Expansion Module Project
The 7800 Expansion Module has been moving at a remarkably fast pace given this is a hobbyist driven product. Moving into the highly ambitious position of actually producing new, from the ground up custom plastic case. As well as 1984 inspired Box Design, the 7800 Expansion Module brings the true spirit and life of what Atari products used to be like when they were original produced and sold...
So on September 30th, the project was kicked off officially and work began...
One very important feature of the Expansion Module was the complete re-design of the High Score cartridge design done by GCC in 1984. The original reproductions stuck precisely to the design for historical accuracy sake, but they came with the inherent battery drain flaw as well. So it was time to do a new High Score design and eliminate any battery. Instead scores would be retained in a non-volatile RAM chip (nvRAM). A new design was created and 8 original 7800 High Score cartridges were modified and thoroughly tested. On November 16th the design moved onto the Expansion Module PC Board layout and those test carts were then made available for sale. Upgrades will be offered to existing High Score cartridge owners after the 7800 Expansion Modules have been shipped.
On November 23rd the dream of creating brand new plastics for an Atari product, based on prototype designs done back in 1984 were realized when the shipment of plastic shells for the 7800 Expansion Module arrived...
Meanwhile, the Nov 30th deadline loomed. . . Still much work needed to be completed. The PC Boards still had to have some redesigns and several components were falling very far behind schedule from the vendors.
By December 10th is was becoming apparent that 2010 was going to have to pass by without the 7800 Expansion Modules making it out the door, too many unfinished items were left:
Aluminum name plates - truly one of the key cosmetic parts of the project just weren't materializing. The S. Carolina firm kept falling behind and so much relied on receiving the nameplates - especially have the final box art shots taken and for the cover the Owners Manual which were completed except for the Cover Page. Plus assembly of the units with the plates.
December 23-January 7. While on vacation, a new nameplate company was hired and the files sent to them. The California firm has been hit with hard times and they were willing to work hard to get the project done in 2 weeks to ship out to us. So on January 7th the Purchase Order contract was signed and we are moving forward with them and with a lot more confidence in them delivering the same Diamond cut appearance and rainbow stripe silk screening, so quality will be maintained to its highest level.
Once the plates are received, the Box Art and Manual's can head to the printers with an expected completion time of 7-10 days afterwards.
January 9th. Meanwhile PC Board work has gone into high gear, many changes have been made to the design to take it from its TTL design that it was going to for speed sake, but now headed back to a CPLD design because of the issues with Fine Scrolling enhancements, so "Opcode" has been asked to complete the CPLD design as original spec'd... "Groovybee" is also hard at work on a more robust onboard firmware to correct High Score Cart memory conflict issues and have a more intelligent cartridge type sensing routine.
If all of the PC Board work can be completed within 5-10 days, the PC Boards can be immediately run and will be assembled, tested and shipped in a stream to get the 7800 XM's out the door as quickly as possible.
January 26th. Nameplates have arrived, big hurdle has finally been overcome. Working on box artwork today and will submit to the printer tomorrow, they said 7-10 business days to complete and will ship out from CA. to us here in NY, so should be 3 days or so for shipping, we should be around Feb 10th to receive the boxes and start packing things up to begin initial shipments.
February 27th. One our team had to move onto other personal matters so things slowed down on the main controller chip within the system, which is now codenamed project "MILO".
March 15th. Okay we are now on an Atmel 1504 CPLD chip, the board has to go back for a nearly complete redesign due to space issues and changes into the control IC and also with the implementation of a more intelligent BIOS system onboard, rev 1.03f is in testing, new CPLD has been upgraded from a 44pin version up to a 68pin version because we have a LOT more control select logic going on and also the implementation of a Scroll Assist circuit which Steve Golson, one of the original architects of the 7800 MARIA chip has reviewed over.
April 26th. New boards completed - again, looks like this will be the final rev changes, boards are being run off for testing and implementation checks with our newest team member - Pontus for CPLD checking and over to Mark (Groovy Bee) for BIOS implementation in realtime. Notices went out to the first round of buyers for a May 5th ship, going to push that back a week, still see if we can start to get a few units trickling out to the first buyers and then start to kick things up and get going so everyone can get their modules...
