And its been sitting in the bottom of my junk drawer since last summer ...
Anyway this is hardly a new idea, but most of the time when people make these out of VCR heads they end up with a bunch of junk glued, taped, screwed, and wired to the outside of the head, making for a functional but unattractive device. I wanted mine to be as nice looking as possible (considering my limited tools and workspace) and for me that meant one thing, all the junk needs to be in the tiny space inside the VCR head
Okay then why do you want one? what does it do? Well it functions just like the wheel on your mouse, except its larger and has momentum which is good for scrolling through long web pages or piles of source code, or heck a large document.
Its also good if you do any media editing as you can scroll around in large video or audio files effortlessly, with the heaviness of the VCR head you can just set it in motion and its inertia will keep it spinning for quite a while, when you come up on something you want to work or read you just stop it with your finger.
Is it for everybody? No not really, but if you constantly wheeling the mouse in some application (and forgot the Home, End, Page Up and Page Down, and cursor keys exist) you might like it
Step 1: Parts and Tools
- A scavenged VCR tape head and its mounting bracket
- A USB wheel mouse, the chip from inside the mouse must fit inside the VCR head, mine was a cheap 5$ Microsoft basic optical mouse, and the whole thing ran from 1 central chip where the camera was located (plus I didn't really like this mouse much anyway, its too light)
- Blank copper clad PCB
- 1.75 to 2 inch machine screws (3 in my case to mount the bracket to the base)
- Some washers, mainly for looks
- 0.75 inch or 1 inch hollow spacers (I used some 0.25 inch wide 0.75 inch tall hollow hex spacers though some metal tubing would do fine also)
- Paint of your choice
- Resin of your choice (if you want to fill in the gap on the top of the head, I used black polyester resin, epoxy mixed with some model paint would work too)
- A chunk of wood for the base
- 1* LM393 (dip 8 probably would work, the one I scavenged was in a SOIC 8 package and it worked great)
- 2* 150 ohm resistor (all resistors used were standard 1/4 watt radio shack variety)
- 2* 470k ohm resistor
- 2* 33k ohm resistor
- 4* 10k ohm resistor
- 2* 0.1uf capacitor
- 2* infrared LED
- 2* infrared photo-transistors
- 2* 330 ohm resistor
- 1* 1uf capacitor
- 1* "mouse chip" (mine was marked with house numbers)
- 1* "mystery cap" (I think its a capacitor the code on it comes up to 24 picofarad, though it also has WS printed on it which is odd)
The usual hacking stuff ... though I used a few power tools at work, you don't have to but its so much faster and easier
- Soldering iron and usual soldering accessories
- Wire cutters, needle nose pliers (I have this craftsman set, though mine are older)
- Flathead screw driver
- Screw driver or hex wrench to take your VCR head apart (mine used hex head screws)
- Hole saw
- Belt sander (actually the right angle disk sander on the side)
- Drill and bits
- Sandpaper and abrasive pads
- Latex gloves
- PCB Enchant (and in my case a old pickle jar)
- Digital Calipers (are a huge plus, though a ruler would work)
- Pencil / Paper
- Hobby Knife / scalpel / Xacto (and a box cutter / utility knife if you want)
- Rubbing alcohol
- Multimeter (and an oscilloscope is nice but not necessary, I use this sub 20$ craftsman meter)
- A new Sharpie or similar permanent marker with a "fine point"
Step 2: Gut the VCR Head
Inside the head you will see 2 cast ferrite disks with magnet wire loops in them, this is how the head connects the spinning heads to the rest of the unit, and 2 loops act as a motor drive, its all done with magnetics and quite ingenious. Take a flat head screw driver and simply pop them out, mine was held in by a minimal mount of adhesive, though the bottom disk required a couple whacks with a hammer since it sat flat against the base of the head assembly.
Once those are removed you can see the actual magnetic heads, mine were held in place by these odd shaped star head brass screws. I tried to use a small flat head on them but they would not budge, so I went in from the side with a large flat head and hammer, bending the heads up and away from the body so I could cut them off with a hacksaw. It made a bit of a mess on the inside, but its on the inside so no big whoop.
