I even bought another modem at a yard sale, thinking that maybe ours was on the fritz – but no… same M.O.: laggy/lossy connection, then no connection, go downstairs and the modem’s really hot to the touch. It probably doesn’t help that it’s in a small closet under our stairs with the router and everything else, but even if we keep the door open with a box fan blowing in there it still does it.

As I was throwing away a bunch of old stuff, I happened upon an old add-on DVR I’d bought to steal the drive out of. I opened the lid just to see if there was anything else in there I may want, and there it is – a small square heatsink just the perfect size for the little ARM (I think? Could be MIPS though) CPU in our SpeedStream 4100. I pulled it off, then used a razor to scrape all the old thermal pad off. I mixed up some thermal expoxy, applied a really thin layer and glued it to the top of the CPU in the modem, as you can see on the left.

We since went about 4 nights with no disconnects, despite the fact the top of the heatsink still got a bit warm to the touch. I was not taking any chances though, I wanted to attach a fan to it as well.

It’s been a while since I’ve had an excuse to get this uhh…. AfroTech on any project. I’m terminally lazy and still obsessed with genuinely attempting to do things The Right Way, so this site hasn’t seen much action as of late. Anyway, I grabbed an old squirrelcage fan I’ve had laying around a while – I don’t know how many CFM this thing is rated for, but let me tell you on 12v it is like a hair dryer. I scrounged up an old 5vDC PSU to power it because I didn’t quite need it to roar that bad.

I cut a hole in the top of the DSL modem for the heatsink to poke through into the front of the fan. I thought about having the exhaust blow inwards, but frankly it looked stupid – not to mention flimsy. In the end I decided to go negative pressure, with the exhaust blowing off the side.

The top of the modem is quite curved, so it really didn’t fit all that well when I affixed the fan to it. So I broke out some latex based sealent and caked it around the gaps. The air is sucked in via the vents under the board, makes its way around the board and over the heatsink and out the exhaust.

So I took a break from authoring this post, and it’s now about a week short of 2 months since I started working on it. The uptime on the modem speaks for itself:

Time Since Last Sync 031 days 20:56:49

I actually rebooted the modem at that time on purpose, it turned out it was a problem at AT&T’s end. For all intents and purposes, the modem’s been up and synced since I plugged it in. Considering we’ve had some really really hot days here, that’s a huge improvement.

I decided to take it apart and try and clean the “pot” – it’s not really a potentiometer, but that’s what many people would call it anyway and cleaning it is basically the same principle. It took me a few minutes to get into the machine and access to the front of the volume control – it involved removing the top cover, then the entire face (after disconnecting a couple of ribbon cables), and then removing the PCB from the plastic face.

Once I got access to the front of the “pot”, I pried open the four tabs holding it together. I grabbed a can of my trusty CRC 2-26*, and sprayed a bit in there before putting it back together and twisted it a bit. After pressing the tabs back together, I carefully put the ribbon cables back on, plugged everything in and gave it a shot. There was still some obvious scratchiness and/or errors, but it was markedly better.

I made sure the tabs were down tight, and reassembled the entire unit. Now that I’ve got a working volume control again (and I don’t have to find what the hell the kids did with the remote), my receiver works perfectly for my PS3 and my PC’s digital output.

* A word about 2-26 and spray “lubricants”. 2-26 is, according to the manufacturer, a plastic-safe precision lubricant that displaces water and can improve the electrical contact properties of things like switches, relays and stuff. WD-40, on the other hand, is advertised as a lubricant, but I’ve had terrible luck using at as such. While it doesn’t seem like it’s billed as such, in most situations I’ve used it it acts as a degreaser. Based on my experience cleaning volume controls and such, I wholeheartedly recommend you make the trip to Lowes or wherever and pick up some 2-26 instead.

But even so, this mouse has served me well. It’s now called the “Microsoft Optical Mouse Basic“, but back when I got it I think it was just called the “Optical Wheel Mouse”. Regardless, it’s out-lived virtually all my wife’s mousing devices, most of which have been the wireless variety. I’m now used to the cord on it, we have quite the rapport and it never seems to get in the way… and it’s nice not having to change out batteries every freakin’ month.

Unfortunately though I have this rather nasty habit of highlighting and un-highlighting text while I read it. It annoys my wife no-end, the constant click-click-clicking while I’m engrossed in something particularly interesting, but I figure at least it’s better than me mutilating my fingers which is my other annoying concentration-habit.