May 17th. Development boards done and built, shipping out first run of systems to the key developer/testers for review and any final design and/or BIOS changes. Going to give everyone about a week to review over, if nothing major needs changing, then will set things in stone, start building out shipment batch #1, first 25-30 units to ship there after
May 31st. :-( Another week, another slipped deadline. Still working out a lot of design and functionality issues that are still cropping up and making this final stretch a bit irritating. The only good side is everyone has been great about the delays and not pressuring to get it released faster then it can be. Got an interesting email over the Memorial Day weekend which I thought was kind of interesting. Someone asked how new 7800 cartridge connectors could possibly be made, were they just poured resin replica's, no they are newly made ones, here are the original 1984 mechanical drawings and some shots of the Solidworks 3D CAD files.
June 23 - Batch #1 assembly work begins! Firs t 14 of 30 for this batch begin assembly work, PC Board assembly starts Monday, integration and testing to follow... getting closer to shipping! :-)
June 27th --- Happy 39th Anniversary to Atari, Inc. !!!
Some people asked if they could see the PC Boards, here are some unassembled ones so you can see all of the component footprints. Every pin and connector is clearly marked, so hacking/modding will be a breeze for those interested in really tinkering around with the XM's (as you know, opening the plastic shell automatically voids the warranty so tinker at your own risk :-)
Yes you see it right... there is a foot print for a PIA on there, its not tied to the Bus, but it has connection points so you could wire it up to the system lines if you wanted and enhance features of the I/O system.
For those who've asked what the manuals look like:
These are actually nicer then the original Atari manuals, the covers are a think 00.18 semi-gloss full color cover with 00.24 premium black & white inside pages. Its a 16 page User Manual. The Technical Manual is going to be so comprehensive, thick and expensive it will be published in PDF format instead for download for those who want to know the full technical details of the units, programming, etc...
July 28th - After several back and forth's on debugging, work is progressing well, POKEY tests are
being performed and the Yamaha YM2151 is running a bit warmer then I'd like so looking at that, getting
a new BIOS update shortly from Mark and Pontus is receiving some updated HW Monday. Added
debug jumpers permanently into the Rev11 boards based off of the Rev9 testing boards that now have
Cuttle Cart 2 and H2 Debug jumpers permanently on them, this will allow for future multicarts and other
more advanced carts to be debugged and integrated for compatibility with the XM's, also of note the
BIOS will be upgradable (how we will deliver upgrades out in the field to those who don't own a CC2 or
H2 will obviously have to be figured out, but this will future proof the design for more flexibility and with
a 29F010 in place, not only is there room for the BIOS, but there is the potential for an onboard game to be
added in a future BIOS update....
October 4th - Long overdue
Sorry for keeping everyone in the dark for so long, and I will make up for this by doing a weekly update on things. Seems a few mischievous trolls were trying to make it sound like I've run off with people's money and that the whole XM project was to pay bills. Well, since I've been showing the actually hardware this whole time and group photo's of equipment, parts, boards, boxes, packaging and professional printed manuals... I think its obvious that those were complete lies by people looking to try and hurt me and in turn hurt everyone who's invested in this project, pretty damned selfish and unbelievably petty.
Box of XM's waiting for final CPLD code...everything else is done, now down to the final bits...
250 Remaining Nameplates to apply to plastic housings to ensure that after all of the Pre-Orders ship, there will still be plenty of XM's for those wanting to buy them.
Testng 2600 games... gotta start somewhere - why not CX-2601 Combat.
If sales get close to 500, I will have the plastic factory and nameplate company run off more material for another 100-200 more.
As you can see up on the shelf there are several test runs of cartridge guides done over the last couple of months. The guide has been widen at the base 2mm in each corner.
XM High Score System Running...
All Activision carts fit with no problems, Zellers, and Coleco carts are still a little tight, Spectravideo carts look like they are going to be a no-go at all, they are just way too big and too tight a fit, even with the new wider cartridge guides...video below demonstrates almost all of the types of casings.
CLICK ABOVE IMAGE TO PLAY VIDEO
So what is the hold up currently? Several key issues with the final main control chip design:
Cuttle Cart 2 (and presumably H2 wont as well) is not working at all currently on XM's
PAL game recognition to shut-off High Score section not working completely
SARA some equipped cartridges causing conflicts with Non-Volatile memory in High Score system
Yamaha Music chip not implemented yet and also need testing code added to onboard BIOS system
Scroll Assist graphics enhancement not implemented yet.
