Project NXTXE

[Matej]

Project NXTXE brainstorming:

 

- Micro - ITX or mATX motherboard

- Real 65c816 CPU (same as in Rapidus)

- 64MB RAM (4MB block XE RAM and 48MB block RAMDISK, 8MB VIDEO data RAM block, 4MB AUDIO data RAM block )

 

GPU

- FPGA "Romeo" (VBXE graphics)

• RGB output providing crisp clear picture using LCD TV or RGB monitor
• up to 1024 colors on screen from 21 bit palette
• graphics resolution up to 640x480i (640x240p) in 64 colours, 320x240p in 1024 colours, and 160x240p in 1024
• true 80 char mode for text display
• blitter with 7 modes of operations capable of zooming displayed data, transparency, collision detection and many other features
• full downward compatible with GTIA chip

APU

- FPGA "Juliet" (Stereo DAC core - like Amiga Paula, Yamaha YM2151 core - from Atari Arcade)

- 2x socket for real POKEY chips

total:

8 Atari Pokey PSG channels

4 DAC stereo channels (2+2)

4 FM stereo channels

16 channels for music

 

- ATMEL / PIC or ARM for keyboard, mouse, gamepad and USB / SD flash reading - > firmware upgradable from SD

- ATMEL / PIC or ARM for booting BIOS + NXTXE BASIC (in ATMEL EEPROM for example) - > firmware upgradable from SD

- 5x USB (2x keyboard, 2x gamepad, 1x USB flashdisk)

- 1x micro-SD (max 64GB or more)

- 1x stereojack (Audio - Out Port)

- 1x HDMI or 1x VGA port

 

- white PCB mask

- blue power LED, white loading LED

 

Compatible via re-emulator with:

- Atari 65XE

- Atari 800XE

- Atari 130XE

- Atari XEGS

 

Easy porting from other 6502 computer platforms...

 

Fileformat:

- .NXTXE (ATR disk equvivalent)

- ROM (ROM) flashing to ATMEL EEPROM?

 

PC/MAC:

- NXTXE Basic (GFA/STOS like Basic with export to NXTXE)

- NXTXE Tracker (RMTTracker like 16 channels tracker - PSG, DAC, FM)

- NXTXE Paint (Deluxe Paint like drawing program with charset / sprite maker and map editor)

 

OS?

NXTXE Basic with build in DOS commands

NXTXE GUI (8Bit GUI but in VBXE graphics)

 

PINS:

- like on Raspberry PI (you can use real XE keyboard for example)

 

CASE:

Mini - ITX PC case or mATX

 

Cartridge:

- something like PCI card slot in dimensions

 

What you think about it? Our own C65...

 

Price:

Under 200Euro.

 

Inspiration Clone-A-Dore:

http://clone-a-dore.blogspot.sk

Edited October 17, 2015 by Matej

[Rybags]

Sounds cool.

 

But I would suggest re video output. RGB at 16 KHz isn't practical in the modern day, there's virtually nothing out there that works with it. A little extra effort for Component video which has widespread support though that's dwindling now (in part due to DRM). Or a bit more for HDMI.

 

Additionally I'd request - if fullscreen video is buffered then provide support for PAL blending on both system types. Maybe even allow PAL operation mode with video buffered and displayed at 60 Hz.

[Poison]

One for me:) grey boxed, please.

[Bryan]

Sounds awesome, but...

 

I just see this as further fracturing of our limited development resources. It's a cool design but why not just write for the Amiga or some other more powerful retro platform rather than creating something which will only be seen by a few people?

[lemiel]

Nice idea.

But who will wrote firmware and code for it when there is still not to much for VBXE. Rapidus is just behind corner and how many code will be written especially for it? For audio capabilities: stereo is almost standard, Slight Sid is still comming, SidAri and Yamari are open projects, probably Evie will come, and there is still never comming SoundBoard but partially available via core for VBXE.

 

How to use it all?

[AtariNerd]

A compatable board, usable on modern displays, with some of the modern extensions built in, sounds good to me.

[pseudografx]

Please add more POKEY cores. Quad-POKEY would be nice :-)

[MEtalGuy66]

If the "DAC" audio doesn't do DMA, it's useless... And not really "Like the AMIGA" at all..

[Matej]

OK 4x Pokey socket (or 2x socket for Pokey and 2x Pokey in FPGA) + FPGA audio - 1x FM + DAC (OK DMA PCM) = 5+1 sound!

