Applefritter - Hardware

Apple USB keyboard repair

Fri, 03 Nov 2006 12:55:25 -0800

I recently obtained a non-functional but nearly new Apple USB keyboard. This is the most recent model, with a transparent base. [thumb:19401] What's appealing to me about this keyboard is the small desktop footprint, yet it has full-size keys - unlike the compact keyboard that shipped with the original iMac. After cleaning what looked - and smelled - like a coffee spill, I found the keyboard worked fine with the exception of the left control key. I decided to see if it could be repaired... The keyboard is held into the transparent case by three very small hex screws. Not having the proper tool on hand, I found a jeweller's screwdriver, sides slightly filed down, was able to remove the screws. There are two ribbon cables that hold the keyboard to the case at this point: [thumb:19399] They pull out by holding the clear plastic "wings" on each cable, and go back in the same way. No clips to remove, etc. [thumb:19397] Looking at the metal backing of the keyboard, there are several dozen screws. Yep, you've got to remove them. A Phillips jeweller's screwdriver fits perfectly. A larger driver may have trouble. There are three raised screws near the top centre. These do not need to be removed. Two of the screws will drop a piece of plastic that sits above the cursor keys. Lift the metal plate off the keyboard. You will see layered sheets of plastic. This is our goal. Carefully lift it off the keys. You don't want to send the little rubber grommets underneath flying. [thumb:19400] If your keyboard has been a spill victim, some may be stuck to the plastic. Carefully remove them. The plastic sheets are the keyboard switches. A top and bottom layer are covered in conductive printed paths to each key, while a middle layer, with holes at each key position, provides an air space between conductive pads. When a key is pressed, the rubber grommet presses down on the top layer, and it is pushed down to contact the bottom layer, completing the circuit. This is a very clever, or very cheap, design. It's certainly prone to damage from spills. [thumb:13398] I tried cleaning the contacts at left control key position, as they were blackened...and smelled like coffee. After continuity testing with a multi-meter, I found the bottom layer had corroded, and the contact pad was no longer connected to the rest of the circuit. The proper fix would be to use a conductive paint, like those used for repairing automotive rear-window defrosters. Having a desire to not spend any money on this, I tried something free. Move the top and middle layers of plastic out of the way. Do not completely separate the three layers. I scraped away some of the green covering on the circuit path near the control key pad. Careful, as it doesn't take much to scrape away the conductive path as well. The orange in the photo below represents where the original circuit path was. [thumb:19395] I cut a piece of tinfoil to shape, and with a toothpick, drew lines of superglue on either side [b]but not on[/b] the exposed circuit path, as well as over the control key pad. I placed the foil overtop, and clamped it to my desk. [thumb:19396] The clamp was centred over the scraped circuit path, as I want the foil to be glued in contact with it. I left this to sit overnight. I checked the connection by placing one multi-meter pin on the foil, and the other on the next available pad in the circuit. There should be little to no resistance measured. Assembly is the reverse of the steps above. Line up the plastic on the keyboard grommets as best you can. Make sure all the grommets are sitting properly before placing the metal plate over the plastic sheets. I pressed the plate down and held it while I screwed in the first few screws. I put several middle screws in first, then worked my way from right to left. The three hex screws back in place, I plugged the keyboard back in, and it works fine. Using keycaps in OS 9, it looks as though a gentle press of the previously bad control key results in several momentary contacts, but under normal usage works without any ill effect. As stated above, conductive paint would be a much more reliable solution, but the instructions above will hopefully help regardless of which method you choose. Apple keyboards are no longer the expensive peripherals they used to be, and many would rather buy a new keyboard than bother with an attempt at repair (which is how I received the keyboard in the first place). Part of the appeal of this project is the joy of getting something for nothing. It's also saved one item from going into landfill, so for someone like myself who enjoys tinkering, the project was very worthwhile. Disassembly will differ, but the black and white versions of the Apple USB Extended Keyboard use the same layered sheet for switches.

