First 5 months report & payback calculations – Grid Tie Solar System

Five months ago we turned on our new grid tie solar system and started to produce power.

Component Failures & Warranty Service
In the first five months, we have had two components replaced under warranty. The first was one of the 36 SolarEdge DC optimizer units (A $75 component with a 25 year warranty) that are mounted under each of the solar panels. The Solar Edge monitoring system had flagged that one of the optimizers was not reporting (and presumably also not producing power) and sent a message to our installer (3Guys Solar). They called me (before I even knew anything was wrong) and let me know they would be sending a crew out to climb up on the roof to replace it the next day, so that particular panel was only down for only 2 days. The other 35 panels continued to produce power.  The crew said that they sometimes have to replace several optimizers on a house, and some houses never have any fail.

The second component to fail was the system’s main DC->AC inverter, a 10kW Solar Edge grid tie inverter (a $2,000 component with a 12 year warranty). On October 23rd I noticed on the phone monitoring app that we had no power produced, so that night I went out and rebooted the inverter.  The next day, 3Guys solar called me to report that they had received a fault code from the Inverter via the Solar Edge monitoring system and were working with Solar Edge to try and resolve the fault. The following day they called again to tell me that they would need to replace the inverter under warranty. Unfortunately, it took close to two weeks to receive the replacement unit from Solar Edge, so we were not producing solar power again until the 7th of November.

It was concerning to have the inverter  fail within the first three months, but it is covered by a 12 year warranty, was replaced within two weeks and we haven’t had any problems since. 3 Guys Solar also sent us a check for $95 to cover the cost of the energy generation lost during this period. This is a limited time program, and was not part of their original install contract, so the check was not expected but appreciated. (By my estimates they overpaid us by 20-30$ for the energy the system wasn’t producing while the inverter was down.) The warranty coverage by 3Guys Solar (the installer) and Solar Edge (equipment manufacturer) left nothing to be desired.  My expectation is that with solid state electrical equipment like this, most of the failures that are going to happen will happen early in the life-cycle (or very late in the lifecycle), and I hope the equipment will be stable now that we’ve gotten the early failures out of the way.

Power Generated & Usage

In this 5 month period, we generated 5096 kWh of solar power, and used 7405 kWh of electricity (paying for 2309 kWh from Duke, at a cost of $337.44 (-$95 credit from 3GS lowers this to $242.44)).


In the same 5 month period last year, we used and paid for 7378 kWh of electricity from Duke, (costing $977.81) so our usage appears to be closely correlated to before we had the solar system.

By averaging production between the week before and the week after, I estimate that in the 15 day period our inverter was down we should have produced an additional 540 kWh of power ( $70 of electricity at 13 cents per kWh). So without the inverter failure, we would have produced around 5636 of the 7405 kWh we used, or 76% of our total electricity usage via solar. [With the inverter failure, we were only at 69% of our power from the sun.] Our goal was to produce 80% of our electricity from the sun, so these numbers are close to our goal, and I hope that the sunny spring (and no more inverter failures) will raise our percentage.

From a cost perspective, because Duke Energy has a fixed customer charge, our solar cost savings is lower than percentage of power generated, and was a savings of between 65% to 75% off our electric bill.

Seasonal Effects on Future Predictions

Estimates made with less than a full year of solar data are going to be wildly inaccurate. The Aug->January time period is cooler than other months, so our AC usage will be lower, but there is also less solar production in the winter, meaning that our generation will be lower as well.  The following numbers are a guess, and are much less accurate than those I hope to calculate after the system has been working for a full year.

Over the same 5 month period last year, the solar system has saved us between $640 and $735 (depending upon if you add in the non-contractually obligated 95$ check that 3Guys Solar sent us).  This equates to an estimated payback period of  11 to 12.75 years.  (The 1.75 years, or 14% difference in payback time is driven by the 13% difference in cost savings that $95 check equates to over the 5 month period.)

