« January 2010 | Main | March 2010 »

February 26, 2010

Perpetuum Mobile and the Ecstasy of Gold

For years, I would hear this one song in commercials and never knew what it was. But it would show up from time to time, usually somehow technology related. It sounds a little like Philip Glass. If you heard it you would recognize it right away. So tonight I was watching TV and a commercial came on for the restaurant Carrabbas and there was the music! I hoped it wasn't too new to Google and searched the internet, knowing there are sites where people identify music in commercials. The answer popped up in no time! And from the responses this post got, I wasn't the only one that had been looking for the name of this for a long time. Anyway, I'd never heard of the song or the group. The song is "Perpetuum Mobile" and it is by The Penguin Cafe Orchestra. You can see them playing it on YouTube. That led me to a YouTube video that makes a movie out 200 pictures that a pretty woman (girl? she can't be that old) has taken of herself each day, set to Perpetuum Mobile. Penguin Cafe Orchestra has some other videos here and here. They made this one using a dial tone (not my favorite).

That got me thinking about another song I really enjoyed lately. Back around Thanksgiving you might remember that I bought the Blu-ray DVD of The Good, The Bad, and the Ugly. The theme to that movie is really good, very famous all by itself, and I knew that as an Italian-made movie (that's why they are called Spaghetti Westerns) that the composer is Italian Ennio Morricone, who Clint Eastwood presented with an honorary Oscar in 2007 for all of his (not Oscar-winning) movie music (nominated 5 times, not for any movies Clint was in). But there is another song towards the end of the movie when Tuco (the Ugly) goes running through the cemetery looking for the grave Clint Eastwood (the Good) has told him has the gold. It's just a really good song and a great scene in the movie and part of why I realized this movie is really all about Tuco and a lot less about Clint Eastwood than I realized. The song is called L'Estasi Dell'oro in Italian, or The Ecstasy of Gold.

So I am watching the winter Olympics the last couple of weeks, entirely too much Olympics, and way more figure skating than I ever need to watch (though I only watch every four years). But I noticed that some of the songs sounded a little like The Ecstasy of Gold, but then would turn out to be something different. And I thought that would be a perfect figure skating song, and even better for the Olympics where everyone is after the Gold! So since I was looking around for music, I thought I would see if anyone else had made the connection and found a commercial on YouTube with American skater Sasha Cohen set to that music, although she has only recently restarted her career after winning silver in 2006 and is not in the Olympics. As I was looking this up, I also found a pretty good version by the heavy metal band, Metallica, if you're in to that sort of thing.

February 14, 2010

Wash Sale

I've known about wash sales for a while and what I always thought it meant was when you sell a stock at a loss and then buy it back within 30 days. The IRS sees this as just a way to harvest some losses without actually changing what you invest in, a move done solely for tax purposes. So what they do is say you have to defer that loss until you sell the shares permanently. But when I have shares bought at different times and then sell some of those shares, I would rather show a gain than hassle with deferred losses from a wash sale.

As the market was dropping in 2008 I would buy more shares of some mutual funds, like Janus Overseas. As the market continued to drop, I kept buying more. Eventually it bottomed out in March 2009 and soon started going up like crazy. Not sure the market would hold, I thought it would be good to sell the shares I had bought at the bottom and lock in a gain. This had worked out really well with some individual stocks, particularly Suntrust which had some crazy swings in the preceding year. Anyway, I bought my last batch of shares of Janus on March 2 and by March 27 those shares were up 20%, so I sold them. But the average cost of all of my shares on March 27 was still less than what I was selling them for, so it would count as a loss. When I got my 1099 from Janus, it said this was a wash sale. I checked to make sure I had not bought the shares back and I had not (in fact the market kept going up and I kept selling off more batches of shares). Why would Janus say this was a wash? So I researched wash sales further and found out that not only is it a wash when you buy back the same shares later, but even if you buy back the shares before you sell them! I guess the thinking is that a person might buy shares at a low point and then sell them at about the same price and declare a loss by identifying the earlier shares they owned. This never occurred to me. The annoying thing is the loss was only $77 total and now I have to distribute that $77 over my remaining shares (which isn't a big deal really because my spreadsheet I use for investments lets me override the average cost of that sale with whatever value I want).