 

Another inspiration Individual C64 Reloaded:\

http://wiki.icomp.de/wiki/C64_reloaded

Edited October 18, 2015 by Matej

[Fratzengeballer]

The c64 reloaded is just a modern Version of the old PCB and no conpletely new system. Hence why the c64 scene is still alive? Because it still mainly relies on a stock system.

 

Gesendet von meinem A0001 mit Tapatalk

[+tf_hh]

Hi,

 

Fratzengeballer hits the point I think. There are some expansions for the Atari systems with only a very small "customer base", so thats not really interesting to write programs for. IMHO a memory expansion for more comfort is nice. A stereo POKEY solution also, but this is not a problem, as I didn´t know any game or demo which requires a dual POKEY.

 

Rapidus and all other speed-it-up solutions are... quite useless (MY PERSONAL OPINION!). I want 8-bit, the real thing. When I want more technical base speed, I take an Amiga. For the most 8-bit fans I know the low speeds and technical options are THE challenge why they code on 8-bit with <2 MHz and 64 KByte of RAM.

 

IMHO the VBXE expansion is something "alien" (I like it, I have it, it´s a great project) - software must be specially coded for it and at this time only a very few games and demos are availible. I´ve built-in over 10 pieces in the last years for ABBUC members, all of them use it primary as a "good quality video output solution" - for a not so small price.

 

A VBXE 3.0 - release with state-of-the-art FPGA and HDMI (!) output would be very nice. Using modern chips like the Altera MAX10 should it make possible to create a newer VBXE with a price around 50 Euros. This is - as a lot of users say - the range very more people would buy it. And when the "customer base" is high enough, more coder would accept it and program for it.

 

So... back to threadstart... a nice idea, not the first. But I see here no real "use case" for it. Real hardware, real (old) machines or something new like FPGA solution using foft´s core.

 

Just my 2 cents...

 

[TXG/MNX]

Cool

 

Would be cool to implement a PBI and real cartridge slot on this aswell :-)

 

Speed switching between 1.79Mhz and full speed whatever that is

 

Also a socket to allow people to use a REAL Atari XL/XE 6502 processor, (switch between that and the 65c816) to be 100% compatible with all existing software because not everything runs on the 65C816

 

Will there be a SparaDOS ( X ) ramdisk driver supporting the full Extended RAM ?

 

Last add an W5100 interface to it, its an ethernet module with embedded TCP/IP, place this HW above the normal 64KB in the XL/XE (maybe at the end of the 65C816 mem space)

[+Philsan]

Hi,

 

Fratzengeballer hits the point I think. There are some expansions for the Atari systems with only a very small "customer base", so thats not really interesting to write programs for. IMHO a memory expansion for more comfort is nice. A stereo POKEY solution also, but this is not a problem, as I didn´t know any game or demo which requires a dual POKEY.

 

Rapidus and all other speed-it-up solutions are... quite useless (MY PERSONAL OPINION!). I want 8-bit, the real thing. When I want more technical base speed, I take an Amiga. For the most 8-bit fans I know the low speeds and technical options are THE challenge why they code on 8-bit with <2 MHz and 64 KByte of RAM.

 

IMHO the VBXE expansion is something "alien" (I like it, I have it, it´s a great project) - software must be specially coded for it and at this time only a very few games and demos are availible. I´ve built-in over 10 pieces in the last years for ABBUC members, all of them use it primary as a "good quality video output solution" - for a not so small price.

 

A VBXE 3.0 - release with state-of-the-art FPGA and HDMI (!) output would be very nice. Using modern chips like the Altera MAX10 should it make possible to create a newer VBXE with a price around 50 Euros. This is - as a lot of users say - the range very more people would buy it. And when the "customer base" is high enough, more coder would accept it and program for it.

 

So... back to threadstart... a nice idea, not the first. But I see here no real "use case" for it. Real hardware, real (old) machines or something new like FPGA solution using foft´s core.

 

Just my 2 cents...

 

I agree all you wrote, apart one thing.

There are plenty of homebrew games, demos and tunes that use double Pokey.

The tunes are awesome!

[Joey Z]

IMHO the VBXE expansion is something "alien" (I like it, I have it, it´s a great project) - software must be specially coded for it and at this time only a very few games and demos are availible. I´ve built-in over 10 pieces in the last years for ABBUC members, all of them use it primary as a "good quality video output solution" - for a not so small price.

 

A VBXE 3.0 - release with state-of-the-art FPGA and HDMI (!) output would be very nice. Using modern chips like the Altera MAX10 should it make possible to create a newer VBXE with a price around 50 Euros. This is - as a lot of users say - the range very more people would buy it. And when the "customer base" is high enough, more coder would accept it and program for it.

