Silicon Valley Redneck

Aw, c'mon...

Once again, the Mythbusters build team is just being goofy.

The clown-car-full-of-balloons thing?

The lab setup is completely non-analogous to the crash scenario.

And, as I watch this, they've discovered that neither easily-popped balloons nor extra-sturdy balloons have the desired effect, so now they're back in the lab trying to get the lowest possible peak accelerometer reading when they release a weighted plunger atop a cylinder of balloons.

Er, no.  Not unless you suddenly have the floor rise two feet.

In the setup they've got?  Lowest peak acceleration would be obtained with a solid cylinder of concrete.  The plunger wouldn't move at all, giving a peak reading of 1G exactly.

This would not work well if the floor suddenly rose two feet... or if, analogously, the whole system were moving rapidly and then came to a stop within two feet of travel.

What you're looking for is something intermediate, which will ease the deceleration such that the clown (a) doesn't just go right through and crash into the dashboard, but (b) also isn't so firmly restrained as to experience the full deceleration of the passenger compartment (which, admittedly, ought to be substantially less than that experienced by the front bumper).

THIMK!

El Reg has a piece about James Bond's new pistol.

He's back to carrying the signature* PPK.

Only it's been high-techified with a biometric lock.  Kind of like the camera gun from a few movies back, only now the fancy fingerprint recognizer is built into a mostly-normal-looking pistol.

And this high-tech super-spook feature is useful because?

Let's see, now.  You're a secret agent operating behind enemy lines.  You get into a confrontation with enemy agents, and they get ahold of your gun.  Which is your bigger worry:

• The enemy agents shoot you with your own gun, or
• The enemy agents figure out who you are?

The way I figure it, any enemy worthy of the attentions of a secret agent probably already has guns of his own, and so do all of his henchmen.  So: do you carry a standard-production beater gun that any thug could have bought on the black market, or do you carry a brand-new high-tech custom piece that you could only have gotten from a major government?

Think carefully.

* Signature: so much so that, as we learn in at least one of the movies, villains can identify him by it, even if he's regenerated again and has a totally new face.  Because, y'know, no one else in the whole wide world carries a Walther PPK.

Finally dusting off that Aard-wino thingy

Got a little project that involves some fairly trivial hardware interfacing and firmware, perhaps leading to more challenging things later.  Arduinos (or whatever the plural is) are involved.

Well... in pursuit of a project that never quite materialized a couple of years back, I actually have in my toybox an Arduino Uno, that having been the cheapest generic-ish model on offer at the moment I decided I needed one.

I vaguely recall rumors of Issues with the Uno.

So: plug it in, run the GUI thingy, load up the blinky example.  Works.  Load up the analog-to-serial example.  Wedgie!  Errors on /dev/ttyACM0.  GUI becomes horribly sluggish.  Serial monitor won't launch.  Regular terminal program can't open the port.

The board wasn't totally bricked; if I played with the RESET button long enough while unplugging and replugging the USB cable, I could at least get it into a state where the GUI would run and let me load a program that didn't try to use the serial port.

A bit of Googlage led me to this fix for the problem.

So now I have a properly-working Arduino, and at least some handle on rudimentary programming.  Guess I should probably buy a Book, but for immediate purposes I can probably get by with just mashing up a couple of the example programs and tossing in a few lines of original code.

Speaking of physics...

Scenario: your crash-test dummy is more or less upright on a surfboard that has a rocket attached to it.

Observation: when you fire the rocket, the board+rocket+dummy system does a back-flip.

Which of the following must be true?

1. The rocket is too powerful.
2. The thrust vector is below the center of mass.

Right... so, for your next test, you:

1. Use a cluster of little rockets with lower peak thrust.
2. Replace the dummy with a set of flat weights strapped directly to the board.

Right: both.  So of course it doesn't do a back-flip, but the test no longer bears any relation to the problem at hand.

And that's meant to be a science program....

Update: oh, this just gets better.  They go and call the rocket-surfer story "busted" on the basis that their one live test run didn't quite achieve the results claimed in the story.  Uh, guys?  Once again: just because you didn't do it in a week, and a single live test, doesn't mean someone else hasn't done it ever.