The end result is that the micro-switch for my left mouse button has reached the end of it’s operating life. It bounces now, and dragging anything with it is a nightmare. I have a new mouse on the way (I actually bought one already, but it’s defective so it went back, and now I’m waiting again), but I can’t keep using my computer like this.

So I took my mouse apart with a view to switch out the left mouse button switch from another mouse – unfortunately they’re different switches and it won’t physically fit in my mouse… so I came up with a hackish solution in the mean time. I simply de-soldered the left mouse-button switch, and the switch under the middle mouse button, and swapped them.

I reasoned that the occasional, unintentional double-click on the middle mouse button probably wouldn’t anger me as much as the much more likely double-left-click… and how often do you ever drag with the middle mouse button on Windows systems?

Never.

So it should live until my new mouse gets here, at which point it’ll serve me well as a backup or for when I’m roped into fixing some asshole’s computer. Also, I apologize for the lack of technical content in this article.

The PS3 is even better because most games support Dolby Digital 5.1 native – the PS2 lacked the processing power to generate this signal dynamically, only pre-rendered cut-scenes had Dolby Digital.

I started out modest – an old “Paramount” Pro Logic decoder, which I eventually traded in for a Yamaha 5.1 receiver. This thing is great – it’ll fill up a huge room with modest speakers no trouble at all, and it’s been in our “great room” since we moved out here. Unfortunately, our “nice TV” broke permanently, and we shut off our satellite TV in favor of streaming NetFlix… meaning my beautiful receiver was relegated to my younger brothers in law playing PS2 on it.

This morning I decided enough of that, I’d bring it up here and use it in our bedroom. Story time over: I decided to see if I could hack digital output from my HP Pavilion a1213w desktop. I checked the motherboard, it’s an Asus K8S-LA “Salmon” board… bit of a piece of shit really, but it does the job.

It sports a RealTek (ugh!) on-board audio with 5.1 output. Awesome – but no digital output in sight. Searching for the manual, I found an S/PDIF output on the board which requires a daughter-board to give you a coaxial/ToSLink output. Check eBay – ~$20… fffffuuuuuu that.

A quick Google search shows plenty of other folks hacking their own so I decided to give it a shot. Ratting through my box of parts, I came up with a 4-pin CD-ROM-Audio cable from years gone past and cut it apart. I also dug out an RCA cable (I went with Coaxial for the PC since my PS3 will be using the only available ToSLink socket on the receiver) and cut it up too.

A simple hack really. Pins 1+2 are ground and digital-out respectively, with the third pin being +5vdc for powering the bits for ToSLink communication… irrelevant for my purposes. I carefully cut the connector to a little larger than the three-pin connector, then used a Dremel with a sanding wheel to smooth it to a perfect three-pin shape. I pulled the unused extra wire out, since I didn’t want to accidentally short my mobo’s +5vdc. I had the black wire on the GND pin and the white on the S/PDIF pin (see diagram at left) and it was a simple matter of soldering colors to colors to connect the RCA. Route it out the back of my PC and into my receiver.

Digging through the RealTek control panel’s “Multi-Channel Sound Manager”, the instant I clicked “Enable Digital Output” the PCM light came on my receiver and I’m good to go. Turn off all the DSP shit the kids had turned on, and my iTunes output is so manly I need chest hair supplements to keep up.

Update: Okay so apparently my RealTek card can’t output Dolby Pro Logic II (or even anything remotely close to surround sound) over S/PDIF – PCM is it (and PCM is limited to two channels). According to a thread on Overclockers, most sound cards are limited like this… if you want surround sound in games, you’re stuck using the analog outputs and the 6ch input on my receiver.

So I scrounged up some 3.5mm to RCA cables, and hooked it up… and low and behold I have full surround on games like Left4Dead. I press the 6ch button on my remote, and I’m switching back to digital output, so I can take advantage of my receiver’s vastly superior DAC for music. Not optimal, but the best I can probably do without a heinously expensive Pro Logic II capable sound card.

So what can we do with this? I’m led to believe that Dish hackers use this interface to take a snapshot of the firmware on the dish and replace parts of it as they need to. We’ll go into no such nonsense – under fear of the DMCA – however I don’t believe that discussion of or probing the JTAG interface falls under such legislation, unless I receive a letter from a company involved. So here we go:

You will need:

• 5 100ohm resistors, at least 1/4 watt.
• Category 5, or some other kind of signal cable
• A female DB-25 connector
• Some software to read the JTAG interface on the parallel port

You will also obviously need a soldering iron and some proficiency in using it. We’re not dealing with any SMT devices or anything here, the pin holes are quite large and should pose little difficulty for anyone that’s ever soldered anything before in their life. You will want your cable long enough to reach from the box to the back of your PC. In my case, I simply chickened out and cut up a printer cable, but OrganizedChaos used Cat5 with success.

