Friday, February 26, 2016

I'm slacking.


I admit it, and I apologize to those who might be looking for tips and tricks.  I actually completed my build two months ago and have struggled to find the time to document things.  I was in a horrid rush to finish in time for Christmas that I had to stop even thinking about blogging the build, as I literally finished the last bits on Christmas eve.

But I did take photos and notes, so I'll try to explain things still.

When I left off, I was just about to start details on wiring.  But before I do that, there are some basics that really need to be covered.  This may be boring for some, but I'm hoping to get enough detail so that those unfamiliar with electronics can still build their own panel.

With that, let's talk about LEDs.  You need to know just a few basics to successfully build your panel.  First off, every LED has an anode and a cathode.  What does this mean to you if you're not into electronics but just want to build a panel?  Not much.  Just remember this - the anode on an LED is the long wire that comes off the back, the cathode is the short wire.  It will be critical not to get them mixed up, because current (electricity) flows only one way through an LED.  If you wire in an LED backwards, it will never light up.

So there's that basic.  Next, you need to know that every LED is a bit unique.  When you connect it to a power source (like a battery), it has a voltage drop across the LED that is often referred to as the forward voltage of the LED.  Hopefully your LEDs had that spec available for you. If not, you can build a simple circuit to test for it.  If anyone needs to know how to do this, let me know and I'll try to add those details.  You will need a voltmeter for that if you want to do it.  Most of the LEDs I used had a forward voltage of about 2 to 2.1 V.  Only the blue LEDs were significantly different, at 3.0V.

Why is this important to know?  Because if you connect an LED to a power source and don't limit the current flow through the LED, it will burn out.  Fast, in some cases.  The 10 mm red LEDs I purchased, for instance, if connected directly to a power source, would burn out instantly or get noticeably dimmer very fast.  So how do you stop this?  With a resistor. Attaching a resistor to each LED will limit the current flowing to the LED.  But what size resistor do you need to have?   Well, each LED also has a maximum current limit.  If you know this, you can simply use a current limiting resistor calculator such as at http://ledcalc.com/ to calculate the value, remembering that the supply voltage will be 5V, the voltage the Arduino operates at.  The calculation is simple, but using the webpage is simpler.  That will give you a minimum value for the resistor.  But you can go higher.  I recommend you actually do go higher, at least until you notice the LED start to get dimmer.  Often you can double the resistance value with no noticeable decrease in the brightness of the LED.  If you don't know the maximum current for the LED, then generally you can simply start with a high resistor value (~680 ohm) and start reducing to lower values until the LED stops getting brighter.  If you need directions on how to do this, just let me know.  Odds are actually pretty good that the resistor values I chose (which I'll give later) will work well for you.

Next up, I'll get to talking about LED matrices.  This is where the detail on the wiring and programming will really start.

Wednesday, December 9, 2015

Panel support structure

It's been awhile since the last post, but I'm still working feverishly on this project - I'm simply behind in posting.

First, a few details on finishing up the individual panels...

I tried gluing the transparency to the cutting board for the status lamps.  This turned out terribly.  The glue discolored and the cutting board warped significantly.  So I decided to simply sandwich the transparency and cutting board up against the clear transparency I had printed for covering the painted wood on that panel - I hadn't cut out the center section, and it holds the cutting board and the printed transparency neatly with the LED holder box behind it.  A little hot glue around the edge of the LED box on the backside (NOT up against the transparency, just between the plastic and the masonite) once it is installed secures it all in place.

Similarly, I left the transparencies over the LED bargraphs and seven segment displays intact, simply pushing the electronic component in from behind and hot-glue tacking it into place.   Note on this procedure - make sure that your components are going in oriented properly.  This is easy with the seven segment LEDs as they have decimal points that will point you to the bottom.  This is not as simple with the bargraphs.  The ones I purchased have a small miter on one corner - that will point you to the side which has pin #1 on it (there are 14 pins overall, 7 on each side).  If you want to double check, apply a 5v load to the two pins closest to that miter, with the positive (anode) attached to the side with the miter (pin #1) and the cathode attached to the side without (pin #14).  The LED closest to the miter should light up red on the components I used.  If it doesn't light up, that would imply you either have your voltage reversed (remember, LEDs are one-way only!), in which case try swapping your leads to the pins  - or you've got pins 7 and 8, which are both cathodes on this particular component.  It is critical that you put the component into the panel the right way so that it is wired properly and so that it does what you want when the arduino tries to control it later.  For bargraphs mounted vertically, I put the miter corner at the bottom of the panel - this will match the LED behavior that Jeff used.  For the cryogenics panel, I put the mitered corner to the left as you're looking at the front of the panel (to the right as you're inserting it from behind).