After every thing is removed you can set the 2 halves of the VCR head aside for now, just keep track of the brass base and its set screws (zip lock type bags work well for this)
Step 3: Making the Encoder Wheel
Reflective encoders shine light up onto a disk, which is printed with a special pattern, directly below that disk is a pair of photo-transistors, if one goes dark before the other the wheel is moving in one direction.
There is a website you can goto that will make all kinds of disks for you, in any shape and form you want ...
Which is awesome, but it has a couple draw backs
- It uses pixels as its unit of measurement, and I don't know about you but I do not have a measure that has a pixels unit on it
- The image it produces is quite rough on the edges which causes a little bit of jitter
You can see a full size (1280x1024) screen shot of my final settings here
To solve issue number 2, I simply outlined the rough edges with a sharpie marker, finally I cut out the encoder wheel and super glued it to the top half of the VCR heads inside surface.
Step 4: Gut the Mouse
Once you know how the thing is hooked up you can desolder all the parts on the mouse, and sort out what you need from the lot, this includes ...
- The main mouse chip / camera
- The 24pf cap that is connected to an area of the pc marked for a crystal clock (hence the X on my drawing)
- One of the 1uf capacitors (though I show all three, was not sure what was needed at the time)
- And the USB cord and pcb connector
Step 5: Breadboarding
- Make sure I understand the circuit
- Figure out the minimum hook requirements of the house branded (and therefore no datasheet) mouse chip
- And to figure out how the ir part of this is going to function
Next I hooked up the mouse chip with just an mechanical encoder to see how much stuff I could leave out from the design notes from the mouse, it ended up saving me 2 fairly large (in this amount of space) electrolytic capacitors.
Then I hooked up one of the IR LED's and a photo-transistor to see how that all worked and how to optimize the signal as being reflected. I had to black out the clear sides of the photo-transistor, as the IR led was going to be sitting right next to the thing and though it worked before, it was much more responsive with only the top being exposed.
On a side tangent, stupid radio shack! on the back of the IR led / photo-transistor it states the max continuous current of the led was 150ma! no way! On their website it states 40ma which is much more in line from other parts with the same spec's, I ended up driving them at 30ma just cause I happen to find 150 ohm resistors first in my parts bin.
Next the LM393 is a dual voltage comparator, you give it a reference voltage (in this case 2.5 volts though a resistor divider) and anything above that voltage is a logic high (in this case 5 volts) and anything below it is a logic low (darn near 0 in this case)
You need the comparator because the photo-transistor never hits below about 1.4 volts due to reflection of light + ambient light + printer ink does not absorb 100% of the IR light hitting it. It also never goes above about 3.4 volts due to similar issues + the resistor needed on the collector side of it to get the thing working.
Also the photo transistor is acting more like an amplifier than a switch. With fairly erratic output that shows up on a scope as a squiggly line at all sorts of voltages between the 2 extremes (so much so I could never get my scope to store an image of it), versus the square signal needed by the mouse chip. Simply put we need something to make the "yes this is on or no this is off" decision and the 393 is specifically built for the task
Finally with a working model on the bread board I drew up a "proper" schematic (full size image here)
I just noticed on the schematic I labeled the decoupling capacitors as 0.01 uf, it should be 0.1 uf, it is correct on the side list of parts, my bad!
Step 6: Prep the Parts for SMD Mounting
bending the leads is easy just use your needle nose pliers and bend them down then out at 90 degree angles.
The clearance issues were on the mouse chip being just slightly too tall (keeping in mind the thickness of the pcb) and was solved by taking a sharp file and filing one end down on a 45 degree (ish) angle, but don't go too deep! The second issue was the photo-transistors being slightly too tall as they have to sit at a 90 degree angle to the encoder wheel, I simply sawed off the domed part of the package and smoothed down with a file.
Yes you could just use SMD parts, but I did not want to buy a handful of components and already had this on hand, and it really did not take a bunch of time or effort.
Step 7: Make a PCB
Next I drilled a hole in the center point, through the PCB and into a piece of wood, un-chucked the bit and used it as an axle. While holding the edge of the pcb to the right angle disk sander I slowly turned the pcb getting a pretty decent round shape. For the inside hole I used a hand drill and a hole saw which was just a bit larger but fine.
