Fyber Flash

Nearly a Decade

It's hard to believe that we're coming up on a decade since I've posted here. My life is totally different in so many respects. I would dare say even my beliefs are a bit different than they once were. I suppose ten years offers plenty of new insight to make a person more cynical!

I admit, I miss the notion of blogging. I really did enjoy LiveJournal back in its heyday, as opposed to modern-day social media posts, where I doubt anyone is interested in reading more than a paragraph before they resume scrolling through their feed. I liked seeing friends' posts in chronological order, allowing me to easily see what I've read. Social media is a garble of ads, videos, images, and then occasionally a real post from someone, but only ones it's decided are worthy of me seeing.

I used to post here about projects I've been working on, but I admit that there haven't been a lot. Most recently I reverse-engineered the protocol that my treadmill uses to communicate over Bluetooth, and was able to control it with a Raspberry Pi. I spent several years modding and reverse-engineering Minecraft, but got a bit burnt out on it a few years ago. I competed in Ludum Dare a few times, including making a DOS game. My current job is actually getting paid to make video games, so you can't beat that. I'm sure there were various other projects worth talking about, but the brain meats don't remember so good.

I just don't get a lot of time for personal projects like I used to. Work takes up most of my time. I prefer to spend the evenings with my wife. Weekends tend to involve doing stuff that didn't get done during the week, or trying to visit/help family. I miss certain aspects of how things were when I was younger, sure, but there's no question that life is better now.

Maybe I'll try to post here more often again, even if I'm just talking into the void nowadays. If not, I'll see you in another decade!

Fyber Flash

Hearst Magazine Tomfoolery

I don't write here much at all anymore, but since I know Livejournal posts can still be found through Google, I felt that I'd make an entry about this in case anyone searches for similar cases.

Basically, I just had to write to a collection agency to tell them I have no intention of paying a $15 Cosmopolitan bill. This is a subscription I was tricked into through a contest on Hearst's website. After submitting the entry form, I was taken to a page thanking me for purchasing Cosmopolitan. I was very aggravated with this, and within five minutes I was on their customer page and canceled it.

You would think canceling within five minutes would be sufficient, but obviously Hearst has this scam all worked out. I got the magazine anyway, along with a bill of course. So I put it where it belonged: the garbage. I received a couple more bills from them, as well as emails. But they certainly didn't waste any time selling my debt off to a collection agency.

I did some research online and found that this is not an isolated incident. Perhaps with magazine subscriptions down these days, selling uncollectable debts is the only way they can stay in business.

In any case, I was very polite with the collection agency, explained the situation to them, but made it clear that I will not pay this bill, and was sorry they had acquired it under such circumstances. I pointed out though that I doubt this is the first time they've had to deal with disputed Hearst Magazine bills, and that they should possibly reconsider doing business with them. I also made it clear that if it goes any further than this, that there's more than enough evidence to indicate Hearst's questionable business practices to absolve me from this nonsense.

If they still want to take me to court over $15, they can go right ahead and try.

So this is a warning to anyone who ends up anywhere near a Hearst-affiliated website: close it immediately, before you end up owing them money.

UPDATE: I went ahead and filed a complaint with the Better Business Bureau while I was at it. There's just way too many complaints from people about this place to let it slide.
Fyber Flash



I keep getting email notifications lately about how I have new comments, but it's just a bunch of drug spams and stuff. I finally took the opportunity to delete all that crap. I thought I would spew some words while I was here.

Does anyone even still read these things? I sure don't! Well, once in a while I'll click to view my friends page, but I wasn't logged in so I wouldn't have seen any friends-only posts.

Haven't really worked on any projects worth noting. A few programming-related ones here and there, that's about it. Though my brother gave me his old Droid when he upgraded phones, so now I finally have the ability to tinker around with the Android platform. I've got the SDK installed (mostly for the transfer tool and emulator), and I'll likely be trying to write an app at some point. Maybe I'll think of something useful to make in the process. Either way, the geek in me must admit, I love the idea of carrying around essentially a Linux computer in my pocket. It's rooted of course, and I have the terminal and SSH client and all that good stuff. I also setup OpenVPN in case I ever decide to use any open wifi. And having access to GPS coordinates is also fun, for plotting out where I go, embedding coordinates in photos, and mapping wifi APs around town.

I've tabbed out of this like five times now, so I suppose I'll submit this post before it's lost into the void which is my train of thought!
Fyber Flash

Greasemonkey Geek

I haven't updated this in ten forevers, but I've been messing with a project of sorts that might be worthy of more than 420 characters on Facebook. Who it might interest however is another story!

