My bamboo table I made, haha. The multimeter used to check if the flash electrolytic capacitor is charged.

Pin from the image sensor connector. Indian SIM card for scale.

Lens block.

Lens block.

Lens block.

Lens block.

Lens block.

Lens block.

Image sensor.

Image sensor.

Image sensor has conductive traces that are thicker than others, which makes that pin that came out no big deal.

Viewfinder and zoom motor separated.

Lens moving in an out fine.

Step down gears that move the lens in and out.

Step down gears that move the lens in and out.

Step down gears that move the lens in and out. Worm gear connected to the motor.

I kind of broke the connector to the image sensor, but it still worked.

Connector to the image sensor.

Inside of the camera without the Lens Block

SY-189 Mainboard

Camera, lens flush.

LCD is broken from the hard fall, just need a new one, $10-30 max on eBay.

Lens comes out fine.

Camera works good even without the LCD, just use the viewfinder. 4x optical zoom, fat lens.

New LCD from eBay. $20

Installed it, works nice. eBay LCD $20

MOSFET Hydraulic Comparison

I made a simple circuit and also drew up a quick schematic in Eagle CAD to show how one is hooked up. S1 is like a momentary push button switch, but in the real world this would probably be replaced by a microcontroller, otherwise you are just connecting a switch to make a switch work, might as well just avoid the MOSFET and go with a switch. R2 is a current limiting resistor for the LED. If the positive supply voltage is +5V, then R2 would be 330Ohms, just use Ohm’s Law. Because a capacitance forms at the gate of the MOSFET you need a pulldown resistor from the gate to the ground. Without this resistor the gate floats on and always remains on even if the 5V threshold is no longer reached at the gate (even if S1 is off). If the gate threshold voltage is 5v, a 10k Ohm resistor works well. Also in the real world you probably wouldn’t be using a MOSFET to switch an LED on and off, MOSFETS are good for high power switching such as turning a motor on and off. Giving a pulse width modulation (PWM) signal to the gate, you can control the speed of a motor. If the Duty cycle is at 100%, motor will be at full speed, 50% duty cycle, half the speed.  Some MOSFETs have a built in protection diode, some don’t. If your controlling a motor with the MOSFET, it would be nice to have this diode. Motors act as a generator sometimes and if the motor feeds voltage back into the MOSFET, it can damage it.

N-Channel Mosfet Sample Circuit

Sometimes when prototyping you might burn out a MOSFET. MOSFETs when burned will either be completely open or closed. Using an Ohm meter if it has infinite resistance across any of its pins or little resistance (practically conductive), its probably a bad MOSFET and should be replaced.

I immediately split apart the LCD monitor to pieces. Bottom right is the LCD screen showing cables for the back-lit display, and the left ones carries all the color information.

Left side shows the mainboard and the power supply.

Power supply board. Bottom right showing combusted electrolytic capacitors. Top of the board are the inverter circuits for the back-lit display. Left side shows 120V in, various coils, a diode bridge, probably a bleeder resister, step down transformer…

I decided to replace all these capacitors. One probably failed and it probably set a chain reaction and screwed them all up. They have cuts on the top so that they break open to release pressure if they fail. Top 4 all busted open. The bottom ones are bulging.

Aerial view.

Close up of the bad capacitors with voltage regulators in the background.

New VGA cable, (6) 470μF electrolytic capacitors. I got one rated at 35 volts and the rest rated at 25 volts. Originally all rated at 25 volts, but I thought the really bad one may have encountered higher voltages or whatever. The smaller capacitor is a 220μF rated at 25v.

Bottom side of the printed circuit board. I desoldered C216, C214, C213, C212, C211, C215, and C222

Random picture from the inverter section.

Power Supply board with the capacitors removed.

Power Supply board with the capacitors removed. Bottom one I didn’t remove yet. Left side shows a header connector which plugs into the main board to power it. This is the power supplies output providing 5 and 14 volts and some other pins to help the thing switch to power saving mode and stuff.

New capacitor placed through the holes ready to be soldered.

One pin soldered.

Messed up capacitor versus the new replacement.

Comes in a smaller package and a nice green color, haha. The pins are too long, so I snip them before it gets soldered to the board.

All the solder joints done!

Capacitors replaced!

Capacitors replaced! Looks nice with the green color instead of black.