It became obvious that I would have been able to access the JTAG port from outside the case, without even opening it – I assume this is so that Dish can test and possibly program the boxes before they leave the factory. The JTAG pins are pins 184-190 on the CPU (as can be referenced from the datasheet found at Datasheet Archive). The traces are absolutely tiny, so it takes some serious time to follow them to the correct pads – an ohmeter with tiny probes speeds things up considerably.

In our case, we won’t be powering the JTAG interface – we’ll simply plug the unit in, leave it off, and let the standby power supply the juice necessary for our probing. I imagine this would be much more reliable than trying to supply 3.3v from our parallel port.

The first thing I did was source 5 100ohm resistors. I managed to find two in an ATX power supply, and another four off the board of a television set. I needed the extra one because I broke one swinging it around on the PCB. I straightened the legs out, and soldered one resistor into each of the pads that require them.

In the diagram below, I’m labeling each of the pins going down in two pairs from what I’m assuming is the “#1 pin” (the one closest to the angle that’s been chopped off the box on the PBC). This pin layout is not the standard 20-pin JTAG! I was probably pretty lucky I traced the tracks and didn’t just assume it was standard, because I’m sure I’d either have killed my printer port or the dish box (the former being the bigger deal) if I’d just blindly hooked it up. By the way, if you check out the full-size photo to the right, the pins are highlighted in mouseover notes for your information.

N/C DB25
N/C DB24
N/C DB23
N/C DB22
DB02 DB21
DB03 DB20
DB04 DB19
DB13 N/C
N/C N/C
DB05 N/C

You will notice that pins 19-25 are connected to the grounded pins on the JTAG interface – this probably isn’t necessary, one would probably suffice (I’m pretty sure pins 18-25 are all connected at your PC’s parallel port), I just got frisky with the soldering iron. Pins 2, 3, 4, 5 and 13 are where the action is, so you’ll want to make sure the connections through the resistors are pretty good. OC put his resistors on the DB25 connector, I put mine on the board, either way… doesn’t matter.

Once everything’s soldered up, plug the DB25 cable into your PC, then plug the box in but don’t turn it on. You’ll need some software – after googling around I found some software called jKeys which seems to work pretty good. I should point out that it’s often used for theft of service and I’d really rather steer clear of that on this website – do not email me about trying to steal Dish Network TV.

Once you point your software at the parallel port, you should be able to find out all kinds of nifty information about the device. I’m led to believe JTAG supports in-circuit debugging… we’re actually toying with the idea of playing around with a Dish x700 “hello world” application, because looking at the datasheets for the STi5500 CPU the OSD modules seem pretty trivial to operate.

But for now, we can experiment with all kinds of neat functions that JTAG allows. Happy hacking!

Update: 2007-08-11

I’ve gone ahead and hosted Jkeys anyway, but I want to make it clear that I don’t support theft of service. The reason I’ve included it is because I haven’t yet found another (free) way of programming the flash after you build something using the STi5500 toolkit. So far I’ve not yet managed to even so much as get the LED to flash, but I did manage to resurrect the box after I thought I’d bricked it.

I’m also including a nifty program called JTango. It’s a JTAG program again, but it’s used for in-circuit debugging. At least, that’s what I gather, I couldn’t make it work with Windows XP. I’ve included the driver that’s allegedly supposed to fix the “Privileged Instruction” error, but it didn’t work for me. If you figure it out, please do let me know.

Now it’s time to hit up more documentation and see if I can at least make the freakin’ box flash it’s LED or something… at least something to tell me it’s working.

I tore it apart, and soon realized that there was way too many wires, and none of them were standard color codes. This is because it has a weird dongle that you can attach a DB-9 serial plug onto instead of a USB. Next I noticed the Cypress Semiconductor USB controller sitting on the board. I managed to track down the data sheet to it, and enumerate which pins were which. Tracing these back to the cable I get:

If you want, you can tie E17 to the shield – or E17 to USB black, and E13 to the shield of the USB cable (the original cable is unshielded however).