Yes maybe you are right, that VBXE could be made for about 50 euros in parts. However, once you factor in licensing for the FPGA core (yes, this is unfortunately a thing, the FPGA core is NOT free) then you'll be above this, not to mention the cost of assembling the boards. So someone would first have to develop an FPGA core that is free/open source.

[ricortes]

A lot of this has been tried with other systems and hasn't been successful if you define success as more then ~100 units. Raspberry Pi was the most successful and it hasn't caught on except with hackers. Specifically you are ~proposing a consumer/gamer type device where others have gone before. If people would be interested in them, then they would be the 'Hot Ticket' item for Xmas and birthdays.

 

Some of the stuff like dual<or more> POKEY, last time I checked they were about $10 each. Glue chips, expanded RAM, et al, would end up with a ~$500 break even point.

 

It's not that I don't think there can and will be upgrades, it is just the price point and direction. For instance there are literally tons of cell phone displays that are being converted to shields for rPi and Arduino type devices. They can be wired into the Atari memory map just like we do with IDE type devices. They would provide ~320x240 graphics with 18 bits/pixel or a 176k ramdisk for ~$6. Unfortunately half the user base is looking to go large screen monitors/TV vs. a 3.5" LCD cellphone display!

 

You also run into problems with just the expanded amount of data you have to toss around. Expanding audio to something more modern, mp3s for instance, well most of them are ~4+ megs and sampled at 128 kbps. The most popular Atari file systems are limited to 16 megs so you can have only 4 songs per mount. Compare that to ~32 gig space you can get on a simple mp3 player let alone a Linux or Windows box and you can see the problem. The most expanded Atari you can dream up will be seen as an overpiced downgrade by the general population compared to what they have available or are currently using.

 

There's been some other interesting developments. Several people have come up with 6502 emulators for the Atmel/Arduino. If someone ever comes up with an Arduino with a couple of shields like LCD display, SD, mp3 player, running a 6502 emulator with an Atari OS, I would be a buyer. Probably couldn't do it with anything less then an Arduino Mega which bumps the price up a bit but would still be cheaper then something like Mist.

[barnieg]

I received my MiST box this week and I have to say I think it and systems like it are the way forward. I do think OpenSource is important for these type of hardware projects to allow the potential of future support. Collaborate on existing cores and there is always the possibility of crowding funding an ASIC conversion for some of them.

[morelenmir]

I received my MiST box this week and I have to say I think it and systems like it are the way forward. I do think OpenSource is important for these type of hardware projects to allow the potential of future support. Collaborate on existing cores and there is always the possibility of crowding funding an ASIC conversion for some of them.

I would like a MIST so very much and have been wondering how not to eat for a month in order to afford it! Humour aside, I would like to see a specifically A8 version of MIST a bit like the ZX Spectrum FPGA system Lotharek has announced.

 

Yes maybe you are right, that VBXE could be made for about 50 euros in parts. However, once you factor in licensing for the FPGA core (yes, this is unfortunately a thing, the FPGA core is NOT free) then you'll be above this, not to mention the cost of assembling the boards. So someone would first have to develop an FPGA core that is free/open source.

 

It is not trivial to properly attach that bastard of a TQFP with its microscopic pitch! At least not without a Mantis scope. I don't think drag-soldering is an option, at least not for my non-mad skillz. Paste and hot air was the only way to go in my opinion, even then I needed much tidying up with an iron and solder-wick Nonetheless, the same FPGA goes for about £40 here, so... I am very surprised the core itself is not open sourced. How did that come about? I assumed Candle'o'sin or Lotharek had designed it themselves.

 

A lot of this has been tried with other systems and hasn't been successful if you define success as more then ~100 units. Raspberry Pi was the most successful and it hasn't caught on except with hackers. Specifically you are ~proposing a consumer/gamer type device where others have gone before. If people would be interested in them, then they would be the 'Hot Ticket' item for Xmas and birthdays.

 

Have you seen the 'LeMaker Guitar' ricortes? That thing blows the Raspberry Pi out of the water in my opinion and when I first looked at it I thought it would make an excellent basis for an A8 hardware emulation.

 

A VBXE 3.0 - release with state-of-the-art FPGA and HDMI (!) output would be very nice. Using modern chips like the Altera MAX10 should it make possible to create a newer VBXE with a price around 50 Euros. This is - as a lot of users say - the range very more people would buy it. And when the "customer base" is high enough, more coder would accept it and program for it.