So still several gating compatibility and implementation issues open that need to be dealt with. I got ragged on pretty hard because there were about 50 or so games that didn't work out of 1,000 on the Flashback 2's... I don't want people to have problems with the XM, especially where they would have to take the unit off each time they'd want to use a specific game because of conflict or compatibility issues. The idea is to install this and never have to remove it except in very rare circumstances as described in the above video where a few cart cases, even with the wider connector are still very snug... they fit much better, some fit now that never fit with the original 7800 cartridge guides.
Stargate/Defender II testing to ensure compatibility with enabled High Score XM logic and no Corruption of NvRAM contents
So more updates later next week, and will keep everyone more informed as we are just a few feet from the finish line and just inching ourselves these last few bits to victory.
Oct 13th - Very intense week, Mark and I have been at this hard and heavy non-stop...
found a couple of the issues, and created some new solutions and also making changes to the CPLD pinouts and functions. The downside is some major PC board changes need to be made :-( So I am going to do jumper fixes to boards for the updates, but ultimately for other batches, it means I'm going to run another Rev of the boards (uggh!) so we are at Rev 12A.
Progress is happening, BIOS is shaping up really nice, and the testing functions are coming along.
So I am hopeful we should be done very soon, plus Mark wants to make sure he can fully demo his games at Replay UK so we are going to make this happen. :-)
Oct 26th -
Just a very quick post - Pokey Audio Test, Yamaha Audio Test later tonight, should also post up a video of the system doing a cartridge sense test on the top slot, recognizing the type of cartridge and then loading it up...
Really getting close to ironing things out.... couple of things to fix and figure out, and stabilize, but the beastie is working, behaving itself a lot more and now just getting down to the nitty gritty...
CLICK IMAGE TO PLAY VIDEO
Oct 27th -
Quick clip of the Top Slot test, detecting a 7800 cartridge installed and then loading it up and a few moments of Pole Position II playing...
CLICK IMAGE TO PLAY VIDEO
Nov 2nd --
Time for some eye and ear candy... Thanks to Perry Thuente for working with Mark and I to get his Pokey Modified Pac Man and Donkey Kong working with the XM Control Registers and Pokey location. I also want to thank CPUWIZ for Express Mailing me out a pair of his 48K boards so I could test these games out on the XM hardware.
CLICK IMAGE TO PLAY VIDEO
Nov 7th - Mark (Groovybee) had a great showing at the Replay UK 2011 Expo this past weekend and showed off some of the 7800XM goodness to the crowds, here is some video from the show:
So as you can see we are getting very close, just working on PAL/NTSC console differences, the BIOS needs some more refining and we are going through as many games as possible to locate the troublesome ones, try to resolve conflicts and complete the additional logic and changes to the 2.9 rev boards and prepare the final production 3.1 boards... Be nice to be done before US Thanksgiving Day..... let's see if we can make it happen.
November 28th -
Sorry for the delay, been sick past week with the flu and only have been on here and there sporadically. Feeling a lot better tonight. Mark sent me over the demo video of the XEGS keyboard test. When I designed the XM to have a keyboard port, I decided the best option would be to use something already in existence that should be fairly easy to obtain for most people, so the XEGS keyboard seemed like the best option to go with and that was what I incorporated into the design.
To be honest, I never really thought I'd even be discussing the keyboard until months after the release of the XM and then people would start to get curious about using the keyboard port. However, thanks to Mark, he found that my keyboard port needed some changes to it and he decided to write some test code to allow the keyboard to talk to the BIOS test program (and hence its ability to talk with any 7800 program that will be coded to use it) and now we actually have a demo of the keyboard working. This was as much a surprise to me the day I received a BIOS update and wired up the changes to the hardware as anyone else. So enjoy this short demo that again, shows the future potential for the 7800 is wide open thanks to the XM and its expansion ports.
March 2 - Wholey cow, has it been THAT long ?!?!? Well, the new 84 pin CPLD design is done, some audio issues have been addressed, we now have a bus divider into the design, a reset circuit, test circuit and the SIO has been fully implemented with lines to a PIA footprint that people can add on, plus there are experimenter points along the PIA for those who want to tinker further. Just shaking out and debugging all the major design changes, if all looks good, Mark is going to squeeze one last bit of features into the PLD code that he's been holding onto until we knew all of the remaining core functionality was implemented.
