Andrew Que Sites list Photos
Projects Contact
Main
Captured

Captured

   Did a series the other day on my incense burner titled "Captured".  Interesting story behind this.  When I was still in high school I bought this incense burner, but it was lost after I moved.  I searched around for an other but didn't find it again.  Then in 2003 I found a picture of the same design and traced it down to someone who made it using a lot of crazy search engine skills.  The site with the picture link had a dead e-mail address, but the name was unique enough I was able to find an alternative address.  When I e-mailed them they were willing to make me one.  Turns out the incense burner was a mold distributed by Scioto Molds.  Those who bought the mold could then cast a piece, fire it and glaze it.  I don't know how popular this design was, but in 2003 it was sure hard to find and I was pleased to be able to have it again.
   As an incense burner it is a little picky.  I don't think it has enough vents and incense don't always stay lit when inside.  However, once it starts up it looks pretty good.  So decided to try various lighting angles to see how I could capture the smoke from Captured.

September 16, 2014

Retry for Operation Lux Housing

   The housing for Operation Lux has been attempted a second time.  This time I used a large piece of glass, different caulk, and a gasket.  The caulk should hold up better under water than the previous caulk.  The gasket acts as a spacer and keeps the caulk out of the hole for the lux meter.  After this has had time to dry it will be time to test the setup again.
Andrew Que in the Lab

Andrew Que in the Lab

   At long last my inspection microscope arrived.  For a first test I went to work on repairing a Kindle that had a bad USB connector.  I had attempted to repair this some months ago for a friend, but didn't have the equipment.  I ended up soldering two USB leads together and gave up.  With the microscope and solder station I had what I needed and tried again.  Although still difficult I had much better luck and succeeded in installing the connector.  The Kindle was charging and seemed function... well, it that was until I knocked it on the floor with the charger plugged in.  That broke the connector again.  So I was able to make the repair even if the device insisted on being broken.
   The inspection station and solder rework station make a good combination.  I've wanted both for a long time and now that I've been doing electrical projects again I could justify the investment.  I found that 20x zoom was good for this project and I have three levels of magnification available: 10x, 20x, and 40x.
   Decided to take some pictures of myself in the lab last night, so here I am.

September 14, 2014

River Sprite Encounters Sahuagin

Riding Aboard the River Sprite

Riding Aboard the River Sprite

   The group was now aboard the River Sprite on our way to Newport, about a days journey away.  There was nothing special on the first day of travel.  Andal was sea sick, and the group discussed our ideas about keeping our true identity secret while in Twilight City.  That evening we awoke to heavy seas.  Reaching the deck of the ship we saw no weather, and there was not a strong wind.  Looking into the water we saw Sahuagin.  They were riding waves in a circle around the ship attempting to create a whirlpool to sink the vessel.  The first to see this was a new character named Scar, a raptor man in leather armor with goggles, opened fire on the enemy with a bow and battle began. 
   In the first round of battle Marv, who only attack at close quarters tried to balance on the bowsprit, but fell into the water.  Jam mannaged to throw a road and lasso Marv's arm.  Andal attacked using Sparky as a lightning source.  This caused Sparky to begin glowing, and then a bolt of lightning jetted out from Sparky to Andel (who was in the crow's nest) and then down to an enemy in the water.  The enemy exploded.  Marv continued to battle in the water and doing pretty well.  Scar used an action point and in a single round killed an enemy with a volly of arrow that would have made Legolas proud.  Ellenoria sang several Songs of Battle to the tune of old sailing songs, giving everyone a bonus to attacks.  One by one each enemy was destroyed. 
   After all the enmies were destroyed, Marv went to inspect the device the the Sahuagin were riding to create the whirlpool.  He discovered they were, in fact, tentacles from a much larger creature that is larking under the water.  The first fight was over, but the battle was about to go into a second phase.

