I pretty much don't write in this.
Slashdot videos: Now with more Slashdot!
We've improved Slashdot's video section; now you can view our video interviews, product close-ups and site visits with all the usual Slashdot options to comment, share, etc. No more walled garden! It's a work in progress -- we hope you'll check it out (Learn more about the recent updates).
Extremely long time since my last entry.
I've been busy, of course. Been trying to draw up a new mechanical scheme for the USB webcam mount, and have a lot of parts for the CNC:
(3) bipolar Crouzet stepper motors
(1) WinSystems LBC-586 single-board computer
(1) 510MB laptop hard disk and adapter for the SBC
(15) UDN2954 full-bridge PWM control chips
(1) long heat sink big enough for six of the power chips
(1) roto-zip style cutout tool...very high speed, we'll see if it works...was cheap
Several cables soldered up to test the single-board computer.
That SBC was a really good deal. I picked it up on eBay for about $40. This board is what controls John Carmack's Armadillo project. It has IDE, floppy, diskonchip sockets, ethernet, 4 serials, keyboard, vga, flat-panel, 48 dedicated I/O, a parallel port, and simm socket. Winsystems still sells those boards...for approximately $600. Some are still left on eBay. Pick up one up now.
I also got the full-bridge chips on eBay. A great find...they can handle up to two amps continous per coil, and 3.5 amps starting. I will need two chips per motor. The great thing is that I don't have to design a chopper circuit. When trying to run a stepper motor at high speeds, you can't just switch the rated voltage on the coil. The inductance won't allow the current to reach the rated value before you have to switch it off and move to the next step. This means that the faster it runs, the less power you get. When using a chopper circuit, the stepper coil is driven by a much higher voltage than rated. This give a very fast current risetime in the coil. When the rated current is reached, the control circuit shorts the coils, and the current recirculates until it falls below the rated current. Anyway, this is all included in the chip, so I can set the desired current with an external resistor, and the chip does the rest.
All I need now is a power supply and some minor logic for controlling the steppers. There is a retrofit CNC endmill at work, which no one ever uses. I'll try to use that to make some of the parts I'll need.
Otherwise, work is boring, and the job market is gloomy. Engineers just aren't cool anymore.
I'm watching the stock market with interest, and have been for the last year or two.
It feels like once the Dow cracks 10,000 again, I'll get a job. Maybe that's illogical. But until then, I'll keep punching out the resumes and working my temp job.
My new workbench is done, and currently the paint is drying. I have at least one more coat of black paint to put on, then I'm overcoating with polyurethane. Hopefully that will leave a nice, durable surface. It's been pretty cold, so everything needs extra time to dry.
No progress on any EE stuff since I've been building the workbench in the evenings. Once the final coat is on, I'll finish soldering the USB cam and find more parts for the CNC workcell.
I wish I had thought of this myself.
Currently, I'm rebuilding my senior project. I had boards made, and every evening, I solder on yet another fine-pitch SMD. All kinds of fun, you can imagine. I've heard rumors that surface-mount soldering can and has been done with solder paste and a toaster oven. They don't even bother painting the solder paste on each little pad, they just run a bead along the pads, squish on the part, when the solder melts it beads up and separates onto the leads and even centers the part with surface tension.
However, I'm stuck with a soldering iron, a length of wick, and my patience.
Anyway, back to the money-making idea. I was playing with my little servo camera, and Steve (guy who works next to me) suggested that I could make it available as a kit.
This makes a lot of sense. While my USB-controlled, integrated hub pan/tilt mount would require some decent manufacturing outlay to get going, a kit form of this other unit would be pretty neat.
I would have to do a little redesign...I don't expect everyone to tear apart their webcams to integrate in, and I don't expect everyone's webcam guts to be the same size and shape. I can design a C-mount screw and provide holes for custom mounts.
So...the kit would include:
(3) Cut and bent lucite parts
(2) R/C micro servos
(1) etched circuit board
(1) Socketed microcontroller and support parts
(1) Serial interface cable
(1) USB passthrough cable (maybe)
(1) CD-ROM containing source for microcontroller firmware and control application
The USB cable I'm iffy on. That would be a blatant disregard for USB standards, since the power using components don't report their existence. USB powering is pretty slick, but anything I'd sell probably would require a wall wart. Wouldn't need people calling up mad because their USB port got fried from a bad solder joint.
This is a pretty interesting idea, I will be seriously looking into it. Once my CNC machine is up, it would make production of the plastic pieces very smooth.
A robot materialized out of my junk box on Sunday afternoon, while I was wondering what to do with the guts of my old webcam.
I found two old micro R/C servos, grabbed a spare PIC16F84, and a sheet of 1/8" lucite. Added a few discrete components, thermoformed the lucite with a blowtorch, and it became a pan/tilt unit with integrated webcam.
It is controlled over a serial port, and the USB port connects to the camera and provides power for the camera, servos, and PIC.
It's pretty small, about 3" square and 4" tall. The servos provide around 180 degrees motion in pan and tilt. I wrote a control app last evening, it sends commands over a user-selected COM port, allows the user to control pan and tilt, adjust the motion increment, and send the camera to a centered position. I'll add the ability to type in "bookmarks" and assign them to servo coordinates. That way I can program all surrounding coworker's desks for quick access (as well as the door to Engineering).
I'll have to take some photos and maybe work up a little construction walkthrough. It was pretty simple to make, and with the lucite bent to form all of the parts (no gluing required!) it does look sleek.
Almost over the cold. Hope it's the last one this year...yeah right
I'm sick. Sore throat, fever, chills, the works. I paid $8.64 at the gas station for a pack of DayQuil, hope it works. I have a meeting later today to investigate guarding for the hole punch units on the presses.
Can't complain much, but some electrical engineering would be nice. AutoCAD is getting old. At least I get some decent freedom in design; we have an application that requires air-cylinder-activated rollers. The wheels were made of custom cut rubber on custom turned steel cores. I'm going to use inline skate wheels instead. A lot cheaper, available everywhere, should save maybe $60 per wheel, which is considerable when you're talking about buying 400.
Fried a robot the other night, I had some new motor control hardware soldered in, and there must have been a short. The PIC got pretty toasty. That board was too old anyway, I have enough parts to rebuild it from scratch. I'll separate the boards too, one dedicated motor control board and one processor board. I have nice setup for motor control, a GAL22V10 monitors three inputs per motor. One input is for PWM, the other two control direction. The GAL makes sure the H-bridge doesn't burn itself up by turning on two transistor on the same side of the bridge. When both inputs are active, the two bottom transistors are activated for regenerative braking. Using a GAL might be overkill, but it means I can reprogram the control scheme whenever necessary.
My eBay stepper motors are on the way, so pretty soon my CNC machine will be in development. Servos might be a better way to run a CNC machine, but I don't plan to have accurate position encoding yet.
Anyway, back to the grindstone.
I'm currently looking for an engineering position. Hopefully, the stock market's climb will have positive results down the road.
Until then, I'm just working on a few projects. I'm rebuilding my senior project (USB controlled robotic webcam mount with integrated USB hub), working on a couple of robots, and gathering parts for a 36"x24" work area CNC machine.
Should be fun.