Jump to content


  • Content Count

  • Joined

  • Last visited

Community Reputation

241 Excellent

About MaximRecoil

  • Rank

Profile Information

  • Gender

Recent Profile Visitors

9,240 profile views
  1. If you're including JAMMA-type power supplies as being hard to source, they aren't, since they've been an industry standard for over 30 years. Plus, they all have a -5 VDC rail, and their screw terminals make it very easy to connect them to anything.
  2. That isn't lag. You're talking about the vertical refresh rate, and there is no inherent specific amount of time that it takes a CRT to draw a frame. The number you mentioned applies to a 60 Hz vertical refresh rate, but the refresh rate is determined by the video signal, not by the CRT. The best direct-view CRT monitors had about 400 MHz of bandwidth, which means that if you fed it a standard resolution like classic consoles and arcade games used (in the neighborhood of 320 x 240), it would have enough bandwidth to sync to a video signal with a 3,500 Hz (3.5 KHz) vertical refresh rate. It's zero because refresh rate has nothing to do with input lag. A CRT responds in what is effectively real-time to the video signal, and it is always responding to the video signal regardless of what position the electron beams are on the screen. Of course, nothing actually happens instantaneously, so it isn't truly zero input lag, but it happens at the speed of electricity through its circuitry, which might be a nanosecond. There is no hyperbole; 9 ms is 9 million times longer than 1 nanosecond. An SED couldn't equal a CRT's lack of input lag. SED was a digital display, so it would have input lag in the millisecond range like any other digital display. In terms of picture quality, the SED could equal a CRT because it's using the same electron-beam-exciting-phosphors method of generating light, and it could exceed it in terms of geometry, purity, and convergence (which would all inherently be perfect on an SED, while close-enough-to-perfect-that-you-can't-tell-it's-not-truly-perfect is the best you can get with a CRT). An SED wouldn't make for a good replacement for a standard resolution arcade monitor for a similar reason that high-resolution CRT PC monitors don't make good arcade monitors. But an SED would be even worse than a high-resolution CRT PC monitor, because it is a digital display, which means it has a fixed resolution. To get a 240p image to fill a screen that has a fixed resolution of say, 1920 x 1080 (or even worse, 3840 x 2160), it has to be scaled, and scales images look like crap. CRTs have no fixed resolution, so any resolution they can sync to, they can display fullscreen without any scaling involved.
  3. Yeah, that type of power supply has been the standard for video arcade machines since the JAMMA era began (1986), and the Happ Power Pro has been a popular one for a long time. The other popular one, at least with arcade operators, is Peter Chou from Betson Imperial: https://www.betsonparts.com/amusement-redemption/power-supplies/15-amp-power-supply.html Those were the original design that others copied, and popular enough that some people use the term "Peter Chou" as a generic term for that type of power supply, like people using the term "Kleenex" for any brand of tissues. Here's some information about them from arcade guru Ken Layton: However, they aren't as popular with regular people because Betson is the primary retailer for them and if you're not an operator with an account with them, you'll end up paying big money for shipping, and they're one of those annoying companies that won't tell you the shipping cost upfront.
  4. A JAMMA type power supply like I linked to above would give you all the amps you could ever want (15 amps on +5 VDC), but that little Mean Well wall adapter power supply only has 2.5 amps on +5 VDC, and the TI's motherboard will be drawing about 1 amp of that at idle.
  5. Or, you might find that that little wall adapter-type, probably made-in-China, replacement power supply isn't as robust as the original. Also, the motherboard in your TI is just as old, and has a lot more likely points of failure, especially those TTL chips, and in my experience with TTL chips from the 1980s, they fail like it's their job. What do you see on there that's likely to fail, especially after you've replaced the 6 electrolytic capacitors? Most of those components, such as the coil, diodes, and resistors don't fail very often. That DIP14 IC is an 89-cent part. Those linear voltage regulators are cheap too, as is that transistor. You could replace every component on that PCB for not a lot of money, with the possible exception of that DIP8 IC, which doesn't seem to be available new anymore and might be hard to find. If I were concerned about it, I'd replace the electrolytic capacitors, the TO-220 package linear voltage regulators, that TO-92 package transistor, and the two ICs (or just the DIP14 one if I couldn't find the DIP8 one for cheap). That covers everything on the power supply PCB that contains an electrolyte or silicon.
