Remembering Tim May

posted Dec 15, 2018, 5:03 PM by Doug Sinclair

I note with sadness the recent death of Tim May, reported here and here.  Before I was a spacecraft engineer I was a Cypherpunk, and Tim was at the core of it all.

During his time at Intel, he discovered the source of Single-Event Upsets (SEUs) in DRAM as alpha particles from natural radioactivity in the IC ceramic packages.  This contribution is deeply relevant to all of us in the space community today.

I will mostly remember him for taking me aside at an early Computer, Freedom and Privacy conference to tell me that my jacket with a dozen pockets stuffed with tech, which young me thought was desperately 1337 and tactical, looked kinda goofy.  I don't think I ever wore it again.  In certain circles, Tim was the arbiter of cool.

A busy week

posted Dec 3, 2018, 12:08 PM by Doug Sinclair

Most of our satellite launches for this year have happened in the past week.  On November 29, Blacksky Global-1 was launched on a PSLV.  On December 3, Blacksky Global-2, SkySat-14 and -15, Landmapper BC-4 and Hawk A were all launched on a Falcon 9.

All together, this represents 9 star trackers and 22 reaction wheels on orbit.

Ruby Lasers

posted Oct 18, 2018, 2:54 PM by Doug Sinclair

Sharing this picture, because they are just so pretty.  These are flight-model ruby laser assemblies for the Darkstar laser downlink terminal.  They are not the main communications lasers (nor technically are they ruby lasers), but are low-power metrology lasers used to measure the attitude of the internal fine-steering mirror for realtime control.  The ruby operates simultaneously as an optical lens and as a mechanical pivot point.

Between these, the diamonds in the reaction wheel bearings, and the gold that coats pretty much every surface, this lab feels like a game of Splendor.

Vacuum pump free to a good home

posted Sep 7, 2018, 4:52 AM by Doug Sinclair   [ updated Oct 18, 2018, 2:43 PM ]

Welch 1399 vacuum pump now on the free to a good home page.  Every Sinclair Interplanetary sun sensor and torque rod was potted under vacuum from this pump.  Come take a piece of history!

Edit: The pump is off to a new happy home!

MOST space telescope turns 15, and a secret is told

posted Jun 30, 2018, 6:18 PM by Doug Sinclair

Today is the 15th anniversary of the launch of the MOST space telescope.  This little mission was ground breaking for both its science and its technology.  Scientifically, it has observed oscillations in stars in a way that no other instrument could.  Technically, it demonstrated sub-arcsecond pointing stability in a microsatellite platform for the first time.  And after all this time it is still working and producing useful science.

I built a rather unusual power system for this spacecraft.  It needed 12 switch-mode converters: 10 boosts for the solar array strings, and 2 Cuks for the shunts that bleed off excess power when the battery was full.  I couldn't afford the area and power consumption of 12 discrete controller ICs.  Instead, I built a 12-channel 4-phase controller out of a 4000-series ripple counter, 74-series CMOS flip-flops and LM339 quad comparators.  It is a byzantine gated-oscillator (not PWM!) system, but it runs in a bounded limit cycle and the total chip count is less than 12.

In the late 90's, when all of this was being designed, there was a global shortage of surface-mount tantalum capacitors.  They were all snapped up by manufacturers of cell phones and Internet routers, and a small project like ours just could not buy them.  So what was I to do for bulk capacitance in the power supply?  Aluminum electrolytics were too big, and unsuitable for vacuum.  Ceramic capacitors were my only option.

But I wasn't the only one thinking this.  Very quickly the good military-grade stacked ceramics also disappeared from inventories.  It was even difficult to find high-capacity parts in wide temperature range X7R dielectric.  I was forced to some very low performance dielectrics -- Y5V and Z5U.  These lose capacitance rapidly when hot or cold, so to compensate I needed a lot of physical volume.  I used massive surface-mount chips the size of thumbnails.

These low-grade parts have terrible piezoelectric properties.  As the voltage on the capacitors goes up and down they get physically larger and smaller.  The gated oscillator carrier frequency is 450 kHz, but the converter limit cycles could be down at 10 kHz or so and are a strong function of the battery voltage and load current.  Each capacitor acts like a little speaker, and the combined effect of 12 simultaneous channels of noise meant that everybody in my work area had to wear earplugs when I was testing.

Now MOST is fifteen.  The NiCd batteries lost all of their capacity long ago, but its orbit keeps it in sunlight ten months of each year and the shunts continue to keep the bus voltage regulated.  And I take comfort that the little spacecraft continues to scream outloud with the pure joy of being in space.

Rainbow Rotors

posted Jun 5, 2018, 2:31 PM by Doug Sinclair

We are now making RW3-0.06 wheels in large volume, on slightly different CNC lathes than before.  The new rotors have a beautiful rainbow iridescence to them.  I believe this is because the tooling marks are now spaced by a distance comparable to a wavelength of visible light, causing them to act as a diffraction grating.

High Water Mark

posted May 14, 2018, 6:04 AM by Doug Sinclair

In July of last year, we noted that the order book was momentarily empty.  On January 15 we announced that the order book has closed for flight models for 2018.  Even so, there has been some expansion as outstanding proposals have been accepted and contracts signed.

Today the order book sits at its highest point ever: exactly $9.99M spread over 171 deliverable units.  We have hardware that should ship out to three customers by the end of the month, and I hope we can bring that number back down significantly.

Unsolicited advise for others who may want to start their own business in this field: be prepared for ebbs and flows.  The shot noise in constellation contracts can kill you -- zero is too few, two can be too many, and the various constellation customers are uncorrelated with each other.  But relying on a single customer puts you at the mercy of program slips.  Keep a lot of liquidity on hand, as you may need to ramp up towards a big piece of work that will not pay out for a year.

A Decade of CanX-2

posted May 4, 2018, 12:54 PM by Doug Sinclair

An important milestone passed quietly last week.  On April 28, the CanX-2 satellite marked 10 years on-orbit.  Onboard is the first RW-0.03 reaction wheel, still running nominally.  CanX-2 was the first 3U Cubesat to demonstrate high-performance 3-axis pointing of an optical payload.

All of our 129 reaction wheels currently on-orbit ultimately trace their heritage to the CanX-2 wheel.  We have scaled the design larger and smaller, as different missions have needed, but we have never strayed far from the proven mechanical design that just keeps working.

Alouette Award

posted Apr 11, 2018, 2:47 PM by Doug Sinclair

We were pleased to learn today that the CanX-4/CanX-5 team, including Sinclair Interplanetary, has been selected for the 2018 CASI Alouette Award.  More details can be found in the press release here.

3 mNms Wheel Now Available

posted Jan 19, 2018, 10:41 AM by Doug Sinclair   [ updated Jan 19, 2018, 10:43 AM ]

Those of you who came to the Utah Smallsat Conference in 2017 may have seen an early engineering model of the 3 mNms reaction wheel.  We finally have literature on the website (here).

This is a tiny reaction wheel, but built with the same quality and design philosophy that underlies our larger parts.  We continue to make all of our motors from scratch, carefully controlling our bearing geometry and materials.

A small number of flight-model wheels are currently in inventory (left over from 2017 production), available for immediate sale.

1-10 of 71