If it turns out that our solar system produces a much higher percentage of our usage in the sunny spring months, or if Duke Energy raises their rates, this payback period could drop.  If we use too much AC in the summer, or Duke lowers their rates (?!?!) the payback period could increase. I’ll feel much more confident about the estimate after collecting a full year of usage and generation data. (And even then, solar generation and electrical usage can still vary year to year with the weather.)

Optimizing Pokemon Go Gift Exchanges / XP from Friendship levels

The addition of Trading / Friends / Gifts to Pokemon Go has added an extra level of play (work?) to the game. These are my thoughts on how to optimize friendship and gift sending and opening.

My starting assumptions:

  1. The goal of exchanging gifts is to build up friendship levels as quickly as possible so that Pokemon trades cost less stardust. [I am explicitly rejecting the idea that the primary reason to open gifts is to restock items.]
  2. You have not yet reached level 40 (or still care about building up your XP) and are willing to use Lucky Eggs to increase the XP you receive from friendship level increases.  [If you don’t care about XP, optimizing Pokemon gift exchanges becomes much easier.]


The Simple System (for Level 40 Players)

Continue reading

Grid-Tie Solar System Pricing in Orlando – My shopping experience

We recently hired 3 Guys Solar to install a 10.4 kW grid-tie solar system on our house. They installed 36 Axitec 290 watt mono panels with DC Optimizers on Iron Ridge racking, and a 10 kW Solar Edge grid-tie inverter.  Our final price was $2.409 per installed watt (Summer 2018).   I say final price, because we had a wide range of quotes from multiple companies, and several companies lowered their initial quote upon seeing quotes from the other companies. Our home,  with a new asphalt shingle, south facing roof with plenty of room for panels was about as simple and optimal as you can get for a solar install, so if your roof is more complicated (tile, metal, gables, vents, skylights etc) you can expect to pay a bit more. Read on for the full story of how we got to this final price.

Continue reading

Ego 21″ mower (LM2100SP) 2nd self propel motor failure & repair report

No, this post isn’t an accidental duplicate of this post from last year. Yes, the self propel unit on my 21″ SP Ego lawn mower failed a second time. (This is 9 months after it was replaced under warranty from the first failure, which occurred around month 4 of ownership, so this failed self propel motor lasted twice as long as the first one.)

However,  I’m quite pleased with Ego’s warranty service compared to the last time when I had to take the mower to Home Depot’s service department (who kept it for 5 weeks).

This time when I called Ego’s customer support number, I only had a seven minute hold time before talking to a person, and the customer support representative said that they would send me out a new replacement mower, arriving in 5-10 business days. [I was instructed to remove the serial number sticker and take the defective mower back to home depot “for recycling”.]

I have read many accounts on Ego’s customer web forums of other self propel units failing, so they may have had a bad batch of mowers go out and are now being more pro-active about replacing them. Alternatively, maybe I’m getting a replacement mower shipped out quickly because this is the second issue I’ve had. Regardless, receiving a working replacement in 5-10 days is much better than taking 5 weeks for a repair.   I’m still going to have to push my mower to mow the lawn (and it’s a big lawn, so this is more exercise than I am looking for….) but at least I could use the defective mower as a push mower until the replacement arrived and didn’t have to hire a lawn service while it’s in the shop.

The new mower arrived 11 business days after my phone call (it would have been 10, except UPS had a delivery exception and delayed it over a weekend, which meant I had to push the old mower around one more time.

After using the new mower with its super quiet and silky smooth self propel, I can say that the self propel unit that home depot repaired/replaced had many signs of problems before it finally gave up the ghost. First, it was louder than the mower itself. Second, it didn’t have enough power to fully propel the mower up a slight incline. Third, every time the mower went into overdrive due to thick grass, the self propel would slow down. These issues were either there from the time I got the mower back from Home Depot, or they started and got worse so gradually that I hadn’t taken note of them, but after using the new mower, it became obvious that I should have known my self propel unit was not working the way it was supposed to. On the new mower the self propel has plenty of power to move up a grade at the same speed it moves on flat ground, it makes less noise than the blade mower, and it doesn’t slow down when the blade mower goes into “overdrive” cutting thick grass.