That should cover most of the population details.

Oh, and use your voltmeter on a resistance (continuity) setting to check that all your SPST switches are aligned properly so that they are open (infinite resistance or no continuity) when the switch is in its "off" position.

From there, it was off to building the support structure.  This will be mostly hidden from view, so all we care about is something that will tie things together affordably.  I used 1"x2"x8' furring strips from the local home improvement store.  3 pieces at about 80 cents each is all it takes.  I cut 4 3' sections out to run the width of the panel - one at the  very top, one at the bottom, and one along the panes at the top of the center panel and one at the bottom.  I cut 2 1' sections to serve as horizontal runs for the panels at the middle of the overall structure (Jeff didn't have a cutout for a monitor in the center, so he could continue that run all the way across the structure, whereas I had to stop it at the center).  I then cut 4 2' sections for vertical runs - one at each end and one for each side of the center panel where the monitor goes, so those 1' sections had something to rest on.  Using the panels as guides to mark the points where their individual edges would go, I essentially dadoed a slot where each wood strip would overlap so that they would fit together nicely with just wood glue to bind them.  This was not precision work - I simply notched it all out on a sliding miter saw.






First, a simple dry fit to test












Then check for square....







Then it was a matter of simply gluing the pieces together and clamping them until dry.  Luckily I have a lot of clamps.  Again, don't worry about this being precision work - my overall structure was a hair too wide, so I simply belt-sanded it down to be flush with the edges of the panels.  And you'll likely need to notch out some bits for panels to fit on.  I did.  Better planning might have alleviated this, but it was fairly straightforward and the structure is still strong enough.







Wednesday, November 18, 2015

Populating the panels

So all of my panels are dry and ready for switches, etc.  This is about as straightforward a step as you can get.  Simply put each switch / device into its opening.  All the arcade switches, toggle switches, and rotary potentiometers and switches have nuts to hold them in place.  Just use them, it's that simple.  You'll want to cut out the opening through the transparencies first, of course.  I've currently left all the transparencies intact over the 7 segment and bargraph LEDs.  I may leave them that way for a neater appearance, if it doesn't interfere with the LED appearance.

I've also noticed that as I glued the transparencies to the cutting board for the status lamps, the cutting boards were warping significantly.  We'll see if I can correct that as it cures.  If not, I'm going to leave the transparency across the opening intact and simply sandwich the cutting board and the label transparency between this outer transparency and the printed LED matrix holders.

Now is also a prime opportunity to switch the LEDs in the square arcade switches - these come as white LEDs, but you'll probably want to change the color.  I've settled on using a 5 mm diffused red LED in the master alarm and the abort switches.  The switches in the mission sequence panel I'm populating with 5 mm diffused blue LEDs.

Switching out these LEDs is pretty easy, but I'll step through the process.  First, remove the switch and LED from the back - just a slight turn and its out.

 Now pull the LED and its holder straight out from the switch.  

Now just straighten out the LED leads from the holder


And then push the old LED out through the front of the holder.

At this point, put the LED of your choice into the holder.

 Then wrap the LED leads around the end of the holder, like the old LED was.  It doesn't matter which way you orient the LED on the holder, but for best contact with the rest of the switch, keep the leads to the outside of the holder. 
Then put the led holder back into the switch.  At this point, you'll want to test the orientation of the LED - you'll need the LED holder put in the right way around for it to light up when you wire everything together, since LEDs only work with power running in one direction.  I chose to align things so the anode (the longer of the two wires on the LED) was connected with the lead on the red side of the switch - this matches the orientation that the round switches were shipped with, so it should work well for soldering things together.

And voila!  The switch now lights up a nice, blindingly bright blue...




Gluing on the transparency templates

As I mentioned previously, I used loctite spray adhesive to attach the transparencies.  A few tips as you do this:

1)  Try not to move the transparency around much after it touches the glue.  It can still smear a bit, but if you hold it still, this isn't an issue.

2)  Be certain to press the transparency under a heavy weight after application

3)  Wait overnight

4)  Then use a credit card or scraper to try to remove as many air bubbles / distortions as you can.  Believe it or not, but most of the glue isn't fully dry yet, but is more like a thick paste that can be moved with a lot of pressure to improve the uniformity of the panel appearance.