The thing I've been working on is an old weed eater. Might sound pretty boring, but it was designated as junk by the previous two people who had it, so rather than send it to the junk yard, I thought I'd try to make it run. And the objective is to do so without spending any money!

It's a Poulan GLT-800, which I can't really even find much about. The places who have anything regarding the GLT line seem to only recognize the 770 and 900. The 770 is pretty similar though from what I can tell. I have no idea when it was made, but it looks nothing like modern weed eaters. There's no plastic housing all over it. It's mostly just the motor, gas tank, and a pale green plastic shroud around the front to cover up the clutch, which also has the bolt to tighten down the shaft into it. You don't hold it by the shaft at all. There's an adjustable handle bar attached to the shaft which is half-way aligned with the shaft, and half-way pointed out 90-degrees. Both ends of that have rubber hand grips. The section aligned with the shaft is also where the throttle control and engine kill button are. I supposed this duel-grip handle which kind of surrounds your body more is supposed to give you better control, or easier on your back, I dunno. Since it has the fancy handle and the clutch, I've assumed it's more of an industrial weed eater. I'm more inclined to think that based on where it came from, too.

It wouldn't really do much of anything initially. I tried cleaning the carburetor out somewhat (it uses a fairly typical Walbro, but with no "primer" or return gas feed), but still not much of anything out of it. That's when my dad said we should just throw it out. But I had a few ideas to try on it. And if those wouldn't work, I'd just take it apart to see inside if nothing else. It'd be a learning experience.

One of the first things I found wrong was the spark plug. It was kinda sparking, but I noticed it was mostly happening down near the insulator and on the sidewall. Tried to clean it out some to see if that would help, but no good. So I swapped that out, saw this one sparking properly across the gap, and at that point it would sometimes sputter up better, but eventually go back to needing a million pulls with no result. I figured it was flooding out or something.

Later on I went for a much more thorough cleaning of the carb, which meant taking out the adjustment screws, the circuit plate, everything. The diaphragms aren't in great shape, but they're not terrible either. So I cleared out all the holes with carb cleaner, carefully scraped any crap off the edges to make sure everything had decent seals, tried to blow out the fuel screener, and put it back together. After mounting it back on, I decided I should check for spark first, just to make sure after all the other troubles so far. And there was none, coincidentally. Tried every spark plug laying around, and nothing. So all that carb cleaning was a waste of time at that point because now I had a new problem. I thought maybe the heavy rain had shorted out the coil or kill switch or something, and maybe it'd be fine the next day out in the sun.

Next chance I got I pulled the recoil off the back, which did have some water in it. But I immediately saw that the points (yes it's an old points and condenser setup) were gunked up. So after fighting for quite a while to get the flywheel off, I was able to clean it out, sand the contacts, adjust the gap, and start putting it back together. That's where things kinda went downhill for a bit, because I dropped a washer down inside the flywheel the first time and had to take it all the way back off. The second time, I realized I bent up the points plate with the screwdriver I'd wedged down in there to keep the flywheel from spinning while tightening the bolt. So off it came yet again. Managed to bend that back into shape, re-gapped it, and FINALLY put the ratchet levers and recoil back on it. Put the plug on, checked for spark, and it was good again. So that had been the problem apparently. I'd kind of been afraid that it might have been the coil, condenser/capacitor, or even the plug wire. So that was a step in the right direction.

Well after a few pulls to get it started, I realized it might be flooded again. I pulled the plug back out, noticed it was kind of wet, but like an idiot I decided to test for spark right then. And yes, I grounded it against the metal near the spark plug hole, like a pro. As soon as my brother yanked the cord for me, I got a fireball shot back in my direction. The cylinder had pushed all those fumes straight out the plug hole just as it sparked. I got up pretty quick! Only burned a couple of knuckles, luckily. Not on the hand holding the plug, though, ironically. Took the hair off of that hand some too. I put the ice on it so it didn't burn for too long, and only had one small blister. Totally a stupid place to ground the plug, where as I'd always been kinda careful about that beforehand. Oh well, lesson learned!

On the bright side, despite all the flywheel troubles and nearly burning my face off, I did make it run that day. It would even idle all on its own, albeit roughly. I was pretty proud to have accomplished that much, even if I got no further. But I was having to turn the mixture screws nearly all the way in, particularly the high screw, which meant it was likely getting way too much gas somewhere. So that'd be the next thing to focus on.

