coqui frog.

Coqui Signature

Coqui frogs are endemic to Puerto Rico. They were introduced into the Hawaii ecosystem several years ago, and have thrived ever since. From what I understand, coquis are a point of pride in Puerto Rico similar to the mockingbird in Texas. But in Hawaii they’re considered a pest. In the densely populated areas on the wet side of the islands, their nighttime calls can be deafening.

This stretch of audio has very few bird calls, but you can see the crickets around 6000Hz. I think only two species are going in this plot. The frogs show up between about 1500Hz and 4000Hz. The coqui frogs show up as a pair of comma/apostrophe marks, the first around 1500Hz, followed by a second that starts around 2500Hz and finishes at 3000Hz.

All of this, of course, is old hat to an amphibian biologist, just as my entire epiphany is old hat to anyone who’s been doing sound for any amount of time. But to me both of these were new.

I’m enjoying this foray into sound. I can’t wait to finish the other ground-based experiments I’m doing and put this to the test in the air.

– Tom

Advertisements

Posted in Audio | Tagged: , , , , , | Leave a Comment »

Finishing the DIY Microphones (v.1.1)

Posted by Tom Benedict on 14/07/2015

The more I thought about the hot glue closure on the back of the microphones, the less I liked it. Don’t get me wrong. It works well. But it’s… permanent. I know the BT-EM172 capsules are only $10, and I know the rest of the microphone is largely scrap-boxed, but I hate to make a thing that can’t be serviced when it needs it.

So I re-designed the enclosure to include an end-cap. It’s drilled out 1/4″ to take a cable grommet, and has three #2-56 screws placed every 120 degrees around the periphery to hold it in place.

BT-EM172 Microphone Enclosure - Exploded View

The end caps took about fifteen minutes apiece to make, and were a comfortable fit in the back of the microphone bodies I made previously. Unfortunately, drilling and counter-sinking the screw holes for the end cap meant I needed to re-coat the microphone bodies along with the end caps. Since I had to re-coat them, I added grooves to each mic body to accommodate a Shure RK183T1 lavalier clip. I’m pretty sure a generic clip for a 9/16″ diameter mic would’ve worked fine, but these turn out to be tough to find. There are several listed on Ebay, but if you look at the metric equivalents, the specs say they fit something around 7-9mm in diameter. 9/16″ is closer to 14mm, so I think something was lost in translation. The clips from Shure will fit. (For shure! Har!)

Countersinking Screw Holes

I wasn’t happy with my previous coating job, so I came up with another way to apply the coating. I shoved each part onto a wooden dowel of the appropriate diameter (3/8″ for the end caps, 1/2″ for the mic body), and chucked it in a drill. I applied the Cerakote with the drill spinning. This gave each part a very uniform coating, and let me hit every outside surface without running into my fixture. I loaded the parts in the oven, dowels and all. On a whim I coated the screw heads, too, so I wouldn’t have shiny stainless screws in a black microphone body. Unfortunately the spray gun malfunctioned, so two of the mic bodies didn’t turn out as nice as I’d like. I slated those for the pseudo-SASS array, where they won’t be seen, and saved the two “good” ones for lavalier mics. Note to self: test the spray gun before loading product into it!

Parts Ready to Cerakote

Once the Cerakote cured it should’ve been a simple matter of assembling each of the mics. But I love to fiddle. I assembled the two for the pseudo-SASS array since I already had that cable made. But I needed more cable for the lavalier mics. Even though I’m already using Mogami W3031 cable for the other mics, I ordered 100′ of Mogami W2697 from Redco Audio to use for the generic lavs (only 20′ of which I plan to use). W2697 is almost identical to W3031, except for the way the shield is constructed. W3031 uses a braided shield. W2697’s shield is served (wrapped). Electrically they’re identical. But a served shield is easier to work with when making cables. I’ll have to wait for the cable and clips to come in before finishing the generic lavs.

Completed Mic Bodies

Rather than waiting like I did with the mono mic I built out, I grabbed my pseudo-SASS array and my recorder, and hiked out to the rocks south of Hapuna Beach. The last time I was there the waves were big, and made big, dramatic crash-bam-booms on the rocks. Of course that was in the winter. The summer wave pattern is a lot more bathtub-like, so the sound was a lot more subtle. Still, I ran several side-by-side comparisons of the pseudo-SASS against the built-in mics on the Tascam DR-05. I put together a set of 30-second clips comparing the two. The recording has eight tracks, alternating between the DR-05 built-in mics and the BT-EM172 array, done at four locations. When listening, keep in mind that the gains are different on the two mics, as are the frequency responses. I did no processing on the tracks aside from cutting and fading, so some tracks are louder than others. That’s a function of my technique in the field (or lack thereof), not the microphones themselves. This test was only so I could tell how well the pseudo-SASS array was separating the two channels.

