[dump]

Loved the zweihundert stunden film. Would love to fly one day.

 

[Toshi]

I've been absent from RM for several weeks. Life is busy. Working 7:30-6 most days, and both days on about every seventh weekend. Moreover, I'm actually busy during those working hours, as opposed to my last half year of quasi vacation.

The commute is just as challenging as anticipated. Steep hills are steep, yo: ~9% grade.

Oh, and I went on a trail ride today with a high school riding buddy and another rider. I was slow. Very slow. Didn't break anything, though, and it'll be better next time, especially now that I know this trail system is more 29er oriented than Turner oriented...

 

[stevew]

29er

my younger brother is a member of that cult.

and his land cruiser has been retired to snow duty only.

has a diesel mode of transportation now.

 

[Toshi]

My baby is cute. And now is six months old (unadjusted for being four weeks early):

https://plus.google.com/u/0/photos/115479414905422234350/albums/5905386098681067553

Matching Ben & Jerry's tie-dye goodness:

Work is still busy. Programming some ****, oddly enough: see http://faculty.washington.edu/tjclark/ to see a few Java projects I've been working on. Commute still is long. Been jogging with the wife weekly in an attempt to get generally less slow on foot and on the bike, and I hit up the same trail system on the Turner last weekend solo and was marginally less slow.

 

[Toshi]

Just came home with this.

 

[Toshi]

My e-bike has **** the bed: Rust on the rotor and stator thanks to crappy Chinese construction has caused the hub motor to nearly seize up. My attempt at a rustectomy was unsuccessful. From here on out I'll be using my U-PASS (bus pass) every day, I guess.

Sucks being stuck with the throngs on the bus, as the leg from downtown to West Seattle is perpetually packed, but it doesn't make sense for me to drop another few hundred $ on the e-bike, either, especially since I still hadn't figured out a pleasant way to cross downtown. For now I'll bide my time and will reassess once/if I get a job and know which city I'll be in, for starters…

 

[Toshi]

The Prius left our possession yesterday, rendering us a one-electric-car family… if you don't count my mother in law's Lexus RX that we could borrow anytime if we really needed it.

In other news, my first iOS application went live on the iTunes App Store today:

https://itunes.apple.com/us/app/li-rads/id687600020?ls=1&mt=8

Booyah. I may even make back my developer program and webhosting fees for the year if I'm lucky.

 

[Toshi]

This is why the e-bike is all borked. Blame rainy Long Island and, now, Seattle commuting + crappy design and construction.

I'm attempting to resurrect this hub motor by attacking the rust with CLR and, now, phosphoric acid. Initial appearances after this caustic treatment are much improved. We shall see whether I can actually put the thing back together properly, though… Time to learn how to press bearings in place!

 

[HAB]

The commute is just as challenging as anticipated. Steep hills are steep, yo: ~9% grade.

Where are you coming from/going to? I've been trying to ride in here and there, but 20+ miles with nothing that even vaguely resembles a road bike kind of sucks. It is pretty damn flat though, particularly for being in Seattle.

 

[Toshi]

Headed from West Seattle, Admiral District (thus the hill) to UWMC down on the water just south of the UW main campus. Crossing downtown and climbing the hill to West Seattle are the big issues.

 

[ALEXIS_DH]

This is why the e-bike is all borked. Blame rainy Long Island and, now, Seattle commuting + crappy design and construction.

I'm attempting to resurrect this hub motor by attacking the rust with CLR and, now, phosphoric acid. Initial appearances after this caustic treatment are much improved. We shall see whether I can actually put the thing back together properly, though… Time to learn how to press bearings in place!

there is an industrial-strenght loctite product i used to buy to remove rust from forklift chains.
i dont remember the name, but its a re-usable water-like blue fluid that comes in a translucent plastic jug (by the gallon).
just drop anything you want to de-rust overnight, and it will be squeaky clean the next day.
i think it costs about $30/gal.

never seen anything eat rust like that.

 

[Pesqueeb]

You need to get a mod to move this thread to PNW.

 

[Toshi]

Its location here dates back to undergrad, not this most recent 4 year long Long Island sojourn. Who knows where I'll be next year anyway… (albeit not likely in the Northeast!)

 

[Toshi]

Things already were looking better after attacking these parts with a wire brush and some CLR this evening, and now they're percolating with a thin coat of phosphoric acid overnight.