Next day I tried bending in the lever for the needle in the carb to lean it out a bit. I thought if less gas was sitting in the metering diaphragm then it might not flood out so much. The diaphragm is slightly stiffer than it should be too, which wasn't helping things. Well, problem was, I practically couldn't get it started then, despite my success the previous day. At one point when I'd put some more gas in it I did get it going better, and got up to some decent power (more than enough to engage the centrifugal clutch and actually hear that familiar whizzing sound of the head spinning fast). But during my screw adjustments it eventually died out, and I just never got it going well again. Wore a blister in the side of my finger in the process. Doesn't feel too great! I just couldn't figure out my intermittent success at getting it to start.

Later that night, I got to thinking. When I'd taken the gas cap off, it hissed, as if there were a vacuum. I put that together with the fact that the couple of times I'd gotten it to run great were after putting gas in it. That made me think that the gas cap was plugged up. Normally they're supposed to let air in but keep gas from going out. The vacuum was probably keeping it from getting enough fuel to stay running. And sitting out in the sun during the day was making that pressure worse, which was probably why I'd been having such a hard time getting it to do anything at the start of each day that I messed with it.

So today, when I got the chance, I went over and twisted the cap off, which released a bunch of vacuum again. Pulled the cord a few times, it sputtered up to speed, and despite not sounding perfect I still got quite a bit of power out of it, all things considered. I twisted the cap back down on it tight, and in a few seconds it sputtered out and died. So, that confirmed my problem, but didn't help me any either. It doesn't use a standard gas cap. So I'd either have to fix that valve in it or just jab a small hole in it.

Later in the afternoon I went back out to try adjusting the screws while running it with a loose gas cap. But I couldn't get it to start. Back to this shit again, it seemed. I decided to bend that lever back in the carb to make it a bit more rich again while I was at it. I also took the recoil off to see if I could figure out what was causing it to catch hard occasionally. Oiled it up and stuff, put it back together, then pulled on it a few times. That's when I noticed that the cord was frayed bad at the very end up inside there, and was starting to hang up. Insert another sigh, and commence ripping the recoil back off, removing the wheel and rope. Looked like it had been messed up in there for a while, since there was a piece of rope really crammed down in the channel on the wheel. So I chopped it off, re-knotted it, rewound it, fiddled with it for a while before I could redo the tension on the spring properly, but finally finished that. Though it still didn't want to work perfectly every time after I started trying to start the thing, since now the cord sometimes hangs out about half an inch or so after some pulls. I dunno what's up with it.

Eventually I did get it to start (which seemed a little easier now), and started tweaking the screws. Got some decent power out of it again at one point. But by accident I noticed air coming out of the carb opening. A lot of air. And sprinkles of gas. Yet another problem! So, turned it off again, and started trying to diagnose that. Took the exhaust off, but the exhaust port looked really clean. Cleared out the muffler as best I could with a screwdriver, stuck it back on, but no difference. So I put it away for today. This is obviously why I can't get it to run smoothly all the time. It sounds like it misses a spark sometimes, especially when it starts to get real choppy, which is usually the result of this kind of back-pressure messing up the carb.

So the potential issues now are: bad fuel pump gasket/diaphragm, bad intake port gaskets (it uses a stupid carb adapter so there's two gaskets, and yes they're really worn gaskets), or a bad reed valve/cylinder in backwards/who knows. I don't know what type of induction this engine uses so I can't determine the latter potential problem at this point. I don't even know if a two-stroke piston port induction engine would run at all with the cylinder in backwards. But based on the intake/exhaust arrangement, I'm more inclined to think it uses reed valves. If they're broken/missing/whatever, then I'm basically fucked, since I didn't want to put money into it.

What I'm going to do is loosen the carb, then slip a piece of cardboard between the carb and the carb adapter to block off the impulse jet. Then I'll pull the cord over and few times real fast and see if I feel any air coming back through the carb opening. If so, then it's probably a problem in the engine itself (unless the air is seeping through the gaskets and into the carb throat). If I don't feel anything though, then it's probably the fuel pump gasket/diaphragm leaking the impulse pressure out through the carb. That's also a problem since I didn't want to spend money, unless I can find a good one on one of the spare carbs laying around.

There's a couple of other issues too. Sometimes when I pull the cord it feels like there's no compression. Then immediately afterward it'll feel like there is again. So I don't know if that could be bad rings or a bad cylinder or crank gasket or what. Also, after repairing the recoil today, I suddenly started to notice that the cord had oil on it. I'm almost positive that it didn't when I as fixing it. That might mean the rear oil seal on the crankshaft is blown. Mighta happened from all that beating on it the other day to knock the flywheel loose. I don't think that would be why I'd loose all compression though, since I should still be feeling it from the cylinder. But, when I cleaned out the points before, it seemed a little greasy in there too, so it may have already been doing that to some degree. If so, it might lose spark again if it gets too gunked up.

