chue

Still looking? There are two on ebay at the moment. https://www.ebay.com/itm/285460008407 https://www.ebay.com/itm/296377248801 Edit I see you got one. Nevermind!

Let me plug @JasonACT's Pico PEB. Full thread here: Here's my summary of the feature set: - Speech Synthesizer implementation - 32KB / SAMS / Myarc Foundation Memory - GRAM/GROM Cartridge + Multi-bank RAM/ROM + MiniMem-RAM - USB subsystem keyboard + joystick + mouse - Disk image (.dsk) + TI Files (on SD) support - Network clock - RS232 - Via Pico's free Serial2 port - P-Code Card implementation - And more... see the file "PPEB2.ino" in the Pico PEB download zip for more details The Pico PEB features are controlled/ enabled via a configuration file as well as TI BASIC commands. Currently this is a DIY project, so nobody is selling these. You have to build one (or have someone build it for you). There is probably a handful of users (4 including Jason, that I know of). Edit Here's a picture of one that I built, it plugs into the TI's side expansion port:

Silver side port. If I didn’t already have a QI, I’d take a chance on it. $30 shipped isn’t too bad if it boots.

Silver & Black: LTA5282 Beige: LTA3983 QI: LTA4383

I know nothing about implementation, but JediMatt's RAM tester sees greater than 1MB on Classic99, as well as JasonAct's PicoPEB: According to the github, the tester will see up to 16MB: https://github.com/jedimatt42/ti994a-memtest

I don’t know anything about your RAM expansion, just taking a guess. Did you try plugging in the Ram first, and then speech synthesizer into the Ram? Does the Ram expansion require an external power supply?

Try a different, preferably non-QI console?

Ok, I did the QI mods: 1. Pulled out the resistor pack 2. Pulled out U32 and put in a "shunt" connecting the data lines The QI is now able to run carts from the PicoPEB (previously it didn't work): I am currently running a SAMS burn-in test. I also tested the FinalGrom99 (without the Pico attached) - it doesn't work with Ralph's 1.3 firmware. It still needs the QI specific firmware to work correctly. Let me know @JasonACT if you would like to see any other testing.

I would be inclined to touch up the pins on the sd card socket as well.

I actually don't have any, other than playing friends' 2600s and 5200s. The video I link was for an Atari machine, but the keyboard was made by Mitsumi. It appears to have the same mylar design as our TI keyboard, and therefore I thought it might be useful to try the same fix.

A history of Texas Instruments (the company) - price wars at about 33:20

Nice writeup, Brad! I recently had a go at this as well, and did not have the success you did. I did not do everything that you did, so I will go back and try it when I get some time. When I received my keyboard, only one key worked. Now, most of the keys work but there are few that do not. Of the ones that do work, there are many that need debouncing. That is, when I press a key once, I get 4 - 5 characters on screen. My research found a lot of the same fixes as described in your post, but I did find one other possible fix. 1. Max's video, which heats the mylar with a heatgun. I tried this but it had no effect for me. 2. Cleaning the carbon pads with alcohol. I did this and it certainly improved things for me. 3. Fixing the traces on the keyboard and mylar. I did not do this, but will go back and do so. 4. Making the keys stiffer, using some heat shrink. I tried this but don't have the right sized heat shrink, and gave up on it. Here's the video: I think that over time, a lot of things can degrade on this model keyboard. Each keyboard may have its unique failure points; so a particular fix for one keyboard might not resolve issues with a different keyboard.

Just curious, can you explain a little more? Does the mod make the side port and the cartridge port behave like the original? Do you see any other potential behavior changes? When I do this, I'd like to do it non-destructively. My thought was to socket both the resistor pack and U32. For connecting the blue lines in your image, I was planning on doing that with another socket, which could be removed when needed. That way I can have the non-QI behavior by removing the components and inserting this connector socket; or, if I need the QI behavior I can re-insert the original components. By doing this there is not a need to cut W7. Does my theory hold up? I am a little short on time at the moment, but hope to get to it in the 2nd half of April. Anyone else is certainly free to give it a try.

I just watched this, it was more a history of TI (the company) than it was about the computer.

Decent read. a lot of great points in the article. I found it interesting though that he says developers shouldn't bother each other because it is a productivity downer; however later he says they should run their UIs by others. I don't know, maybe he's saying it's ok to bother people in the hall ("hallway usability testing"), but not at their desks? Anyway, just questioning out loud... probably only "Joel" can answer that one.