Let's investigate and repair the turbo functionality on the Packard Bell PB430 / PB410 mainboard. Music by Model Povedeniya modelp.bandcamp.com/ Patreon: / necroware
God-tier repair skills as always. Most people would have given up at this point. You should definitely archive your custom firmware until somebody dumps the real deal.
Certainly not an easy fix, especially when you first have to root out the cause. That said, you don't absolutely need a PAL/GAL chip here. An 8-bit microcontroller could probably be made to do the job. The clock division alone could probably have been done with a multiple flip-flop IC chip.
Excellent work, you never cease to impress with your repair skills. While I have no need of the firmware personally, I do think the firmware you wrote should definitely be released, these old boards are going the way of the dodo and having the ability to keep them going for as long as possible is hugely important imo
I think this is the first time I've encountered anyone programming a replacement PAL from scratch. That's some next-level work! It was also interesting to hear an explanation of the different implementations of turbo functionality.
Most people would just connect 5v to the affected pin permanently. Actually I take that back. Most people would know how to even find the problem. Please keep making these videos. They are very impressive and entertaining.
I really have to respect the skills and patience required to pull a repair like this off. From the last few episodes I was really looking forward to how you were going to repair these boards. Thanks for sharing!
This was a great explanation of Turbo-Switching. I was never aware, that it was not just slowing down the clockspeed! At "rest in peace old PAL" i chuckled 😂
Wow reverse engineering the PAL chip and writing a replacement firmware is next level! I loved seeing how you worked out the logic, great visualizations! Also it made me realize that the PIO cores on the Pi Pico RP2040 are basically embedded PAL chips 🤯
The analysis and problem solving is what i love about your channel. And even if i dont know a lot about electronic/electricity, i'm still learning and understanding HOW it works and im starting to follow along more easily!
A dead PAL chip is usually where most people figure it can’t be fixed. Reverse engineering one is incredibly impressive! From reading some of the FPGA recreation forums, it’s usually the custom PAL chips that hold up so many, that they end up trying to just use the original chips salvaged from ruined boards.
I'd recommend archiving that reverse engineered firmware. Something tells me this particular implementation of a turbo controller has a high failure rate, due to choosing the cheapest IC they could get their hands on, based on the age of the hardware when it was manufactured.
Wow. Just wow. I don't want to say that your skills are godlike, you' said in some previous videos that you kind of involved in software development and this skill to "dig in" is essential for every good programmer. But you move forward further and go mess with hardware. And i think fighting this issue (in such elegant manner) is a great personal achievement. And... from retro-hardware folks perspective what you did - is total magic. And... my hat tips and respect to your research and repair work. Nice show, thanks for your video!
Sorry for the loss of a pal of yours ;-) Great video! Hoping it may come useful, here's a technique I use when investigating a logic signal on a PCB which is not behaving as expected: instead of shorting it to ground or to a supply rail, I always use a resistor. In this way you get more information about what's actually going on and, for example, you can discriminate between a dead short (like a solder bridge) or the lack of a driver winning over a pull-up/down resistor. It also avoid doing more damage or overloading the supply. For example using a 10k from 5V, 0.5mA is enough to pull the signal to ground and this current should not be a problem for anything. If the signal is lifted to 2.5V you know that the net is connected to something equivalent to 10k. Again, really impressed with your perseverance and skills!
I've thought about this video multiple times today after watching it yesterday. Fantastic. This is excellent content, and very relevant. There are fewer and fewer of these machines, and the problems will continue to happen. So please, more content like this!
Terrific work on the reverse engineering! So much knowledge about these custom programmed logic chips is lost to the annals of time, and it makes repairs like yours all the more difficult! Very well done :)
This is the kind of content we're here for. Other channels be like: "Yeah, motherboard iz half-ded, it's for parts, too bad". Necroware be like: "Lets reverse engineer the PAL and write new firmware." And he proceeds to do it. WHAT. Amazing.
Yes it would for sure be useful to have access to the programming file for the PAL/GAL for archival for fixing others of these motherboards and also learning from the source.