WOW, I got it to power up!

Message says no signal, because the guy that threw away the LCD monitor, stupid guy cut off the VGA cable, I don’t know why. Probably thought the monitor was garbage since it wouldn’t power, but I got it to power. So now I have to attach a new cable.

My glass table, top are the packaging from the capacitors. Left are the capacitors. Center are snipped pins from the new capacitors and right shows splatters of solder.

Messed up capacitors. Bottom right you can see it was bulging from the bottom.

Stripped the outer sheath and the shielding of the new VGA cable. I cut off the female connector since I didn’t need it.

Old half of the VGA cable.

Stripped the wires. Stupid thing doesn’t match up. None of the colors of the old cable match up to the new one. How stupid, it was incredibly difficult to get this part to work when it shouldn’t have been broken in the first place.

My testing setup. I used a sharpie to hold up the LCD. LCD cable, and inverter cables are attached, and VGA cable goes to the mainboard, and other end to the computer. Then I splice the wires together and test. Bottom right shows the power cable and the button panel for the LCD.

Started to research that chip, the TSU16AK.

Removed the barcode sticker to desolder the quartz oscillator.

Tried to count the pins to see which is which it goes counter-clockwise starting from pin one indicated by the indented dot on the chip. Pins are incremented by 5 separated by small white dots and markings on the PCB.

Quartz oscillator was in the way, I wanted to follow the traces. Ultimate goal is to figure out which wire to splice with which wire for the VGA cables because the stupid colors don’t correspond.

I looked up the chip online by using a google trick to search for the datasheet. Datasheets and schematics are often in PDF format so I googled: TSU16AK filetype:pdf. I found a service manual with schematics and pinouts of the TSU16AK. This manual was meant for an Acer LCD monitor but the circuitry was very similar and useful. Thank god for copy cats.

Using a continuity tester and the Pinouts for a male VGA connector I know which pin goes to what for the new cable. But the old cable I didn’t know which pin was for horizontal and vertical sync. So my genius idea was to follow the circuits. From this pinout diagram we see that Pin 37 is for Horizontal Sync and Pin 38 is for Vertical Sync on the TSU16AK chip.

Quartz oscillator removed and desoldered. so we can follow the circuits better. Top chip is the TSU16AK the last circle is Pin 35, previous mark would be Pin 30, etc. So the second to last pin is pin 37 which I identified as the Horizontal Sync. So what I did now is follow the circuitry. Its super small and difficult, but the photos blown up helps. You can see it goes through a surface mount resistor R17 then goes further down…

…it then splits off to go under that diode (D3), and then goes to that surface mount resister R24. From there I used a continuity tester to see which pin it was on the above connector. I quickly got to know it was the white wire. Did the same for Vertical sync. Turns out there is similar circuitry and it connects to the neighboring diode and resistor R23.

Another view.

Tested it and it worked! Genius.

The final test splice. The bottom unconnected wires are kind of useless. Grey one is unconnected, and the other ones are like ID pins and stuff. In old monitors depending on the combination of which is connected to ground, a computer can determine if a monitor is capable of supporting 1024×768 or whatever.

I set up a dual monitor setup, plus I have my other laptop I use for testing and stuff. Main laptop uses 1920×1200 and then the new monitor is 1280×1024 additional screen area.

Some quick notes.

Notes on my whiteboard. Left is the cable, right is from the monitor.

Schematic from online plus some notes.

1. Power up the PCB so that the ATmega328 or ATmega168 gets 5v of power.
2. Plug the USB Cable into the AVRISP MKII and the programming cable from the AVRISP MKII into the ICSP of the board.
3. Make sure the chip has the bootloader on it.
4. Open up C:\Program Files (x86)\arduino-0022\hardware\arduino\boards.txt (file path may be different) and add the appropriate code below. You can add both if you want.

