The decline of RSS

Slowly over time, RSS has declined in utility. Not for technical reasons, but miserable market driven reasons.

Google is canceling Google Reader – so there goes my sync clearing house. I’ve also noticed this happening over time:


Dilbert readers – Please visit to read this feature. Due to changes with our feeds, we are now making this RSS feed a link to

The feed providers are moving useful content OUT of the feeds. This pretty much kills the aggregator reading model. Sigh.

WebKit Crash. WTF.

Well this amuses me – were it not for the fact that it is keeping me from getting work done, it would be even more funny. Safari just vanishes in a puff of magic SEGV smoke. I need some more of that StringHash, WTF? [Posted from Chrome]

The stack trace:

Thread 0 Crashed:: Dispatch queue:
0 0x00007fff8d72f611 ExternalCookieStorage::combineWithSessionCookies(CookieSifter*, CFArrayEnumerator<__CompactCookieArray const*>*) + 189
1 0x00007fff8d72f246 ExternalCookieStorage::newCookieArrayEnumeratorLocked(CookieSifter*) + 414
2 0x00007fff8d727e7c HTTPCookieStorage::copyCookiesMatching(__CFDictionary const*, unsigned long, __CFArray const*) + 172
3 0x00007fff8d12006d -[NSHTTPCookieStorage cookies] + 42
4 0x00007fff8ff8c7cd WebCore::getHostnamesWithCookies(WTF::HashSet<WTF::String, WTF::StringHash, WTF::HashTraits<WTF::String> >&) + 77

That last line again?

WebCore::getHostnamesWithCookies( WTF::String, WTF::StringHash, WTF::HashTraits…)


N-i-i-ce. Symbols matter. People see this stuff. (Ok, not many people).

Starcom1 Digital to Yaesu FTM-10SR Radio Interface

I recently decided to finally install my FTM-10SR radio on the Multistrada 1200. The only catch is that I wanted to integrate it with the Starcom1 Digital intercom system I had purchased.

[more notes to come]

I had quite a bit of trouble getting this working and it isn’t finished just yet. I still have to tune the trimmers for best audio and finish moving the antennas to their final home, but in the interest of starting to document all this – here is the schematic for the build. The intercom manufacturer indicated that the radio was completely incompatible with the intercom – which I refused to believe but I did have to spend quite a bit of time figuring out how PTT was signaled from the intercom to the radio. (reference single-line PTT pages here -todo).


To add / write:

  • Interface cables and pinouts
  • Pointers to single line PTT reference pages
  • Pointer to autocom interface circuit built by others
  • Pointer to manhattan construction technique
  • Photos of build
  • T2-301 installation notes
  • Photos of installation (ugly!) and the antenna mounting — not so good yet.

PWM Controller Hacking

Enough people have asked about it so I figured I would write up my PWM controller that I’ve hacked together. Ultimately I’d like to make it ATtiny powered and I might have made a mistake by getting the analog version working – as it works well enough and has removed the pressing need to build the digital version.

I managed to source everything at Radio Shack except the MOSFET. I picked that up at a local, terribly overpriced component supply store. I also purchased the pot there, simply because I liked the physical size of it – RS had a reasonable replacement. In choosing your N-channel MOSFET, make sure you select one that is fully ‘on’ at the level generated by the 555 circuit (not a problem as the signal is around 12V peak-to-peak) and has a very low Rds(on). If you are planning to use an ATtiny for the controller (instead of the 555) then you need to select a MOSFET with an appropriately low threshold voltage, such as the STP36NF06L.

The schematic is a modification of another circuit I found online at There is an alternate PWM 555 circuit (similar) at, but I did not like the lack of EMF/flyback protection in that design – effectively they are very similar circuits.


I modified the schematic slightly – as we are not controlling a large inductive load and I installed a much smaller capacitor for C3, just to take the edge off the PWM wave-form.