replica 1

Fri, 03 Nov 2006 12:52:23 -0800

[image:6731 absmiddle] The Briel Computers replica 1 is a fully functional clone of the Apple 1© computer created by Steve Wozniak in 1976 that started Apple Computers©. The replica 1 is available as a kit that you solder all the components onto the circuit board, or preassembled. If you have any questions feel free to email me anytime at [email]vbriel@yahoo.com[/email] A new circuit layout was done to implement a more simple circuit and replace older no longer available IC’s. This new design allows for a smaller board yet still gives all the functionality of the original design. Just like the Apple 1, the keyboard and power supply are not included. Simply add your own PC 386-Pentium© class power supply, a composite TV or monitor and a PS/2 keyboard or Apple II© style ASCII keyboard and you are set. [thumb:4859] replica 1 revision c [b]The new replica 1 SE is now on sale.[/b] Please visit briel computers for more info at [url]http://www.brielcomputers.com[/url] [thumb:6732] The serial I/O board is a unique design made to work with both the replica I and the Apple 1. It is a piggy-back board that goes in the 6821 socket and is transparent to the computer. With this design anything that goes to the video display is sent to the serial I/O board and anything that comes in through the serial I/O is treated like a keystroke. This method reduces the speed of the I/O board to 2400bps but gives the advantage of being able to save BASIC programs as a text file. No setting switches or running software required. Just install it, set up Hyper Terminal, connector your serial cable (not included) and your all set. [b] EDUCATIONAL VOLUME DISCOUNTS[/b] If you want to use the replica 1 as an educational tool, email me at [email]vbriel@yahoo.com[/email] and we can discuss volume educational discounts. [b]SYSTEM INFORMATION[/B] [thumb:6734]____________________[thumb:6733(middle)]____________________[thumb:6735] [b]PICTURES[/b] [thumb:6737(left)]This is the first Replica 1 prototype in its early stage on breadboard. This is when I first got a working version. At this point I had no video display. All I could see was that there were 7 ASCII bits on the PIA. Reading the bits I decoded the binary to ASCII and realized it was trying to output a ‘\’ backslash character. [thumb:6738(left)]This is the breadboard version of the Replica I. At this stage I have hooked up a Basic Stamp microcontroller board to the video out section so that I could display text through a terminal or a computer with a terminal program like HyperTerminal. I still have this board but no IC's are currently installed. I'm hoping to get it up and running again just for fun someday soon. [thumb:6739(left)]The final prototype board. This board has the video circuit implemented displaying 40 X 24 text in the same format that Woz’s terminal section did. At this stage of my prototyping I didn't have the ps/2 interface circuit done so it was not on this or any prototype. I actually added the design of the ps/2 interface with the prototype to make the first revision a boards. Please visit the Applefritter forums under Apple 1 for the latest news and events about the replica 1, apple 1 and other related material. I hope you enjoy the replica as much as I enjoyed creating it. Vince Briel Briel Computers