How to Burn the Bootloader onto ATmega Chip

Click Tools > Burn Bootloader > w/ AVRISP mkII

Board Code for ATmega168 and AVRISP mkII


###############################################

avrispmkii.name=AVRISP mkII w/ ATmega168

avrispmkii.upload.protocol=avrispmkii
avrispmkii.upload.maximum_size=14336
avrispmkii.upload.speed=19200
avrispmkii.upload.using=avrispmkii

avrispmkii.bootloader.low_fuses=0xff
avrispmkii.bootloader.high_fuses=0xdd
avrispmkii.bootloader.extended_fuses=0×00
avrispmkii.bootloader.path=atmega
avrispmkii.bootloader.file=ATmegaBOOT_168_diecimila.hex
avrispmkii.bootloader.unlock_bits=0×3F
avrispmkii.bootloader.lock_bits=0×0F

avrispmkii.build.mcu=atmega168
avrispmkii.build.f_cpu=16000000L
avrispmkii.build.core=arduino

###############################################

Board Code for ATmega328 and AVRISP mkII


###############################################

avrispmkii.name=AVRISP mkII w/ ATmega328

avrispmkii.upload.protocol=avrispmkii
avrispmkii.upload.maximum_size=30720
avrispmkii.upload.speed=19200
avrispmkii.upload.using=avrispmkii

avrispmkii.bootloader.low_fuses=0xFF
avrispmkii.bootloader.high_fuses=0xDA
avrispmkii.bootloader.extended_fuses=0×05
avrispmkii.bootloader.path=atmega
avrispmkii.bootloader.file=ATmegaBOOT_168_atmega328.hex
avrispmkii.bootloader.unlock_bits=0×3F
avrispmkii.bootloader.lock_bits=0×0F

avrispmkii.build.mcu=atmega328p
avrispmkii.build.f_cpu=16000000L
avrispmkii.build.core=arduino

###############################################

After adding the code you should be able to change the board to use the AVRISP mkII programmer. See the below screenshot.

A group of my friends are DJ’s known as Desi Jammers, and did DJing at my Birthday BBQ. One of their turntables had an issue with the LCD reading “no disc” The CD would never spin up and its like the laser had issues reading the disc.

First step was to remove the laser pickup assembly the CD loader. This is a front loading CD ROM drive.

At first I saw that the right metal connector was touching the nearby motor which probably grounded it out possibly causing the problem to begin with. Photo doesn’t show this, photo is showing after I already bent it where it should be. This switch is like a sensor that tells the system when the laser sled has reached the beginning of the path that it travels. I put it back together to see if that was the issue, that was not.

If the motor wasn’t spinning, I was thinking this chip could have been faulty since it is connected to the motor, so it drives the motor to begin with. No other component has any noticeable damage, but this one may have had internal damage not noticeable from the outside. Also I suspected it overheated since it has a not so great heatsink on top. My friend thought the motor was burned, but I ran voltage through it from a battery and the motor spins fine. Plus it passed the continuity test that I did earlier.

BA5979S schematic and pinouts.

Reverse side of the mainboard and power supply. Top right is where the BA5979S chip is soldered.

I used a clear CD that you find at the bottom of a spindle of CD-Rs to troubleshoot, so I can see whats going on.

Clear CD, and one desi CD to check for playability.

Looks like a photo diode, at first I thought it was the laser diode.

The laser diode shines, and the laser optics read the pits on the CD, the pulses of the pits are then detected by this Photo diode. Which sends the data back to the controller.

Photo diode.

Spring that pulls the top portion of the CD loader down when the CD is in and keeps it firmly in place.

Laser diode shining through the optics. This laser diode is from a CD Clock Radio that I found in the trash several weeks back.

Engineering plans to fix this bloody thing.

Notice the laser sled assembly on the left closely resembles the one on the right which is from the CD clock radio I found in the trash.

Both assemblies were really similar so I decided to try the laser pickup assembly from the CD clock radio I found in the trash into the DJ turntable. Only problem which I did was I didn’t put the paper all the way on the bottom which allowed the motor to touch the metal thing and spark and it messed up again. Stupid. However, I got the motor to spin which proves that the BA5979S chip was never the problem.

I tried it again after that thing touched the metal, but it didn’t seem to work. I thought maybe I messed up the laser diode, because it didn’t’ seem to be shining anymore. So I decided to get one from a CD burner I found in the computer which I also found in the trash.

CD burner from the trash.

CD burner from the trash, different laser pickup sled. The gearing is kind of cool too. If something is forced for whatever reason, like some debris gets stuck or whatever, it has some kind of spring loaded gearing thing kind of like a clutch which is wild, look closely. Its like a gear on top of a gear.