Megatree: Materials & Costs

So, what do you need to build a large digitally color controlled LED outdoor Christmas tree display? (Commonly called a “megatree” by people in the Christmas display community.) And how much will it cost?

First, you need strings of color controlled RGB LED lights, wires to connect them, and a few power supplies. I bought 16 strands that have 50 lights each, with 6″ spacing (PixaBulb w/ Strawberry lens) from (I also got a 17th strand as a spare, and am using it as my tree topping star right now). This cost me $650 (including extra connection wires & two 12 volt power supplies).

Then you’ll need a light controller. I bought a Falcon 16V3 from for $210, and a CableGuard CG-1500 outdoor enclosure to protect it from the elements. (This enclosure holds the controller board, but is not large enough for the two power supplies, which I have under a better ventilated anti-rain plastic storage box)

Then you’ll need a lot of mounting hardware (lag eyes, quick connect links, etc) and wire ropes (small cables, cable thimbles & crimp connectors) plus zipties, lots and lots of zipties. I used stainless steel hardware from except for the galvanized wire rope that I bought at Harbor Freight (shipping a 500′ spool of stainless 3mm cable was prohibitively expensive). Including miscellaneous pieces of wood I used for my mounting ring and star tree-topper, a few extension cords, tent pegs and a 100′ run of Ethernet cable, all of this hardware cost me around $245.

I’m not including the cost of all the tools needed, plus a laptop to sequence and control the show.

So all in, for a “mid-sized” megatree (20′ tall, 16′ diameter at the base) you are looking at a little over a thousand dollars (plus a hundred hours of work). On the plus side, this cost can be amortized over multiple years, so it’s cheaper than multiple years of fireworks. Plus, with some creative work, you could re-purpose the lights for Halloween, weddings, parties, etc…


A more specific list of mounting hardware:

4x 1/2″ x 6″ Stainless Steel Lag Eye Bolts – Mounted in my tree. Completely overkill for supporting my relatively lightweight megatree, but may be re-purposed in the future for heavier loads. Holds up 4 support cables that allow me to raise/lower my top support ring.
20x 3/16″ Stainless Quick Links – for connecting and disconnecting the four support cables and top ends of the light strands to the top ring. I bought a 50 pack from a Chinese seller on ebay, as stainless steel quick links were quite expensive otherwise.
40x 1/8″ Light Duty Stainless Steel Wire Rope Thimble – Used at the top and bottom of each light strand, plus for the cables that lift the top ring.

40x 1/8″ Zinc Plated Copper Sleeve – Crimp connectors that hold the cable in place around the thimble. (Buy the proper crimping tool for these.)

16x 3/16″ x 1″ Stainless Steel Lag Eye Bolt – For connecting the top of the light strands to the top ring.
4x Eye bolts, washers & nuts to mount in the top ring for the support cables to connect to. (forgot the exact size, bought them at Lowes)
4x 4″ Stainless Steel Flag Pole Cleats – My jury-rigged solution for holding the four cables that support the top ring. I’m sure you could come up with a better solution.
16x 9″ tent pegs (bought at Walmart) for staking down the ends of the light strands.
8 packages 100′ 3mm galvanized wire rope from Harbor Freight (16 26′ light strands, 4 50′ support cables, leaving several 20′ seconds left over…) I seriously considered buying a 500′ spool of 1/8″ stainless cable from, but the added shipping cost made it prohibitively expensive. I’m willing to pay double for stainless, but not quadruple to get it shipped to me. Plus I figure the LED light strands will probably fail before the 3mm wire rope rusts through….
800-900 black zipties from harbor freight (to hold the lights to the light strand cables.
Misc deck screws to hold the top support ring together (made of 2×4’s) and mount the flag pole cleats and cable guard enclosure to the tree.

Wifi Printing to HP Envy 4500 printer chops off top of page

If you are printing from Ubuntu Linux to an HP Envy 4500 printer and try to print a document with a very small top border, you may find that the printer chops off the very top quarter inch of the page.

