Air Sampler
 
      Main           Construction           Operation           Schematics & Code           Pollen      Other Projects  







 


The idea is simple: a timer-controlled fan would produce constant airflow over a sticky tape and hopefully will make some of the airborne particles stick; attach the tape on a slide, pop it under the microscope and you are in business.

One of the great pleasures in starting a new project is going on a shopping spree in the junk electronics store. It is so relaxing to spend a couple of hours digging through piles of discarded electronic stuff and looking for just the right box for a project!

My initial requirements were simple - any box that's big enough to fit a small CPU fan and some electronics would do. Then, I found this beauty and fell in love:

  

I don't know what it is(*) - if you can figure it out, please send me a note. Whatever it is, it already has a display that I can use, six buttons (exactly as many as needed), and something not to forget - a price tag of three dollars. From RadioShack I can't buy a bag of resistors for that much!

(*) Update: I gave up trying to see what the original purpose of this device was. Then I got an e-mail from Keith Ward, who told me it is a point-of-use gas detector - similar to this one. Quite a coincidence - I had unwittingly converted one air analyzer to another air analyzer! Keith, thanks for the info!

     

Getting back home my first job was to see what's inside. It's a bit like opening a Christmas present - you're never sure what's in the box. This time I found two PCBs, one on top of another, and below them - that really cool LCD display module with backlight and all bells and whistles:

     

     

Time for some hardware work - I had to drill a hole opposite to the funnel (the "funnel" being the silvery cylinder where that weird cable used to be), and to make a slit for the sticky tape holder:

     

     

That thing done, I had to make some tough decisions. Using the original graphical LCD display - the one with the bells and whistles, - to show just a couple of digits would be a real shame! Luckily, I happen to have on hand a simpler text-only LCD module (prize from a previous raid on the junk store). It turned out to be a bit of a disappointment - it didn't fit the window...

     

The final option was a frequency indicator from an old PC - two green LED digits. Remember those? Used to show "33" that turned to "66" when you press the "Turbo" button, without this having any effect whatsoever on the actual CPU frequency. Ah, those were the days...

The first tricky part about using that LED display was reverse-engineering the schematics of the board on which the display sits, and figuring out a way to use it without unsoldering the LEDs. The second tricky part was explaining to my wife what am I doing for two hours straight looking at a tiny gadget with a 10x watchmaker magnifying glass stuck in the eye, instead of coming to dinner. The third tricky part was sitting through dinner while thinking how in the world the LEDs can be securely attached in position behind the display window.

The solution I finally came up with involves a serious "upgrade" of the original PCB to a LED-capable-version; an upgrade performed in a really professional manner with a hacksaw and a glue gun:

     

The main thing the board does after the upgrade is to support the LEDs and hold them in the exactly right place, a couple of millimeters behind the display window.

There are two other things I decided to use from the board. One is the small LED that would show on the front console if the fan is running - you can see it on the first picture above, to the right of the 7-segment display. The other component that is still in use is the blue connector to the left of the LED displays on the second picture. This one is intended to plug into the buttons embedded into the front panel - it can still do its job, with the help of a small additional board with a few resistors.

In the extremely unlikely event that you are the person who originally designed this thing, please know that: (a) I deeply respect you and your work, and (b) I totally hate you for making the keyboard cable that short. You try plugging it back in the connector after taking the PCB out!

Anyway, this is the resistor board that makes the keyboard work after the "upgrade":

     

The PIC sits on a small and quite cramped board, keeping company with the power supply and the LED & fan power control electronics. I've left a tiny amount of unused space in one end, just in case I'm forgetting something:

     

You can imagine my elation when after a couple of hours I got the LED display working, and the bitter disappointment a minute and a half later when I plugged it in the wrong place and half of the segments burned out. (Did you know you can make a green LED glow yellow - for a short while?)

Replacing the LED display with a new one took some time, but at the end I was happy with the results. In fact, now it looks better than the old display. To improve the things even further, I attached a small piece of green plastic underneath the front panel window and painted parts of the board black, so they won't be visible through the window:

     

With the fan, the PIC board, and all those wires, the box will be quite cramped. Opening it and closing it up every time I need to reprogram the microcontroller would be very inconvenient. For that reason, one thing that I do for most of my projects is to install external connection through which I can program the PIC without ever opening the box. A floppy cable comes pretty handy and looks neat from outside:

     

The only problem is remembering which way the programmer should plug in. That's what the red dot does - it marks pin 1.

Everything fits with just enough space in the box. It is a surprisingly tight fit - I was expecting that I'll have more room available. The good side of a tight fit is that there is no need to find ways to affix anything - there is just no place for things to rattle around the box.

     

Note the metal plate hiding the PIC on the second picture. This is an improvised heatsink that, as it turned out, works surprisingly well to keep the voltage regulators cool. At least, they haven't burned out yet.

Finally everything is ready for the first real test. Usually this is the time when I carefully tighten every last screw, turn the power on, and watch the screen go alive for half a second before faint wisps of smoke start coming out of the box.

No smoke this time. Phew!

After fixing a couple of bugs in the code, the air sampler is fully functional. Below you can see the holder for the scotch tape, and the device in action:

     

Time to go out and start collecting pollen. You can check the next pages for details on the air sampler's schematics and code, sampling procedure, and could also enjoy a few pictures of what got caught on the tape.




      Main           Construction           Operation           Schematics & Code           Pollen      Other Projects