Now knowing about this pre-wash sale business, I went back to see if I had any of those on Suntrust shares that I had been playing around with. And sure enough, I had some shares that I showed as a $600 loss when I had bought shares less than 30 days prior. The only way to avoid a wash sale was to say I was selling the shares I had just bought and show a $200 gain. Ouch. I was able to counter that somewhat by selling some of those expensive shares at a loss later in the year when wash rules wouldn't apply, but I couldn't wipe out all of the difference. Because I was showing a loss on the year and offsetting income, that means I would have to pay taxes on 25% of the difference, so it was over $100. And then, as I was writing this, it occurred to me that I might have something similar on shares other than Suntrust, and once again, the same thing had happened with some Microsoft shares. I had no idea. Then I went back to 2008 and noticed I had declared a few losses that really should have been wash sales too, but don't tell anybody.

In the end, it all works out. If I can't declare losses now, then it just increases the cost basis of the shares I have left meaning I will have smaller gains later on. So it's all kind of a game. But I was trying to always sell the most expensive shares in order to show a loss now and then hope that when I sell the other shares at least they will be long term gains taxed at 15% instead of 25%. I guess the advantage of a wash sale for the taxpayer is that I could defer the loss but owe nothing in taxes (since it doesn't count as a gain or loss), whereas if I show a gain, I definitely have to pay right now. I'd still rather not worry about it and just bite the bullet on the gain right now. Unlike with the Janus shares where they know whether I have a wash sale or not, my broker has no idea since they don't know which shares I'm identifying as being sold. So it easy to mess this up.

February 12, 2010

Snow Showers

Today we were forecast to get snow showers. I don't know how or if snow showers are any different than just getting snow. But eventually we did get some pretty good snow. Downtown we watched as it snowed more and more and even started to build up. At about 4:30 our boss came through and said he hadn't gotten the message that we had been dismissed at 3:00. In the city, the snow was nice, but not that impressive. But as I rode the train home I started seeing how it had built up in the trees. Here is a picture I took from my front yard looking across the street. It was really pretty.


Last Light of the Ox

Last Friday I ordered my first programmable flashlight, the Akoray K-106. This light has a lot of fans at Candlepower Forums and it was the first direct-shipped budget Chinese light I thought about getting back in November when I was thinking about getting some lithium ion batteries. You may recall that the first LED flashlight I got was a Fenix L2D, which could be converted to a L1D that takes 1 AA battery. That light has six modes but is not as bright as the L2D with its 2 AA batteries. However, if you up the ante further by inserting a single 3.6V lithium ion battery, you lose all of the lower modes and the light is stuck on High. For this reason, I wound up buying the budget Uniquefire S10, but I blew it out when I tried a lithium-ion battery. So I was still looking for a 1 AA light that could handle a lithium ion battery properly.

Anyway, I didn't get the Akoray for a few reasons. One was the light was advertised on DealExtreme as a 6-mode light, but people were getting a 3-mode programmable light. But then some people started getting actual 6-mode non programmable lights and couldn't really complain since that is what it was advertised to be. Another problem is that many of the lights people were getting had a very bluish tint to the beam, which I didn't want. And lastly, what blew people away about these lights was they were very, very bright on a lithium ion battery, but this was done by badly overdriving the LED. But it seems like some versions didn't overdrive the LED as much and weren't all that bright. There was no telling which light you would get, and at $20 that was a big risk to take.

But then some people said they were able to get a light on KaiDomain that advertised the 3 programmable modes and the light wasn't blue (there is also a similar PDC AK-16 which is a little fatter and is able to take a single CR123 battery or a single AA battery by using plastic rings in the top and bottom of the flashlight that the AA battery fits inside of, but that the wider and much shorter CR123 sits in between with springs filling the gaps at each end). So I thought maybe it was time to try it out. Also, since DX was selling people 6-mode lights, I feared that maybe supplies of the programmable light would be drying up. The advantage of the programmable light is that it has 3 modes but you can set the 3 modes to any brightness you want, including flashing modes and any strobe rate that you want. It would be nice to have one mode as the maximum brightness, then a low mode, and maybe a very low that could be used for reading close up in the dark.

However, I knew that Chinese New Year was approaching. It takes a few weeks to get a light from China under the best conditions, but China pretty much shuts down for the first couple of weeks of the new year. The date of the new year changes from year to year based on phases of the moon, so I made sure I looked up the date which is February 14. I also checked to see if they take time off before the holiday or after or both. They are supposed to take off after, so last week on the 5th, with a little more than a week to go, I ordered the light. It usually takes a few days for KD to actually get the light from their supplier, then a couple of days to package it. However, the next day KD posted a message on their home page saying they would be closing their shipping department from February 7 to 21. However, they must have worked hard to clear out their orders because on February 7 I got a notice that the light had shipped and they gave me a Hong Kong Post tracking number.