Physics department? Really?

OK, so I saw the story on Slashdot this morning, thought it silly, and just now got around to clicking through to TFA.

So: a great innovation in wirelessly charging electric cars involves using grid power to spin a magnet in the garage floor, which spins a magnet in the car, which runs the generator that charges the battery.

Unwrapping this, we have:

Grid power to floor stator creates rotating magnetic field.

Rotating magnetic field acts on rotor to turn a shaft.

Shaft rotates floor magnet to create a rotating magnetic field.

Rotating magnetic field acts on car magnet to turn a shaft.

Shaft turns generator rotor to create a rotating magnetic field.

Rotating magnetic field interacts with car stator to produce electricity to charge the battery.

Um.

The way I see it, everything between the first and last rotating magnetic fields is extraneous, and adds weight and failure points.

Also, with that nice big on-board magnet, your car will pick up any ferrous debris you happen to drive over.  (Which is maybe OK if your other car is a Harley, and you can use the spare parts.)

And: frequency 100 times lower than what's used now?  I assume this is compared to competing inductive charging schemes and not, say, 60Hz grid power.

And this comes from a physics department?

The things you spot during bring-up...

...when you didn't have a design review!

Started stuffing & testing the purple board that arrived yesterday.

Power section first, of course.

Now I note that the 5V regulator doesn't have a ground fill area under it, let alone vias to the ground plane.

Also, while I'm at it, I observe that some of the other ground connections are stupidly autorouted, optimizing for minimal via count instead of shorter traces.

Oh, well.  Should work anyway, though some parts may be seeing more noise than they should.

3.3V and 1.8V came up just fine.

5V didn't come up at all, until I replaced the regulator chip.  I'm guessing the original had gotten damaged during the multiple iterations of reflow soldering (both of the little DFN-packaged switchers floated out of place the first time around, and needed to be nudged along with the solder melted for a second or third time).

Anyway, all the power stuff is working now, and the lack of proper heat-sinking under the 5V regulator just means I won't be able to do as much as planned with the 5V output from the board - everything on the board should be fine (it's all low-power stuff).

Next up: install the rest of the passives (mainly bypass caps) and the MCU, and see if I can talk to the MCU with the emulator pod.  Then the sensors and incidentals, and it'll be (from a hardware perspective) all ready for show'n'tell.

Update: in among other stuff, I found time to install the MCU.  Hooked up the debug pod, and thence the SWD from the Discovery board. Wouldn't talk to the SWD.  SWDIO was pathological... like it had a huge capacitive load on it.  Hmmm?  Ah: too much solder paste on the debug connector; it had wicked up and shorted SWDIO to, indeed, a power line (fortunately not one that was actually in use, but one with considerable capacitance present).

Since unshorting vertical FFC connectors never seems to work for me, it was time to pry the housing off, unsolder all the contacts, clean up the pads, apply a thin layer of paste, and install a fresh connector.  After that, the debugger came up.

Alas, when I try to run one of the test programs for the Discovery board, it chokes partway through initialization.  Um.  The Discovery board appears to have an 8 MHz crystal on it; my board has a 16 MHz MEMS oscillator.  Obviously, the difference between quartz and MEMS technology is to blame.  Or, y'know, trying to run a 168 MHz processor at 336 MHz, assuming the demo code is trying to set the critter for full speed.

Ah, well.  The day is officially over, and the board is at least showing signs of life.  Between now and Halloween, I want to get some sort of code running and the sensors installed (and maybe even tested in some minimal sense).

Update 1a: a little more tracing reveals that, just after the demo program sets up the PLL for high speed and selects it as the clock source, the whole thing goes BLATZ!

Sounds like I need to figure out which register needs dividing by 2, and which #define corresponds to that.

At least there appears to be a good excuse for the malfunction.

Update 1b: found the two, count 'em, two places that "8" needed to be changed to "16" - one in a header file,  and the other in a nominally-immutable source file.  Change those, rebuild, debug: the debugger is now happy and seems to be behaving itself.  So, a working CPU core, at least.  (Also working flash, and RAM, and so forth.)

Last of the summer gadget boards

That board I ordered from OSH Park just over two weeks ago?