Installing the Drivers on Windows XP

This part sucks – it’s not very intuitive at all. I’ve found it easiest to install the driver prior to plugging in the tablet, otherwise you have to do some juggling to get it to ever function. Instructions:

• Download jamstudio-tablet-win95-98-me-2000-xp.zip and extract it to a folder someplace. This mightn’t be necessary, but I did it anyway so I could burn the driver to a CD just in case.
• Run setup.exe from inside the driver folder, and follow the instructions. This will install the driver (but not attach it to the hardware in Windows XP) and the control panel applet.
• Plug your tablet in. Windows XP will detect the tablet, but it won’t function yet.
• Under Device Manager (Right-click on My Computer, choose properties, Hardware tab, then choose Device Manager), look for the two devices that represent the tablet. They should be something like USB Human Interface Device and HID-compliant device. Open up HID-compliant device.
• Under the Driver tab, choose Update Driver. Tell it you want to select the driver yourself, and eventually you’ll convince Windows you really do want to select the driver yourself. By this time, it should pop up the list of similar devices, one of which should be KBGear Jam Studio Digitizer. Select it, then dismiss all dialogs.
• On both my computers, the digitizer now worked for positioning but didn’t reckognize mouse clicks from the pen touch. If your’s does this too, simply open the control panel applet and toggle the settings to heavier then back to light. Experiment with these settings.

So there you have it. A cheap ass tablet, salvaged thanks to the fact that the semiconductor company who made it’s controller saw fit to put the information on the Internet for all to see.

Links

• Download KBGear JamStudio Tablet Driver for Windows 95/98/ME/2K/XP.
• Download the Datasheet for CY7C63001Am USB controller.
• Search for a decent Graphics Tablet on eBay.

Note: I take no responsibility for any damages that might result from you reading this article. You might damage your DVD Player even more from trying to re-enact the steps described below. Think about that before doing it.

The first thing you’ll want to do is open the drive tray on your DVD player. Sometimes it’s tough to get the unit apart when the tray is already all the way closed, so eject the disc, and while the tray is out, unplug the unit.

Next, find a nice comfortable place to sit down where there’s something grounded nearby – a computer desk is fine because you can touch the metal case of a computer (assuming your PC is grounded) and discharge any static electricity that might be in your body. The MPEG decoder and encryption-related ICs inside your DVD player are very very sensitive to electrostatic discharge. I’ve had at least a couple of junk devices where I have taken the lid off, stared at it, poked one or two things, put the lid back on, and the thing’s even worse off – the only explanation I can think of is ESD.

So ground yourself, and try not to scuff your feet on any carpet or anything while the lid’s off your DVD Player – just in case.

Next, pop the lid off the unit and set it and all the screws aside. Locate the power supply area, and stay well away from it. While the unit is unplugged, your chances of dying from an electric shock with the residual charge still in the PSU are pretty slim – but you probably don’t want to take the chances.

See if you can locate the lens and see it without removing anything. If you can, great – if you can’t, you might be able to get to it by moving the “eye” all the way to the end of it’s track, where it would sit while reading the outside portions of a disc. To accoplish this, look for the sprocket which is closest to the motor (furthest from the “rack” gear, or the eye) and give it a gentle turn with your finger. The eye should start moving out.

Some DVD Players (and indeed CD players and other stuff, the GD-ROM in a Dreamcast console comes to mind) use a Worm gear instead of a few gears and a rack. On those, you really don’t have much of a choice except to carefully nudge the eye along it’s track. It will probably make an ugly crackling noise during this time, and double check to make sure nothing’s broken. If you can see the motor that drives the eye, you can also use an AA battery and some wire, after finding the correction polarity, to drive the eye outwards. This mightn’t be a good idea while the motor is in-circuit though (still hooked up to the DVD Player’s “brain” so to speak), so I didn’t go into too much detail with this.

If you still can’t see the eye, you might need to take the actual reader apart. Your experience will vary, with mine I didn’t have to take it apart, so you’re on your own here. A tip though: be very, very careful with the ribbon cables – just the slighest lateral force on them is enough to rip them.

Now hopefully you can see the lens, and it’s time to give it a good cleaning. If you have a steady hand, I thoroughly recommend some strong Acetone to clean the lens with. Acetone is nice because it eats oil and everything else, but care must be taken to ensure you don’t put any of it anywhere except on the lens itself. If it touches plastic, it can misfigure it and cause a mechanical problem. If it lands on the ribbon cable, it can quite possibly eat the ribbon cable and ruin it’s ability to transmit a signal. Use extreme care!

Your other choices are Isopropyl and some dollar-store lens cleaning fluid. I would use the Isopropyl first, but that’s just me – I have no real reason or thinking behind this. It’s cheaper, how’s that.

