Book Review: Goodbye to Berlin

Goodbye to Berlin
Isherwood, Christopher
Vintage, 1998 ed, of 1939 original

Per the author’s introduction, this collection of loosely connected sketches featuring a first person narrative from a character sharing his name, his background, and many of his experiences is fiction and was intended to be something larger than it came out to be. It is a convenient, and perhaps accurate, approach to classify the gray area between autobiography, fiction, and reflection. It also makes it possible to write the stories an author anticipates for themselves but which may never have had a factual base.

The beauty of this book lies not so much in this format, nor it’s direct timeframe, but in the detailed and generally factual description of life as an English writer in Berlin as the Nazis came to power. With contacts ranging from a wealthy Jewish family where he works as a private english tutor, through the English ex-pat community with all it’s variations, the matron of his middle-class boarding house through to the working class family whose son invites him to move in with them when his finances take a turn for the worse, the fictional Mr. Isherwood sees a broad swath of humanity in one of the leading cities of Europe just as it’s light is about to be extinguished. From neighborhood dives filled with pick-pockets and prostitutes winding down their day to high end nightclubs, tenements to mansions, rural holidays to urban centers, the book slices a cross section of the society just as it is about to change in a way that would make these scenes unrecognizable for decades.

The book ends as the author leaves to return to England, having decided that it is too dangerous for him to stay in Germany any longer. The timeframe is approximately 1933, and while war is still a few years away the daily decline has already begun.

Posted in Uncategorized | Tagged , , , , , , , , | Leave a comment

Book Review: Black Wings Has My Angel

Black Wings Has My Angel
Chaze, Elliott
New York Review Books, 2016 ed of 1953 original

Basically a pulp fiction novella with a bit more depth, this is a good period piece of mid-1950’s America. The story revolves around an escaped convict’s obsession for carrying out the perfect heist which he and one of the other prisoners developed while in prison. His partner was killed in the escape, but after a fling with a hotel prostitute in a backwater town leads to an unlikely affair he finds in her a perfect and willing accomplice. The plan takes time to develop, and as they travel the country in search of the perfect place the tone shifts from pulp crime to road novel to social observations of time and place.


Eventually, however, the book shifts back away from descriptions of how neighbors water their lawns or the dangers of being on the backside of office politics in a pre-OSHA factory, and the long planned heist takes place, but despite the apparent success of their plan the couple find themselves caught in the uncertainty of their relationship with one another and the various worlds they have created around themselves. The book ends as any good 1950’s story must – but it leaves a reader pondering how it would have ended had it been written in a later period.

Posted in Book Review | Tagged , , , , , , , , | Leave a comment

Through the Camera: The Flower Eater

I was wondering why my purple lantana had noticeably less flowers on it than the yellow one next to it… And then I caught the chuckwalla in the act of selectively munching down the purple flowers.

Posted in garden, Through the camera | Tagged , , , , , | Leave a comment

The Busy Skies

Several weeks ago I purchased a device called a SDR Dongle to try and catch the signals coming from my wireless temperature transmitters and import them into a computer application    . I quickly found an easier means of doing that job, but by that point I was hooked on the possibilities opened by having the device. SDR stands for “Software Defined Radio” and it essentially functions as a broadband receiver that decodes / encodes radio signals into something a computer can work with. Combined with some impressive (and free) software it can be a very powerful device for identifying and understanding what is happening in the airwaves around you.

Well and good, but it probably would have ended up as an occasionally used tool had I not discovered that the frequency on which ADS-B messages from aircraft are transmitted also happens to be in the SDR reception band. Beyond that, several of the popular flight tracking services are actively looking for ADS-B receiver stations in order to increase their coverage and accuracy, so finding the necessary information and software was very easy. I signed up with flightradar24 and, since I didn’t want to leave my desktop on all the time to run it and I already have a Raspberry Pi running continuously for the swamp cooler control, I used their setup package for the Pi. A handful of mouse clicks and copy and paste lines later I was streaming ADS-B reports to the flightradar24 servers, as well as had the data available for my own use.