However, Hong Kong Post wasn't ackowledging the tracking number. And even though I had made it through KD barely, the package still had to make it through the post office and customs which would both be affected by the holiday as well. It wasn't until the 10th that HKP started tracking the light, which they said they had been "posted" on the 9th. On the 11th, they said the package was being processed for export. Then on the 12th they said they were processing it for shipping that day. However by this morning with the 12th essentially over in China, they were now saying that as of the 11th, the package was being processed for export. Soon after that however, the message was that it was being processed for shipment on the 13th. That is Saturday and the eve of the New Year. Now I wondered if it would really make it or if it would wait for the inauguration of the year of the tiger (probably a couple of week into the year, actually). But tonight, I was glad to see that they showed the light as shipped on the 13th. In the past I have gotten the light a week after it actually ships from China, so I hope that next weekend I will get one of the last flashlights to make it out of China in the year of the ox.

February 11, 2010

DX and KD reviews

Here is a list of product reviews I have written on DealExtreme and KaiDomain. They don't have a way of looking them up on their site (or editing them once you have written them, which isn't so great), unlike Amazon which KaiDomain bases its web design on.


AKOray K-106 3-mode, KD

Trustfire XP-E F23 1xAAA flashlight, DX sku.24291

Ultrafire WF-502B 1x18650 flashlight with XP-G R5 drop-in, DX sku.35074

Ultrafire WF-504B 1x18650 flashlight with XR-E R2 drop-in, DX sku.24200

Uniquefire S10 1xAA flashlight, KD ID 10303

Other things

18650 Charger, DX sku.6105

Universal NiMH/Li-ion Charger, DX sku.14885

NANJG 112A driver, KD ID 10243

February 5, 2010

Cheap 18650 Li-ion Charger

I bought a cheap charger from DealExtreme for my 18650-size lithium ion batteries. My other cheap charger will charge most 18650's (though some are probably too long to fit), but it uses a lower charge rate that would take forever to charge up an empty battery. For instance, the small charger uses a 300mA charge rate and an 18650 battery is 2400mAh, so it would take 8 hours to charge that battery at a constant 300mA. And my big flashlight uses two 18650's and the little charger can only charge one at a time.

So I found this other charger for $7.92 shipped that has two independent charging bays, supposedly charges at 1000mA, and stops charging at 4.20V like it is supposed to.


Here's a review:

This charger looks cheap and is cheap, but it works. The bays are long enough to charge even oversized 18650's. Mine are 67mm long and there are still a couple of more mm to spare. The cone-shaped springs are really weak so when you put a battery in, instead of getting pushed down, they just get pushed over.

The lights on the charger are bad, though you get used to them. For instance if you plug the charger in, the lights both light up green even without a battery. When you put a battery in, the light goes to red, indicating it is being charged. Once charging is complete, the lights will turn green. But if the battery somehow wasn't making contact, the light would also be green. Oddly, when the charger isn't plugged in and you insert a battery, both lights turn red. So if you weren't paying attention and put the batteries in the charger unplugged, the red lights might make you think you are charging the batteries, when, in fact, they are slowly being discharged by whatever voltage it takes to light up the lights. I'm not sure why it wouldn't have been set up so that the lights wouldn't be on unless a battery was inserted, like with any other charger.

The charger is smart in that it uses a higher charging rate on an empty battery and then lowers the charging rate as the battery gets full. The charge rate is supposed to be 1000mA, but because it varies depending on the voltage I measured a charging current of 850mA with a battery charged down to just below 3.6V and it varied to less than 100mA as the battery was being topped off. Here's a graph of time in minutes and the charging current and voltage.


This is a good thing, but this charger seems to take it to an extreme. I charged two batteries to 4.20V and let them sit. Two days later they measured 4.12V each. So I put them in the charger to top them off. One of the batteries took 30 minutes and the other took 50 minutes just to top off 0.08V. So this isn't real fast. However, the gentle rate at the end of the cycle means I can use it to safely top off even my 600mAh 10440 cells after the smaller charger cuts off at 4.10V.

Charging 18650 batteries from 3.6V took it about 6 hours. That isn't that bad except that you are supposed to keep an eye on li-ion batteries while they are charging in case they explode or catch on fire. So it shouldn't be used as an overnight charger or while you are out of the house.

More often than not, I measure 4.21V on charged batteries instead of 4.20V just as the light goes green, but that is probably within the error of my multimeter. After topping off, there may be a 10mA trickle charge for a little while, but I couldn't detect any current after 30 minutes.

