Improve your photos by using professional framing choices

The two photos below are of the same subject. The first photo was taken by an amateur photographer, making rookie mistakes. He decided to use a vertical layout to capture the pool, but this chopped off the near corners. He does not include the sky or a visible horizon, and the background is an uninteresting wall of green. The center of the photo falls in the water, with no interesting details.

IMG_20160507_103348

Below is the same subject photographed by a professional photographer. Notice how the landscape orientation she has chosen complements the vertical layout of the pool. By moving to the other end of the pool she includes sky and an implied horizon. Notice also how she has placed the vanishing point at the splash of color inside the pool shed to act as a focus point. Simply by changing the focus point, orientation, and camera location the subject is made to sparkle!12 1293 Berkeley Pool

Sexist Google Image Search filters

I was sitting next to my wife who was goggling for images of haircuts. (I have to admit,   I’ve never performed this type of search before tonight….)   She showed me a few of her search results, and I noticed that she had some super cool search filtering options across the top of her Google image search results that I had never seen before. (I had thought I knew all the tricks of how to use Google image search…..I was wrong.)

As it turns out, I just hadn’t been using the right search terms, like “for women”. Here is an image of what I’m talking about.

sexistGoogle

 

That’s right, when you search for “short wavy haircuts for women” you can filter by “over 40”, “over 50”, or face shape!   But why can’t I filter images of men by age or face shape when searching for   “short wavy haircuts for men”?

For that matter, how come I don’t see any men when searching for “short wavy haircuts”… Continue reading

Sink Drain Sizes

Repeat after me: 1-1/2″ for kitchen sinks, 1-1/4″ for bathroom sinks. Or, you could construct an adapter like this….

IMG_20160702_095225

Follow along with me from left to right:   we start with a 1-1/2″ to 1-1/4″ drain and trap connector, which is required to connect to the steel drain pipe. But then things go downhill, as we have a 1-1/4″ to 1-1/2″ adapter, a 1-1/2″ (kitchen sink) P-Trap, followed by a 1 1/2″ to 1 1/4″ adapter to connect to the bathroom sink drain.

I’d like to think that somebody decided they really wanted a larger P-Trap on the master bath sink…but I suspect that the plumber only had a kitchen sink P-Trap in his truck and didn’t want to make a special trip

Since I needed to change the length I went ahead and used 1-1/4″ throughout. I think it looks a bit nicer. (I also sealed the hole into the wall with expanding foam while I was under the sink.)

IMG_20160704_085417

Current imbalance with dual chargers – TSM 2500 / EVCC

AvsBchargers

This graph shows the temperature (in C) of my two chargers, as well as the watts of power they report providing over time (in minutes) as they charge the Nissan Leaf battery pack in my truck.

For the first 23 minutes of the charging session, both chargers are putting out 12.5 amps (at around 129.5 volts) or around 1600 watts. The 12.5 amps is limited by my MaxC (maximum current) setting of 25 amps total by the Thunderstruck motors EVCC (Electric Vehicle Charge Controller). So in the first 23 minutes, the amount of power drawn is limited by the chargers.

From the 23 minute mark until the end of charging, the battery bank begins to accept less than 25 total amps (because the cells are getting charged, and the pack voltage is getting closer to the MaxV (maximum voltage) that the chargers are putting out.

So for example, at minute 32, charger A is still supplying 12.5 amps (1600 watts), but charger B has dropped down to supplying around 4 amps (500 watts).   (The difference in power output also has an effect on the temperature of each charger, shown in the yellow and blue lines…) I have been having overheating issues causing my chargers to shut down, hence the detailed logging….

My hypothesis is that the power difference is due to one charger providing a slightly higher actual voltage (e.g. it’s voltage sensor reads just slightly less than the other charger) so although they both claim to be limiting voltage to 129.5 volts, one of them is actually putting out a slightly higher voltage than the other, which causes more current to flow from the “hot” charger to the battery bank.

This graph is evidence that the ThunderStruck motors EVCC does not do “active” balancing of the chargers. I suspect that the EVCC probably just sets the “maximum” amps and volts for both chargers and then turns them loose.

The chargers report back the amps/voltage/watts they are providing, so the EVCC could in theory dynamically change the maximum current settings on the “hot” charger to only provide an amp or so more than the “cold” charger, more evenly balancing the load. However, this would add complexity and require extra code and testing. And as you limited the current provided by the “hot” charger the “cold” charger would start providing the extra current, and over time the identity of the charger providing more current would switch back and forth.

This active balancing would only be needed (or be useful) when the battery pack was drawing less than the maximum current capacity of your charger array (near the end of the charging session, in the 80-100% SOC range).