Thank you for all your efforts. If possible, please archive your work. This may help others even with different motherboards that use similar chips. Great video.
that was very interesting how you worked out where the problem was, and even more interesting was that you managed to build up the PAL with no knowledge of what needed to be programed in there, only what it should produce on the outputs, I guess you have a fair bit of background knowledge of programing these devices.
My friend and I both worked at PB writing tech docs during the time of these boards. That document you found was very likely written by him or he was involved in its creation. What you found was probably a webified version of the one created in the early 90s before PB even had a website - the info would be sent to service centers on 3.5" floppies that we hand created. I remember this board but it was replaced shortly after with the PB450. If you ever run across one, many had a quirky BIOS that caused them to slow down when you added the cache chips! :D :D
Absolutely great detective work. :) I might have socketed the PLCC chip when it was desoldered anyway... but then the chance of that same defect happening again in the same place is low. Although, since 2 boards have failed on the same chip, it may be more common than one would think, hmmm...
Wow, your patience and technical know-how are amazing - makes me believe I could achieve the same one day. Most motherboards are a black box to me, so this was a very interesting and easy to understand video, thank you.
Incredible work, I enjoyed this very much! The PAL stuff was very interesting, first time I've seen someone reverse-engineering one of those. Keep up your wonderful videos man, you're a gem. :)
Being able to reproduce the code on these PAL chips is a huge deal, since it's frequently the point of failure that takes these old boards down for good. A lot of old Macintosh computers have multiple programmable logic chips that can die, and their opaque contents also impedes reproduction board efforts, something that's ongoing in the retro mac community.
👏 Very interesting. I had been wondering how the turbo button worked. Now I have a pretty good idea. If anyone can benefit then I think it is worth while publishing the PAL firmware.
Great work, brilliant detective work on the Turbo functionality, thanks for sharing I think that will be very useful for other people struggling with broken Turbo functionality on their machines. Also makes me wonder if implementing 'Turbo' (slow down) on other machines that don't have it might be possible so that some speed sensitive games could work better.
My very first pc was a 386SX25 with a turbo button. If I remember rightly it dropped from 25MHz to 14MHz when it was used. It also had 2 Meg of RAM (soldered to the mobo as 8 256k chips) and a massive 50 Meg hdd.
As a (software) reverse engineer with only basic hardware/electronics skill, I love these vids and so much of the process & approach is the same as with software. Your vids also taught me not to be as intimidated by 'chip blackboxes' as I used to be. Since I have a very talented 'hw buddy' I always let him do his thing for complex stuff, kinda selling myself short in hindsight.
I have a soft spot for the turbo button and went to great lengths to find a turbo capable motherboard for my primary PC. Turbo in general is a poorly documented and esoteric topic, this video was enlightening.
This is fantastic work! I'm so glad you figured this out! As I said on the previous video I have a particular interest in these mainboards as my first PC had a PB410 mainboard in it and I learned a LOT while using the machine for several years.
Просто нереально крутое видео: здесь и скилл реверс инжениринга, и объяснения принципа поэтапного делителя частоты, и самостоятельное написание прошивки для PAL/GAL. На моей памяти это, верятно, лучшее видео на тему ремонта ретро-платы! Я считаю, что выложить прошивку для GAL на theretroweb - отличная идея. BTW, насколько я понимаю, в некоторых случаях возможно считать прошивку из PAL/GAL. В данном случае, вероятно, производителю было нечего скрывать, и считать прошивку со второй платы возможно? С целью бэкапа, разумеется.
Спасибо! Рад, что видео понравилось. К сожалению, большинство ГАЛов закодированы и вычитать прошивку не вариант. При записи на ГАЛ шифровка - стандартная установка, которую активно надо отключить. Большинство производителей не парятся и зашивают так, чтобы вычитать было не возможно.
So, if I am understanding it right - it wasn't too clear in the video - the 6 MHz, 3 MHz, 1.5 MHz, and 750 kHz signals output from the PAL weren't used by anything else on the board, and were just a vestigial artefact of the PAL dividing down (by half each step) the 12 MHz clock signal to get the 750 kHz FLUSH pulses?