One thing I'm tempted to try is cereal box gaskets. I've heard they work in various circumstances. I might put'em on the intake, and maybe even on the head and crank if I pull the entire engine apart. Might even fix the compression issue!

Anyway, this is a huge post, but I was catching up on many days worth of work. Any follow-ups I do won't be nearly as long.
Fyber Flash

(no subject)

I suppose I can post the videos I made recently of my AVR video project. They're not much to look at, particularly since I have a hard time filming it without the picture looking overexposed (which it still does a little even in these, because the TV image has crisp edges). But they're enough to demonstrate what my code is doing for now.

The first one is of my 16x8 text display, using a 4x6 font. I only put in 128 possible characters, and some of those aren't even there so I just used black and white blocks for the time being. But the alpha-numeric ones are the most important anyway. I'm just incrementing through all the possibilities, at a slow enough rate to view them. It's also showing off the text scrolling capability I added.

The later iteration still had the ability to draw text and everything, but it's not doing it anymore. It's basically a simple Pong game, using a potentiometer (a knob) to control a paddle on the screen to bounce the ball around. There's even sound, using a roughly 1300hz tone through the TV speaker. You can hear it faintly through the terrible buzzing sound of the camera.

I thought I would have a problem cleaning up the pot input, because the paddle was jittering around at first. But a smoothing capacitor took care of that. A neat thing I found was that using a cap that was too high in value would actually make the paddle glide to position rather than respond instantaneously. That could be neat to implement in something in the future.

The sound is a simple square wave generated by one of the PWM channels of the chip. Not much needed externally for that, aside from resistors to limit the output and a capacitor for coupling. Though since it uses the same pin as one of the connections used for programming it, I had to stick a diode in too. Every time I program it now I hear a neat little electronic spurt through the speaker briefly.

This is still using the 64x48 pixel display as from before, just to clarify. It takes 768 bytes for video memory, since each pixel takes 2 bits (allowing four possible colors, of black, white, and two shades of gray). The ATmega8 I'm using to do all of this only has 1KB of RAM, which is a huge limitation in the quality of the video I can generate. I mean, it's possible to generate higher resolution video lines on the fly if I wanted, but that's too much like Atari 2600, where you figure out what to draw on each line when you get to it. In other words, too much work. I just wanted a bitmap-based display, where I could have a code loop running which just writes pixels to memory locations, and then the video interrupt handles displaying that independently.

If I went to 1-bit-per-pixel, that would still just allow something like 128x48 resolution with the way things work now, and limit me to just two colors. Even using every single byte of RAM and limiting the functionality of the game code would manage 128x64. Just not worth the limitations it would cause.

Anyhow, I want to put this all in a project box, running from a battery, and with a couple of RCA jacks for audio and video sticking out of the top. I have a box like that laying here which I ordered years ago. Just needs a hole drilled into it for the pot's stem to come through. I also need to find a knob for the pot. It's almost worn a blister on my finger from turning the stem so much, since it's kind of stiff. I stuck a piece of plastic on it for now which makes it a bit easier. I'm just not sure who even sells individual knobs that will fit on that, other than like Ebay. I'd rather buy it local though so I can make sure it fits first well.

I still need to clean the code up too, maybe add in a scoring system, maybe an AI paddle, I dunno. But I still feel like I accomplished my original goal(s). Not only of a generating video, but of a text display, as well as a simple Pong game. So anything else I add is just icing on the cake. It's been a good learning experience and confidence booster, and I feel like now that I'm comfortable with AVRs enough that I can do most anything I want in the future with them now.
Fyber Flash

Later, Juno

I just can't in good conscience continue to pay Juno the yearly fee for their MegaMail service anymore.

I've probably complained here about them before. When I was trying to gift a Steam game to my brother once, who at the time still used a Juno account as his primary address, the link in the email to redeem the game simply wouldn't work for him. And Juno offered no way to view the email source in some manner so that we could manually fetch the link. It was literally impossible to do what we needed to do, until we finally were able to forward the email to his Gmail account and do it from there. Even forwarding took a few minutes to go through. That whole debacle really aggravated me with them, and made me see just how primitive their webmail service was compared to every other free email provider out there.

For the record, the only reason I've continued to pay them the $10 a year up till now was purely for the pop3 email access. I wanted to pull my emails into a client of my choice (Gmail these days, with my other mail) rather than using their oudated Juno client or the aforementioned terrible webmail client.