Based on some recommendations I opted to install the Virtual Radar Server on my desktop and link it to the datastream coming out of the Pi. This software takes the decoded ADS-B messages and portrays it in a much more easily understood format – a table of the message content as well as a real-time updated map for the aircraft which are sending position reports. It also uses the airplane registration to automatically go out and obtain public record information such as the year it was built, ownership information, and pictures of the airplane. In addition to this, it can also be configured to create a database of the recorded information.

ADS-B is a line of sight transmission system, and being in a valley with mountains around me in nearly all directions I wasn’t expecting to see much of the traffic around me since mountains are in the way. That generally is the case, but I also happen to be under the main eastbound departure track from LAX so there actually is a fair bit of commercial traffic overhead, and depending on timing I sometimes get messages from westbound departures as they transit through a window between mountain peaks. I’ve set up the virtual radar server to launch when I power up the desktop, so even though it’s generally only recording for a few hours each evening, in roughly a month of operation of the receiver I now have a reasonably large database of what’s gone by overhead and decided to take a look at it today.

I was fairly surprised by the results. There were 5,576 unique airplanes which operated 17,980 distinct flights. Of those, 2,165 airplanes were only recorded for one flight, meaning 3,411 airplanes were picked up multiple times. 3 of the top 5 repeat aircraft were helicopters, which isn’t overly surprising given that the local airport is the base for 2 law enforcement helicopters and there is a forest service firefighting helicopter also based in the area, and it’s been a busy season for them so far. What I found most surprising was that the second most frequent airplane was a Spirit A321, which had racked up 41 distinct flights on the database. In terms of commercial aircraft, there were 1,101 unique Airbus aircraft vs. 2,236 unique Boeing aircraft

Some charts..

Leading manufacturers, airframe count:

Leading types, airframe count:

Leading foreign registrations, airframe count:

Leading types, by flights:

Leading operators, by flights:

Posted in Uncategorized | Tagged , , , , , , , , , , , , , , , , , , , , , , | Leave a comment

Shop Projects: Building a better swamp cooler control

When I moved into the house the swamp cooler had a very basic control. It was a dial switch with several of the pre-printed markings scraped off and left with the settings “Off” “Fan” and “Cool”. That was pretty much how it operated – set the switch to off and nothing happened, on fan it just turned the fan on at high speed, and on cool it turned the fan on at high speed as well as ran the water pump. Better than nothing, but not very good.
As I got to know more about the system I wondered why the low speed settings didn’t work given that all the documentation I had showed I should have a high and low speed setting. On investigation I discovered that for some reason the system had been installed with only 4 instead of 5 wires, and the installer decided that they could dispense with hooking up the low speed side of the fan motor. I wanted the low speed option which meant I’d need to open things up and run another wire, and since I was already going to be going into the system and would need to replace the dial switch for one with the low speed markings I decided I might as well upgrade the controller.
My plan was to get a programmable thermostat like those in common use for air conditioners, but I quickly found that unlike most air conditioners, swamp coolers generally use a line voltage controller, they have different numbers of outputs, and the selection is pretty basic. There are a few low voltage options on the market, but they were well off the cost / performance curve of line voltage models and my system was already set up for line voltage. I was unable to find any line voltage systems, so I opted for a basic Dial brand 7619A swamp cooler thermostat. Since I wanted a programmable time function, I also picked up a 1 Hp motor capacity programmable timer. After running the additional wire up to the swamp cooler, I connected the timer in series with the power input to the thermostat and had what I thought I wanted.

In general the system worked as designed. At the times I had set, the timer turned on or off the power to the swamp cooler control. When the power was on, the thermostat determined whether to turn on the cooler based on the temperature it sensed. So far so good, but ….