The fix I found was to change the paper size from “US Letter (11 x 8.5)” to US Letter Borderless (11.14 x 8.72)” in the print drivers.

I have read several places online where other users with Mac’s and Windows machines have the same issue when printing over wifi, so I feel the issue is with the printer itself (and it’s wifi to printing bridge) as opposed to the Linux drivers.

FB1-4001A motor brush photos

Here are some photos of the brushes installed on my DC FB1-4001A series wound motor, rated at 72-144 VDC, (19kW) 25HP continuous ( 100 ft/lb of torque / 48hp @ 500 amps ) which is  9.1″ in diameter.

My motor has a curved protective metal sheet that wraps around the front of the motor (the aluminum with diamond cutouts that has a cutout for the A2 post in this photo). It has a spring catch that releases it (not shown) on the other side of the motor.

I use the metal sheet to hold down fiberglass screen material underneath the diamond cutouts to provide extra protection from road debris. The brushes are accessed after removing the sheet.

The brushes are mounted in pairs, and each brush is held against the commutator by a metal spring and has two copper pigtails to transfer the current.

Here is a closeup of two brushes where you can see the commutator which rotates under them.


Ego 21″ mower (LM2100SP) self propel motor failure & repair report

After four months of ownership, around 40 hours of usage, the self propel motor unit on my Ego 21″ Self-propelled electric lawnmower (Model LM2100SP) failed. Before this failure I was very happy with its performance, and although it was repaired under warranty, the procedure took longer than I think was reasonable.

Ever since I purchased the mower, I have been monitoring the Ego community forums, and I knew that reaching a customer support representative would sometimes take extended hold times, and I had heard that taking a mower to Home Depot for repair could be an extended procedure, so when my self propel motor failed, I was relatively well prepared on how to handle the situation.

August 29th, 2017 – The self propel motor fails. I finish mowing my front lawn pushing the mower by hand (which makes me realize that paying extra for the model with the self propel motor was the right choice.)

August 30th, 2017 Time to call customer support. After re-charging the battery, and letting the mower sit overnight to cool, I tested it again (yep, still no motion) and then called Ego customer support early in the morning (to avoid a long wait on hold.). I only had to wait a few minutes on hold, and then a helpful customer support representative walked me through a few simple questions (yes, my mower blade would turn on and spin, so the battery was good, and the folding handle interlocks were correctly latched, etc…) to verify that the drive motor had actually failed.  After that, she “made a note in my file” and told me to take it to Home Depot.

What I wish she had told me: 1. Home depot will charge you a $20 deposit, just to look at the mower (and you authorize up to $150 worth of repairs upon drop-off). They will refund this deposit to you if the work is covered under warranty (mine was).  2. You can only take the mower to a Home Depot that has a tool rental / repair clinic (call first to check). 3. The Home Depot repair clinic doesn’t have their own EGO batteries for testing, so be sure to leave your battery in the mower. [This makes sense in hindsight, but I was hesitant to leave 7.5 AH battery that costs $400 to replace at HD, so this required me to make a second trip back to HD with the battery a few days later.]

Also, after talking with her I received an email from Ego that stated my “case had been closed” (this is the case associated with the phone call to Ego only…but the wording of the message didn’t inspire confidence.)

Because I had heard horror stories about the Home Depot repair clinic taking a long time to repair EGO mowers, I made sure I called them every week to check on the status of my mower, just to make sure it hadn’t fallen into any cracks. I also posted an update every Monday on the ego customer forums, which may have also helped things behind the scenes.

Sep 18th 2017: They are still checking on the mower, but the HD technician felt that they could repair it locally instead of shipping the mower to Atlanta.

Sep 25th 2017: In progress, waiting on a part which usually takes a week or two (presumably the motor/gearbox unit).

October 2nd 2017: HD is still waiting on the part.  [After this update to the Ego community forum, April from EGO said that they were tracking the shipment and that the part should be arriving at the HD store within a few days.]

October 9th 2017: HD claims to still be waiting on the part.  [I mentioned this on the Ego community forum.]