Today's update:

Post-phosphoric acid, pre-rigorous wire brush scrubbing

Post-scrubbing

State of the magnet world

Overview

I then popped out both bearings (using the gear puller and a large flathead + hammer as a pseudo-punch, respectively) and verified their sizes, as different production runs of this same hub apparently use different bearing widths, oddly. Now the parts are all sitting outside for another day or two with a fresh, additional coat of phosphoric acid on them, and I'm going to see about replacing the bearings.

 

[dan-o]

Being a hippy is hard work. Who'd a thunk?

 

[Westy]

You want to be careful and make sure none of your cleaning causes the plates the coils are wrapped around to short together. Motors and transformers use thin sheet metal plates insulated from each other. A more continuous mass of iron would allow large magnetic eddy currents to form. These currents sap power and generate heat.

 

[Toshi]

Are you talking about the outer plates on the rotor on this pic?

(click to embiggen)

I'm pretty sure that I spy little sticks of wood, a la matchsticks, separating each metal plate… At least now those gaps aren't bridged by rust and general crap that had accumulated in there.

 

[Westy]

Are you talking about the outer plates on the rotor on this pic?

(click to embiggen)

I'm pretty sure that I spy little sticks of wood, a la matchsticks, separating each metal plate… At least now those gaps aren't bridged by rust and general crap that had accumulated in there.

Yep, those plates. The magnetic field that the coil generates aligns the magnetic poles of the iron atoms creating a stronger magnetic field for a given electrical current. Ironless motors can be made much lighter but require significantly higher currents to generate the same amount of force.

Since the iron plates also pass through a magnetic field from the permanent magnet a sympathetic electrical current is generated inside the plates and therefore an opposing magnetic force. This creates a resistive force and generates heat. Breaking up the iron into smaller plates effectively reducing this resistive force while having little impact on the previously described increase in magnetic flux from the coils.

Are the matchsticks between the gaps where the coil is wound? If so they are probably used to wedge the coils in place. If the coils can move around they are likely to short out. The plates are insulated from one another by a thin layer of a lacquer type material.

 

[Toshi]

After a few more cycles of phosphoric acid + buffing + rotary wire brush-head thing the rotor is looking pretty snazzy:

The magnets on the stator still look like crap, in part because of the fine iron dust that's attracted to it. I may hit it up with a lint remover right before reassembly.

New bearings in the side covers, and over 100% more holes! (tiny ones, albeit)

Small and large speedholes. I'm not sure if the small ones are peripheral enough to really serve as drains. The large ones are meant to vent steam after water inevitably makes its way in there and is cooked by the hot internal bits.

 

[Toshi]

All parts now have a sloppy coating of red insulating varnish on them, so I reassembled the beast, including reconnecting the 5 phase wires and 3 power wires once the covers were back on… and it turns under electric power! Praise be to the FSM.

Still have quite some work to do: brakes are all sorts of off since I flipped the axle (intentionally), and there's a wobble that makes me think the bearings haven't completely seated on the axle. I'll bust out the hammer and pound on it a bit tomorrow, because why not?

 

[Toshi]

E-bike is revived. Sort of. Feels weaker than before… I shall try it out this upcoming week and see how it and I fare.

In other news, CHAdeMO rocks:

I drove to a distant trailhead and stopped off for a quick 17 minute boost to 80% state of charge at the dealer on my way back from the ride. Nothing quite like seeing over 100A at just under 400V on the charger's screen and hearing all the machinery behind the scenes whirring away madly.

While having twice the range (a la RAV4 EV or Model S) would also solve most day to day range issues I think quick charging will be a must-have feature if we buy or lease another LEAF-esque vehicle in future years.

 

[Toshi]

E-bike is revived. Sort of. Feels weaker than before… I shall try it out this upcoming week and see how it and I fare.

The e-bike took me to and from work* today without issues.

* To be precise, it took me about half the distance each way, with the balance made up by a diesel catamaran water taxi (because Seattle, duh) and a diesel-electric hybrid bus (because big hill, trek across industrial wasteland, and long day).

My own calibration must have been off when I wrote the above quoted post, or else something burnt or scraped off in the interim as it felt as strong as ever. I pedaled along in my usual gears on usual stretches of road. I had my usual set of antagonistic encounters with drivers, too, but that's not the bike's fault but rather a shortcoming of my route.