The bottom line is that it is pretty cheap and charges li-ion batteries correctly. The fact that it has some quirks with the lights and takes a while is secondary.

February 2, 2010

Ultrafire MCU WF-1200L

My first impression is that it is smaller than I thought, but also heavier. This is a serious hunk of metal! One nice thing about getting it second hand from a CPF member is the guy I bought it from treated it very well. There aren't any marks on it that I could see. Plus he made sure to lubricate the o-rings and actually modified the tail switch so that you can halfway press it and the light comes on momentarily (all of my other lights have reverse clickies that you have to click all the way on and then let go before the light comes on; this light has a forward clicky). One of the reasons I got this light is that it had very good heat sinking and comes completely apart. The LED is mounted to a large brass pill that screws into the throat of the flashlight (threaded portion is 29mm diameter and the part that holds the driver is 23mm). The brass carries the heat from the LED directly into the body of the flashlight near its heaviest part. People talk about the light getting fairly hot after using it for a while, and that is a good thing. What would be much worse is the body staying cool while the LED cooks itself. Still, I left the light on for about 15 minutes and it only got warm. It would be pretty neat to be able to buy an extra pill and add my own 21mm driver and LED (maybe a warm P7 or a SST-50), almost like having a drop-in.


Battery: 2x18650
Switch: Reverse clicky
Modes: High-Low-Strobe
LED Type: SSC P7
Lens: glass
Tailstands: not really
Price Payed: $35 (used)
From: DealExtreme



  • Bright
  • Long runtime
  • Good heat sinking
  • Great build quality


  • No real low

Build Quality

My first impression is that it is smaller than I thought, but also heavier. This is a serious hunk of metal! One nice thing about getting it second hand from a CPF member is the guy I bought it from treated it very well. There aren't any marks on it that I could see. Plus he made sure to lubricate the o-rings and actually modified the tail switch so that you can halfway press it and the light comes on momentarily (all of my other lights have reverse clickies that you have to click all the way on and then let go before the light comes on; this light has a forward clicky). One of the reasons I got this light is that it had very good heat sinking and comes completely apart. The LED is mounted to a large brass pill that screws into the throat of the flashlight (threaded portion is 29mm diameter and the part that holds the driver is 23mm). The brass carries the heat from the LED directly into the body of the flashlight near its heaviest part. People talk about the light getting fairly hot after using it for a while, and that is a good thing. What would be much worse is the body staying cool while the LED cooks itself. Still, I left the light on for about 15 minutes and it only got warm. It would be pretty neat to be able to buy an extra pill and add my own 21mm driver and LED (maybe a warm P7 or a SST-50), almost like having a drop-in.

(All images in this review can be clicked to open a window with a picture twice the size shown here).

All of the pieces seem of high quality. The bezel seems like stainless and the reflector is aluminum (some lights have plastic reflectors). Reviewers on DX complain that the glass lens seems kind of thin, and I would agree with that. Because the lens is larger in area, it probably needs to be a little thicker as well. Another mm of thickness would be good. I was a little suprised that the body doesn't seem to screw off separtely from head. At the silver ring you would think there would be a threaded joint, but there doesn't seem to be (actually the pictures on DX showing it coming apart there, but I can't budge it and don't see any reason to force anything). Instead it is primarily three pieces: the tail, the tube with head, and the bezel. Actually, the tube has an extension in it so that it can take the two 18650's instead of 3 CR123A's. With a maximum of 18 volts allowable, the light should be able to use 4 16340's (reviews on DX say they won't fit; again I see no reason to try). I don't have any 16340's or enough CR123A's to do a test. So the extension tube adds an extra part. Still, nice o-rings everywhere. The modded tail switch has a rubber boot, but I don't know if it would be waterproof or not, though given the care the previous owner took with the light, I wouldn't be against it being fully dunkable. The tail piece itself feels like it might be made of plastic, though I don't know why it would be.