Building a Case for the Replica I

Tue, 22 Feb 2005 07:35:09 -0800

By Larry Nelson [center] As a person with multiple hobbies, I finally found a way to combine at least two of my part-time interests: namely, woodworking and the Replica-1 computer by Vince Briel. After buying the Replica, and adding the serial board, I made a box to put my Apple-1 Replica into. A few months later, Vince came through for the whole group of Replica enthusiasts with authentic ASCII keyboards. My old Replica case wouldn’t take the keyboard I bought from Vince, though, and I decided to build a case that would comfortably hold the new (old) keyboaof my system. Start out with a base of MDF (medium-density-fiberboard) cut to 18 1�?�4 inches by 14-3/8 inches. If 1�?�2 inch plywood is available, that could be used just as well. (The MDF was cheaper at the Home Depot.) Then make two sides from 15 X 5 1�?�2 X 3�?�4-inch pieces of walnut. These boards I cut on a band saw to the shape shown in the drawing. Lacking a band saw, one could cut the boards to shape with a handsaw, a scroll saw or a jigsaw. Carefully sand the rough-sawn edges smooth. Route a bull-nose on the two sides, using a 1�?�4-inch round-over bit. See the pictures for the placement of the round-overs. Don’t route the inside of the sides on the 1inch vertical front edge or the inside of the back edge. If you lack a router and round-over bit, use a wood rasp to round over the edges. [center] Now add a 1/2-inch rabbet to the bottom edge of each side piece. The rabbet is 3/8-inch wide (half the thickness of the side) and 1�?�2-inch deep for the base to fit into. Rabbet the back edge of the sides 3/8-inch by 1�?�4-inch to inset the back into the sides. Remember, the two sides are mirror images of each other. Be careful to correctly orient the two sides before rabbeting the edges, or you’ll wind up with two right or two left pieces. Don’t ask me why I bring this up, but my scrap-box seems to have a spare side in it. [center] The front edge is a piece of walnut wood 19 1�?�4 X 1 1�?�4 X 3�?�4. Shape it as shown in the drawing. Use a rasp to round over the outer edge of the front edge piece. The 15 o bevel on this piece will not properly shape with a round-over bit in a router. Notch the front edge piece on both ends as shown in the drawing. At the top of the back, add a piece of wood 18 1�?�2 X 2 X 3�?�4 to hold the hinges for the top. A 1�?�4-inch by 3/8-inch rabbet on the back side of this piece allows the back to fit flush. [center] The back of the case is a piece of 1�?�4-inch birch plywood, cut to size to fit into the rabbets. Before fastening the back in place, test fit the power supply and Replica circuit board in place. Mark the back to cut out access for the power cord, the PS fan, the monitor connection and the serial plug. Since placement of these holes may vary, I am not providing dimensions. After carefully marking the necessary holes in the back, drill and cut as necessary. There are three more pieces of wood needed to complete the case. Since I have a planer, I planed the wood for these three pieces to 1�?�2 inch thickness. If allowance for the difference in thickness is made, the top and front cross piece could be 3�?�4 inch thick. The key board however should not be thicker than 1�?�2-inch since the keys of the ASCII keyboard do not extend through 3�?�4-inch boards far enough. An alternate key board could be 1�?�2 thick plywood. [center] The top is 17 1�?�2 inches by 7 inches. The key board is 17 1�?�2 by 7 1/2 inches. These two pieces are maple in the original and are made by edge-gluing oversize boards together until you have sufficient width, then cutting the pieces to the correct width and length on the table saw. The vertical cross piece between the top and the keyboard is also 17 1�?�2 inches long. I started with a width for this board of 3 inches, but trimmed it to 2 1�?�4-inches after experimenting with the fit. The angles involved make calculations hard to do. Into this board you will mount the power (on-off) switch, the RESET switch, and an LED for power indication. [center] You will have to drill three holes centered in the board for these switches and light. First mark the location of the three holes, drill 3�?�4 inch holes to within 3/16-inch of the front face, then switch to the finish bit size to complete the holes. This will allow the locknuts to be recessed into the vertical cross piece. Drill a 1/4 inch hole, 1/2 inch deep in the center of each end of the top back piece. Drill matching 1/4 holes, 3/8 inch deep in the sides. We will use two pieces of 1/4 dowel, 7/8 inch long, to hold the top back piece in place. Dry assemble the base, the two sides, the front edge and the top back piece to check for proper fit. Then glue the five pieces, clamping the assembly and checking for square in vertical and horizontal directions. Set the assembly aside. Round the top of the front edge of the top piece using your 1�?�4 inch round-over bit or wood rasp. [center] Cut two triangle pieces of 1�?�2inch scrap wood or plywood into triangles with a 15 o angle. Glue and clamp them onto the inside of the two sides of the assembly. These triangles go to the front of the cabinet and support the key board. The key board is not fastened to the case itself but just rests on the triangles. I found that the best way to cut the key board inlet was to make a cardboard template, test-fitting it over the ASCII keyboard until satisfied with the fit. Center the cardboard template on your case key board, draw the outline, and carefully cut out the pattern with a scroll saw or jigsaw. Smooth the cut edges with a rasp and sandpaper, test-fitting the key board onto the ASCII keyboard until it fits with enough clearance to allow all the keys to move freely. When satisfied with the fit, set the key board aside. Don’t attach the ASCII keyboard until after you have applied your finish. [center] The vertical cross board attaches to the top piece with 4 - #6 X 1-inch screws, glue and the support piece cut as shown in the drawing. If you have cut it to the proper width, the vertical cross board will contact the key board with its bottom edge and hold the key board in place. Now add hinges to the underside of the top piece to attach it to the top back piece. I inset the hinges, using a chisel to cut the inlets, but this is not necessary. With all the pieces finally done, you are ready for sanding and finishing. Sand away all scratches and rough edges with 80 grit sandpaper. Then progress through 120, 180, 240, 320 and 400 grit sandpaper until the surfaces are as smooth as possible. Clean off all dust and fingerprints with mineral spirits. Allow the surface to thoroughly dry, and then rub on two or more coats of Danish Oil or other finishing oil. Allow the oil finish to dry for 24 hours. Assemble the ASCII keyboard by screwing it onto the back of the finished key board. Mount the switches, LED, power supply and circuit board as previously explained. Connect all the wiring. This is not a wiring manual. If you have questions about wiring the Replica, refer to the instruction manual. Contact Vince Briel if necessary. Add four feet to the bottom of the case. I used stick-on felt pads available at the local hardware store. Fasten the back to the case with small screws. Plug it in. Load Microchess. Have Fun. [b]Materials:[/b] [list][*]Back: 18 1�?�4 X 4 1�?�2 -- 1�?�4-inch Birch Plywood [*]Base: 14 3/8 X 18 1�?�4 -- 1�?�2-inch MDF or 1�?�2-inch plywood [*]Front: 18 1�?�4 X 1 -- 3�?�4-inch Walnut [*]Sides: (2) 15 X 5 1�?�2 -- 3�?�4-inch Walnut [*]Key Board: 17 1�?�2 X 7 1�?�2 -- 1�?�2-inch Maple [*]Top: 17 1�?�2 X 7 – 1�?�2-inch Maple [*]Vertical Cross: 17 1�?�2 X 3 – 1�?�2-inch Walnut [*]Back Top: 17 1�?�2 X 2 – 3�?�4-inch Walnut [*]Triangles: (2) 6 1�?�2 X 2 – 1�?�2-inch MDF or 1�?�2-inch plywood [*]Hinges: 1pr. 2-inch X 1 3/16-inch, Brass[/list] [b]Miscellaneous Materials:[/b] [list][*]1�?�2-inch X 4 wood screws (6) [*]1-inch X 6 wood screws (4) [*]Wood Glue [*]Sandpaper in 80, 120, 180, 240, 320 and 400 grits [*]Danish Oil Finish [*]Felt feet for case (Set of 4)[/list]