My only complaints are that the seat post I swapped in can't hold a given angle for the life of it, and that the chain tensioner I added makes a racket of its own. (The chain tension never has been good since I went from 36-18t to 48-16t gearing, iirc. Stupid not quite horizontal forward facing dropouts without screw tensioners...)



Example view snapped via my iPhone from said water taxi:

 

[Toshi]

http://www.ntsworks.com/

Latest from Neal Saiki, as in dude who founded Zero Motorcycles. I'd like to try one. It's a pedelec so no throttle, and the steering is via a funky linkage. 2x4 is a stupid name, though, as 2 of 2 axles are driven, not half of them as in the pickup namesake.

I'd also like to try an electric motorcycle, or I could just cave in in the future and commute in our LEAF or its equivalent. If only traffic and parking weren't so painful here in Seattle... Unless I land a job here for future years I probably I may get my wish.

 

[ALEXIS_DH]

toshi, what are your thoughts on the 2010 g650gs and the 2011 g650gs (apparently facelifted).

last saturday, i went to a bbq at bmw motorrad of lima, and they were showing me these bikes (2010 has 3000 miles, 2011 has 4000 miles) at relatively ok prices (50% off for the 2010, 40% off for the 2011).

are they worthy toys for 100-150mile day-trips in the dirt?
my experience is mostly with small (250-450) mx/enduro bikes that need engine rebuilds every 80 hours... so i´ve have no reference when it comes to cast-wheeled, once-a-year service machines

 

[Toshi]

Those were the single cylinder ones, right? Never rode one or paid much attention. Figured one would go with a DR or KLR if a cheap thumper was desired or a twin F 650 GS if one felt rich.

 

[Toshi]

For Westy, some fruits of my MTF-project labor in ye olde Matlab:

Swept sine and square waves x 4 simulated runs through a contrast enhanced MR scan, with each run varying in how quickly the contrast was administered

Subset of those four sine wave runs near where the MTF drops down, with [0,1] image data plotted as just that instead of mapping to [0,255] pixel intensities. 0 on this x axis corresponds to 0 on the above, and it extends to just over halfway over.

This one is like the above but with only the peaks minus troughs, as it were, so as to bring everything down, flat against the y-axis. 0 on this x-axis corresponds to about 1400 on the second chart, 1:1 after that. This is the more standard MTF curve as from the camera/lens literature. All I need to do now is to translate these x-axis dimensions back to the original sine waves, figure out the wavelength at that distance over (by measuring the handy square waves), and then translate that through our model into a physical spatial frequency.

The eagle eyed will note that the blips in the data happen every 512 pixels. This is because our simulated MR matrix size is 512 x etc. so I had to run things in 512 px chunks. When you have a half-sine-wave up against the edge of the matrix it gets modeled differently than an otherwise equivalent frequency full-sine-wave one wavelength over. I don't think these figures will ultimately be needed for publication, only the final 10% and 50% MTF spatial frequencies, so I'm not bothering to work around this at the moment.

 

[Toshi]

Mariko turned 8 months old yesterday:

https://plus.google.com/u/0/photos/115479414905422234350/albums/5928755481655161041

 

[Toshi]

Time to engage in some car musing since I have a spare minute and the balls on my multiple ongoing projects are in my collaborators' courts, as it were:

I think autobrake, as the IIHS puts it, is something I'd like to have on my next car. This would imply that my '06-07 Land Cruiser dream be shelved yet again, which is something that has been in the works mentally during this Seattle year. I haven't felt a pull for a Land Cruiser while on my commute, as what I need here is less traffic and better parking options, not a nicer, larger vehicle in which to drive. Even if I end up somewhere semi rural with little traffic buying a $30k 8 year old vehicle seems a bit daft.

Anyway, this autobrake thought comes from two sources. One is that my wife's grandparents, of all people, bought a 2014 Forester with EyeSight. The other is that the IIHS published tests on autobrake and forward collision avoidance systems this week, with their press release picked up and rehashed by autoblog, since that's what they do:

http://www.autoblog.com/2013/09/27/iihs-introduces-new-test-for-avoiding-crashes/

The IIHS blurb is interesting because Subaru's system did great, stopping the Legacy and Forester at both 12 and 25 mph. Cadillac's and Volvo's systems did ok (stopped at 12 mph but hit the target while moving 6-10 mph when starting at 25 mph), and others sucked. Non- or minimally-functional systems in the Infiniti JX, BMW 3 series, and Toyota Prius V were specifically called out, but as long as they even worked a tiny bit the vehicles can still be Top Safety Pick Plus eligible, which seems just a bit dishonest.