Here we go. Nice and shiny, if incompletely depanelized (a few seconds with pliers took care of the rails).

Continuity check shows the inner layers connecting the things they're suposed to connect.

Finish, like Kryten's final words, is ENIG.

I was going to spend the afternoon stuffing first the power section and then the processor, but a busness lead just came over the transom, so I may be chasing near-term revenue instead.

Ding!

That dinky radio module whereof I posted a picture a few days back?

Just sent the letter Z to it.

Better still, it received the letter Z, all wrapped up in 802.15.4 formatting.

I spent much of the afternoon bashing my head against this.  I was sending from an XBee Pro, at a distance of a few inches (no antennas connected on either end).  Send a character to the XBee, read the FIFO on the AT86RF231 (for such is the radio chip on my board; that and some Skynetworks front-end thingy); get zero length if nothing had been sent yet, or a reasonable-looking length if I'd sent a character... and consistent packet contents... but the contents were consistently wrong.

All of which pointed to the actual RF portion working, but the protocol being somehow FUBAR.  Almost like there was a scrambler somewhere, or the XBee was using a proprietary protocol, or they weren't speaking the same Baud rate, or....

Hey.

Last test was Continuous Transmit with Spreading.

In high-speed data mode.

Folks, when you're trying to get an AT86RF231 and an XBee to talk to each other: don't have the '231 in high-speed mode.  You will not like the results.

With TRX_CTRL_2 now set properly (i.e.: zero), the packet gets through in what appears to be a reasonable format, with intelligible address fields.  I can even find the character I typed.

Still not getting the RX interrupt, and I'm not sure about various other status... but it is currently in promiscuous mode, and may not be well behaved.  Maybe for my next trick I should try setting an actual ID, and sending addressed packets to it.

Update: that lack of interrupt status?  My initialization table still had some magic settings from the transmitter test mode.  Take those out, leave the chip at mainly default settings, and the status and the interrupt show up as expected.

Still need to do a receiver sensitivity test, which means scaring up a string of attenuators, which I hope I can borrow (though at some point I really ought to have my own set, and the ones from Mini-Circuits don't cost all that much).

Update 2: haven't rounded up a string of teenyfiers yet, but hooking up an antenna to each end (with a notable lack of proper RF grounds, mind you), and putting the XBee Pro in the next room (with much metallic clutter in between) and at minimum output power (+10dBm), I'm seeing Energy Detect of 48 (nominally meaning 48 dB above the minimum useful receive power) on my board with the LNA enabled, and 31 with it disabled.

Which says that the LNA is actually performing a useful function, implying that my board doesn't have a big noise problem.  Which, given that there's a 72 MHz processor less than an inch from the LNA, wasn't exactly a given.

Looks like that mixed-signal layout lesson I had from an RF guru back in the mid-90s is still useful.

Still TBD, how much signal loss is being caused by my layout and choice of component values.  Though, given that TX and RX share a common path between chips, and the TX test gave pretty much the expected output power, the RF path can't be too lossy.

Update 2a: sanity check.  Not quickly finding my old cardboard* path-loss calculator, I look for one on-line: so.  Sez 2.45 GHz and 0.005km should have a loss of 54dB, unrealistically assuming no signal-blocking clutter.  With LNA off, ED=31 should mean I'm getting roughly -60dBm at the PHY, which means I'm suboptimal by roughly 15dB.  LNA on, ED=48, so -43dBm at the PHY; LNA gain is supposed to be about 12.5dB, so call it -55dBm at the antenna terminal.  Think clutter, antenna placement/alignment, and lack of grounding could account for a 10dB discrepancy?

Oh, and background RSSI is basically zero, so the radio isn't picking up lots of interference from either itself or the WiFi.

*Not cardboard-and-string.  Sorry, Bluebottle.

Update 2b: hmmm.  The antenna I'm using on the transmit end is quite a bit longer than the one on the Gadget.  It's one of a pair I bought somewhere-or-other, quite some time ago.  Has no markings on it.  Does have an RPSMA connector, but... I wonder if it's meant for 915 MHz.

