Tuesday, September 17, 2019

diy lamp update

Some learning from five years outside. I have taken down my lamps to make changes. Let's see how they are weathering.

My concept was to build simple safe cheap reliable rugged outdoor lamps using common materials: wood, found glass, cheap LED chips and reused "wall wart" power supplies.


This one (Model upside down L) is shown attached to a corner of my deck under the railing. I mounted it low, about 75cm off the ground. The light shone downward on the path and was limited upward to the horizon by the top of the lamp. There was no glare from these lamps as you walked toward them, just illumination of the path. Would the design be classed as "full cutoff"?

I am pleased with the result. They were bright! Each used only 2 watts at 9 volts from an old printer power supply. The cost was about $5 each for new and reused materials.

[click any pic to enlarge]

A gray connector box is barely visible on the other side of the post. The wire carrying low voltage DC to the lamp passes through a hole in the post to the connector box. I tried to hide the wires and plug entrances for rain and bugs. Not very well it turned out.

I liked using old glass telegraph insulators for the lamp lens although I tried all sorts of glass. Some, like the insulator, were difficult to mount, especially for outdoor year round use.

The idea was that the glass object would protect the LED chip from the elements yet pass the light in a pleasing and interesting yet effective dark sky kind of way.

A common material I used was 2x4 lumber, usually without any protective coating so that the wood weathered natural grey. I avoided cracking of the wood by selecting evenly grained sections with no knots or visible cracks. I pre-drilled all the screw holes. Cracks did form over the years but they were minor.

I had intended for the lamps to look rustic in a modern way if that is possible?

I had not run these lamps last winter so I didn't know if they still worked.

On the bench connected to a power supply!

Perhaps I shouldn't sound surprised?

Let's have a look inside. To do that, I have to destroy them.

One of the lamps has been a bug home for a while, the other is clean, apparently a better seal.

The bug is alive. No bugs were harmed in this research!

The seal was a bead of silicone around the skirt of the insulator where it touched the aluminum sheet. I had trouble getting the silicone around where the wires went through the sheet.

The thin aluminum sheet is a piece of natural finish roof flashing. It serves as reflector, heat sink and support for the bale wire. The glass hangs from the bale wire so that the silicone has backup. The glass insulators are relatively heavy. The aluminum sheet with it's attachments form a sub-assembly that gets screwed to the wood frame enclosing the electrical part from the weather. There is a cavity in the wood to fit.

It was crudely done.

The back view before removing the bale wires and peeling the aluminum sheet away from the glass. I can see that I used both steel bale wire (the thinner rusty wire) and aluminum wire (thicker and shiny). I recall that the aluminum was easier to bend to the required shape. The steel, although thinner was more stiff.

The smaller rectangles are pieces of aluminum bar to improve heat transfer to the thin sheet. The LED chip is mounted on the other side with two screws. There is thermal compound between the chip and the bar and between the bar and the sheet.

They don't look the same because I tried a couple of variations. The one of the left was a later model. Notice that the silicone fills the holes for the bail wires but not the electrical wires!

I'll bet that's the bug entrance!

In this pic, the bug house LED is the bottom one. Surprisingly clean.

The top one has suffered some rusty corrosion to one of the two bolts but not the other.

Both LED's still work as I showed with the bench power supply. I wouldn't count on the top one lasting much longer just by appearance. There seems to be some darkening at the top corners of the COB array potting.

This LED is from the dark sky shielded lamp. As you can see, it did not do well at all. The LED no longer works.

I had used a small glass jar as a lens for this lamp. I found that it almost immediately showed signs of moisture inside the glass. None of the screws I used were rust resistant and the lid of the jar rusted badly.

The connector boxes were ok although 2 of 4 became bug houses. Corrosion of the cover screws was a problem. The screw used is an unusual fine metric thread (M4x0.5 15mm long) so finding a suitable replacement is difficult. I like these boxes otherwise for outdoor use. The one has been in direct sun and rain for five years.








This autopsy of my lamps after their "life test" shows what needs improvement. It's never perfect the first time!

Your comments are welcome. Please browse the other articles in the series.

Thank you for over 55,000 pageviews of this DIY Lamp series so far!



George Plhak
Lions Head, Ontario, Canada

diy landscape lamp reading list
a very bright 1 watt diy led garden light
making a lamp from a 2x4
best light at least cost - about testing bright diy leds at home
diy testing of led lamps
diy 1 watt led update
diy garden lamp progress
a shielded low power diy garden lamp
diy lamp update - this article


You might also be interested
shop lighting series

Tuesday, September 03, 2019

digital clock



Some additional detail. A short photo essay to expand on the above FB post. [the FB link does not seem to be fully live. Try this instead]

The circuit working on my bench yesterday. Link to the video. It keeps time from the 60Hz line frequency so it is pretty accurate unless the power goes off in which case it needs to be reset. It doesn't know about daylight savings time. It does not dim at night. It's big and heavy. But still pretty cool to watch.

In the very early 70s, while a student at UofToronto, I found giant neon nixie tubes at a surplus shop in the electronics market then around Yonge and Wellesley.

I was told those tubes came from the stock market display board at the TSE (now the TSX). The display tubes for this vital board were changed regularly and these were the rejects. I think they were expensive, maybe $5 each?. I was a student so couldn't afford new ones then. In the 80s, I replaced those tubes with these real NEW ones that were still available then, about $30 each. Advertised on EBay now for US$200 each for USED ones.