Laser diode enters from the bottom, goes through a series of lenses and mirrors.

The mirrors and prisms for the laser optics, its pretty awesome.

I Desoldered the laser diode.

The laser diode from the sled assembly in the DJ turntable looks way bigger, but I figured it may still work.

The board that connects to the laser has two metal discharge points on the board to the right. I read somewhere that you need to short circuit this and discharge every time you are trying to work with it. Maybe for electrostatic reasons, or that surface-mount capacitor that’s also located on that board which is connected in a parallel circuit probably needs to be discharged.

To pull the sled off its tracks, you have to pull the bar that it slides through off. To pull the bar off you have to bend the plastic thing over to the right and pull on the bar.

Laser diode removed, penny shown for scale. Shows how small the stuff I’m working with.

Look how closely the two laser pickup assemblies look alike. The right one is from the Turntable, the left one is from the CD Clock Radio I found in the trash. I wanted to try and make that one work again in the DJ Turntable.

I took the balls and springs off. These help to keep the CD snapped onto the spindle. However that’s from top loading CDs, and this DJ turntable is a front loading one.

CD wasn’t being gripped to the plastic. So I added a padding I got from the CD burner. I tried a rubber gasket I found at CSU East Bay on the ground of the parking lot one day, but those few extra millimeters of thickness caused the laser optics not to focus properly on the disc.

I tried switching the sleds since the CD didn’t work well with the spindle from the clock radio. However the sleds don’t swap easily. When you use the CD clock radio entire laser pickup assembly, it doesn’t fit into the DJ turntable because the motors are too big even though they have the same specifications.

The mess of trying to fix it. Chipotle in the back. Clock radio, compressed air from that right guard deodorant, solder, soldering gun, cd burner, dj turntable, laser shit, monkey wrench.

Split to pieces.

Drilled a hole through the top CD spindle in order to undo the screws underneath.

Drilled large holes to remove metal to try and pull the motor off. The top spindle thing doesn’t come off.

The drill accidentally ripped off the wires connecting to the motor. Superbad.

Looking at the bottom of this, I figure I can bend the 4 metal tabs (2, 4, 8, 10 o’clock positions), open up the motor and solder new wires on. However that was difficult. I discovered something even better. Additional metal connectors on the bottom which have been snapped off.

I managed to solder wires at those points. I used a continuity tester to determine the anode and the cathode and soldered it to the small printed circuit board (PCB). At first I was going to have one big motor and one small motor since the small motor with the small gear wouldn’t match all the gearing for this sled assembly. However, I realized that while the spindle is not removable on one motor, the gearing is removable on the other. So I managed to get both small motors on the assembly so that later when I put it all back together, it won’t have issues fitting in. The sensor at the right is the next thing to fit on to this assembly.

I drilled another pilot hole and then the screw to get the sensor on. I stripped the screw badly though. So then I used thick wires soldered to the PCB to make the sensor rigid. I used another solid wire soldered to another point to hold the motor wires away.

Completely re-engineered laser pickup assembly. New sensor position, old spindle. New metal frame, new shock absorbers, new gear, old motors, new ribbon cable.

Another View

Testing

CD playing, DJ turntable now works good. The only bad thing out of all this, is when I was pulling out the spindle motor from the old assembly, I accidentally bent the spindle a tiny tiny bit which makes CD reading bad in the later tracks after like track 8. When the CD spins, there is more gyroscopic stability in the center since its closer to the spindle, which means early tracks play fine, but later ones not well. The small amount of wobbling doesn’t allow the laser to focus well. However with this working now, it allows to the computer to hook up without any issues and even without that, a two turntable system makes the whole idea flawless.

Blackberry Torch with the screen turned on.

Damaged BlackBerry digitizer with the Screen off.

That’s when I ordered a new touch screen digitizer on eBay for the BlackBerry Torch 9800. Swapping it out was very difficult so I’m making a tutorial here today on how to change the touch screen and the disassembly/teardown of the torch.

First remove battery plate, extract the battery, mirco sd card and sim card.

Remove the four T6 Screws that are visible. One of them will be under a blue sticker, not sure if that's a water sensor or what, but your warranty will be voided as soon as that's removed or broken.