But yeah, after that episode with my brother's account, I had decided that maybe I didn't really want to continue paying for the MegaMail service, and that I'd just finally transfer everything over to Gmail and let Juno revert to a free account (aka no pop3 access anymore). Well, not so fast. Ebay won't let you use the same userid for their site as your email address, which I guess I can see the reasoning behind, but it's nonetheless very annoying. So I thought I might be stuck with Juno. But when it was time to pay the yearly fee, my previous credit card had expired and I never put the new card number into their site. So they started emailing me to update my information. I just never updated the info. And Gmail continued to pull in my Juno emails via pop3 for about five more months..!

One day I realized I hadn't been getting emails from that account anymore. Juno's pop3 service is stupid and doesn't actually give you an error when you don't have pop3 access; it just always says 0 new mails. So I checked the Juno webmail, and sure enough, there were a bunch of mails sitting in there which I had missed. So, since I couldn't transfer my Ebay email setting over, I kind of reluctantly went and updated my credit card info at Juno, then messaged their billing department (even though there was no actual "email" product to select in their shitty request form) to ask if they could reinstate my Megamail service.

I also pointed out to them that there was no longer a way to change your password. I'd tried to do this once before, actually, but it never took. Now there is literally no place on the website to do it; the one link you can find by digging just redirects to the password request form, and the help is still outdated and tells you things which don't exist. This all really bugged me. I pointed out to them in my message that things like this made me question whether I wanted to continue business with them. But I did so in as polite a way as possible, because I didn't want to sound like a jerk about it, but that I had valid concerns and all.

So a couple of days later I get two separate emails from two different people at Juno. Neither of them appeared hand-written, and seemed to just be form letter help responses. One of them told me what I already knew, that my account had been return to "free" status since my credit card had failed to charge. It said I would have to sign back up for MegaMail, and at the higher yearly rate now than what I used to pay when I originally signed up. The other one told me that there is in fact no way to change your password on the website anymore, and that you had to download their client to do that.

What. The. Fuck. You seriously have no way to change your password on the website? Who is programming this shit? I was writing code that could do that when I barely had a grasp of MySQL.

So I decided fuck this, and I was going to just associate Ebay with a separate Gmail account if I had to. I wasn't going to pay these people for this terrible shoddy service anymore. Though I then found a trick I could use to tie Ebay to my existing Gmail account after all. And so that's what I did. Along with the other things I currently know of that are still connected to Juno. I think I got most everything changed, but I guess I'll see over time.

Out of curiosity, I did download the latest version of the Juno client. Which, mind you, is still version 5, which is what it's been for the last 10 years I seriously believe. They've probably only done jusssttt enough to keep it running properly in newer versions of Windows, I guess. It looks identical to the version I used over 10 years ago, I know that for a fact. It also operates the same. It has no filtering capabilities, no spam handling, no nothing. What a piece of garbage for this day and age. I did manage to change my damn password with it, at least.

Since Juno offers no way to export emails of any kind (neither in the webmail nor the client), I'm going to manually go through and forward every email worth keeping to my Gmail. I'm probably over half-way through with that.

I'm just really annoyed by all of this. I've been a customer of theirs for a long time. I used free email starting in about 1997 or so, then we actually paid them for internet starting in about 1999, and when we got cable internet around 2004 or so I paid them for the MegaMail service ever since. And this is still the shit they offer their customers. Ten-year-old products with less features than their free counterparts, and a shitty support center on top of it. Hell, even to request your password, you create a four-digit pin and have to email them and they will manually help you. Even the dinkiest of free blog and forum software packages have a password reset functionality.

So yeah. I should have just never paid for Megamail to begin with, since it was basically just me rewarding them for having an otherwise terrible email service so that I could still use my email account in a better third-party product.

It's all really just a shame, because I still have fond memories of using Juno back before I had internet. It was a big stepping stone for me between the BBS and internet periods. But nostalgia and brand loyalty aren't enough to make up for their shortcomings. If your pay product isn't even as good as free alternatives, then something is seriously wrong with your business model.
Fyber Flash

(no subject)

I was going to type up a thing documenting the video project I finished, in case anyone stumbled across it and wanted to learn how to do such things themselves. But ehhh, I'll work on it later. I'm more in just a mood to ramble from a personal point of view.

Using only a microcontroller chip and custom software I wrote, I was able to generate a composite NTSC television signal.

This is something I've wanted to be able to do for a long time now, and due to a combination of things (including laziness), I never did. I tried once a long time ago, but it didn't work, and thinking back on that now I probably didn't even fully understand what I was doing. I've learned a lot about NTSC video from various sources over all this time, so I realized I had the concept down, at least. But I had to teach myself how to use this particular brand of microcontroller to even do this (since it's the only one I had handy), which was another reason I never did it sooner.