Okay, now for some fun stuff and this is really cool.... have a listen to this audio, this is a full blown example of the YM2151 Audio on the the XM:
Sept 6th - I mentioned that with the book project behind me, I'm back working with Mark on some open issues. I noticed on Atariage a few people suggested "Just ship the damned thing as is to save your reputation" Mark and I can't and won't do that. There are some show-stopper items that are left open to implement and to just say, I'm done, lets go to assembly right now would mean 250 units would ship and there would be problems and they'd all have to come back for modifications and patches and that would be a logistical nightmare for all.
I think its important for everyone to understand, basically the design was scrapped and restarted 3 times, we are on Rev 3.1-"X" of the design, it is a COMPLETELY different design from the past two, the system has been partitioned with bus dividers, the addressing and function calls have been changed and expanded, we tried to incorporate Xboard compatibility, we had to redo the whole audio section to also allow existing POKEY equipped carts to work, the SIO interfacing was changed from being a very bare bones implementation to a fully capable design, the PIA option has been added for that as well, and some test point and hacker points have been added. The BIOS is field updatable and the list goes on...
I actually had to sit down and spend about 3-4 days just completely refamiliarizing myself with the whole design again and go through my notes so I can get back to where things left off. I'm going through a bullet point list of to-do's and changes now on one of three sheets and I'm hoping in the next couple of days to pass things by Mark to make sure we are both back in sync, we also need to test out the new 84 pin CPLD design and Mark has been secretly coding some extremely incredible demo's of the scrolling capability of the new system. Everyone has been exceptionally patient and gracious through all these bumps and tons of delays, and I know I had to take a 6 month complete hiatus from this project for the Atari book, but I am solely back on this. I actually have a whole new USB joystick/paddle project as well as a new arcade interface project, both are totally sitting to the side until the XM is done, this is my only focus now.
Thank you, Curt
Sept 15th - I have completed running through the bullet list of open to-do's. Going to spend today and tomorrow rechecking the circuits and making sure the calculations for components and then run things by Mark so he can doublecheck my circuits. Then I can start to layout the changes and components and reflect them on the PC board layout. So that will be my task this week.
Sept 25th - Layout changes and additions have been made, going to doublecheck everything and send over completed updates to Mark shortly so he can cross-check my work and see if anything further needs to be added or changed, otherwise if all looks good then I will look to make another run of the Rev 3.1.x board samples, build out and populate 2 of them and ship one out to Mark so we can start to install the JED code and go through a thorough testing of the system.
Oct 12th - Sorry get sidelined by a nasty upper chest cold, took me a week or so to fight it off. Mark was pre-occupied a bit with a personal project of his which is now been made public, he took my Atari ST Star Raiders source code and ported it over to the Jag64, certainly an interesting turn of events. Okay well back to the board, I've printed out the top and bottom layers and I'm running through visual checks of the traces, things look good so far, no errors, so I'm hoping to place the order for the test boards and begin section by section population of the board, running tests until fully populated.
Nov 13th - Okay, just spoke with Mark on a bunch of things, last 2 1/2 weeks were completely lost due to Hurricane Sandy and then the Nor'easter that hit. Office power was down for a while, some minor damage to the roof and looks like I've got 2 leaks coming into the workshop downstairs, but fortunately its on the opposite side of the shop so all of the boxes of XM cases, cartridge parts and other things never were in trouble. The Legacyengineer.com server is down however, so I am working to try and get things fixed with that. So I'm trying to get things back on track again and need some comments back from Mark this coming weekend, so I'll get another update up early next week.
NOV 29th - Design is locked down. No further changes, or modifications whatsoever. With the holidays coming, I will begin to order in the new boards. Mark and I will implement the firmware and after the New Year I will prepare the workshop for assembly, and begin to build out the XM's to begin shipping January 30th. So this is it, no more changes, no more attempts at further compatibility with any further games, the design is locked down and now its time to get past the holiday madness and focus on preparing to build XM's. I should be able to do 5-6 daily and get them out to first ordered first and go through the list. I will email people in batches to verify addresses, if I don't get a reply back, they get bumped to the next batch and if I still don't hear I'll keep bumping them until I'm at the end of the list and if I have to start making phone calls or posting on Atariage asking if anyone has heard from such and such a person, I will but I'm not holding anything up any further, time to move forward, everyone has waited long enough and too much time has been invested into this project.