Once I got home I traced it onto a sheet of paper, held the mouse chip down and made little dots where the ends of the leads were and drew little lines for the pads, then I did the same with the cord connector, then the LM393. Next I started connecting it all with lines on the paper adding pads for passive components where needed based on my schematic.
Pretty quickly I had everything in place and a very good idea (but not a final version) of how everything needed to be laid out, I then took my sharpie marker and traced over it, and circled potential problem areas due to the difference in thickness between a .7 pencil lead, and a felt tip marker.
From this point until the board is etched you should not touch it with bare fingers, so put on your rubber gloves!
Next its time to clean the board, using rubbing alcohol to wash off finger prints, and then a nylon abrasive pad to remove oxidation.
Using my pencil + sharpie test pattern I repeated the process on the board using just the marker, making minor changes in the problem areas where appropriate (sometimes its just where the marker ink soaked in the paper). Mistakes can be removed with rubbing alcohol, or with a hobby knife scraping the dried ink off the area
Once satisfied (as much as you can be with a left hander drawing a pcb with a marker) its time to etch, I placed just enough enchant in a old pickle jar to cover the board. I also microwaved a bowl of water to just under boiling (2:30 in my microwave), and then set the pickle jar in the water for about 10 min, to heat the enchant up (never heat directly). Removing the jar from the water and wiping the bottom off its now time to place the pcb in the enchant, once in I swirled it around nearly constantly, and the whole process took no time at all.
The board was removed and washed then inspected (the leftover enchant is still in the pickle jar, as it still has a few more small boards worth of power in it). Sharpie marker is not the worlds greatest etch resist but it does the trick, unfortunately I did have 1 little tiny break in a trace where it was not thick enough, and that has to be fixed (in this case with a blob of solder)
Step 8: Populate the PCB
I started by soldering the 4 places that needed jumper wires (well 3 I forgot one at first) and then soldered the chips, and so on
I did make a mistake on the PCB, there needs to be 2 330 ohm resistors in between the USB data lines of the chip and the cord, I did not leave a gap for them in the trace, but its nothing that a sharp knife and a little bit of scraping cant fix
Once done inspect your work visually, check for shorts and continuity before plugging it in, once you are sure that all is well you can put the board in place and start wrapping up. On mine I had to feed the usb cord into the hole on the side first, plug it into the board then place the board in the base. It is held in place with a bead of super glue.
You should be able to test it out, but don't button it up quite yet as there is still stuff to be done with the top half
Step 9: Paint and Assembly
I grabbed a decent looking piece of plywood, cut it to size and sanded it. Next I held the original head mounting bracket to the wood and marked where its mounting holes were located.
I drilled the 3 holes just slightly larger than the screws diameter, flipped the board over and drilled halfway though with a bit slightly larger than the nut that I would be using so there is nothing poking out of the base and everything gets nice and secure.
Once done with the drilling and sanding of everything I used the paint booth at my work to paint the metal parts black, the wood base got a coat of sanding sealer, was re sanded again, then sprayed down with lacquer that had a cherry tint mixed in with it.
The space on the top of the VCR head tween the edge and the center hub was begging for something different (and boy I got it). Since the encoder wheel is glued to the underside its a little bit liquid tight, but not enough for polyester resin, this stuff makes water look thick! So I put a few drops of super glue down each hole and let dry overnight, once the holes were sealed I poured the resin. (sorry no pictures it had to happen very fast, but its similar to all other resin stuff, mix goop with hardener pour and pray)
At first the polyester resin was awesome looking, then the hardener started kicking in, it causes heat which goofs up the surface and a few hours later I had a lumpy top. I should have left it, you can barley tell in the pictures, but no I had to go and touch it with sandpaper ...
3 SOLID hours of sanding later, my arm was dead and my finger is still sore the next day, but it looks much nicer, had a silky smooth touch. Its not perfect, there are still a couple small dips in it, but if you care about it that much, you can sand it (that polyester is some hard stuff!)
Of course all that sanding ate up the metal in multiple directions, so I put the top in my drill and held it to my abrasive pad to make a uniform(ish) satin finish.
Assembly is pretty easy, put the VCR head back together, install the brass bottom and tighten up the set screws. Screw the head back to its bracket, use the washers, spacers and long screws to mount the bracket to the base, and your done!