I used an ATmega8 microcontroller from Atmel running at 4mhz. I did't use any interrupts in my code, so everything is carefully instruction-timed. I managed to squeeze in a 64x48 pixel resolution, using three possible colors: black, gray, and white. It's even technically a bitmapped display, because the pixel data all comes from RAM, and can be changed during retrace periods and the code will then display the new content appropriately at the start of the next field (at 60 fps).

To give you an idea of just how little time you have to work with, a complete horizontal line of NTSC video plus the retrace period takes 63.5 microseconds. At 4mhz, each instruction executed by the microcontroller is 0.25 microseconds. And you need more than one instruction to be able to toggle output pins of the microcontroller to control the color/signal being output. So I was lucky to get those 64 pixels horizontally!

The only reason the screen is vertically limited to 48 pixels is because of RAM limitations, and me using a bitmapped display. I designed it to use 2 bits per pixel. So that's 4 pixels per byte. At 64 pixels per line, you're using 16 bytes. A normal TV is capable of 242 lines of video, but just say I decided on 192 (since some of those lines are off-screen). 16 x 192 lines is 3072 bytes, or 3KB. The microcontroller I'm using has 1KB of RAM. Not gonna work. So what I did was to clone each line 4 times, effectively dividing that 192 lines by 4, giving me the 48 vertical resolution. 16 bytes per line times 48 lines is just 768 bytes, which will in fact fit in our 1KB of RAM. And best of all, 64x48 is a 4:3 aspect ratio, just like the TV.

Somebody somewhere reading this might be thinking thinking "2 bits per pixel allows 4 possible values, so why did you only use 3 colors?" And the answer is that technically there IS a fourth value. But that value outputs a sync signal on the line, which is only used in retracing the beam. So yeah, you could specify that value in RAM for a particular pixel, but you would just succeed in messing up your display.

I tested my timing in the simulator that comes with the assembler I'm using, and it was perfect. I was pretty confident that this would work on the first try once I got the chip programmed and connected to the TV. And I was kind of right; I just didn't realize it at first! I had an issue with the reset line of the programmer keeping the chip in reset, and me not realizing what the problem was for a few minutes. Once I remedied that though, it sprang to life, and I was a bit surprised at just how good it looked on the TV screen. I could never take a good enough photo of it.

So yeah, I have to say, I'm proud of it. Sure, it doesn't actually do anything visually, aside from show the Fybertech logo in a couple color shades with a border around it all. But there's something very awesome to me in being able to control a CRT. I'm literally telling it when to intensify a beam of electrons at a particular microsecond in time, when to retrace the beam across the screen for each line, when to retrace the beam back to the top of the screen, etc. LCDs are just not cool to me in that sense, because they don't actually do anything exciting.

Anyway, I didn't do color because that's a whole 'nother ball game. I doubt I'll even try to do it without assistance from another chip which will handle the encoding for me, because you start getting into quadrature encoding of signals. Fortunately for me, the chip I want to use is the same chip I plan to use for converting the output from my Karate Champ arcade board to work on the TV. So I should be able to rig up a single converter board and use it for various things like that.

I guess that's it for now!
Fyber Flash

(no subject)

I was just mentioning in chat how my tiredness has left me not wanting to work on any of the projects I've had in mind lately. Dunno if it's from not sleeping well or what, but even when I take naps, I still feel tired afterward. It sucks. But anyway, that made me want to make an LJ entry to post some of the stuff on my mind lately, project-wise.

- On-hook audio. I wanted to make an interface to listen to the phone while on-hook, to record caller ID tones, but never had the parts. I desoldered some of them from that cordless phone I was messing with, but not all of them. That was because I also had another project in mind, which I thought might be more interesting to use the phone for.

- Wireless serial connection. I wanted to use that cordless phone and try tapping into its data connections (which it uses to send messages between handset and base) and see if I could make a bidirectional wireless serial interface, which could be used with like my Palm or the Mailstation to get data in other parts of the house or out in the yard. With the Mailstation, I could just send modem tones for that matter, but that would require a modem on the other side too, and a way to initiate the dialing. So it seems simpler to just use the data capabilities of the phone. But it's not something I've messed with recently, so it might be on the back burner for a while.