September 13, 2014

Selecting a Holding Capacitor

Test Passes

Test Passes

   As stated before the solar powered web server will have a power source selection relay.  This will allow the server to switch to a power over Etherenet source should there not be enough solar/battery power to continue operating.  The relay I have selected states it has a 10 millisecond switch time.  During that time the server will be without power.  To compensate for this I plan to use a capacitor to store enough energy to keep the server running during the switchover process.  The size of this capacitor is what I need to calculate.  It must be large enough to keep the system powered for 10 ms.
   This calculation could be done mathematically, but it can also be done experimentally which is how I chose to do it.  The setup is fairly simple.  A capacitor is placed on the supply voltage just before the DC-DC converter.  An oscilloscope placed on the voltage fed into the DC-DC converter, and a second probe on the output.  The scope is setup to trigger on the falling edge of the input voltage and for a long interval. 
   The results of a test using a 100 µF capacitor are shown in this scope trace.  Each vertical division represents 10 milliseconds of time.  The trace at the top is the output of the DC-DC converter which should be 5 VDC.  The bottom trace is 12 VDC feeding into the converter.  To preform the experiment, the scope is setup to trigger for a single trace.  The holding time is the interval between when the 12 VDC begins to fall, and the 5 VDC begins to fall.  That is, the amount of time the setup is running without power.  Vertical cursors have been setup marking the location of the start of the 12 VDC drop (i.e. when power was removed) and the start of the 5 VDC drop (when power is too low to run the DC-DC converter).  In this experiment the 100 µF capacitor holds power to the running Raspberry Pi for 34 milliseconds.  This would be sufficient to sustain power during the 10 millisecond relay switch over.  I used the Pi because it's idle power is higher than that of the Odroid, and the Odroid is currently being a server for one domain and should not be switched on/off.

September 12, 2014

Test the Weather Proof Fixture

   Here is ππ's new case undergoing a submerged water test. The pipe wrench is just for weight to hold the case under water. On the top is a glass window for the lux meter.
   The results: pretty good, but there were problems. The case was mostly water tight. After more than an hour submerged there was a little water inside the case. The caulk I used on seal the glass did not hold up. Either I didn't allow the caulk to dry long enough, or it didn't like being submerged. The glass came right off after I pulled it the test out of the water.
   The new weather proof housing for ππ has arrived and it is time to put it together. I have to get an Ethernet cable inside the enclosure. I wasn't impressed with any of the weather proof connectors I found online, and decided to make my own. Here is what I came up with. This is part of a contact lens case. The lids seal water tight. I drilled a hole in both the lid and through the bottom of the container. This gives me a connector I can mount on the case. On the cable side I plan to use a bit of cork (from a wine bottle) and the lid of the contact lens case. When matted this should create a compressed seal around the Ethernet cable.
   The first step was cutting a hole and attaching the modified contact lens case to the enclosure. A lot of super glue does this. After the super glue dries, I will add a layer of silicon caulk that should create a water tight seal. We will have to see if this works after it is completed.
   Our new soldering station arrived today. It is a nice setup with both an iron and a hot-air rework wand. I have a couple of projects in mind that will need the rework abilities. After assembly I placed the station on the shelve with my power supply and oscilloscope.
   Our under-bench tool shelve is now installed and in action. I ran a strip of LEDs along the top for illumination and have placed all my basic electrical tools on this shelve. This allows me to work efficiently in this area of my bench on electrical projects as I have access to anything I would generally need within arms reach. The shelve is nothing fancy made of plywood, a coat of black paint and a couple coats of varnish. It should, however, be quite functional.

September 08, 2014

Odroid synchronization

   As the Odroid now has all the hardware components necessary for it to be a web server it is time to set it up to be a server.  However, testing is needed before any switch takes place.  The Odroid's 128 GB micro SD will be storage for all the websites hosted by DrQue.net.  I setup an rsync script to synchronize the Odroid to the Micro Dragon.  One of the problems with this is permissions.  The web server runs as the user www-data and since the users of the Micro Dragon are not the same as those on the Odroid I can't preserve ownership during the synchronization.  All the web pages are owned by www-data anyway so I wanted a way to force this user to have ownership for new files.  I accomplished this in a rather complicated manner.
#!/bin/bash

ID_FILE=`mktemp`
sudo chown www-data $ID_FILE
sudo chmod 600 $ID_FILE
sudo cp id_rsa $ID_FILE

COMMAND="rsync -rltv --no-p --no-g --delete -e 'ssh -o IdentityFile=$ID_FILE -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null' user@drque.net:/webpages/ /webpages"

echo "*************** Sync webpages ***************"
sudo su - www-data -s /bin/bash -c "$COMMAND"
sudo rm $ID_FILE
    What this script does is first copy an SSH key that allows a no password log in from some location to a temporary file.  This file is then assigned to www-data with no read access given to anyone else.  This will allow www-data to log into DrQue.net over SSH without needing a password.  Then a command string is setup that uses the temporary file to the SSH log in for rsync.  Since www-data cannot normally log in it has no user directory, so the known hosts and strict host checking need to be disabled.  Once the command is setup, su is used to switch the user to www-data and run the rsync command.  The rsync command is setup to not preserve permissions or ownership, meaning all the files transferred will be owned by the user doing the transfer—www-data.