Now, for those final two panels - the mission sequence and the mission timer/display panels.  These two measure 6"x12" and 12"x12", respectively.  That's too large in both cases for using a single transparency sheet.  So you'll need to divide the print among 2 - 4 transparencies, depending on the panel.  This division can be done simply by importing the label image into an image editor, cutting out a section and pasting it in a new window and printing the new one and the old one after it has been cropped so the blank area is removed.  Just choose locations to divide the label that are logical - such as between letters/words/ buttons, not through them.  This isn't hard.  Now you get to match the multiple pieces together.  Because you'll have margins, you'll end up with transparencies that overlap to get the labels right.  Don't worry, though, there's a woodworking trick we can use here.




First, take the sheets, and line them up  so that the printing is properly aligned.


Then tape across the edges so the two pieces are held together, but be sure to leave the line where they match up as visible as possible




Now simply cut down that dividing line where the two panels are supposed to meet, being certain to cut through both transparencies in a single pass.  You don't have to be perfectly straight - you can even make a jigsaw design if you so desire.  The only important thing is to cut somewhere in the non-printed area between the edges (on the mission timer panel, you'll have to make sure to pass through the points where the dark boundaries around the cutouts meet, but other points don't matter).







Now remove the tape and dispose of the waste material from each transparency.  You'll have two perfectly matched pieces that have a single neat splice line where they meet


 Now glue these pieces onto the panel just as you did the others.  When you get to rubbing the bubbles out, you'll likely get a little bit of ooze at the joint that won't look terribly appealing - but this is easily removed with a little Goo Gone.  If you do this all carefully, those joints will be nearly invisible when everything is fully dried (not fully dried below)













Monday, November 16, 2015

Assorted Updates

So over the past few days, I finished cutting out all the mdf panels.  For the rectangular openings, I highly recommend getting a square file if you don't have one already - This simple cheap one from Harbor Freight worked well for me:

http://www.harborfreight.com/8-square-file.html

That will allow you to fine tune the fit of any rectangular pieces.

There's no real advice to give on this portion - just follow the templates that you glued onto the back of the mdf (remember, use the reverse image templates!) and when you get close, try to fit the component, and then file where necessary. Things don't have to be perfect, because you can use the transparency overlays to cover gaps.

When you're happy with the fit of everything, its time to prime and paint.  This is simple - choose your preferred color spray paint.   I'd recommend against metallic paints for uniformity issues and against the primer+paint in one can - those don't do a very good job on MDF in my experience.

Once the painting is done, give it a good solid day to cure at a minimum.  Then you can get to gluing the labels on.  I considered trying to use an inkjet on the laserjet transparencies to do an ink transfer - the ink won't absorb into a laserjet transparency, so if you press the panel into the transparency immediately after printing, it may transfer.  Of course any shifts will lead to smudges and a very undesirable appearance.  So I chose to stick with laserjet prints.

Now here comes another detail to pay attention to - If you choose transparencies for your labels, inkjet transparencies probably need to be printed on the exterior surface of the transparency to prevent smudging from the glue.  They might be a little less scratch prone than laserjet transparencies facing the same way, but moisture can still damage them more readily.  I chose to use laserjet with everything reverse printed.  That way, when glued on, the printing is actually underneath the transparency - things will look the right way then, but you won't have to worry about accidentally scratching the surface of the transparency and taking off the label.

That said, I found the ink can still smear some in the glue, so try not to move the transparency once you've applied it.  And before you even spread glue, double check the fit and alignment of the transparency.  Printers have a nasty tendency to change aspect ratios ever so slightly.

Now, for gluing them on - I tested three different glues.  Simple Elmer's glue stick, Cyanoacrylate glue, and Loctite spray adhesive.  Only the Loctite was labeled as acid-free, which was a bonus in its favor.  Glue stick simply had almost no holding power.  Cyanoacrylate instantly smeared all the prints, so don't even think about using it unless you print with the ink facing out (ie not reverse printed).  Loctite seems to hold ok - it doesn't end up with a perfect appearance, however - tiny imperfections in the surface show up with a somewhat mottled appearance and you can notice some streaks that formed in the paint (I'll put up pictures later).  I noticed the same exact appearance in Jeff's design, so if it is something you're ok with, by all means use this approach.  If not, you'll need to do more experimentation.  After applying the transparency (I didn't trim them first), I put them under stacks of heavy books to keep pressure on while drying.  I've finished all but a few of the panels, including the main screen and the mission sequence panel.  Those are all too large to fit on one transparency, so some splicing techniques will be used.  I'll try to grab pictures of that approach - gluing the transparencies on I simply didn't have enough hands to capture in process.  You'll also need to adjust the default print settings from Word or copy the images into another piece of software so they print in their entirety for these two panels.