The only other definitely-2.45 GHz antennae I have handy fit into one of two categories: (a) not an RPSMA connector, but something else for which I have no adapter, or (b) currently connected to my WiFi router, which is currently in use.

Update 2c: I think the big antenna must be Spark Fun #WRL-00558, which is meant to be a 2.45 GHz item.  Also, I had a second one, and substituting that for the dinky one that presumably came off an old WiFi router picks up 2dB.

Update 2d: borrowed a bunch of SMA attenuators and a good SMA cable.  Installed U.FL-to-SMA cables on both UUT and XBee Pro.  Modified the UUT firmware to flash a green light on receipt of a packet with valid CRC, red for invalid CRC, and black for no packet.  Problem: there's enough leakage on both ends that, with each U.FL-to-SMA cable terminated and no approved signal path, I still get consistent green flashes about 4-5 feet away.

So, I guess, I need to change the code to drive external LEDs, and stick the UUT in a metal box with nothing protruding but the business end of the SMA connector and the tips of a pair of added LEDs - maybe I can scare up LEDs with long little round noses that can poke through tiny holes in the box.

Update 3: Did as planned above, except that I could only get one of the external LEDs working - I figure the fiddly little DFN-packaged filter/ESD clamp, for which my footprint was suboptimal, probably still has Soldering Issues.

Anyway, one blinky LED is at least somewhat informative, and with the radio in the box. the lid sealed, and an attenuator on the connector to avoid a dangling impedance mismatch, it was non-blinky with no cable connected.

Stack up attenuators until XBee transmit power 0 (+10dBm) gives no (or, anyway, few) happy blinkies on my gadget.  That adds up to... 3x20 + 4x10 + 5x6 = 130dB, so I can't cope with a -120dBm signal.  No surprise there; -113 would be optimistic.

Set XBee output power to 1 (+12dBm).  Blinky mostly comes back.

So... -118dBm?  I'm not quite sure I believe that.  Still: there's a pretty good suggestion that I can get some serious range with this puppy, without even resorting to high-gain antennae.

BTW: there's no metal can on this board.  Everything hanging out in the open, as shown here.  Well, apart from the whole thing being shielded for test purposes, but that just locks the MCU in a reflecting noise cage with the radio.

The principle, I have found it!

My brain's Rant Center has been so frothy of late that nothing coherent actually comes out, but something just did, in response to overhearing, a few days ago, someone going on, vehemently and at length, about the Citizens United ruling, or at least the version of it that had been presented to him by one of the left-wing lobbying corporations.

Yes, you see, all the noise about that ruling - "How horrible!  How can corporations possibly have legal rights?  They'll be able to spend their infinite money* to influence elections!" -is being spread by politically active organizations that are themselves corporations.

So... what's the difference between the nasty, evil, plutocratic corporations that pollute the political system with their vile propaganda and the good, public-spirited corporations that elevate public discourse by raising awareness of vital issues?

I finally figured it out.

The bad ones are for-profit corporations that pay taxes.

The good ones are nonprofit corporations that are tax-exempt.

Hence the new rallying cry:

No representation with taxation!

* Every corporation has infinite money.  When it files its corporate charter, it automatically receives a special dish to put on the roof of its corporate headquarters so the Ben-Bernank's Money Satellite can beam unlimited funds directly into its bank account.

Contrast, and the Old Boys' Club

As the momentum shifts from Zaphod Hussein "Somewhere in the vicinity of Betelgeuse, a planet is missing its idiot" Beeblebrox to Mittens "Crazy rat guy" Romney, and Zaphod amps up the class-warfare rhetoric while continue to shower future taxpayers' money on super-rich supporters, and Mittens shows no public sign of actually understanding how this is a bad thing....

I ponder the totally-over-the-top, foaming-at-the-mouth political rants that I've been too upset to type up and post lately, regarding crony corporatism, tax farming, and the sale of the working class into serfdom, and I think:

It would be really easy to word one of these rants such that the reader would have great difficulty in determining whether the perspective was Libertarian or Communist.

Because, while Libertarianism and Communism are almost perfectly opposite, the system we have now is, somehow, almost perfectly opposite to both of them.

(Well, in principle, anyway.  What we have now looks remarkably similar to Communism-as-practiced.)