LED1 is optional (along with R1) – and in my case I used a Radio Shack 12V ‘LED indicator’ (see photos). If you wanted to do this yourself, select R1 to match LED1 at a 14V supply voltage. I recommend R1/LED1 as it permits a quick glance in the mirror to confirm that the controller is active and the power level setting. D4 is simply providing a path for any collapsing field energy, but I am not sure there is much in a resistive load like a heated vest. I used a very small cap for C3 – simply because I didn’t have a 470uF cap lying around and I was impatient. Seems to be working fine for me.

The NTE2395 data sheet  mentions that Rds(on) is 0.028 Ω so it doesn’t dissipate much power as heat – an important consideration.  [ Todo - power dissipation calc. ]

Power enters via the right-hand pair of contacts (Vcc/GND) and the vest / grips plug into the upper right contacts (Vcc/PWM_Return). The design is switching the power return circuit – effectively modulating the ground connection.

At these voltages and power levels, I have no concerns around that approach – and it a common technique (vs high-side switching).

Here is the parts list for the build, all caps at 20V rating and all resistors are 1/4W:

Part Value RS Part No
C1 .1 272-135
C2 .01
C3 470uF  272-1030
D2 1N4148
D3 1N4148
D4 1N4001
IC1 NE555P
LED1  See text
Q1 NTE2395 N/A
R1 1K (see text)  271-1321
R2 47
R3 1K  271-1321
R4 50K


The schematic was built up on Radio Shack breadboard styled protoboard, with point-to-point wiring. Nothing to write home about, in some cases, I put insulation on leads and used the components to assist with the hook-up. (See obvious diode D2/D3 as examples). The assembly was housed in a small Radio Shack plastic enclosure.

Here’s a shot of the nearly assembled unit:

Note the generous overhand knot and ties for strain relief where the power cables exit the case.

The Radio Shack indicator LED is in the lower left, beneath the potentiometer. I haven’t assembled anything in years, so it is a little messy in there.

The connectors are a 2$ towing extension cable from the local auto parts store – SAE connectors on either end happen to fit the vest I’m controlling. A single piece of red heat shrink tubing on one of the connectors is used to mark the ‘hot’ side (vs the vest side). I haven’t thought it through, but it looks like hooking it up backwards would not do any damage and just leave it inoperative until swapped around.

Here’s a top view for scale:

If I was doing this again, I’d consider using a multi position rotary switch with discrete resistor values instead of the pot. It would then be a case of clicking a setting and knowing it would stay set. I would also consider a bypass switch that forced the unit into bypass – assuring 100% power transfer. This might not really be needed, but it would make me feel better. To work around the ease with which the knob turns, I plan to jam a large o-ring under the dial to drag on the case. This would also add weatherproofing. The entire unit fits snugly under a waist adjustment buckle on my riding suit so is held in place with the blue velco in seen in the photo. A small pass-through in the suit pants pocket allows this to live outside and control the heated vest on the inside. I highly recommend the Aerostich Kanetsu electric vest.

Overall it is ugly, but it works. And even at Radio Shack prices, it was far cheaper than a commercial product – not to mention satisfying to put together.

Amazing Video and Sound – SRB POV STS, etc.

Wow. Love the ethereal and ambience in this video. It is full of amazing footage that is at once beautiful and inspiring – as an aerospace buff.

Amazing footage and sound put together by people usually dealing with Science Fiction. I just love the footage of the orbiter peeling away after SRB separation. Amazing in that this is real, not CG and not Sci Fi. Sad in that it is footage from a program we don’t have anymore.

I also recall reading Richard Feynman’s discussion (in the context of the Challenger incident) of dynamic overshoot at SME start and how the SRB ignition is timed such that the orbiter-stack is properly aligned for the release. This effect is very very clearly shown after main engine start and you can see the stack rock away from vertical, back once again and then liftoff, moments after SRB ignition. Wild stuff!

New kind of flying machine

The Japanese Ministry of Defense has come up with an amazingly cool looking flying machine. Story available on

It’s basically a hovering, flying sphere.


Courtesy of YouTube