I'm still not enthused by Subaru but might come around in time, especially if they introduce, say, a less ugly Outback down the road. I am still open to something large, too, and will get to test my tolerance for the same soon. At the end of October I'm headed down to southern Oregon for a job interview, with a side trip to see my parents in Coos Bay. I will be driving a rental SUV, probably a Tahoe, since our LEAF clearly isn't making that journey in a timely manner. Multiple days of seat time in a Tahoe will give me some guidance as to whether I'd be comfortable long term in that size of vehicle. (I did have the parents' '99 Expedition in the period between my WRX's death and my purchase of the RX-8, but that wasn't a good sample since that vehicle blew goats dynamically and was ridiculously thirsty even for its era. I hit 16 mpg at 55 mph on a flat highway. Once.)

 

[stevew]

spotted in the tetons....same day as the LT28 camper i posted earlier....

california edition westy eurovan....diesel...

 

[Toshi]

Even if I end up somewhere semi rural with little traffic buying a $30k 8 year old vehicle seems a bit daft.

My wife took this idea and ran with it. Now the apparent plan, assuming I end up with a job outside of Seattle, is to pick up a 2008-2010 Toyota Sienna AWD for her sometime before June 2014, with the LEAF turning into my commute car come July or whenever I start. The LEAF is better suited for that role, anyway, especially as I have no intentions whatsoever of having a commute longer than its 84 mile EPA range - 20% buffer.

If we stay in Seattle by some chance then the plan will be to stay a one car household but move somewhere where my commute sucks less.

We could do much worse, especially given my neuroses about visibility and all. (Minivans have great visibility, if nothing else.) ((Thing 2 is not on the way yet, but may be by the end of the year.))

 

[Toshi]

I want an electric ultralight. 51 dB!

It's first up in the video.

 

[Toshi]

For Westy, some fruits of my MTF-project labor in ye olde Matlab:

Swept sine and square waves x 4 simulated runs through a contrast enhanced MR scan, with each run varying in how quickly the contrast was administered

Subset of those four sine wave runs near where the MTF drops down, with [0,1] image data plotted as just that instead of mapping to [0,255] pixel intensities. 0 on this x axis corresponds to 0 on the above, and it extends to just over halfway over.

This one is like the above but with only the peaks minus troughs, as it were, so as to bring everything down, flat against the y-axis. 0 on this x-axis corresponds to about 1400 on the second chart, 1:1 after that. This is the more standard MTF curve as from the camera/lens literature. All I need to do now is to translate these x-axis dimensions back to the original sine waves, figure out the wavelength at that distance over (by measuring the handy square waves), and then translate that through our model into a physical spatial frequency.

The eagle eyed will note that the blips in the data happen every 512 pixels. This is because our simulated MR matrix size is 512 x etc. so I had to run things in 512 px chunks. When you have a half-sine-wave up against the edge of the matrix it gets modeled differently than an otherwise equivalent frequency full-sine-wave one wavelength over. I don't think these figures will ultimately be needed for publication, only the final 10% and 50% MTF spatial frequencies, so I'm not bothering to work around this at the moment.

The new hotness: Now works in all 3 dimensions (we model a 512 x 512 x 100 voxel volume), the x axis is labeled with wavelength as measured in pixels/voxels, and the 512 px artifact is solved--I did bother to work around it, after all. Getting ready to write up and submit!

Now with 10 runs. Takes about 15 minutes to run the simulation for these 10 runs, with input files that now span 5012 pixels to get down to about 2.7 pixels per wavelength (against a theoretical maximum resolution/minimum wavelength of 2 pixels).

MTF. I don't know why that blip at around 5.1 pixels wavelength exists but my best guess is that it's something to do with the sampling/aliasing of the original input files.

Here's how the MR ultimately works.