There were a few shortcomings still to be addressed. First off, the thermostat was designed for constant power to it and had no setpoint memory – so whenever the power to it turned off (which the timer did several times a day) it next powered up to the default setpoint of 77 degrees F. Second, the thermostat relied on manual settings for the fan speed and water configuration. Third, there was no accommodation for humidity settings, so on one abnormally humid day I came home to find it happily adding yet more water to the already damp air.

For many users these would be trivial issues. For the way I use my swamp cooler though they were close to deal breakers and I wasn’t at all satisfied with the setup. Where I live in the desert it is not unusual for night lows to be in the 50’s while daytime highs are over 100, and my house is fairly well insulated. On a typical day I like to keep the system off until the house starts to heat up in the early afternoon, then run it through the afternoon and evening. Once the outside temperature drops below the inside temperature I’ll shut the water off and use the fan to cold-soak the house overnight. The house also has an air conditioner, but I have it set at a relatively high setpoint so that it only comes on if we have an extremely hot day or on the few days a year when it is either too humid for the swamp cooler to be effective or outdoor air quality is particularly poor.

I had been thinking for a while about how to improve the situation, but it took a mistake on my part while overriding the timer one day to turn that into action. On that day we were under an excessive heat advisory (and the high at my house that day ended up being 116) so I decided rather than using the timer to keep the system off all day while I was at work it was probably worth letting it run on the thermostat. I remembered to hit the override button on the timer, but I forgot to set the water selector switch on the swamp cooler thermostat to “on”. I came home expecting a relatively cool house, and instead I walked in and found the swamp cooler running, the air conditioner running, and the inside temperature at 105.

It was time to act. I bought a Raspberry Pi computer, a couple of temperature / humidity sensors, some suitably sized relays, and a box to put them in. After a few hours spent learning the basics of the Python programming language, I hooked up the sensors and relays (but did not connect them to the swamp cooler), launched the code, and let it run for a few days in the background to confirm that it was operating as I expected. Following a couple of minor tweaks, I took the plunge and connected them to the cooler. Since I anticipated I would want to be able to override the control at times, I also included an isolation switch to allow going back to the existing timer / thermostat system. The basic setup is as below:

As the system currently stands, when operating under the Pi control the system looks at indoor and outdoor temperature and humidity and the time of day. Based on the values and differences between inside and outside temperature and humidity as well as the rates of change, it determines when to run the fan at high or low speed and if water is needed to meet the desired setpoint for the conditions. During the portion of the day when I am typically not at home the system is generally off, but will turn on if the house becomes unusually warm. Likewise, if the temperature continues to go up it will turn off just below the air conditioner setpoint to avoid having the situation where it is sucking hot outside air in while the air conditioner is running. Overnight, when it is colder outside than inside, it uses a reduced setpoint to coldsoak the house.

I was happily surprised by the capabilities and ease of setup and configuration of this Raspberry Pi based system. For a total cost of under $100 and a few hours of time I have a very capable adaptive swamp cooler controller with the flexibility to change and update as new situations arise.

Posted in Shop projects | Tagged , , , , , , , , , , | Leave a comment

Swamp Cooler 101

I’ve received a few questions about the process by which a swamp cooler takes hot air and water and creates cooler air rather than warmer water. The key process involved is evaporation, and a way to quantify this is through the enthalpy of vaporization, which describes how much energy it takes to turn a specific amount of a liquid into a vapor at a constant temperature. For water, this is roughly 970 BTU / lb. Since a BTU is defined as the amount of energy needed to raise 1 lb of water by 1 degree Fahrenheit, this means that evaporation takes 970 times more energy than raising the temperature of a given amount of water by 1 degree F.

In a swamp cooler, hot and relatively dry air brings thermal energy into the system. For reasons associated with the discipline of heat and mass transfer (and encouraged by the design of the swamp cooler), the primary energy transfer involved as the hot air meets the thin film of water on the cooling pad is evaporation. During this process, 970 BTU’s of energy are transferred out of the hot air to evaporate every pound of water used by the swamp cooler. This results in cooler, moister air as the energy used to evaporate the water lowers the air temperature and the evaporated water increases the humidity of the air.

