The PBI Device We Need

[Mathy]

Hello Bill, guys

 

12 hours ago, BillC said:

So an MIO/Black Box may be capable of 115,200 baud by adding a 1.8432MHz external clock connected to the 6551 ACIA(may also require a W65C51N), and modified firmware.

 

Bob Puff confirmed years ago that the BlackBox can be run at up to 115k2 baud.  He even wrote some drivers to get to speeds up to IIRC half that speed.  Not sure, it's been years.  You can read more about that (and find the drivers) on my BlackBox Page.  (Just below the picture of the BlackBox)  The existing 6551 can be used.

 

Sincerely

 

Mathy  (who'd love an external 80 column device, preferably with color, maybe based partly on the Sophia 2 and an external memory expansion with at least 1MB addressed like the MIO does.)

 

 

Edited January 9, 2021 by Mathy

[Alfred]

On 1/8/2021 at 1:48 PM, Stephen said:

 

If we don't have software for 2 devices, having 5 different devices is not going to help the situation.

This. I think people seriously overestimate the number of good systems level coders available for things like this. Want to know why there isn’t a version of say Action! that runs on Rapidus and has a VBXE display, try looking in the mirror and asking why *you* haven’t done it. I bet there isn’t ten guys like Drac030 on here and they already have their own pet projects. Then, after you’ve written a few hundred thousand lines of assembly language, you tend to lose a little of the old enthusiasm.

 

 I get that lots of people just want to load their multicart and Joust all day, and that’s great. But before you bemoan the lack of a decent version of SPF/PC for the Atari you need to understand that unless somebody *you* steps up and does something, it ain’t going to happen. The days of a commercial software house writing code for this platform are long gone. KMK is busy, you’re going to have to find someone else. And not Rybags either, I think he’s busy too.

[Gunstar]

I think the PBI "device" we need most is an expansion box. I'm not talking about the 1090 or something similar, just a box (similar to a 1090 only as far as the case goes) that has multiple PBI slots and PCB edge connectors for multiple PBI devices (so we don't have to string a bunch of devices out the back like the TI 99/4A did out the side (and which they eventually brought out the expansion box for this very reason. And Atari's own 1090 was meant for too).  I've been planning on making one DIY for myself anyway, for my 1200XL. I've nothing on paper yet, just ideas swirling about in my head. But I only make things for myself and have no intention of trying to make some to sell, as I don't have the facility or money to lay down upfront for production costs. But if someone else who has the proper facility and finances to do so, then I'd buy it.

 

My idea is just a box with multiple PBI slots and edge connectors that many of the current PBI devices as well as future ones can plug into: the plan is to also incorporate ribbon cables for current devices which already have a PBI pass-through, like the one I used from my Incognito to a PBI edge-connector out on my 800. But much shorter ribbon cables since slots/edge connectors would be right next to each other. I am thinking something like the first is a card edge that has a device plugged in, then a ribbon cable with two PCB connectors on either end, coming off the pass-through of the first device, to the second device which would connect via cable, and it's pass-through plugs into the second slot, the third a card edge, and so on, i.e.,This means the first spot would be a card edge, the second a slot PCB connector, the third a card-edge and so on. either alternating like that, or all PCB slot connectors but then have double-male cards like how the Atari Jaguar's CD unit plugs into the base Jaguar to allow chaining of PBI devices that way.

 

The only other thing besides multiple slots would of course be a built in PSU and maybe  PHI2 signal generation for PBI devices without them.

 

Of course I plan on forging ahead with my DIY project within a year or two if no one else makes something similar for sale.

Edited January 10, 2021 by Gunstar

[Dmitry]

1 hour ago, Alfred said:

Quote

If we don't have software for 2 devices, having 5 different devices is not going to help the situation.

 

This. I think people seriously overestimate the number of good systems level coders available for things like this. Want to know why there isn’t a version of say Action! that runs on Rapidus and has a VBXE display, try looking in the mirror and asking why *you* haven’t done it. I bet there isn’t ten guys like Drac030 on here and they already have their own pet projects. Then, after you’ve written a few hundred thousand lines of assembly language, you tend to lose a little of the old enthusiasm.

