wierd_w

In anticipation of my fancy tipi showing up, I picked up a pi 3 b at micro center. I need to get some female ended jumper wires, so I can route the pi's audio inot the ti's audio rail, but this is mostly a software project. Also picked up a roll if black petg filamentry for printing an enclosure.

It's a genuine commodore composite monitor, with the branding and everything.

I am visiting a friend for the week, and he has an old C64 color composite monitor chilling in his basement. He says it has an issue with the magnetic deflection plates needing to be adjusted, as it has a slightly canted image. I told him that it was worth money, and he was interested in parting with it, but wanted to know its value. He says he is the second owner. If I can figure out how to get the flash to work on my phone, I will take pictures. Just curious what the going pricing was. I think he has an old C64 system tucked away as well.

Hurray! I have been looking forward to this delivery for awhile now. (I should be able to do the "Can the TI send midi messages to the Pi over telnet?" experiments soon!)

Mainbyte is linking to a dead link. Archive.org to the rescue. https://web.archive.org/web/20001009013014/http://www.stanford.edu/~thierry1/ti99/titechpages.htm

I also have an Aureal vortex 3D floating around, but it's PCI.

I still have mine! (and a few IDE internal ones) I apparently hoard old electronics. I also have an old Voodoo 2 floating around, and an authentic soundblaster AWE32.

The internal zip100 can boot DOS. It does not need drivers. It impersonates an HDD, but with the "removable" bit set. (And, Again, CF is a specialized subset of IDE ATA. For reals yo. That the nanopeb does not support IDE devices indicates that it is not a standards compliant CF socket. )

It's fundamentally an ATA-ATAPI device. It *should* be supportable by any device that supports PATA standards. Again, if the nanopeb fully supported the CF spec (which it does not), it would support having this drive stuck on in place of the CF module. (Just power the drive seperately.) As-Is, it COULD be slaved onto a TIPI using a USB to IDE enclosure. Linux totally has drivers. EDIT Example-- A combination of this product https://www.newegg.com/p/0VN-00M5-000U1?Description=external dvd drive enclosure&cm_re=external_dvd_drive_enclosure-_-9SIAHBUAA03702-_-Product With a SATA to IDE bridge https://www.newegg.com/p/1Z4-00YN-00C87?Description=sata IDE bridge&cm_re=sata_IDE_bridge-_-9SIAHBUAAA2012-_-Product In the 3.5" to 5.25" bay adapter it is already living in (which gives room for the adapter!) Plugged into a USB port on the Pi, which is attached to the TIPI. Personally though, I would do that with an LS-120, like I said. That gives me a floppy drive.

Agreed. Zip was overpriced and slow. LS120 at least can read and write 3.5 in floppy diskettes, making it an interesting idea for the nanopeb. (Or, in a USB enclosure, docked on a tipi).

If proper support could be arranged, an LS-120 on a nanopeb would be interesting. It was back-compatible with floppy diskettes. Sadly the CF slot on the nanopeb does not properly support the IDE spec. (True CF is an ATA IDE interface with additional signals and power rails. Because the nanopeb does not work like that and only writes half-sectors, it probably would not work with internal zip 100, or LS120. Scsi might be an option in actual PEB though, or ltp zip100 on CF7.) LS120 support would be interesting. Nice hybrid target between permanent storage and floppy drive.

It's more a function of "Plastic hinges wear out fast" coupled with "Yeah, the LCD can be kinda heavy". There's a reason why laptop hinges are metal inside. If you don't mind using "kitchen cabinet" type hinges, with some kind of printed force-arrester so it can't just flop/slam, that might work, but would not be very attractive... Maybe it could be concealed.. hmm.. Still, the tolerances to mate up with the original chassis needs injection molding. (I am noticing that there are no dimensional drawings of the chassis on mainbyte, or elsewhere! I may have to remedy this!)

The printer needed to print something that large, at the tolerances needed to accurately mate up with an injection molded case, is outside of my price-range (and probably most people's). The 99a has a very large volume compared to most 3d printers. It would be cheaper to design a suitable CAD model, and have it injection molded as a service. Get better product too. Hinge design needs to be considered too. Weak plastic hinge will be a problem.

DOS systems are 1980s through early 2000s. It is nearly 2020. How old do they have to be to be considered old? Specifically, I know how to get a vintage computer (say, 1995 through 1998 era early pentium) to run win9x out of a ramdisk. Such computers are 20 years old today.

Hmm.. Does the scope of this thread also include old DOS system wisdom? If so, I could together a few informative posts on how to do some fun stuff with old DOS machines in here. Any takers for that? I know of at least 2 different ways to get windows 9x to run out of a ramdisk, for instance.