Apple I Processor Section

Sun, 06 Jun 2004 12:09:41 -0700

[image:2938] I have attempted a redesign of the Apple I processor section. It is available in both PostScript and PDF: [node:2940] [node:2941] I recommend the pdf file, unless you're planning to edit the schematic. The schematic was designed in [url=http://xcircuit.ece.jhu.edu/]XCircuit[/url] using [url=http://fink.sourceforge.net/pdb/package.php/xcircuit]Fink[/url]. The design includes a 6502 processor, 6821 PIA, CY6264 8Kx8 Static RAM, and a 2716 EPROM. This is the first design I've worked on that uses a processor or memory, so it is quite likely I have made some glaring errors. I am very appreciatived of any feedback. Comments can be posted here. [h2]The Design[/h2] [b]EPROM[/b] The ROM monitor is at addresses FF00.FFFF. In place of two PROMs, a 2716 EPROM is used. On the EPROM, address lines A0-A7 are connected to the processor's address lines. A8-A10 are tied to ground. /CE is low when A8-A15 are all high. /OE is connected directly to the processor's R/W. All data lines are directly connected to the processor. [b]RAM[/b] The Apple I has 8k of RAM, provided by 8 8kx1 dynamic RAM package. These are replaced with one 8kx8 static RAM package, the [url=http://www.cypress.com/cfuploads/img/products/CY6264.pdf]CY6264[/url]. This eliminates the need for the RAM refresh circuitry. The RAM is at addresses 0000.0FFF (for sysem and user) and E000.EFFF (for BASIC). A0-A11 are connected to the processor address lines. A12[sub]RAM[/sub] = (A12' • A13 • A14 • A15). CE2 = (A12' • A13' • A14' • A15' + A12' • A13 • A14 • A15). /CE1 is tied to ground (why are there two chip enables?). /WE is connected to the processor's R/W. /OE[sub]RAM[/sub] = R/W[sub]proc[/sub] and /WE = R/W[sub]proc[/sub]'. All data lines are connected directly to the processor. [b]Processor[/b] Ø0 connects to the time base generation circuit provided in the Rockwell data sheet. /NMI, /IRQ, /ML, RDY, and /RES are all tied high. SYNC and Ø1 have no connection. Much thanks goes to Michael Rothe for help with the clock. [b]PIA[/b] The 6820 is replaced with the externally identical 6821. /IRQA, /IRQB, and CA2 have no connection. CS1 is high. CS0 is connected to the processor's A4, RS1 to A1, and RS0 to A0. CS2 = (A15 • A14 • A13' • A12). All data lines are connected to the processor's. R/W is connected to R/W[sub]proc[/sub]. E is connected to Ø2. The keyboard connection (PA0-PA6) is wired for an Apple II+ keyboard. PA7 is wired high. The terminal section is presently configured to use a printer for output, which will allow testing of the processor section independent of the terminal section. The printer wiring is based on this schematic. CB2 sends the "Data Ready" signal to the printer. CB1 and PB7 are low when "Data Accepted" is high.