  6. I don't know if there are any new power supplies made specifically for the TI or not, but the original one only outputs +5 VDC, -5 VDC, and 12 VAC, so you could use any JAMMA-type switched-mode power supply (like this one - http://www.arcadeshop.com/i/804/switching-power-supply-15-amp.htm - for example), which would mean you wouldn't need the external brick anymore. However, unless you're having problems with the original power supply, why replace it? The original power supply is a very simple design; not much on there to fail: http://www.mainbyte.com/ti99/hardware/power_supply.html If you wanted to replace the electrolytic capacitors just for good measure, it wouldn't take long to do because there aren't many of them on there. Use first-tier (e.g., Rubycon, Nichicon, United Chemi-Con, Panasonic, Sanyo), high-temp @ long-life capacitors and it will probably last for a very long time.
  7. That's awesome, though I don't think his was that clever. He typed it in while we were at lunch or on recess or whatever, and then just typed run when we were watching the screen. The rocket scrolled up and off the screen and that was it. A bunch of blank print commands after line 180 will make that happen.
  8. The first place I ever saw a TI-99/4A was in fifth grade (1985); there were two of them in the back of the classroom, connected to a pair of 12" B&W TVs. It was strange because there were no other TI computers in the school; there were only Radio Shack TRS-80s and Apple IIs. It turned out that they weren't school property; they belonged to the teacher; a fact that he pointed out whenever students felt entitled to use them. We did get to use them sometimes, but not very often. Parsec and some text adventure game that loaded from a compact cassette tape were the only software I ever saw him run on them. He also typed out a BASIC program once that created an animation of a crude-looking rocket taking off, and challenged us to figure out how he did it, but no one ever did. In hindsight, I'm pretty sure he did it purely with print commands, i.e., something like this... 10 print " /\ " 20 print " / \ " 30 print " | | " 40 print " | | " 50 print " | | " 60 print " | | " 70 print " | | " 80 print " / \ " 90 print " | | " 100 print " | | " 110 print " | | " 120 print " | | " 130 print " | | " 140 print " | | " 150 print " / \ " 160 print "| |" 170 print "| |" 180 print "|_/\__/\_|" ... followed by enough blank print commands to make the "rocket" appear to scroll up and off the screen. We knew how to make the crude rocket with print commands, but none of us could figure out how to make it "take off."
  9. When tracing, you're inherently going by eye rather than inputting values, because there's no way to know the exact mathematical values of say, a bezier curve in a raster image of a letter or number. A raster image isn't even particularly precise to begin with, especially one that's a photograph/scan of already-printed letters (as opposed to one that was created from scratch in a raster program such as Photoshop). When you zoom in on any raster image you quickly see that its outline is very vaguely defined and jagged, due to the nature of pixels, so human interpretation is necessary/inevitable when tracing with vector lines. Additionally, already-printed letters (scans of which being what you're normally working with when doing reproduction work) already have several generations of loss built into them, and your scan creates yet another generation of loss. In some cases, such as with elements of letters that were obviously intended to be straight lines or circles / circular arcs, geometric perfection is easy to achieve with any vector or CAD program (though dimensional perfection is impossible using only information from a scan, because of the outlines of a raster image being vaguely defined), but you'll never get exact duplicates of the bezier curves that the original designer of the letter drew (except by pure one in a "zillion" chance) because the raster image you're working from isn't precise enough. Since tracing of raster images is inherently done by eye with a fair dose of human interpretation, AI is plenty precise for that, and its pen tool is the ultimate tool (IMO) for tracing bezier curves and other irregular shapes by eye. Many, if not most, commercial fonts were drawn in AI to begin with. That's especially true for Postscript fonts, since AI was the original GUI for Adobe's Postscript language, which was the original standard language for creating modern vector fonts (as opposed to the old raster fonts which are rarely used anymore).