Data is read in k-space, in the frequency domain. This is a line through k-space starting at the center (left of x-axis), where low spatial frequency data are read, heading out to the periphery, where high frequency data are read. y-axis is the signal resulting from contrast material in the bloodstream at that moment in time--k-space is read over time, not all at once, of course. The different signal intensity at the instant when different spatial frequency data are read is what leads to the different MTF curves.

Although the resemblance is already clear to the MTF curves, the real comparison would be to MTF normalized by the ideal run that we model (of sampling all of k-space perfectly, only limited by inherent MRI resolution). That would be what I illustrate below:

Those curves are dead ringers for each other, +/- noise, which validates my work. Always a nice feeling, that.

 

[Toshi]

In other news, I swear I wrote up a long-ass post about the latest car-thought for me (post-LEAF-lease, which I would inherit if Jessica gets a Sienna AWD): the 2015 Chevrolet Impala CNG. I either never hit submit or it was moderated... for causing extreme boredom, no doubt.

In any case, I wrote an abridged version for Google+, which I did post: https://plus.google.com/u/0/115479414905422234350/posts/TDSA2f8ijCU

Rationale, in part:

We're planning on getting a used Sienna AWD for my wife, so I wouldn't be constrained to an AWD or 4WD vehicle (see various and sundry prior posts by me about Land Cruisers), and this factory-sold and -supported bi-fuel CNG/gas option somehow tickles my fancy.

It's somewhat amusing to note the strict negative correlation between the number of cars-and-other-crap posts I make and the amount of real work I have to do on my ongoing projects.

For now I'm basically in a holding pattern on both of them, thus this above post.

I haven't started writing MTF abstracts for the above MRI physics work yet since my collaborators need to grok my code and confirm that what I do is correct. My other ongoing project (www.liradsapp.com -- buy my iOS app for 99 cents! or use the Java version) has its paper back on the reviewers' desks after I revised and resubmitted it a week or two ago.

 

[Toshi]

9 months today:

 

[Toshi]

 

[Toshi]

I interviewed in Ashland/Medford, so we naturally stopped by in Coos Bay on the way back for a day or three. My baby is cute when in Coos Bay, too, go figure:

 

[Toshi]

I got a "test drive a 2014 Silverado and get $50" offer in the mail, so I went and did just that yesterday. For comparison purposes I also recently spent five days and close to 1000 miles behind the wheel of a GMT900 vehicle so had a good internal reference.

CN: new Chevy platform is indeed quiet, just as the magazine reviews claim.

First the old, the GMT 900/930 Suburban, a 2013 Suburban LT 4WD with the old port injected 5.3L V8:

- logged a surprisingly good 19.5 mpg overall over several mountain passes, and almost 24 mpg in cylinder-deactivation friendly 60 mph-land
- turning radius isn't half bad given length
- backup camera is mandatory, IMO, and decent resolution albeit small (in mirror)
- switchgear feels flimsy and blue-green LED stereo interface hardens back to mid-90s
- not swayed a bit by wind gusts from passing semis
- road noise well damped and well suppressed
- moderate wind noise, and poor suppression of outside noise from other vehicles--thin glass?
- driving position felt good, and I became quickly used to the bulk

Onto the new, the GM K2XX Silverado, a 2014 Double Cab LTZ 4WD job with the new DI 5.3L V8:

- needlessly high cowl and huge center console
- off-center steering wheel didn't bug me at all
- stereo interface has great graphics, is responsive, but not entirely intuitive in that hard buttons are still necessary despite the touchscreen
- switchgear felt better, and glove friendly large controls are well executed
- backup camera resolution is sub-par, the effect exaggerated by juxtaposition with nice sharp touchscreen graphics
- double cab (think "extended cab" or Ram's Quad Cab and you're on the right track) has near-useless rear seats. Crew Cab or bust if you need rear seats
- wind, road, engine, and other-vehicle noise very well suppressed, making this a legitimately quiet vehicle, noticeably quieter than its predecessor
- driving position ok (fantastic headroom! I raised the seat a bit, even) but shorter/more raked windshield and domed hood limit forward visibility. Not a deal-killer but nonetheless a step back in this aspect

 

[Toshi]

For the record, I did opt to pay $1.50 extra to offset the Suburban rental's carbon emissions for the week. That figure is in line with carbon offset pricing for the full year per prior posts in this very thread.