The limited potential benefits to this added complexity could be:

  1. Your chargers (and wires) would presumably stay closer to each other in temperature (only really an issue if you are suffering from overheating, which is why I was logging this data in the first place).
  2. If your chargers are more efficient at a lower current, it could save you some energy / heat. (Conversely, if your chargers are most efficient at full current, it may be better to run one charger at full current and just use the other charger to “fill in” any extra current needs.)

In short, I can understand why the ThunderStruck motors EVCC does not perform active balancing (at least with the version of the firmware I’m using — v2.3.1), but I would certainly be willing to test a current balancing feature in future firmware releases.

 

TSM2500 (CH4100) Chargers Overheating

Twin Chargers in Box

Now that it is summer, and outside temperatures are reaching 26-35 C (80-95 F), my dual TSM2500 (Rebranded CH4100) chargers are overheating. After about an hour charging at full power, they reach around 74 C (165 F) and shut down. The ThunderStruck Motors EVCC records this as a “normal” end charging event (because the Amperage output goes to zero), and for some reason it triggers a ground fault on my EVSE (perhaps they have a thermal switch that shorts the charger to ground to shut it down, or maybe my JuiceBox Pro 40 is just overly sensitive?)

I guess the overheating is to be expected, as the chargers are in a five sided box (with only the top open) and mounted to a piece of (thermally insulating) plywood. Although there is a tangle of wires in front of them, the wires really don’t interfere with the airflow as much as it looks like from this top view.

In my defense, the charger’s manual (v. 1.05) specified that I should leave a 50mm (2in) gap in front of the charger for proper ventilation and I left around 8 inches. It also noted that the “Working temperature” for the chargers was -25 to 55 C (-13 to 131 F). It didn’t mention anything about thermally bonding the charger to a heatsync.

As a temporary solution, I have re-configured my 80% charging profile to only run at 1.2 kW (8 amps total, or 4 amps per charger on a 128-131 volt pack). This is about 25% of the 15 amp max power that the chargers are capable of in cold weather. At this relatively low power, each charger is outputting just over 500 watts, and even in 32 C (90 F)   weather the charger temperature hold steady at 50 C (122 F).

Charging at one kW may not sound terribly fast (it’s not), but this workaround is actually fine for 95% of my charging needs, as I rarely need to refill more than 8-10 kWh (20-30 miles) per day of use, and L1 charging overnight works fine for most of my needs.

However, I purchased the dual charger setup so that if I was necessity charging away from home I could charge at a 4 kW rate, so I want to make improvements to my cooling so that I can run the chargers at full power (without them overheating after an hour) if needed.

ThunderStruck Motors suggested that I mount the chargers to an aluminum heatsync, which is a good idea, but difficult and costly to implement.

I have decided my first order of business is to drill two 4″ air intake holes into the bottom of my charging enclosure and duct them to the top of the chargers right over the fan using dryer hose. This will allow the fans to draw cool(er) outside air directly over the vanes on the charger, and keep the heated exhaust air from mixing with the cool(er) incoming air. Since the top of the box is open, the heated output air should have no problems escaping, as convection will assist the fans in exhausting the hot air upwards.   If adding intake air vents doesn’t solve my problem, then I’ll worry about making an alunimum heatsync plate to take the place of the plywood.

DC Fast Charging (DCFC) Pricing models

Ben complained that he “got the short end of the stick” because EVGO charged him   $10.51 for a single fast charging “session” where the electricity cost him about $2 per kWh.

But the simple fact is, if you have an EVSE at home,   95-99% of your charging is done at home. Remote L2 and DCFC sessions fall into one of two categories:

  1. Opportunity charging – Just picking up some extra power because you are at a location anyways.
  2. Necessity charging – You need to go somewhere that is farther than your current range will allow, either due to a small battery pack, or a low state of charge.

When you are opportunity charging, you don’t want to pay much more than you would at home. In some cases, stores offer free charging to get you in the door, or as a public service, and in other cases the costs ($1-2 an hour, or $0.20 to 0.50 a kWh depending upon your charging speed) are reasonable and/or in line with parking fees that you would already be paying.

In Ben’s case, he was definitely in the necessity charging category (making a drive to downtown Milwaukee right at the very edge of his range).   So the $10 he paid allowed him to go somewhere that he wouldn’t normally be able to get to on a single charge. This is a convenience fee, not a “fuel” fee. Also, the fact that it was a “fast charge” session should not be overlooked, as that is much more convenient than a L2 charge for necessity charging (when you are not planning on spending time at a location for reasons other than charging).

I recently spent $3 to charge for an hour and 20 minutes outside of a Walgreens at a L2 charger (in 85+ degree heat), getting around 5 kWh   (allowing me to get home without seriously depleting my pack). In addition to the $3 I paid for the charging, I spent $5 in the Walgreens for two sodas and a candybar, (making use of their restroom at the beginning and end, partially due to the two aforementioned sodas). I had planned ahead and brought a magazine with me, but if I hadn’t I would have also spent another 4-5$ on some reading material.