Whatever you use, dip a Q-tip (cotton swab) into the fluid, then tap it on your pants, a piece of cloth or something else to make sure there’s no drips. Ideally, dip just the tip in, and hopefully the fluid will travel up the rest of the Q-tip through a fascinating process I believe is called Osmosis. You don’t want too much liquid on the swab anyway, but you definitely don’t want it to be dry.

Now swab the lens carefully – one complete turn should be plenty. Resist the urge to use the dry end to mop up if the lens is glistening, I believe you can scratch the delicate surface and scatter the beam worse than dust would. If it’s still glistening, leave it to dry – Isopropyl and Acetone both are somewhat volatile and will evaporate quickly without your help. Leaving it a good five minutes or so should be plenty if you’re concerned.

Once you’re convinced the lens is clean, take the player back to a TV and hook it up, dropping a disc in to see if it’ll play. It helps if you’re still near something to ground yourself on here, so you might want to bring a small TV to your work area. With any luck, the machine will play the disc perfectly. It shouldn’t realistically take any more than 5 seconds for the player to read the “Table of Contents” on the disc (for example, for the “DVD” light to light up). If you’re content with how things are going, button it up and call it a day.

Help! It still won’t play!

If your player is still dead (like mine), then it’s time to recalibrate and see if that helps. There are a few things you should know about blind-recalibration. Firstly, the method I describe here is not the correct way to do it – you need proper equipment and stuff to do it right. Second, it mightn’t be the problem to begin with. Third, it pisses off technicians if it wasn’t the problem, the machine malfunctioned due to some other problem, and you’ve twiddled with the adjustments to make their life harder.

Suffice to say, you should only attempt this on a machine that – if unsuccessful – you are going to throw away. If you have a $900 american-made, progressive scan DVD player – take it to a professional and don’t fuck with it.

Let’s go.

Look on the PCB attached to the lens assembly. Hopefully, you can see it without taking the device any further apart. Some DVD players will have one pot, some two, and very rarely you will have three. To understand how to adjust these, a little background is in order. It should be noted, that this is basically how I theorize it works – it might be incorrect.

The “eye” shoots a laser at a given wavelength (your DVD player will adjust the wavelength to read various “layers” of the disc, but this is unimportant for our understanding) at the disc. The disc then bounces the light back, or diffuses it, for a 1 or a 0 – I’m not sure which way around it is, but again, this is unimportant for the purposes of this text. Next, an eye “reads” the beam of light, and tries to decipher if it’s a 1 or a 0 bit.

This is where we have a problem, because the DVD player needs to know how much light constitutes a one or a zero, and where to draw the line. That’s what the calibration is for – it’s a no-holds-barred discrimination point between one and zero, black and white, on and off. It’s the line in the sand.

Sometimes, the line gets moved. The laser’s properties change. Maybe a component weakens, maybe the power supply weakens and the laser isn’t so strong. Whatever, if the line moves too far, the reader won’t be able to discern between ones and zeros, and you effectively are looking at a blank disc – from the binary point of view.

So, we can adjust the reader and calibrate it again, by moving the tiny potentiometer(s) on the PCB. If you’re really lucky, your unit only has one. If it has two, hopefully they’re labeled (mine was, see blurry image above). What you’re looking for is the pot which discriminates DVD signal. My unit has two pots, one is labeled “DVD” and the other “CD” – so I’m golden. The unit still plays video CDs, so I know to leave the CD one alone.

If you can, the easiest way to adjust it is while the drive is trying to read. I recommend a purpose-built plastic driver used for driving tiny adjustment pots, something with some bend in it. Push it into the pot, holding it very loosely so when the head moves the driver moves with it. Put a disc in the drive, and start rotating it back and forth tiny, tiny amounts. A little goes a long way here, ladies and gentlemen.

If you luck out, the player will act like most I’ve ever messed with, and “flick” the disc in small short bursts of acceleration, followed by the disc decelerating during the periods the motor has no power. This means it hasn’t picked up a bit stream yet, so keep fiddling. With just a small amount of patience and a little bit of luck, you’ll hit the sweet spot and the disc will spin up quite fast, and quite possibly even start playing.

You can either quit here, or try to find the two extremes where the unit stops being able to read the disc, and then aim for the “middle” of the adjustment. I quit as soon as the disc played, but it’s entirely up to you. The theory behind aiming for the middle is should it fall out of alignment again, you have a little bit to go before it’ll stop working. It’s your call.

Hopefully, it works for you. It’s worked for me, several times, and I’ve even used the technique to make a Dreamcast play CD-RWs, which is excellent fodder for a follow up article. Have fun!