Andrew Que Sites list Photos
Projects Contact
Main
   Pictured is a poorly put together paniramic view of the progress thus far.  Most of the main bench space is complete, as is the hanging shelve.  What is missing are the book shelves, and a small 4' section of bench added just above the box door.  Xen was kind enough to clean out the inside of the box, so I've been leaving the doors closed to keep out dust.
   Today seemed like it went very quickly.  I installed rope light under the shelve, started the wiring inside the box, and ran lines for the switches that will turn lights on/off for the overall bench setup.  I started on the shelves, but didn't end up liking the results.  So more planning needed there.  After some touch up paint, we were well into midnight and I called it a day.
   Worked until around 2:00 am with Xiphos on the setup in Madison. Xiphos worked on fixing his bed who's metal structure has suffered being used far beyond it's feeble design criteria. I started by adding the shelve above the main box. This hangs from the ceiling from 2x4, and has a 1/2” deck. Underneath it has two 2x4 beams for support. On the front side of this will be an array of outlets. This is accomplished using two double-wide gang boxes, each with two outlets. Each outlet will feed one 6 outlet power strip, which will be mounted in the spaces between the gang boxes, for a total of 48 outlet sources.
   I had just finished my cuts for the suspended shelve when Xiphos arrived with a new tool: a mitering saw. It was promptly put to work and has been a most welcome addition to the arsenal we have available.
   Wiring was started today. I had a lot of 14 gauge 3 conductor wire for whatever reason, so I designed my wiring setup around this fact. My setup will run from 2x 15 amp circuit breakers, most of which come to a gang box on the corner of my bench. I finished wiring up all the outlets on the hanging shelve and made sure they worked.
   I also fixed the left side door of the box. It didn't close very easy, so I removed it, sanded it, and put it back. Now sanded, it moves much easier. And now that I have a router, I decided to forget using the sliding door finger tabs, opting for routed groves in the door. This worked very well, and I am most pleased with the results.

July 06, 2011

LED Rope Light Dimmer

Schematic

Schematic

   I designed a schematic for a simple LED dimmer that should work with the LED rope light I have.  It uses a 555 timer chip to create a Pulse Width Modulated (PWM) that drives a Field-Effect Transistor (FET), based on this design.  This charges a filter capacitor that feeds the LED string.  The trick behind this circuit is the input power.  The LED rope light effectively runs off 170 volts of DC.  The 555 circuit need between 3 and 30 volts DC.  Rather than getting complicated and using a transformer based power supply, I decided to see about a supply based on a Zener diode.  This type of supply is not as efficient, but I don't need a lot of power.  In fact, I found a 7555 timer, which is a low power version of the 555 timer, and it uses 200 µA peak.  I found this document that describes a circuit using a Zener diode for just this application.  I downloaded a schematic capture program called TinyCad and quickly had the circuit drawn up.  What was great about this program is that most parts included a field for the DigiKey part number, and that's who I plan to order from.
   My plan is to house the circuit in a gang box with a blank cover.  I'll drill a hole in the cover for the potentiometer, and this can be mounted where ever I feel like having the control.  The values for the power supply I have chosen should be able to supply 1 mA, more than twice what I need, but still not wasting a lot of power.  It should be a lot of fun to build :)

1 comment has been made.

From Noah

Oh, around...

July 11, 2011 at 9:57 AM

Oh, the 555 and its' brothers. Is there anything they can't do? Good luck!
   I did some experiments with a string of blue LED rope light. The string is made by Sunbeam, and like many LED lights I've play with, they flicker. This is really only noticeable when the lights are the only source of illumination, and when you are in motion. The flicker is more of a nuisance then a problem, but nuisance nonetheless.
   My first task was to figure out what voltage was fed into the string of LEDs. My guess was there were two options: full/half wave rectified AC at 120 volts RMS, or rectified 12 volts. Using my oscilloscope, I found the answer was fully rectified 120 volts RMS. I found the voltage peaked at 166 V, and dropped as low as 75 volts—a change of some 90 volts. This looks like the output of a rectifier circuit with a small filter capacitors—too small. But the nice thing about filter capacitors is that you can add more. I searched around and found I had a 47 μF capacitors rated for 200 volts. So I wired it in parallel to see what would happen.
   I was a little surprised as the effects were slightly dramatic. The LED string brightness increased at last 25%. The voltage ripple was now only 166 to 143 (22 volts)—about 25% less than without the capacitor. Since I was doing the experiment in daylight I wasn't able to tell if the flicker had gone away, but the scope sure showed an improvement so it must have at least been reduced.
   I have placed this string of lights on a dimmer before, and it did work. So I decided to see what happened if I tried to dim the lights now that it had extra capacitance. The results were strange. If there was an additional load on the lights (in this case, my desk light), dimming worked. However, once I removed the load from the desk light, everything began to act strange. The string was blinking when the dimmer was all the way up, and the dimmer range was strange. My guess is there is some minimal current the dimmer expects, and the string of LEDs lamps isn't enough.
    So I am able to reduce the flicker using a 47 μF capacitor. If I increase the size of the capacitor, I should be able to reduce the flicker further. I went to DigiKey and found I can easily pick up a 200 volt capacitor over 1000 μF should I decide I want one. But I will still have to figure out a system for dimming the string.
   Pictured is the same string of LED rope light, but the frame to the right has the added capacitor.  The shutter speed for both shots is the same, and you can see the brightness has increased.

1 comment has been made.

From Noah

Janesville (still)

July 11, 2011 at 9:55 AM

I noticed the LED flicker effect, and played around with it - a motion blurred camera shot of LED lights will produce a string of flashes. *sigh* I would grumble about engineers not properly filtering their power inputs, but the sad thing is the engineers probably got steamrolled by someone in corporate who wanted to save a few cents per light.
   A full day of work today, and a lot of work accomplished. Started by finishing the sides of the box, which include the doors. Then I cut the top of the box. The glue which holds the hardboard to the 3/4” plywood still had not dried completely yet, which I was not pleased with. However, we did get the top cut and set in place. Before securing the top, I decided it best to cut the holes for the air vents. Xiphos had recommended I use a router with a flushing bit for these cuts. I didn't have a router in my tool collection yet, but I have wanted one. When ever I find a tool I want, I wait until I have some project that could use said tool before I buy it—and now I had a reason to have a router. I had never used a flushing bit before, or a router, but it was impressive just how easy it was to make this cut. Using the flushing but, I quickly had perfect vent holes cut.
   Pictured is the largely complete sleeping box.  Most of the components now exist, and what remains are finishing items: caulk, electrical, handles, ext.
   It wasn't easy to get to Madison today—traffic was backed up on Interstate 90 going north as it appeared every person from Illinois was coming into Wisconsin. Once I had arrived, however, the work began. Xiphos was a great help today. We finished the air intake plenum, the and the floor. He had to leave for a few hours, and I cut the sides and sanded them. My new belt sander made this task pretty quick, and the sanding should allow the doors to slide smoothly. Xiphos came back and helped attach the sides, and the 3/4” door spacers. At this point it was quite late. We caulked, painted, and called it a night.
   Today begins the construction of my sleeping quarters. I only had a couple hours to work, so I didn't get a great deal finished, but it was a start. I started by gluing the 1/8” hardboard to the 3/4” plywood what will make the top of the bench. This way, I could stack a bunch of weight on it and allow the glue to try while I worked on other tasks. The first items to be constructed are the light blocking air plenum. And the exhaust plenum was completed today.