If I could have spent $10 and been charged up in 10 minutes I would have gladly done so, and the time savings (and lowered sugar consumption) would have made it well worth the fee. But, I don’t have a DCFC port on my S-10 truck conversion.

So for a one time necessity charging session, $10 seems very reasonable.   It is comparable to spending $6 for a service station sandwich that you could have made yourself for $1 at home.

If you are going to be making long trips where you need DC fast charging frequently, you sign up for the monthly plan, which greatly reduces the price per session. You can think of this as a toll road type system, where you know you need to charge up every day on your daily commute, so you just budget the extra costs as part of your transportation overhead. (Although in that situation, an ICE vehicle may make more economic sense.)

JuiceBox EVSE Pro 40 – Second Unit

installed

I’ve been using an Electric Motor Werks JuiceBox Pro 40 to charge the truck for the last six months. It worked just fine most of the time, but suffered from intermittent GFI trips that would cause the wifi module to turn off entirely (going offline and stopping the charge) very occasionally.   After flashing the firmware didn’t fix the issue, they sent me a replacement unit. (Pictured above).

You can read all about the old unit here. The new unit has a slightly newer version number (v8.14.07) and the power cable going out to the J1772 gun is shorter (looks to be 18′-20′ instead of 25′) and MUCH thicker (marked as 8 GA wires for power, which matches the wires used to supply my NEMA 14-50 outlet). I have some photos of the differences here:

old_vs_new_guns

old_vs_new_length

 

Poking around the Electric Motor Werks website, the “thick” cable and gun appears to be what the current JuiceBoxes come with by default, but they have an option where for an extra $29.99 you can equip your JuiceBox with “Flexible J1772 Cable”   described as follows:

“Outfit your JuiceBox with a flexible & thin 40A 25FT non-UL J1772 (output) cable!

For those who value portability, or ease of cable manipulation, we are glad to offer a highly flexible and thin output cable option. With the addition of this option, your fully warrantied and assembled JuiceBox will come ready to go with a 25FT 40A portable spec output cable.
Please note that this cable is not UL rated, and not designed for compliance with national electrical standards. It is tested as safe to use for the power rating presented. Please note the cable you receive may not appear exactly as pictured – though dimensions and cable materials will be consistent.

So it may be that my original unit came with the “flexible” cable by default, but the new unit comes with a larger UL listed cable. (The cord I got on the replacement unit is shorter than my original by a good five to six feet.)

 

GFI cause update:

I have identified the cause of the GFI events (my TSM2500 chargers cause a GFI event when they shut down due to overheating.) The new charger still gets knocked “offline” when a GFI event happens, but it appears to recover on it’s own without needing me to power cycle it.

 

How far can it go?

Summary: I drove my truck 46 miles on one charge (and had some juice left over).
Instant-1457835710687

When you have an electric vehicle, everybody wants to know how far it can go.
I typically tell them “19,800 miles so far.”

But then you have to answer their real question, which is “What’s your range on a single charge?”. If you have a commercial EV like the Leaf or a Tesla, you can just refer to the EPA range figure for a nice apples to apples comparison. But when you have a conversion EV, the number is unique to your particular vehicle, motor, controller, battery pack and testing methodology. (And changes as the pack ages…)

I used to know the answer to that question for my truck with a (new) lead acid battery pack (“25-30 miles without killing the pack”), but I haven’t fully characterized the trucks’ power usage and range with the new (lighter weight, more powerful) pack made up of Nissan Leaf cell modules. My truck is heavier and has more air resistance than a stock Nissan leaf,   the motor/controller is slightly less efficient, and the (big fat!) tires have quite a bit more rolling resistance. I figured “half the range of a Leaf” would be a good ballpark estimate.

Continue reading

Why you should NOT buy FOSCAM cameras

I used to own two FOSCAM cameras, which used to (mostly) work with Linux. They would ask you to install a proprietary plugin (that doesn’t work in Linux) to view their camera feed, but if you selected the MJPEG feed everything would still work with Linux or other OS’s without installing the plugin.   (And you could manage them via the web interface with no problems.)

HOWEVER, after their latest firmware upgrade, their web management interface will NOT allow you to log in unless their proprietary plugins are detected. This means that you can not manage them from Linux. I asked them for a way to downgrade the firmware and they refused.

These cameras are now useless for use with Linux, and they are also useless unless you are willing to install foscam’s proprietary software on our (windows/mac) system.   And if you trust FOSCAM to put code on your computer, you should read this: Bug Exposes IP Cameras, Baby Monitors or look here: Foscam : Security Vulnerabilities and reconsider letting them install software on your computers.

And then there are things like this:This is Why People Fear the “˜Internet of Things’