- Floppy drive audio. This was kind of a bust, since I couldn't ever get a good recording or playback. I might need to erase the media first, but that would require an oscillator of a hundred kilohertz or so, which I just never bothered rigging up. I also thought that an AC-biased recording source might help with recording it better on the disk, but again, you need an oscillator for that, and well as a mixer circuit. So for the time being, this project is mostly over, until I get some hankering to mess with it again.

- AVR video generation. This is something I've wanted to do for a long time but just never did. Or well, I tried once on another chip, but I apparently messed something up. I'd like to start with just some columns of white on the screen to see if I'm doing the timing right. From there I'd like to try making a Pong game, and eventually a simple video game system. This is the sort of project I'm most likely to work on next, based on my mindset as of late. There's just something about being able to control a traditional CRT display that's neat to me. But, like with most things, I have no dedicated programmer for AVRs, so that means rigging it up to my parallel port and hoping it works properly.

- NES computer mod. This is that thing I started like 2-3 years ago and never finished due to the dozens and dozens of wires that needed to be soldered. It's where I was basically going to add a BIOS and some other functionality into an NES to make it more like a computer, but still let it play cartridges. In recent weeks I've considered changing out a ton of that logic (which required so much soldering) with one or two of these GAL chips I ordered from Ebay when I saw some on there for cheap before. I've even "compiled" some of the logic necessary for putting on such a chip. The problem is in programming them, because they use various voltages between like 8 and 16 volts, depending on the chip, and so you need a variable power supply set just right after you determine the chip type (via software). I have a variable voltage regulator now, but just hooking this mess up to try writing to one of the chips, and then to rig up another mess to test the chip out, has just left me meh about it all for now. A pre-built programmer would be great, but they're like $50 minimum for a cheapy one. But the NES computer thing is still something I want to do. I even recently ordered some PS/2 keyboard/mouse controller ICs which would work great in such a project.

- Arcade game replica. The idea started with wanting to build my own copy of the Galaga hardware, which could run the original arcade ROM images. It requires at least three Z80 cpus (four if you use the bootleg design of a fourth Z80 in place of the custom IO chip), which I actually have that many of, but that would take so much wiring and parts. I'd almost definitely end up buying stuff. Then I looked at Pac-Man, which seems easier. Galaxian might be even easier than that. I think it would be a cool project, since recreating things is always a lot easier for me than making something from scratch. But it would still take some work, including adapting it to work on a TV rather than an arcade monitor. I'd also love to have a cabinet to go with such a thing, but I simply have no room. So just like a mini console system would be fine, even.

- Fybertech redesign. Yes, the site has used that same stale design for a long time now. I used to change it up a lot more often, but I just haven't had much urge to do anything with it. 4get is really the only thing used on there on a daily basis, so I thought about trying to make the website based more around what people want and use rather than stuff like news posts which never get updated. I'm not really sure of a design as of yet, but it's been rattling around in the back of my head lately, so maybe I'll come up with something sometime.

That's all the projects that have been in mind as of late, I think. I want to do all of them, but to be honest I would rather go take a nap..!
Fyber Flash

(no subject)

I seem to have abandoned my long-winded posts here, as less frequent as they even are lately, for 420 characters or less at Facebook over the last couple of weeks. I dunno why. I think the ads and stuff have just kind of turned me off to LJ lately. Not being logged in when checking the site lately (ever since installing Windows 7) has put lots of stupid overlay ads on top of the screen, which is combined with the already existing sidebar ads. I've stuck with this place for years, even when most other folks I know have long since abandoned it, but the ads are getting a tad out of control. When I'm logged in I don't seem to see them, but that doesn't change the fact that they're there for anyone else who stumbles across my journal when searching for any of the project-related stuff I post on occasion. I think their attempts to pull more money out of this place are just going to backfire in the end.

Anyhow, speaking of projects, the main thing I've worked with on and off lately is telephone circuits. Primarily, I've been trying to modify this old cordless phone I had laying here to be able to hear the line while it's still on-hook. That should let me hear caller ID tones, which I want to be able to record and try decoding in software on the PC. I may have mentioned it before, but caller ID uses the Bell 202 standard, which is basically 1200 baud modem tones. So I should be able to figure it out in software, and then parse the data it sends.

I used to think that even when you blocked your number with *67 that it still went through to the other side, and that the Private flag sent in the stream was supposed to be honored by the device and just not show the name/number. But I've found out since then that it's usually stripped out at the phone company before it reaches you. In most, but not all companies, that is. I doubt mine will still let it through, but it'll be interesting to see. And regardless, I think it'll be neat just to be able to decode the data myself. And to be able to hear the line without actually answering it, especially! I've heard that you can actually carry on a conversation with the other side between rings. Or at least it used to work that way, dunno about now with the updated phone systems and all. And I would be willing to bet it wouldn't work with the cell phone network at all.