After removing the bottom two T6 screws, you should be able to start prying off the bezel which is weaved around the phone. After removing the bezel, there will be two more T6 screws which were previously hidden by the bezel. Removes those screws, then pry off the back plate. Use a plastic screwdriver to help you along the way. Plastic screwdriver allows you to pry it all apart without damaging or scratching the housing.

After doing the above steps you should be left with everything at this point.

Next remove the three ribbon cables connected on top of the main printed circuit board (PCB). One part is stuck on like a sticker, slowly peel it off. Remove the black antenna wire on the reverse side of the PCB as well.

Remove the black sticker on the back of the BlackBerry where the battery usually goes. Underneath you will find 6 hidden screws which are of size T2. Remove them.

You should be able to separate the two halves of the blackberry from this point and it should look like this.

Remove even more screws, some might be T4 while others might be even smaller like T3 or T2. You might need to slide the slider up or down to get some screws visible. After this, you should be able to remove the slider assembly.

Pry off the touch screen digitizer slowly without breaking the LCD screen below it. Also do not pull it all the way off as there is still a ribbon cable you have to disconnect carefully on the reverse side.

on the other side of the LCD you will find the ribbon cable for the touch screen which loops around. Remove the brown tape that covers it and lift the clip upwards 90 degrees shown at 30 degrees in the photo. This allows you to pull the ribbon outwards. The ribbon cable is also stuck on like a sticky. Slowly peel it off.

When sticking back on the new digitizer make sure there is no dust or fingerprints on the side that will be closed up permanently because you will never be able to clean it up unless you open the BlackBerry again. Also when putting back the ribbon cable for the digitizer make sure the white line on the ribbon is fully pushed against the connector and flush, it may even be invisible if fully inserted, then push the clip down to hold in in place and stick it on with the sticky. Also during the whole process do not remove the call/end/menu/back buttons as this will be hard or impossible to put back later properly.

Remove the back cover of the blackberry. Extract the battery, sim card, and micro sd card.

Pry the clip on the back of the phone with a plastic screwdriver. Also remove the T6 Screws, there are 4 visible, and 2 which are hidden underneath that clip. One of the visible screws may be hidden underneath a no entry symbol sticker.

Next pry the backplate off the phone with the plastic screwdriver.

Next, move the printed circuit board (pcb) out of the way. Remove the three ribbon cables. The white antenna cable can be left alone.

Here’s a video to help you guys out in taking apart the BlackBerry 8900.

On the reverse side of the blackberry on the trackball PCB, I noticed some corrosion on the bottom right corner. I cleaned it with rubbing alcohol. Allowing the trackball to scroll properly in all directions.

Corrosion on the bottom left corner contains dissolved salts from the water which are conductive preventing the trackball from working properly.

However even though the trackball was scrolling, it was not clicking properly. I thought that its possible there was some corrosion or damaged electronics that went unnoticed so I ordered a new trackball on ebay. To remove the old one you have to pry out the metal plate where the trackball PCB sits. Be careful not to break surrounding components.

Here you can see I bent an edge connector that goes to the LCD by prying open the metal plate for the trackball PCB and using the edge connector as the fulcrum for the lever/screwdriver.

Luckily I didn’t bend that one too much, so I just bent it back as much as I could and it was fine. However with the new trackball from Ebay it would not scroll in all directions or click properly. I had a feeling it was not getting a proper ground connection as the new trackball did not have the foil sponge stuck on it as the old one did. When you put the PCB back, the sponge is supposed to make a proper ground connection.

The Part where the metal sponge connects to the ground plane on the PCB.

I tried everything including a ball of foil, that’s when I figured its probably cheap china made junk. Compare the two PCBs.

The original trackball is on the left and the trackball from ebay is on the right. Looks poorly made and drenched in flux.

That’s when I was like, well at least the original trackball could scroll in all directions the ebay one couldn’t even do that, so then I decided to check one more thing on the original trackball, and that was the conductivity of the tactile button. My battery was dead on my multimeter, so I decided to do a manual look at the button by taking the button apart. There is a small piece of tape over the button. That tape is yellow in color and I believe is heat resistant. This tape is used to protect heat sensitive components during solder reflow processes.

Taking apart trackball tactile button I noticed corrosion under the metal dome and on the contacts below it.

After cleaning it up, you'll probably notice a big difference. After putting it all back together the original old trackball worked perfectly fine.