PowerBook 3400 Pictureframe

Sat, 26 Mar 2005 13:50:37 -0800

[center][thumb:2359][/center] After seeing projects like the [url=http://www.applefritter.com/hacks/duodigitalframe/]Duo Digital Frame[/url] and others online, I thought I'd try my hand at it. Instead of hunting down a Duo or Powerbook 100 for a 4x6'' frame, I decided to go for an 8x10'' size (or 800x600 resolution) to display my travel photos. I have, or had, a PowerBook 3400c that had been surplussed from work due to a dead CD ROM drive, the replacement cost of which is about the same as buying another used 3400. The trackpad button also had some issues, but otherwise, the 3400 was in pretty good repair - no spots on the screen, for one thing. And so it began, as most of these project do, with disassembly. As with every Mac model, Apple provides assembly and repair instructions for the 3400 [url=ftp://ftp.info.apple.com/Apple_Support_Area/Misc/Service/servicemanuals/powerbook_3400c.g3.pdf]here[/url]. Don't lose that Torx #6. Once the 'book was apart - completely apart - I set aside all the parts I didn't need, everything but the motherboard, power supply, RAM, hard drive and LCD. If anybody needs a 3400 keyboard, trackpad, bezel, case, speakers... call me :) [center][thumb:2360][/center] [i]update: In a recent overhaul of the 3400 frame, I put the speakers back in, so I can now, potentially, have the tiny little speakers sound off alarms or something. I don't have any real use for putting them in, but I thought I'd find something eventually.[/i] With all that taken out, I needed to be sure the PowerBook still worked for what I wanted to do with it. So i hooked everything that was left back together (on the coffee table in my living room) and fired it up. With thoughts to making the frame wireless, I popped in a WaveLAN Silver PC card and configured it so I could control the 3400 from my iBook via AirPort with Timbuktu, load new photos onto it, etc. It's a lot easier this way than trying to hook up a keyboard and mouse to the ADB port which, at this point in the build, I couldn't be sure was going to be accessible. [i]update: if you look at the [url=http://www.applefritter.com/node/2356]detail[/url] you can see where the ADB port is, next to the AC power in the upper right. it's accessible when the lid is opened, and comes in handy for soft reboots and making occasional tweaks.[/i] [center][thumb:2357][/center] At this point, I also wrote an AppleScript to manage the tasks of 1) randomizing the photos to be displayed, 2) copying the photos to a RAM disk, as I could never get the contents of the RAM disk to be saved between boots, and 3) kicking off the slideshow. At first, I had intended to use JPEGView set to display the photos at random, fade between, etc. But I noticed one problem with JPEGView: the cursor! So I poked around amongst my old utilities and dug up KPT QuickShow, which does slideshows with configurable fades and delays, but doesn't do random, hence that step in the AppleScript. The JPEGs and slideshow software are on a RAM disk, so the hard drive can spin down and the frame can run almost silent. [center][thumb:2356][/center] [i]update: Instead of a RAM disk, I've updated the script to copy images to a 16 meg PCMCIA flash memory card that acts as a little hard drive, just big enough to hold about 25 photos. The advantage of this, in addition to being flash and thus maintaining its contents between reboots, and the continued silent operation of the pictureframe, is that the limited RAM of the powerbook can be used for the system and slideshow program. cutting into the RAM for the RAM disk posed too many problems (a memory leak somewhere, the setting reverting every time power was cut out). It also comes in handy to shuttle files to and from the machine, now that the waveLAN card is being put to use elsewhere, in one of my Newton MessagePad 2100s.[/i] [i]update again: I've posted my applescript in [url=http://www.applefritter.com/blog/3795]my blog[/url]. share and enjoy.[/i] Then came the woodwork. Quarter-inch poplar (probably my favorite material to work with) was cut, sanded, stained, burnished, and attached with hot glue and some scrap quarter-round for strructure. (Remember, measure twice; cut once.) I lay the screen into the frame and attached it with double-sided tape, then built the rest on top of that with a combination of double-sided tape and white masking tape. [center][thumb:2363][/center] I left the top panel of the frame free so the necessary ports on the motherboard could still be accessed. You can see the AC power plug, the ADB port, and PCMCIA slots. The wireless card is removed during normal operation so I can use it elsewhere, but its a simple matter to pop it in and update the scripts or upload photos. On the far right in this shot, you can see where the hard drive is mounted. [center][thumb:2361] [thumb:2362] [thumb:2358][/center] I took some pains to get the little details on the outside of the frame to look right, despite the mess that's on the inside. For one, I needed a way to power up the 3400 without a keyboard and without taking half the frame off. So I drilled a small hole in the frame opposite the reset button on the motherboard and inserted a... um.. well, it's a chopstick from some carryout Chinese food. I didn't have any dowel on hand, and the chopstick tapers perfectly to hit the button on the motherboard. I also cut a notch in the frame to thread the power cable through, so the frame can sit flush to the wall it's hanging on. The final result is a nice addition to my living room, and leaves a lot of room for improvement. For one, I'd like to keep the wireless card in place, and put together an applescript to grab images off my network, rather than having to pre-load them onto the hard drive myself. There's also the possibility of using it to display things like the weather map or scrolling headlines, but showing off my vacation photos is enough for now. Maybe I'll use the PowerBook 190 I've got in the closet for the other stuff, once I find a power supply for it.