 

I actually think more important than HDMI for VBXE 3.0 would be on-board upscaling/sampling/whatever to give a VGA/DSUB output compatible with 'modern' CRT monitors. That way we would not be stuck trying to find over-priced and 30 years old multi-scan units that can lock on to the super-low frequency of the current design.

[Joey Z]

It is not trivial to properly attach that bastard of a TQFP with its microscopic pitch! At least not without a Mantis scope. I don't think drag-soldering is an option, at least not for my non-mad skillz. Paste and hot air was the only way to go in my opinion, even then I needed much tidying up with an iron and solder-wick Nonetheless, the same FPGA goes for about £40 here, so... I am very surprised the core itself is not open sourced. How did that come about? I assumed Candle'o'sin or Lotharek had designed it themselves.

I seem to remember the FPGA being about $22 from a US distributor. I was able to find it in quantity from the 'far east' for just $7/chip several years ago.

 

As far as soldering, yeah only way to go is paste or at least a few tens of times mag. microscope, preferably a stereo one. I soldered parts at my summer internship of similar pitch. If you've got a decent microscope, an iron with a pointy tip (or thin chisel type tip), and most importantly good solder wick then it's not bad at all. Only issue is that it's time consuming, and if your hands shake too much it could be a big issue. But, as always, the right equipment can make a hard job easy. I really would like to get a decent microscope setup for soldering work, but it's likely outside my budget.

 

EDIT: Oh, and the core was not created by Candle or Lotharek, it was created by Electron, and as far as I can tell, he retains all rights on that, so it's not freely distributed. Also, Candle wrote some firmware for an ATMega which loads the FPGA bitstream out of an EEPROM, and I think Candle retains his own rights to that (though I'm not sure) so that might be more licensing concerns.

Edited October 29, 2015 by Joey Z

[morelenmir]

I seem to remember the FPGA being about $22 from a US distributor. I was able to find it in quantity from the 'far east' for just $7/chip several years ago.

 

As far as soldering, yeah only way to go is paste or at least a few tens of times mag. microscope, preferably a stereo one. I soldered parts at my summer internship of similar pitch. If you've got a decent microscope, an iron with a pointy tip (or thin chisel type tip), and most importantly good solder wick then it's not bad at all. Only issue is that it's time consuming, and if your hands shake too much it could be a big issue. But, as always, the right equipment can make a hard job easy. I really would like to get a decent microscope setup for soldering work, but it's likely outside my budget.

 

EDIT: Oh, and the core was not created by Candle or Lotharek, it was created by Electron, and as far as I can tell, he retains all rights on that, so it's not freely distributed. Also, Candle wrote some firmware for an ATMega which loads the FPGA bitstream out of an EEPROM, and I think Candle retains his own rights to that (though I'm not sure) so that might be more licensing concerns.

 

That is a really good price for the chip if you can still get it! Currently Farnell list the same for £40.29 - so about... $60? However it is no longer available, so maybe it was some kind of end-of-line sourcing problem that bumped it up.

 

I think good magnification - especially in stereo - has more impact on successfully 'doing electronics' than even an expensive variable soldering iron. The work I did on the VBXE chip before going to hot air was through a jeweller's loop, which meant a very short working distance and constantly touching my iron against the thing... I grew to absolutely hate that burnt plastic smell. A stereo scope from Mantis would be an amazing luxury but they are well over a £1000 and I need an oscilloscope before anything else. Not that I can afford that either! I keep hoping someone local wants to give me an old HP logic analyser for the cost of carting it away for them; no luck so far!

[Joey Z]

 

That is a really good price for the chip if you can still get it! Currently Farnell list the same for £40.29 - so about... $60? However it is no longer available, so maybe it was some kind of end-of-line sourcing problem that bumped it up.

 

I think good magnification - especially in stereo - has more impact on successfully 'doing electronics' than even an expensive variable soldering iron. The work I did on the VBXE chip before going to hot air was through a jeweller's loop, which meant a very short working distance and constantly touching my iron against the thing... I grew to absolutely hate that burnt plastic smell. A stereo scope from Mantis would be an amazing luxury but they are well over a £1000 and I need an oscilloscope before anything else. Not that I can afford that either! I keep hoping someone local wants to give me an old HP logic analyser for the cost of carting it away for them; no luck so far!

The high price of that FPGA is probably due to it being out of production, if that's the case. If the HDL files for the VBXE cores were open source, the fix would be relatively simple: just generate the bitstreams for a newer (currently produced) FPGA.