Finally, on Thursday night (Oct 11th), Home Depot calls me to tell me the mower is ready for pickup. Because the home depot with the repair clinic  is 12 miles away from me, I delay pickup until Saturday.

So, it took Home Depot / Ego about five weeks to get my mower repaired, and this was with me keeping on top of Home Depot and making sure that Ego knew what was happening at each step. Reading several other reports on the ego community forum makes me believe that my experience was actually on the faster side of things, as 8-10 week delays are not unheard of if the mower gets shipped to Atlanta for repairs.  I paid $180 to a lawncare company and took a break from mowing my yard, but there are at least three other options to keep mowing if you find yourself in this situation.

Take advantage of Home Depot’s 90 day return policy

Many others in the Ego forums have bought another Ego lawnmower and used it while theirs was in for repair, returning it under HD’s 90 day return policy once their mower was repaired. I felt that this action would be ethically questionable, but after waiting five weeks for HD to repair my mower, my ethical resolve is beginning to weaken, and should the mower fail again, I will seriously consider this option.

Buy a second mower

I did consider buying a second Ego mower (and keeping it), mostly as a way to purchase a second 7.5 Ah battery (it is almost as cheap to buy the battery and mower together as to buy just the battery, plus you get a “hot spare” mower). Other than the fact that you have to store the 2nd mower, this does have certain advantages. If one mower fails, you can just switch over to using the second mower while the first is in for repair. And, you get the advantage of having twice as many batteries and chargers.  Unfortunately,  at the time Home Depot was not offering the same $50 discount on the mower as when I initially purchased mine, otherwise I may have done this.  [My wife points out the questionable logic of using the failure of a product to justify the purchase of a duplicate of the same product…]

Burn hydrocarbons

Of course, you could also buy a cheap used gas mower and several gallons of gas for less than the $180 that I paid for lawn care service, and probably be able to resell it at almost the same price you paid when finished.

Final Recommendation

Due to the fact that my self propel motor failed after 4 months of ownership, combined with a 5 week repair time, I can’t recommend the EGO electric mower to everyone. At least with a gas mower your options for repair are numerous and much faster. However, if you have decided that you will be going with an electric mower, I still think that the Ego line has the best performance. (I have also posted a review of the mower.) I also own the hand-held leaf blower and chainsaw from their Power+ line, and have been quite happy with them.

The new self propel unit installed by Home Depot  failed again (after 9 more months of usage) but I’m much happier with how Ego handled the warranty repair (via a direct replacement), which you can read about here.

Denford Micromill 2000 January 2003 dispatch date – SGR location

Cliff Burger is part of a makerspace ( ) which had a Denford Micromill 2000 (January 2003 dispatch date) donated to them. When referring to my four part series( 1, 2, 3, 4)  about how I got mine working under CNC control, they noticed a few differences with their model and wanted to share that information.

Instead of having a custom made relay & power board, their mill has it’s relays mounted to a DIN rail (bottom left of the case in the image below).  The spindle go relay (SGR) is located in the 2nd from the right position.

A quote from Cliff:

On the DIN rail, the spindle activation relay is the second one in from the right. It’s a 12v relay with the ground for the coil being controlled by the C6 pin. However, currently the relay never sees a 12V signal either. Not sure if it’s something wrong with my board or it’s waiting for another command signal before it sends the 12V out as well. Either way, I’ll likely just get a 5V relay and switch it right off the BOB, but for the time being I’ve moved the orange wire from the “14” position to the “12” position to supply power to the board at all times.


Cliff also sent along his mach3 config file, which you can download here (note, you will have to remove the .txt extension from the file to use it.)   Denford.xml.txt

He has the following caveats:

Things to note about the mach3 config:
1) My limit switch are on different pin numbers due to me chopping 1 wire a bit shorter than I should have (oops!).
2) default units are in inches so the steps per INCH are correct, but may need slight tweaking for each application.
3) backlash settings will need to be measured for each mill, or disabled.
4) I’m running a UC100 UBS adapter board so Mach3 may give an error message the first time you open it with this config file.