The reality is that this board can only use the newest V2.5J firmware due to the lack of connection to 4 of the 6 key scan counter bits on the Pokey chip. For a better idea as to why these bits are not connected on this particular TK-II board version please check out the ARROW2JOY-XLD page for the details.

All other TK-II boards besides this one and the ARROW2JOY-XLD can use either the V2.4 or V2.5J firmware, depending upon what features you want to have. See the Manual for more info on those differences.

Edit: I just wanted to add that the PokeyMax FPGA based replacement product might not be 100% compatible with the V2.5J firmware. So until that issue is completely resolved, it would be better to stick to an original Pokey when only V2.5J firmware is a possibility. Despite what you may see elsewhere (eBay), NOS Pokey chips can still be purchased for $20 from BEST Electronics.

- Michael

Just checked the JLCPCB site today, and the good news is that the boards have completed the production phase and are now awaiting pick-up by DHL.

Based on previous boards that I've gotten from them in the past, the shipping should only take 2-3 days. So if all goes well I should see them arrive no later than this Friday. Then comes the most exciting part of assembling and testing at least one of each over the weekend.

- Michael

Mr Robot had been seeing a persistent although random noise occurrence during deactivation of the FujiNet. This combined with the fact that I nor the other members of the beta test team were able to deal with the very tiny SMT parts being used on the last iteration of the NUC-FujiNet board design, set me up with a new project. This project turned into a mission about making some SIO to ESP32-WROVER interface circuit changes that had been in discussion for quite some time, hopefully solving the audio noise issue, while also creating a board that even I could assemble.

The reason for another PCB run in lieu of doing a point to point wired breadboard for prototyping  and proving out the changes, was that in this day and age of cheap PCBs, bread-boarding is no longer the best way to prototype designs. And although this board looks all the part of a final PCB layout, it really isn't, and only serves to hopefully advance the FujiNet interface circuit, as well as give me and others something we can actually assemble by hand, while furthering our education on how the NUC and FujiNet need to play together in the best possible way.

Why the odd shape and the bigger than needed board?
This board was patterned after the previous board designed by Mr Robot that is meant to drop into the 576NUC+, lining up the switches, LEDs and such with the front and rear panels of the custom NUC enclosures he's designed. Mimicking that previous board footprint insured that it could be used in a similar enclosure if so desired, and gave me lots of room to layout the chips and traces for a very easy routing of the PCB.

What has fundamentally changed in the interface circuit design, is that all SIO signals are now disconnected from the WROVER module via the the ON/OFF switch SW1, including the SIO Audio-IN. Isolation of all the WROVER digital outputs are still handled via a 74LS07 open drain hex buffer which has been retained from the previous design. However for all the WROVER inputs as well as the Audio output, they now get their isolation via a 74HCT4066 quad analog switch having all the individual analog switch enables also coming from SW1. So in all cases the SIO signals get disconnected either through the 74LS07 or the 74HCT4066 chips when SW1 is placed in it's OFF position.

Wherever possible, thru-hole components were implemented, making for a more DIY friendly assembly. Exceptions: the WROVER solderable module was retained instead of opting for a break-out version, since a few of us beta testers already had these on hand. And the SMT SD socket was also kept, although it's relatively easy to solder. Finally because there wasn't an appropriate thru-hole replacement for the USB Serial Bridge chip (CP2102N-QFN24), I opted to go with an external FTDI (USB to Serial) cable connection to facilitate WROVER programming. An even cheaper option for this cable is to be found in the PL2303TA based cables that sell for very cheap on eBay, essentially rendering a similar USB to Serial function, assuming the Prolific driver is installed (already present in most Linux distributions).

As I write this blog entry, sample boards are currently being manufactured by JLCPCB in China, and should be arriving late next week. When that happens, one board will get assembled and tested with the hopes that the new ideas it encompasses will actually work as proposed. Whatever is learned will then be applied to a final version of Mr Robot's FujiNet board.

Exciting times ahead,
- Michael

TKNUC V1.4 Firmware

Today I'll cover a bit more about the special function 'hot' keys when using the PS/2 keyboard with the NUC.

TKNUC V1.4 Firmware

I thought I'd cover where things are at with the PS/2 keyboard 'hot' keys, and how the 576NUC+ is controlled by them.

First of all there is no actual power switch on the NUC, so that function gets implemented via special key press combinations which can be seen in the table below. Primarily this consists of using the 'ALT' key simultaneously with another key. Normally that would be a number key, although in the case of booting without a built-in language or game selection, the 'tilde' (~) key would get used, with CTRL+ALT+~ allowing a boot with a stock XEGS OS. Although in most cases the High Speed SIO OS with Fast Math routines active would be the way to go, via an ALT+~ key press.

When not pressed with the ALT key, the tilde key functions identical to the Atari inverse key.

I think that's a wrap for now. In my next post I'll cover some of the other 'hot' keys, addressing things like extended RAM control and attract mode defeat.

- Michael

Well pretty soon I should be able to hold one of these in my hand (only one hand required), assuming everything checks out on my final inspection. The Beta Testers haven't reported any issues for a long time now, so I'm feeling pretty confident that this will work out of the box on the first try.

Manufacturing of a few sample boards will be done by JLCPCB, leaving it up to me to stuff them with parts and solder it up. Got my fingers crossed on it passing all the tests.

What's changed from the previous V1.1 prototype?

• Audio Video jack changed from DIN-5 to DIN-8 in order to provide some additional pins for stereo audio applications utilizing the 3 pin header next to it. Nice thing is that a standard Atari DIN-5 cord will still work (great idea by tf_hh of AtariAge).
• The P-Channel MOSFET (Q3) has been changed from a TO-220 device over to an SMD part placed on the board bottom. This kills two birds with one stone, giving us a readily available replacement for an obsolete part, and also gets rid of a device that formally stood quite high on the top side of the board. This device serves as the DC power switch via a custom TK-II and PS/2 keyboard.
  
• Added CSYNC buffering to alleviate a GTIA failure problem.
• Changed to larger all SOIC SMD components for an easier hand soldering possibility.
• Broke out a couple of connections that would open the door to adding a Cart Port in a future add-on board.

Enter - FujiNet

AtariAge Mr Robot's FujiNet add-on board is getting ever closer to a final version and looking real sweet.

I have one of the bare boards to assemble, and am looking forward to seeing it in action once I do.

Pretty soon I'll have to put on my thinking cap, view some of the FujiNet tutorial videos, and see if I can get up to speed on this new wonder of the 8-bit world. It's gonna be so cool to fetch stuff from the Internet wirelessly - who would have thought this even possible back in the day.

That's all for this post - stay tuned for more to come.

- Michael

I decided to add a blog to my site, since creating topics in forums tends to get pretty disrupted due to other people's posts. While this is an integral feature of a forum, and useful for getting people to participate in the discussion, it doesn't always work well for documenting the progress on a new widget. Hence the reason I have established my own blog.

Hopefully this new endeavor will prove useful, and provide some very good information on some of the strange and wonderful gadgets I'll be coming up with.

Let the fun begin!

Michael St. Pierre