Ask for 2, 20sec pauses in succession. Problem solved.

If I had a mini 3-axis, I would totally be down for that. (I know how to program milling machines!) However, I do not, nor am I particularly wealthy.

PLA is easy to print, yes, but I find that the prints are rather brittle. Nylon prints just about as nicely (as long as you keep the filament dry, and or-- bake it in the oven at 250F for an hour before printing), and produces parts that are extremely tough. For durable projects, I have become rather partial to the Kodak 6-nylon filament. Just nylon is notorious for having issues with developing static on the surfaces, and that could be dangerous for electronic components. ABS is a PITA to print.

I'm just curious what you guys' preferred material of choice is when making enclosures. Do you just run with the PLA because it's easy on the cheap FDM machines, do you go with ABS for vapor smoothing, or do you go with something more like Nylon for strength? I am gonna have to print a TIPI enclosure at some point, just wondering what the norm is.

The bus is only rated for 500mA, according to the manual. If I were to add more power to the 5v rail there, I would do it behind a beefy isolation diode. That way the injected 5v rail would not course its way back up into the motherboard where it does not belong. Ironically, a modified speech sidecar would be an ideal place to do the power injection, due to it terminating the +5v rail. Instead of doing the bridge mod, you would add a +5v rail on a diode. Then your hungry device (like a TIPI) could get regulated 5v right off the bus without issue, and not pose a hazard to the upstream devices. Either way, the picoATX has +5v rails in copious abundance (seriously, it has 5 of them on the ATX header, and a big beefy one on the HDD power rail), and wiring up a few USB ports just for power would be trivially easy. Bear in mind that USB power spec is maxed out at 2A. Depending on the device you are attaching, it might actually need more than that. (Especially if there are rechargeable batteries getting topped up in there, such as say, if you were to wire up a ramdisk on the sidecar.) As a consequence, do not use the port to charge electronic devices like phones unless you appropriately rate limit the amperage. I mostly favor the picoatx idea because good quality 12v barrel type power bricks are a commodity item, and the picoatx itself has a strong mass-production supply chain, making them very easy to obtain inexpensively. (Compared to the more unusual power supply favored in the above postings, which I understand can be hard to source.) The only issue it has is the missing -5v rail, which can be resolved with a single IC on the breakout board. Since it provides multiple +5v rails, you could wire up all kinds of power-only USB ports on the back with it, without causing voltage sag on the rail you use to run the TI's motherboard.

I personally feel than an adapter for a PicoATX (that has a simple IC that generates the missing -5v rail) would be the ideal solution. The devices are in wide circulation with high demand, and are able to provide upwards of 120W for the premium models. MORE than adequate to power lots of sidecars and other 5v hungry devices.

Agreed, but consistency between glyphs is important all the same. I can get pretty reasonable results all the same, and have perfectly parallel lines, perfectly consistent stem and stroke (even with curves!), etc. Regenerating 2D and 3D data from old hand-drawn blueprints was one of my prior careers you see... (Also, reverse-flatpattern derivation.)

Dassault CATIA is my CAD of choice-- It is internally BSpline (and or NURBS) based, and is VERY precise. It has several useful tools for manipulating spline curves, such as the "affinity" tool. (It lets you input a 32 bit precision floating point value, to precisely bias a curve according to an axis system-- User definable.) Instead of "control nodes" on bsplines, it uses a 32bit "tension" value for tangent lines. (DWG is a terrible format btw. It uses quadratic curves instead of bsplines, and has a nasty habit of turning splines into polylines. Nasty nasty nasty.) I usually use 2d IGES for 2d data exchange with CAD software for this reason.

I prefer my CAD software for that, because I can use actual measurement constraints to assure the geometry of the curves is mathematically perfect. I can save it in a suitable format, then convert it to .AI with something like PStoEdit. I can assure that the curves are EXACTLY where I want them in the glyph's EM space, I can make sure that the stem widths are EXACTLY the right spacing, that the tops of letters and the depths of descenders are EXACTLY right, etc. (this obsession with mathematical perfection is important, because of how truetype fonts do hinting instructions. Without that kind of perfection, you are gonna be working yourself to death making sane hint instructions for the different point sizes.) Illustrator kinda just wings it. The file format is useful as an import intermediary, but I prefer exact geometry.

Ironically, I keep a Yahoo mail account explicitly for all those sites that demand an email for nebulously defined reasons, exclusively so it can service as a trash bin for all that unsolicited email.