It may be easier to see this as an example. Suppose it’s a hot and dry day at sea level with an outside temperature of 110 degrees F at 10% relative humidity. Using a psychrometric chart identifies the following:
• Specific enthalpy of dry air: 33 BTU / lb dry air
• Specific volume of dry air: 14.55 cu ft / lb dry air
• Wet bulb temperature: 68 degrees F
• Dewpoint: 41 degrees F
• Specific humidity: 0.0055 lb water / lb dry air

For the sake of illustration, we’ll consider a steady state analysis and ignore transient effects either outside or inside.
A swamp cooler works along lines of constant enthalpy. Since the wet bulb temperature equates to a relative humidity of 100% at the same enthalpy as the starting point, that identifies the lowest a swamp cooler can possibly cool the air to. For most cases 100% relative humidity would not be desirable, so taking a more comfortable (particularly in the desert) indoor relative humidity target of 55% gives a temperature of 80 degrees. The specific humidity at that condition is 0.012 lb water / lb dry air, so 0.0065 pounds of water would need to be added to every pound of air to cool to that condition. To put that in perspective a 2,500 square foot house holds about 20,000 cubic feet of air, or a bit under 1400 pounds of air. That would require about 9 pounds of water, or a touch over a US gallon.

Compare that with an air conditioner. Whereas on a psychrometric chart the swamp cooler moves along lines of constant enthalpy (increasing humidity while decreasing temperature at constant energy), an air conditioner moves along lines of constant moisture content while changing energy to decrease temperature. For the same case of cooling to 80 degrees, an air conditioned space only increases the relative humidity to 25% and decreases the specific enthalpy to approximately 25 BTU / lb dry air. That requires 8 BTU / lb dry air energy removal, which for the example 2,500 square foot house would need 11,200 BTU of heat removal. Air conditioner cycle efficiency in the US is denoted by the SEER / EER system (cooling BTU / Overall power input (W)) and in the desert environment the EER is the more applicable value. The minimum EER currently allowed for new construction in this area is 12.2, so 11,200 / 12.2 = 918 Watts are required.

In the real world this example analysis is only a starting point. Outside temperatures and humidities are constantly changing. Houses naturally heat up in the daytime and cool at night. There are blowers and ducting to consider. A swamp cooler needs to exhaust the air displaced by that which it pulls in, so unless the system is completely optimized a fair bit of the cool air gets blown outside before it has exhausted it’s capability for cooling. Air conditioner coils can become less effective due to blockages and debris buildup. Activities in the house increase the amount of cooling needed, ….. Despite this, it is helpful to have an understanding of the physical processes involved in even an ideal case.


Posted in Uncategorized | Tagged , , , | Leave a comment

Shop Projects: Cleaning the Swamp Cooler Spray Bar

As the summer heated up I began to notice that my swamp cooler wasn’t cooling as well as I recalled it doing last year.  Everything was in good shape mechanically and electrically and water seemed to be flowing when commanded, so I struggled to come up with a reason for the change.  I finally used an IR camera to try and see if there was anything amiss, and that highlighted that there was a significant temperature gradient across the pad. The only part of the system I hadn’t checked was the spray bar, and if it were clogged than it could cause that type of gradient.

Sure enough, once I disassembled the unit and got the spray bar in hand it was easy to compare the top and bottom holes in the bar and see that it was clogged for the lower half.  My cooler has the bar installed at a slight angle to help it drain, and over time the lower holes had accumulated scale and debris.

A nail and a hammer turned out to be the best tool to remove the debris from the clogged holes, and once I had finished clearing them out I decided to take off the end fitting and empty out the small plugs of debris that had been pushed into the pipe while cleaning.  What I found was beyond my expectations – nearly the entire tube was packed with flakes of scale, probably from the inside walls of the tube itself.  Aft flushing out the tube and reassembling everything the swamp cooler now works better than it ever has since I moved in.

Posted in Shop projects | Tagged , , | Leave a comment