Here's the big brass pill next to a P60 drop-in:

A look into the throat where the pill screws in:

Battery Life

Some people on DX were claiming 1.5-2 hours, but on two gray 2400mAh protected Trustfire 18650's I only got one hour. Turning the light off to measure current and voltage every 15 minutes at the battery, I got 1.31A at 4.2V and pretty steadily to 1.51A at 3.58V when I stopped the test. Here are the results:

Time Amps (High) Amps (Low) Volts
0:00 1.31 0.44 4.21
0:15 1.37 0.45 3.97
0:30 1.42 0.47 3.82
0:45 1.49 0.48 3.69
1:00 1.51 0.49 3.58

Repeating the test with 2 gray protected Ultrafire 2400mAh 18650's that I got with the light, I got another 15 minutes, so the Trustfires maybe aren't so great:

Time Amps (High) Amps (Low) Volts
0:00 1.34 0.44 4.20
0:15 1.37 0.46 3.94
0:30 1.41 0.46 3.91
0:45 1.42 0.46 3.86
1:00 1.45 0.47 3.75
1:15 1.54 0.50 3.57

The light output seemed pretty constant the whole time. Based on Low being about a third of the draw of High, you could get 3 hours. At a 1.4A draw, 2400mAh batteries should yield an hour and 40 minutes, but I wasn't running the battery all the way down either. If the driver is supplying 2.8A, then it would be running pretty efficiently, but I doubt I'm getting that. While leaving it on for that time, most of the body became quite warm, but not too hot to hold. For the first 15 minutes it was only a little warm. Anyway, I'm a little disappointed in the runtime. I think with two batteries, it should last longer than an hour. But I am impressed that the brightness didn't seem to fade and the results seem to indicate it is well regulated, but with a buck converter that is a lot easier to do and should be more efficient as well.

Light Output

Turning it on, it didn't seem insanely bright inside the house. The sun was setting but it wasn't quite dark, so it wasn't all that impressive outdoors either. I did notice a slight greenish tint to the light. The sequence of modes is High, Low, and a killer Strobe. The mode memory seems to work fine. The momentary on of the modded forward clicky will change modes if you cycle quickly, but if you leave the light off for 1 second or more, you stay in the same mode. So no good for morse code, but fine for "tactical" use (whatever that is).

I waited a little longer for it to get dark and then took out my Ultrafire WF-504B that I just got last week to do a comparison. Now this isn't really fair. The 504B has a Cree XR-E R2 LED in it that can produce about 200 lumens. This light can produce 600 lumens (DX says 900 lumens, but I doubt it). The R2 has a 26.5mm diameter reflector head, this one has a 52.8mm one. The 504B has one 18650 battery, this one has two.

Here are some shots indoors of both beams, the 1200L on the right. The big reflector does a good job of concentrating light into a very tight hotspot. If the light is within a few inches of the wall, you can see a + in the middle of the P7 hotspot because of the separation of the 4 different LED dies. But at this distance of 50cm, you can see the + is gone. If the light is further than that from the wall, you get a small donut hole which persists at just about any distance. I tried tweaking the height of the pill and how much the reflector is screwed into the head, but couldn't get rid of the donut hole. It's really not that bad.

It isn't an enormous difference as I reduce the shutter time down to 1/100th and 1/1600th of a second.

Here is the 504b on medium and the 1200 on "low":

Outdoors in the dark, the differences really stand out. This light lights up everything around with some decent flood, but really has great throw. For reference, first is a control shot of the darkness:

Now a shot of the WF-504B on high with a Cree XR-E R2 LED. This is a good thrower so it does a decent job of lighting up the fence post 120' away and to the left of the upright forked stick 25' away.

But here is the WF-1200L and you can see it is lighting up areas past the fence that weren't even noticed before.

I had to turn it to low so the stick wouldn't burst into flames (just kidding). So here is another shot of the 1200L, but now on "low":

The Low isn't all that low and still almost as bright as the 504B on high and it still has a lot of throw. Really there should be the incredibly bright high at 600 lumens, a more reasonable Medium of about 200 lumens that would be like high on most lights, and then a true Low down around 20 lumens. It might even be worth having something around 80 lumens.

I don't have a light meter, so I can't measure the lumens. Nor do I have any experience with other 600-lumen lights. But somehow this doesn't quite look like 600 lumens to me. I know a lot of DX lights with P7's only get about 400 lumens or so. This is significantly brighter than the 200 lumens of the R2, but I was expecting a little more somehow.


I like having a bright light and it is nice that it has so much throw since some P7's are all flood. But, honestly, the light probably isn't going to be used that much because it is so big and heavy. I am impressed with the quality and weight of the light. I'm not sure what more money would buy from a more reputable manufacturer. A lot of DX lights need some tweaking when you get them, and because this light is second-hand by someone knowledgeable, that stuff has been done (or maybe wasn't needed).

The light retails for $73, which is still a bargain compared to non-budget brands with P7 or MC-E emitters. But there are 2x18650 lights like the Aurora for only $47 that look pretty similar. I feel like I got a really good deal getting a big discount on a well-maintained and probably lightly used light, but I would never pay $73. So based on the as-new price, I don't think I can give it 5 stars.