Digital clocks were a very hot item then and there just weren't any BIG Digital Grandfather Clocks, so I decided to build one. Because of the used tubes, this is a reuse project!

The presentation needed to be impressive so I enlisted the help of a friend who was a better woodworker. He helped me with the cabinets made from mahogany (which you could buy then) and a huge sheet of 1/4 inch smoked plexiglass. Until yesterday this clock was almost six feet tall and very heavy and awkward. It was not very stable on a floor but it never fell over! It did not have levelers at the bottom so needed shims to be vertical on a old slanted floor.

I had moved this clock wherever I went for the past 45 years. I moved it maybe ten times. It was coming apart at the bottom and needed to be fixed.

The big space under the clock was intended to house an electronic pendulum which never got built. Nor the electronic chimes. So there was really not the need for the big space. Yesterday I shortened the cabinet and made some repairs and improvements. Now it fits on a desktop or shelf.

Built with six fancy Burroughs special 17 pin sockets, hand wired and supported with a piece of wood.

The bottom of the clock board.

Mains power supply. The nixie tubes need 180 volts. The rest of the circuit used low voltage. This made both. I cannot believe it worked for 45 years!

The board layout.

Then in 1976 Heathkit introduced a digital floor clock GC-1195 and digital shelf clock GC-1197. The first commercially available digital "grandfather" clock. Popular but not a huge success.

Their display used wedge based incandescent lamps that burned out frequently. Heathkit offered a Winchester chimes option. No pendulum.


I got busy with school and never made another clock.

Thanks for your interest

George Plhak
Lions Head, Ontario, Canada

Wednesday, August 28, 2019

battery testing

Some of my lithium rechargeable cells weren't lasting as long as others.

Sorting my batteries was made easier with a simple inexpensive battery capacity tester.


I found a wide variation in my own assortment of various 18650 Li-Ion cells. Now each cell has a number, the milliampere-hours (mAh) the battery achieved on the tester.

[click any pic to enlarge]

The one pictured had been running on the tester for over an hour. At the time of the picture, that 2600mAh battery had achieved 1716mAh (the display shows amp-hours ie 1.716 Ah). The final capacity is determined automatically by the tester when the battery voltage falls below 3 volts, the test is stopped and the final Ah shows flashing on the display. For this battery, the final was 2.758 Ah (2758 mAh).

This one battery is my BEST battery and the ONLY one that achieved a result on my tester greater than what was branded. None of the rest achieved branded capacity out of 15 cells tested. Bought from various sources, my 18650's cost me between C$7 to $15 each.

The better ones are going to be used more often in my flashlights and USB power packs. The weak ones are going to help guide my purchasing decisions after they go to electronics re-cycling.

I am using the ZHIYU ZB2L3. Mine came from Banggood product ID: 1112859 in a little over a week after I ordered. The units I received were v2.3a.

ZB2L3 is a small circuit board which discharges a test battery into a resistor while measuring battery voltage and the current delivered to the resistor. ZB2L3 calculates the mAh. While testing, the display cycles three numbers and an LED at the side of the display shows alternately amp-hours, amps and volts. You can see a video here.

The micro USB cable only supplies 5 volts to the ZB2L3, not data. Since this is a complete discharge test, the ZB2L3 must have an alternate source of power hence the USB connection. Almost no energy is drawn from the USB. You can plug it into any USB charger or port. It will not charge the battery.

There is no data analysis available from the ZB2L3 other than from the display. I'd love to know that there was a way to capture a discharge curve such as this one, to explore different discharge rates and temperatures.

Two power resistors came with my ZB2L3. I used only one for all of this first round of tests. By using one of the supplied 7.5 ohm 5 watt resistors, the battery at 3.7 volts was seeing about a half amp of draw. This is (3.7X0.5=) 1.8 watts which makes the little square 5 watt resistor very hot, too hot to touch. That is why I have a cardboard sheet under the test, to protect the desktop. In hindsight it should have been a fire-proof insulating sheet. The battery gets slightly warm as it's working pretty hard.

I got the 18650 size battery holders from Banggood also. These I won't recommend. Although the battery should be held firmly against the contacts, with these battery holders it is difficult to remove the battery. The wires are very flimsy and probably contribute some voltage drop to the battery voltage measurement. Wires to the battery should as short as possible and a larger gauge wire. The contacts in this holder probably wouldn't handle soldering well. I will order better holders.

This test is a vivid demonstration of how much energy is stored in these little cells. The resistor is too hot to touch for several hours.

The ZB2L3 can be used for a range of batteries up to 12 volts and measures up to 9999 Ah. The value of the resistor can be adjusted to change the discharge rate up to 3 amps maximum but you must keep safety in mind of course.

Thanks for your interest,

George Plhak
Lions Head, Ontario, Canada

These cells, branded Ultrafire 6800 mAh show that the test gives consistent results on one type although the value is way lower than the branded capacity?!

The battery flipped over to show the branded capacity tested at only 795 mAh, the lowest of this batch of "6800 mAh" batteries.







As is my custom, this is not a paid review nor do I receive or solicit any product in return for writing. I am not selling these products. Just ideas. I don't get a commission. I don't show ads. I don't have a fundraising or patreon page.
I write in the hope that this is interesting or useful to you.
If you would like to show support for my independent work,
please consider commenting or buying one of my books? Thank you. George