Templates for printing transparencies

Wednesday, November 11, 2015

Building Status Lamp Panels

In the videos Jeff posted on his original project, he explains how he made some LED holders out of scrap aluminum and plexiglass, then lines the insides of each led compartment with reflective tape.  The result works just fine, but I didn't have scrap aluminum or plexiglass lying around.  Buying some could actually end up to be fairly pricey and then there was the labor in piecing it together.  Instead, I decided to put my 3D printer to work for this part.  If you don't have access to one, there is nothing wrong with Jeff's approach - it will work just as well.  You can also go to a website such as 3dhubs.com or Shapeways.  I've never used any of these services, so I can't vouch for their quality.  It would also be worthwhile checking with your local library for services - ours has 3D printers in their main branch that you can use at the cost of materials, which would make this a very affordable approach.

If you build it yourself, my cutting diagrams called for a 4.5" x 4.5" box.  You'll need to make two (unless you're modifying the design), each with 18 compartments for LEDs - 3 across, 6 high.  I've found 3/4" high is sufficient.

If you choose to print them, I've included two different part files.  The reason for this is that I found that a single 10mm green LED was pretty dim when I tested it next to the red ones I bought:


When I added a second green LED, I got much more uniform results.  The "ultrabright" green LEDs simply aren't that bright, it seems, or they didn't meet their spec.  I tried the 5mm ones I bought as well - I still needed two and the diffusion wasn't as good, so I stuck with the 10 mm ones.  So therefore, each side needs a different part design to account for the different number of green LEDs on each side.



(apologies for the poor cellphone camera work)

The print takes awhile, but its easy work.  I did have some significant warping on the part I printed last night - it will still work just fine, as the warping is entirely contained on the bottom of the part which will be below the control board surface, but it is less than desirable.  That's one of the frustrations of 3D printing - it isn't easy to know ahead of time what printing strategy will work well for each part.  Large parts with sharp corners, as this, can be nightmares in ABS.  If you're printing them, I might try PLA, which is much easier to work with.  I'm going to print the second part tonight, with a little more "ABS juice" (abs dissolved into acetone to make a glue to put on the build plate prior to printing).  I also added mousears to the print to remove the sharp corners and help stick the print to the plate.  These I'll just cut off after printing. 

First LED panel stl file
Second LED panel stl file

Wednesday, November 4, 2015

Cutting out the panels

It's been a few days, but I've been slowly making progress.  I've been working on cutting out all the openings in each panel for the switches and readouts.  This is rather tedious work.  I've got a drill press, so all round holes were easy.  The rectangular ones are slow going, though - I don't have a scrollsaw (what did I say earlier about not being afraid to do this without the best tools?), but I do have an oscillating multitool and files.  The multitool is simply too large to use for anything but the largest cutouts (around the status lights, the LCD screen and the 7 segment displays), so for those it is a matter of using the drill press to open up a hole and then a square and flat file to finish the opening.  If you choose this method, be aware that the multitool (or scrollsaw, for that matter) are not precision tools.  You'll want to leave the line in the pattern and file down to it, test fitting your pieces as you go.

For the round openings, here are the best drill sizes I had to make things work and fit well:

Holes for the #6 mounting screws  - 9/64"

10 mm LED mounting clips - 9/16"

Square and round arcade push buttons - 1"

Potentiometers - 17/64" pilot hole for the shaft, then countersink the backside of the panel with a 3/4" bit - the shafts simply weren't long enough without that countersink.  You'll need to go about halfway through the panel on the countersink.  Be careful not to go too far.

Pyro switches and all SPST/SPDT toggle switches - 15/32"

headphone and mic jacks - 1/4" followed by a 17/64" countersink most of the way through the panel from the back (if you go too far, the only harm done is a little extra clearance around the jack, which might not even be noticeable after you glue it in)

Rotary switch - 3/8" with a 1" countersink on the backside of the panel, just like the pots.

5 mm LEDs - 11/64"

So far I've managed to finish a fair number of rectangular cutouts, but I have quite a few more to go.  It isn't speedy work, by any stretch of the imagination, and with other projects going on, I'm only getting 3-4 done per night.