 

You should be able to find an analog scope *fairly* cheap. They'd be much cheaper than a digital one, or a new scope anyway. As far as the logic analyzer, that'll be harder to find, and sometimes the issue with logic analyzers is finding probes.

[ricortes]

Re: LeMaker Guitar

 

Looks like a nice hunk of silicon. I'll read a couple of reviews before I bite.

 

I liked the rPi for what it tried to be and in spite of some failings. Still hard to get around: When you Google 'rPi USB problems' and come up with just over 800k hits! I read through two years of message board complaints w/o a real solution. Weak processor, weak voltage supply out USB, noisy circuitry on audio.

 

It looks like the LeMaker may solve a lot of the problems. I wonder how much of that 5-12 Volts at 2 amp recommended power supply is actually used by the main board. Android and LCD out of the box would mean a hand held emulator.

[morelenmir]

Re: LeMaker Guitar

 

Looks like a nice hunk of silicon. I'll read a couple of reviews before I bite.

 

I liked the rPi for what it tried to be and in spite of some failings. Still hard to get around: When you Google 'rPi USB problems' and come up with just over 800k hits! I read through two years of message board complaints w/o a real solution. Weak processor, weak voltage supply out USB, noisy circuitry on audio.

 

It looks like the LeMaker may solve a lot of the problems. I wonder how much of that 5-12 Volts at 2 amp recommended power supply is actually used by the main board. Android and LCD out of the box would mean a hand held emulator.

It really does look nice - and it seems to be going for an amazing price as well as around £40-£50.

 

The high price of that FPGA is probably due to it being out of production, if that's the case. If the HDL files for the VBXE cores were open source, the fix would be relatively simple: just generate the bitstreams for a newer (currently produced) FPGA.

 

You should be able to find an analog scope *fairly* cheap. They'd be much cheaper than a digital one, or a new scope anyway. As far as the logic analyzer, that'll be harder to find, and sometimes the issue with logic analyzers is finding probes.

 

Yeah - an analogue scope seems to go for about £200-£300. Its not bad at all in comparison to the modern single-shot digital offerings, but still well out of my range. I need sub-hundred. Really I think electronics is too expensive a hobby for me, which is a shame as the actual parts and components are super cheap - but you absolutely need a good true RMS meter - again £200-£250 and you need some kind of scope - and I have neither. Never mind.

[Joey Z]

It really does look nice - and it seems to be going for an amazing price as well as around £40-£50.

 

 

Yeah - an analogue scope seems to go for about £200-£300. Its not bad at all in comparison to the modern single-shot digital offerings, but still well out of my range. I need sub-hundred. Really I think electronics is too expensive a hobby for me, which is a shame as the actual parts and components are super cheap - but you absolutely need a good true RMS meter - again £200-£250 and you need some kind of scope - and I have neither. Never mind.

Really? Analog scopes here can be easily found for ~$100, which would be a good amount less than 100gbp. Got any craigslist type sites for britain? (that is, local pickup only kind of for sale ad site) Maybe it's just a matter of looking around until you find a good deal? Maybe it depends where you live though, and you're just not going to find a reasonably priced scope in GB, analog or digital.

 

For example: https://chicago.craigslist.org/nwc/ele/5284001285.html

[flashjazzcat]

Quite a few on eBay UK for around sixty quid.

[HiassofT]

Yeah - an analogue scope seems to go for about £200-£300. Its not bad at all in comparison to the modern single-shot digital offerings, but still well out of my range. I need sub-hundred. Really I think electronics is too expensive a hobby for me, which is a shame as the actual parts and components are super cheap - but you absolutely need a good true RMS meter - again £200-£250 and you need some kind of scope - and I have neither. Never mind.

Forget about those old analog scopes, especially if you want to diagnose digital circuits. Digital circuits mean non-repetitive signals so you absolutely want to have single-shot capabilities, segmented memory and advanced triggering features that analog scopes just dont have.

 

You can get a brand new Rigol DS1054Z for 300GBP and that would be a lot more useful than an analog scope.

 

As for DMMs: no need to spend that amount on a true RMS meter. Sure, these things are nice (I have an Agilent U1272A myself), but for a start a sub-50GBP meter will do as well. I used a simple 3.5 digit DMM for more than 20 years and it got the job done.

 

Actually, it's better to have 2 cheap DMMs instead of a single expensive one so you can check voltage and current at the same time.

 

so long,

 

Hias