  10. You can do all of that with AI too, though AI doesn't necessarily have built-in tools for doing some of those things. There are various ways to get precise and consistent spacing, such as by setting the nudge value to the exact spacing you want, or by making squares of a given size to use as temporary spacers that you can snap to, or using the distribute spacing tool, etc. If you need a perfect arc you can just use the circle tool and then cut out a section of the circle to join into your tracing. If you have multiple letters that have parts which are supposed to be identical, you can copy and paste the identical sections to ensure they are identical (there are other ways too), and so on. I've done countless tracings in AI, many of which have included tracing letters/numbers/glyphs, both commercially and for myself, for about 16 years, and when you've spent that much time in AI you naturally find ways to work around its ostensible limitations. When it comes to tracing letters/scripts which have bezier curves rather than just circular arcs (such as the Coca-Cola script logo), AI's pen tool is far better for that than anything in any CAD program (it's better than any other vector graphics program's version of the pen/bezier tool too, for that matter, in terms of overall functionality/usability). Also, most of the popular commercial fonts aren't of the geometric type (an example of a geometric font is Avante Garde, like the font used for the classic Adidas logo). Most of them, like Helvetica, use subtle bezier curves. For example, this is a closeup of the top right-hand corner of a capital letter R, Helvetica: In any case, making an actual font file from the tracing isn't usually needed or even beneficial for reproductions, because you can just print out the tracing directly, either as the film positive in the case of a screen-printed reproduction or as the actual print in the case of a digitally-printed reproduction. I can't remember if I've tried PStoEdit or not. I know I tried quite a few ways to convert AI/PDF files to .DWG the last time I was working in AutoCAD and wanted to use some of the elements that I'd created in AI. AI's own .DWG export function worked somewhat, but had issues. I eventually settled on pdf2cad, which is an expensive commercial product, but the free trial lasted long enough for the project, and it worked perfectly; zero issues.
  11. That's what I do when I can't find an exact font match when I want to reproduce something, except I trace the letters in Adobe Illustrator rather than CAD software. There's no need to make an actual usable font file from the tracing, though I did do that in one case when I was replicating an old screen printed T-shirt that I had as a kid in the '80s. Being able to type the letters made replicating the text layout easier, because some of the text was in an arch (Illustrator has tools for typing along an arched path). I used an ancient version of Fontographer if I remember right. It's just a matter of importing the Illustrator files; it can use them natively because modern fonts and Illustrator files are both vector; in some cases, the exact same vector language even (Postscript). By the way, did some of them come without the white label strip above the number keys? I have one that doesn't have it (it doesn't have the "Solid State Software" label either), and I've seen pictures of others that didn't have it too, like this one. On mine, there's no adhesive residue to indicate that a label was ever there, plus the way the plastic is molded in that area is significantly different than on my other one which has both the label above the number keys and the Solid State Software label.
  12. Yeah, that's why I'd only do those things if I got one for free. The total cost of all the stuff I mentioned would probably be around $100, which includes $60 for a new Happ 2¼" trackball (the cost of real arcade controls would be far less on say, a Pac-Man; just 1 joystick and 2 buttons which would be about $15). I already have several spare 17" CRT PC monitors lying around, but those are easy to find for free last I knew.
  13. Karate Champ, when I was 9, in 1984. That was the first arcade game I ever played more than once, and the first one I ever got good at. It didn't stay my favorite for long though, because pretty soon a brand new Punch-Out machine showed up, also in 1984, and that quickly became my favorite. That remained my favorite until Super Punch-Out showed up at Fossa's General Store in 1987. Karate Champ hasn't aged well for me, but Super Punch-Out is still my all-time favorite arcade game, and Punch-Out is my second favorite.