The pseudo-SASS performed well enough I want to build a real one out of some 1/4″ baltic birch plywood I have in the shop. I still haven’t tested my prototype from the air, but it’s easy enough to include 1/4″-20 sockets top and bottom so I can mount it either way. More photos and sound samples to come!

– Tom

P.S. I’m not keen on the way clips from Soundcloud show up on my web site. I’ve seen other people include Soundcloud clips on their sites that are nice, small, and easily worked with. This thing is ungainly! If you know how to fix this, please let me know.

Posted in Audio, Engineering, Machining | Tagged: , , , , , , , , | Leave a Comment »

Sitelle – Here at Last

Posted by Tom Benedict on 08/07/2015

A little over four years ago I posted a rendering to Flickr.

Cryogenic Detector Internals

I’d been tasked with designing and building new internals for a pair of matched cameras. The rendering was what I thought would be the final design of these parts. I try to save my Flickr stream for photos, but it was an important milestone in my professional life. It was also visual, aesthetically pleasing, and geeky, so I figured it was fair game.

I was wrong, of course. Shortly after posting that rendering I was told we couldn’t use the cryostats I’d design these to fit into, so I had to design two entire cameras. I kept most of the internals the same, but the outer part was completely new. Over the course of several months the rest of the camera design took shape.

Sitelle Camera September, 2012

We held a final design review, I got the go-ahead to start cutting metal, and a few months after that we had our first prototype camera.

Sitelle Camera Pumping

Early in the project we decided to build three cameras. The first was the prototype. The other two were the final cameras. This let work continue on the prototype as the final cameras were manufactured and finished. As each part of the prototype camera tested out ok, we made those parts for the final pair of cameras.

Sitelle Parts

Once we had complete sets of parts, they all went off to be anodized. Marc, the guy designing and building the electronics for the cameras, liked gold. So gold they became.

CFHT Sitelle Parts Anodized

Since it’s a porous surface, anodizing doesn’t work well in vacuum. So the insides of the camera bodies were machined clean and polished. The internal components were never anodized at all.

CFHT Sitelle Camera Assembly

Two internal components went through a number of revisions along the way: the cold strap and the getter – the cryosorb pump that maintains vacuum while the cameras are in use. We went through a number of designs for each of these. The first generation is shown above. We went through two more before we finalized the design for the getters, and almost that many cold strap designs. In addition to the box of unused getters and cold strap assemblies we now have in our clean room, we eventually wound up with two fully functioning cameras.

Two For Sitelle

We boxed the cameras up and shipped them off to Quebec for integration with the rest of the instrument. We had a number of issues that still needed to be addressed, the largest of which was that the hermetic space-rated connectors we used for the feed-throughs leaked once they were potted. We’re still dealing with this, though we’ve got what I hope is a real solution for the next generation of feed-throughs.

The cameras lived in Quebec for almost two years. A little over a month ago the team building Sitelle bid it farewell, loaded it into an airplane, and shipped it to Hawaii. After several weeks of unpacking, re-assembly, and testing, we finally put Sitelle on the telescope.

One Instrument - Two Cameras

And there at the end of each beam of the instrument is one of the cameras Marc and I built, beginning with that early design of the internals a little over four years ago. Designing and building the cameras was an exciting, if exhausting, process. But seeing them installed as part of a new instrument is indescribable.

Home At Last

As one might guess, I did a bunch of photography of the installation for first light. Toward the end, after balancing, after checkout, while we were testing the long instrument limits on the telescope, Marc and a number of other people involved in the project stood, watching the clearance between the instrument and the south bearing. I was struck how similar it was to the final scene of Ocean’s Eleven. I had to smile as I tripped the shutter. It seemed a fitting end to a wonderful caper.

Sitelle's Eleven

And the beginning of an even better one. Sitelle had a successful first light last night. Marc and the rest of the team are taking it through its commissioning tests, which will take several more nights. After that it will be unleashed on the sky, taking four million spectra at a time.

I can’t wait!

– Tom

Posted in Astronomy, Engineering | Leave a Comment »