Jan 3rd - For those who may not frequent the Atari 7800 forums on atariage.com I had posted an announcement to everything. I have been in the hospital for several weeks due to congestive heart failure and tissue fluid overload which nearly brought my heart to a standstill as it was so overloaded with fluid it nearly couldn't beat properly. I've got a severely leaky mitral valve once again and I am awaiting a decision from the heart surgeons n their plan for me as this would be my 4th heart surgery and the risk is very high so they are looking at ways of operating on my by going up through my groin and doing the surgery in a non-invasive manner. Needless to say I have been unable to do any work on the XM's and I am hoping to have my surgery by mid-Jan and then spend a few weeks recovering, so I can't put a firm date when I can complete the XM's to ship out at this point until I can gauge my recovery and know when I can return to my office to work. Right now, my best estimate is perhaps end of March, I'm very sorry for yet another delay, but this is certainly not something I anticipated nor can do much to change.
May 27th - As most have heard, I'm finally feeling better and I've gone back to work on the XM, sadly my partner on this project - Mark (Groovybee) has moved onto some other work and is no longer involved in the XM it would seem. Mark has been great, he's been a real champion in helping on this project, but it has been a very long process and we all do these things for the love of the hobby, no one is obligated to stay on a project. However the Rev 3 design was based on idea's and features Marked wanted to incorporate into the XM design, and the BIOS is all Mark's work. Without Mark's involvement I cannot implement the logic he wanted to put into the design since I don't know what he wanted to lay it out as nor where he wished to tie in. Also some of the features may have required a further modified BIOS to support these features. I am not a 7800 Coder and I can't write the BIOS. Mark also has all of the sources to the BIOS so I couldn't alter it anyway. So I have decided on a compromise - I am going to take many of the benefits from the Rev 3 designs, go back to the Rev 2 which was a working platform and incorporate those Rev 3 improvements where feasible and then use the last stable BIOS Mark had delivered for the project and put everything together. Perry has a friend who is an Analog Audio guy looking over the Yamaha portion of the XM schematics to see if he can fix some of the shortcomings with the design.
So basically the final design will be a Rev 2.5 design. I've also decided that once the XM's are released I am putting all of the schematics into PD so those who'd like to tinker, hack and improve upon the XM design. Perhaps if Mark decides to get back involved in 7800 development at a later time, maybe he will release the BIOS as PD, but that is solely his decision and I would respect either way he would wish to go with that. My releasing of the schematics doesn't obligate him to release his BIOS. So I've started working on wedging the 68Pin 1504AS back into the Rev 3 boards, and re-routing the traces and removing all of the additional Bus partitioning and secondary Address Bus lines out of the design and getting to a Rev 2.5 'Hybrid' design which will be functional and do what was initially intended before feature-creep added to the design.
I will post more info in 2 weeks, I'll also see about bringing a working XM to the Too Many Games show June 14-16th in Pennsylvania.
Sept 10th - The XM's have shipped... okay don't get too excited, only a limited couple went out to developers for testing of their games for the upcomng Portland Region Classic Gaming Expo this Oct 5th and 6th. Albert over at Atariage.com also received a non-working Display Unit that he will be putting out on the table next to a 7800 console for a working XM plugged onto of it that will be demo'ing Donkey Kong XM and Bentley Bears Crystal Quest at the Atariage booth at the show.
TEP392 (Perry) recently posted a video of his DKXM running on the developer XM unit sent to him - video is available here:
For those not aware, DK XM is not a simple hacked version of Donkey Donkey with Pokey support, Perry and several coders/testers have been working over a year trying to make this version of DK as pixel accurate and animation accurate as possible to the original Arcade version. I've been given the opportunity to watch the conversations between them and the level of meticulousness is incredible. This game does NOT disappoint.
Already some very useful feedback has come back from the developers and more is expected, so these will be implemented into the design to iron out things further. No promises on an exact ship date, but this is one very close step towards building and shipping, so keep checking, more updates to follow.
Oct 4th - Okay, some great news - Albert from Atariage sent over a photo this morning, he is busy settng up the massive Atariage booth at the Portland Retro Gaming Expo for this weekend and he reports that the XM's arrived and he's setup one and has Bob's (pacmanplus) Bentley Bear's Crystal Quest running on it. He'll also have Donkey Kong XM, Froggie and I think also Berzerk/Frenzy (With Voice!) running on the XM display at the show booth. So if you want a chance to try out the XM and the incredible games running on it, here is a great chance to try them out first hand.
April 18th Update:
Say hi to the new REV 3.4 XM Motherboard. This is the production ready board that incorporates a major redesign of the entire system literally from the ground up. When compared to the earlier Rev 1.5 and Rev 2.0 boards, this design is an entirely different animal.