  14. Yes, it's usually a major project, but there's no known risk of death, i.e., the people who subscribe to the old "CRTs can kill you" wives' tale have never been able to produce a documented case of it ever having happened. The biggest danger from getting shocked by a CRT is not the shock itself, but your reaction to it. For example, while reacting to the shock you might trip over something and fall down, which could result in injury or death. But that could happen if someone sneaks up behind you and yells "boo" too. I wouldn't want a Sony for an arcade machine. Their Trinitrons had cylindrical, and later, flat-faced picture tubes, and an aperture grille. Neither of those things look right on an arcade machine, because they used standard spherical picture tubes (made by companies such as Sanyo, Philips, Zenith, RCA, Orion, etc.) with a standard RGB triad shadow mask. CRT displays were a very mature technology when they stopped making them about ten years ago. You can't make them slimmer because of the way the CRT itself works; it needs that long neck to achieve proper deflection, as well as house the other elements, e.g., heater, electron guns, focusing coils. None, or at least, nothing significant. It's only been ~10 years since CRTs were last made. I bought a brand new Happ Vision Pro CRT arcade monitor in 2009 for my Missile Command cabinet, and that was right around the time they stopped making them. BNC connectors are fine, so are RCA connectors, or VGA connectors, or a Molex connector, for that matter (which is what arcade monitors use). They all do the same thing in the context of an RGB video signal, which any hypothetical new CRT should definitely have an input for. S-video and composite inputs could be included for legacy purposes. I'd rather it be Sanyo, Wells Gardner, JVC, Mitsubishi, etc., than Sony, but if you remove TV tuning then it is a monitor-only by definition, not a TV. As for sizes, 19" and 25" will fit the vast majority of classic arcade machines, and 32" to 42" would be great for the living room to use with home video game consoles. All 15 KHz CRTs can inherently sync to 240p, and yes, any new CRT should definitely be 15 KHz-capable, and they should be available in only 15 KHz for arcade monitors, since ones that can sync to higher frequencies/resolutions (multisync) have a finer dot pitch / less coarse shadow mask which doesn't look right with the classic arcade games. Multisync would be fine for the 32" to 42" models, say, 15 KHz to 64 KHz, like the big "presentation monitors" that various companies used to make (Mitsubishi MegaView Pro 42 for example). HDMI is meaningless to a CRT. The electron guns in a CRT are driven by RGB signals no matter what you feed it. Even a lowly composite or RF signal gets converted to RGB in order to drive the guns. HDMI is a digital signal which can not drive electron guns, so while an HDMI jack could be included on a new CRT (and indeed, some widescreen HD CRTs did have them), it would just be a built-in HDMI-to-RGB converter. RGB is the best possible analog color video signal.
  15. That isn't even possible. The best LCDs have 9 ms of input lag, while the electron guns in a CRT respond to the video signal in what is effectively real-time, i.e., the speed of electricity through its circuitry. LCDs have literally millions of times more display lag than a CRT. It isn't possible for a CRT to have 9 ms of input lag, because it has no means of delaying and storing a video signal at all, let alone for 9 ms. A CRT only works with analog video signals and those aren't easy to store. They can be, for example, recorded to video tape, but there's certainly no VCR inside a CRT intercepting and recording the video signal and then delivering it to the electron guns 9 ms later. There's also no analog-to-digital converter inside a CRT intercepting the video signal, storing it in digital form in nonexistent memory, then sending it to a nonexistent digital-to-analog converter which then sends it to the electron guns 9 ms later. Claiming that an LCD can have lower input lag than a CRT indicates a lack of understanding of how a CRT works. Nothing can ever be faster at responding to / rendering a video signal than a CRT. The best that could be done is to equal it, and LCDs aren't even close to doing that. Any tests that ostensibly show otherwise are flawed by definition. In the case of the test you linked to, something in his PC is obviously delaying the signal before it reaches the CRT. Tests should be done with a video source that definitely doesn't delay the video signal at all; an Atari 2600 for example.
  • Create New...