Gameshow System

Sun, 06 Jun 2004 12:14:05 -0700

[center][image:1539][/center] I built a simple gameshow card for the Apple II. The system supports up to 32 users. A buzzer is wired to each student's desk. [center][image:1537][/center] Four 74LS244 buffers connect the buzzer lines to the data bus. A PAL is used for the chip select. The buzzers are located at addresses $C0C0 through $C0C3 (8 buzzers to an address). The code scans the addresses until it detects a buzzer going off, then displays the number of that buzzer. [center][image:1541][/center] The board was done primarily with wire-wrap. Here's the code I wrote for it: [tt]100 C = 0 105 FOR V=49344 TO 49346 110 A = PEEK(V) 115 GOSUB 200 120 IF (R <> 0) THEN GOSUB 300 125 C = C + 8 130 NEXT V 135 GOTO 100 200 B = 128 205 I = 8 210 FOR Q = 1 TO 8 220 IF (A >= B) THEN RETURN 225 B = B / 2 230 I = I - 1 235 NEXT Q 240 R = 0 245 RETURN 300 PRINT R 305 INPUT =T$ 310 HOME : RETURN[/tt] This was originally meant as just a test program, but I never got around to writing something more advanced. There are only 20 students in the class, so this code doesn't bother to scan the last eight buzzers. There's no good way scan bits in BASIC, which is why we have this inelegant routine with the division. If anybody would like to rewrite this in assembly or improve the interface (maybe add some graphics and sounds), that would be fantastic. [center][image:1536][/center] It looked really nice before I hooked 60 cables up to it! This is [i]not[/i] the way to do a patch panel. To save money, I used telephone wire, which wasn't a lot of fun to crimp. [center][image:1540][/center] The buzzers are three-terminal lever switches, duct-taped upside down. This works pretty well. [center][image:1538][/center] Wires snake across the floor and up the arms of each chair.