 

[Toshi]

Today I test drove an Acura, as a local dealership was offering a $20 gift card and a raffle entry for a tv in exchange for one's time. As the gift card offer was specifically for the new RLX, a somewhat interesting car, I bit.

The car I drove was a 2014 RLX Advance Package, top of the line. It was the fwd model as the SH-AWD hybrid isn't out yet.

CNs are that it's too expensive.

Slightly more complete bullet point review:

Pros:

- great salesman--a car guy, unlike the Chevy salesmen from last week, one who was too aggro, one who was clueless
- great seat height adjustability up front combined with tilt/telescope wheel and decent glass area made for a very comfortable driving position
- lots of rear legroom
- nice engine note given that it's merely a V6, and it did feel 310 HP strong
- 30 mpg highway rating
- quiet
- lots of techy safety features: autobrake, lane keep assist, dynamic cruise, blind spot warning

Cons:

- way overpriced: $60-70k pre-TTL for the higher packages. A Krell package one on the floor was $60 plus another $7k in market adjustment/markup plus accessory crap. No freakin' way.
- no V8, especially at this price
- oddly missing features, namely no heated steering wheel, and no right-mirror camera as found on the new Accord
- center console and dashboard are still the usual cluttered Acura mess, with the two screen layout not really helping
- lane keep assist only worked for me maybe 20% of the times I tried it. Not reliable enough to be anything but a gimmick, IMO, at this point. Dynamic cruise worked reliably, at least
- usual complaint about my head being up against the headliner when in the rear seat

My price calibrator might be out of whack in today's market, but the RLX did not feel like a car worth $75k out the door. It felt nicer than the Genesis sedan, on par with Infiniti's used-to-be-called-M37, but subpar compared to the Lexus LS and even Hyundai Equus, both of which are in the same pricing ballpark.

 

[Toshi]

In other news, I swear I wrote up a long-ass post about the latest car-thought for me (post-LEAF-lease, which I would inherit if Jessica gets a Sienna AWD): the 2015 Chevrolet Impala CNG. I either never hit submit or it was moderated... for causing extreme boredom, no doubt.

Another few weeks, another set of plans to no one's surprise (including my own). Heeding Shared Skittles' advice, perhaps we don't need AWD on Jessica's future minivan. This would save us from replacing the Sienna AWD's horrid run-flats, not to mention that model's rougher and noisier ride. We could then look at the whole minivan field including the Nissan Quest that we test drove in NJ a year or three back.

This brings us back to my car, post-LEAF lease end. I have many things I'd like it to do:

- lack of ridiculousness in all scenarios is desired, including at work and at the trailhead with a hitch rack
- relative frugality in total cost of ownership, either through buying used or reasonable operating costs
- non-penalty box interior with both rear doors and a car-seat friendly rear seat
- intrigue me, and not in the sense of the Oldsmobile model of that name but rather by being good at something, whether crazy efficiency, off-road ability, alt fuel, or... performance?

Given these goals I was returning to my old saw of a series 100 Land Cruiser, or perhaps even a 2009ish Xterra Pro-4X/Off Road as my parents have along with an EV like the BMW i3. Both of those ideas have their merits but seem extravagant, the Land Cruiser for its size (when we'll have a minivan, again) and the Xterra/i3 idea because of the extra vehicle aspect as well as the i3's price tag.

Well, I had a different idea tonight, to come full circle, as it were: no, I'm not suggesting I get another Nissan Z (as my first car was a Z32), but rather another WRX. The 2015 WRX could be the ticket, bought new with the IMBA discount...

http://www.wrxforums.com/forums/5-subaru-wrx-news/45872-another-2015-subaru-wrx-photo-leaked.html#post477180
http://jalopnik.com/heres-the-2015-subaru-wrx-in-its-natural-element-1467695137

The new-Forester-esque glass area and dashboard layout look pretty decent, eh?

I'd buy it new, keep it more or less stock for reliability (in sharp contrast to my last, ill-fated WRX!), and would both commute and play in it (biking, solo skiing, potentially odd autocross or track day). I would deliberately not wait or opt for, once available, the STI version as I now acknowledge that a bone crushing ride and high NVH are undesirable traits.

I kind of like this idea. It may have legs yet, and seems more reasonable than my recurrent Land Cruiser and alt fuel dreaming.