Anyhow, so far all I've accomplished is making noise. I thought that the low-ohm resistors on the secondary of the transformer (in the base station) were what was making such a high load and taking the phone off-hook after the relay activated, but removing them and either leaving them out or substituting a higher value just makes the feedback in the handset. It wasn't even plugged into the line, just turning on the handset caused it. I think I know why though, and adjusting the receive level screw on the base station's board fixed it for now (until I replace the original resistor values). Regardless, the phone was still going off-hook.

So the next thing to try is a resistor in series with the line on the primary side of the transformer. But I don't have any resistors higher than a quarter watt, and I'm kind of afraid to put those on there. The phone line gets upwards to 100V during ring, and could potentially hit like 250V during line testing, so they need to be able to handle that. I don't have very high values handy either, so if I double them up to increase the wattage then I lose a lot of the resistance value. And don't think I have enough for that. I may also need a higher value capacitor too, because I think the one the phone already uses is intentionally too low to let audio frequency pass. Originally I was just going to pull all the parts off that I needed and make the phone interface separate, and at this rate it may be what I have to do. It would just be much handier if it were still in the phone though.

But even if I do pull those parts off, the phone itself will still be handy for other things. I've been looking over the schematics (I found the service manual on the internet, since it was made back when people still repaired things apparently) and I think I can remove the loopback circuitry, making the send and receive channels act independently. That would let me transfer data between the base and handset wirelessly to something like the Mailstation much more easily than trying to modulate it like a modem.

One neat thing about this phone is that it has a test mode built into it. The service manual details how to get into it and what you can do with it. The handset is particularly handy, because once in test mode it spits out anything on the channels that it's capable of receiving, and you can switch between the channels at will, including four which are usable only in test mode. With a bigger antenna and the oscillator adjusted, I might be able to receive 6-meter amateur radio, since this phone operates directly below there in the 43-49mhz range. The schematics seem to indicates that it would be somewhat easy to add a speaker and mic connection to the base station, too.

Anyhoo, I probably would have made more progress but I've just been tired a lot lately. Bored and tired are a terrible combo. There's things I want to do but just lack the energy and gumption to do them. But oh well, what can you do. I'll do it when I do it!
Fyber Flash

(no subject)

Bored bored bored.

I've felt too tired/blah lately to work on much in terms of projects. I did end up fixing up an old machine I had laying here and installed Windows Server 2003, though. It's just 1.1ghz with 512mb of ram, but it's fine for 2K3. I'm not sure what I meant to do with it, though. Kind of a secondary computer I can leave video streams running on and not tie up my main machine when I want to run games and stuff, for one. My E-Vectra is just 600mhz and runs Linux so it's not as equipped for such things. I used to stream TWiT to my TV with it, and it just barely kept up with the stream.

I installed Windows 7 on my brother's PC for him earlier, too. Had some problems resizing partitions on his drive, though. GParted was being a dick about the partition size I wanted, presumably because there was data still in that upper section and it's too primitive to know how to move it down into the lower part of the drive. Kind of amazing that Linux has gone this long and still doesn't have a decent non-destructive partitioner. I tried going to the trouble of downloading the newest GParted LiveCD even instead of using the one on the Ubuntu install disc, and it still gave the same damn error. I ended up installing Partition Magic on his machine, and had it repartitioned just fine in the same amount of time it took for me to dick around with the open-source junk. Now he's able to run Just Cause 2, which is one of the first games to REQUIRE DirectX 10+. Lame, yeah, but oh well.

Earlier I realized I still had an OSX partition on my drive, so I dumped that. I also have an Ubuntu 8.10 partition, so I booted into that again too, and thought I'd do the distro upgrade while I was at it. I dicked around for 45 minutes or so, messing with Facebook and whatever else to kill time. It finally finished, I reboot, and now the X server is totally fucked up. It goes to garble and freezes the entire machine every time. I tried deleting the config file, but it just freezes without showing the garble instead. What makes matters worse is that it was already fucked up from the LAST time I updated it; totally broke the network connection, and the fix I found on the internet (after rebooting to Windows) was to break the normal connection manager Ubuntu uses and set it all up via the console (like I do on Debian and elsewhere anyway). So fuck it. I think I'll just move all my personal data to my Windows partition and dump the whole damn mess. I don't need Linux on this particular machine anyway.

With the extra space, I might give Windows 7 a try myself. I can't sit on XP forever, after all. Not sure how it'll do with just my 1GB of ram, though.

I spose that's enough rambling for now!