If you just want to quickly check something on your mobile and want to use the trackball, you can do this temporary fix until you get the time to fix the trackball entirely. First lock the phone or put it on standby. If you have it, drop a tiny bit of isopropyl (rubbing) alcohol on the ball and use a cloth or clothes to rub the ball around vigorously, up and down, up and down. After the ball appears to be dry, it should be good to go. You can also try to do this without the isopropyl alcohol.

How it Works

The blackberry has magnetic proximity sensors, or reed sensors that basically detect a change in magnetic field. There are four sensors for up, down, left, and right.

As the ball moves, it presses against rollers which rotate a magnet right over the sensor. A ball mouse uses a ball and rollers to rotate a circular disk containing holes where an interrupter sensor detects pulses of light. The thing is, the ball is never the thing that collects the dust, its the actual rollers which gather the dust. The newer blackberry trackpad is actually the same as an optical mouse, detecting the movement of your finger along the trackpad.

Permanent Fix

To fix the trackball permanently is relatively. Easy, first you need to remove the trackball. I believe for blackberry curve’s its the easiest as you just remove it directly from the top. I’ll give you step by step instructions for the Blackberry Bold 9000. Other blackberries are probably very similar.

Shut the phone off, and remove the battery, sim card, and MicroSD card. Before when opening a Blackberry Tour, I snapped a MicroSD card in half while doing a teardown of the chassis, so its best to play it safe.

Remove all 6 screws with a hex bit, I believe its a T6, but double check. That black plate should then easily remove. Push out the PCB by pressing a bit on the keyboard and prying it out on all sides, then pull it out away from where the LCD is, its wedged in there. Do not use regular screw drivers to pry it out as it may damage components, use a plastic screwdriver if necessary.

From the front side of the PCB, you’ll need to remove the metal plate holding the trackball in place. You can see there are four metal clips holding it, pry those out with a metal screwdriver.

The circuitry under the trackball must be very dirty, I would use isoproypl alcohol and a cotton swab (Q-tip) and clean it out.

You can actually see the trackball in action by moving the ball and watching the rollers moving. When rolling the ball in one direction, only one of the rollers will move, the other will not. If you are moving it in one direction and both rollers do not move, then that roller is the problem one, this will be more clear later on.

The trackball will have to be taken apart, separate the metal clips on all four sides and pry it off, its very tiny. You may notice a severe buildup of dust on the white baseplate. Especially on the particular roller that was not working. Also see on the individual rollers. The problem one probably has the most dust on it.

Pull the rollers off one by one and clean them completely. You can use isopropyl alcohol to help you. Take note at which orientation all the rollers and parts are so that you can easily put it back, the parts usually only fit in one way anyways. As normal, if you’re forcing something together, it probably was not meant to be, take your time and don’t rush it.

Feeding Power through USB

After this is done, and you power up your 5 volt source, the blackberry should turn on and begin charging.

Blackberry Powered on and Charging

Here is a pic showing power being pulled from a 5 volt source. This PIC programmer which I built happens to have two PCB test points for which I can check and pull power from.

Pulling 5 volts from PCB Test Points

Laserjet 2600, 3800, 3500, 3700, 4700 Mirror Configuration

If you look at the mirror configuration on these HP Laserjet Printers, you’ll notice that a few of the mirrors are at a very low angle, so gravity naturally brings dust on to them distorting the laser optics which effect print quality. On my 1600 Laserjet magenta was the first to go since it was on the bottom and collected the most dust. Then yellow would soon become a problem. So I simply opened it up and cleaned all the laser optics. Here’s a somewhat boring video I made showing whats going on.

Here is an upclose pic of the mirror optics I took. Its interesting how one polygon mirror takes care of two lasers.

Laserjet 1600 Laser Optics

Rather than me re-inventing the wheel, I’ll give you guys this PDF document which I saved on my computer a while back when I had the issue. It contains detailed instructions on how to open the 2600n, 1600 and 2605 to get to the laser optics. If its your first time opening this up. Take your time and be patient. It should take around an hour to do it.

Laserjet 1600, 2600n, 2605 Repair Guide

You probably wont need this for this problem, but here is the service manual which contains 300 pages of everything you possibly need to know about the Laserjet 1600.

Laserjet 1600 Service Manual