 

 I get that lots of people just want to load their multicart and Joust all day, and that’s great. But before you bemoan the lack of a decent version of SPF/PC for the Atari you need to understand that unless somebody *you* steps up and does something, it ain’t going to happen. The days of a commercial software house writing code for this platform are long gone. KMK is busy, you’re going to have to find someone else. And not Rybags either, I think he’s busy too.

 

I think that the "2 devices" are XEP80 and VBXE?  XEP80 is a rare find these days, it is not available to purchase.  VBXE is still available, for now.

 

So we have at least four devices that I know about, Austin Franklin 80, Bit3, XEP80, and VBXE.  The first 3 cannot be purchased, and the last one...is hanging on for now.

 

I guess what I'm saying, the number of HDMI capable 80-column cards we have is zero, not 2.

 

From my perspective, what the community centers around as a standard is important.  When I went to buy my ultimate system back in 2016, I spec'd a 256k classic memory upgrade, a rapidus and vbxe.  I priced the upgrades at 4 places, and had 2 people do the install.  Every vendor told me they weren't going to do the 256k and were going to install a u1mb instead.  I could've insisted on a price for what I wanted, but anyway, I'm just sharing, that is how centered this community is on the u1mb.

 

Anyway, I got a bit unlucky because the u1mb and rapidus are not compatible, and my ultimate atari's didn't really work.  So this year, 2020, I just buy some 130xe's and price out getting vbxe, rapidus - no memory upgrade. 130xe was purchased for already being at 128k anyway, right.

 

I priced it out at 4 dealers, and to my surprise several of them told me not to get a vbxe, but to go with a sophia instead. 

Of course, I knew I wanted a vbxe, so I insisted, and I am having a vbxe installed.  In that sense, I'm definitely a vbxe fan.

But its popularity is going to wane now.   There is no device with hdmi output and 80 columns, but there is a device with hdmi output, and its winning mindshare.

That's not good news for an 80 column afficianado, imho.

 

oh well, I'm sure we mostly agree, we have to get out there and code, no doubt.

[Alfred]

4 minutes ago, Dmitry said:

 

I think that the "2 devices" are XEP80 and VBXE?  XEP80 is a rare find these days, it is not available to purchase.  VBXE is still available, for now.

The XEP-80 is a dead end. It's good for a DOS CLI display and that's about it. You want good 80 column editing, it needs to be a VBXE. Watching the XEP-80 scroll a full screen of 80 column text is like watching paint dry.

[Faicuai]

7 hours ago, Alfred said:

The XEP-80 is a dead end. It's good for a DOS CLI display and that's about it. You want good 80 column editing, it needs to be a VBXE. Watching the XEP-80 scroll a full screen of 80 column text is like watching paint dry.

That's not true, of course:

 

 

[danwinslow]

11 hours ago, Alfred said:

This. I think people seriously overestimate the number of good systems level coders available for things like this. Want to know why there isn’t a version of say Action! that runs on Rapidus and has a VBXE display, try looking in the mirror and asking why *you* haven’t done it. I bet there isn’t ten guys like Drac030 on here and they already have their own pet projects. Then, after you’ve written a few hundred thousand lines of assembly language, you tend to lose a little of the old enthusiasm.

 

 I get that lots of people just want to load their multicart and Joust all day, and that’s great. But before you bemoan the lack of a decent version of SPF/PC for the Atari you need to understand that unless somebody *you* steps up and does something, it ain’t going to happen. The days of a commercial software house writing code for this platform are long gone. KMK is busy, you’re going to have to find someone else. And not Rybags either, I think he’s busy too.

This is true, software is much more effort than hardware usually.

Although, if someone were to make, say, an adaptor board for the PBI/ECI that would let you directly interface with it using a Raspberry or something similar, I would probably take a whack at trying something.

[Mathy]

Hello Alfred

 

15 hours ago, Alfred said:

Watching the XEP-80 scroll a full screen of 80 column text is like watching paint dry.

 

You must have very fast drying paint then.  The XEP80 isn't slow, the software written for it is mostly waiting for something where it's not necessary to wait.  See the XEP80 section on my Special Stuff page.  (The three files at the top plus the part where the pictures are)

 

Sincerely

 

Mathy

 

[bbking67]

I personally used my XEP 80 for years with an amber composite monitor and I thought it was pretty decent for word processing and with bobterm among other uses.  But it did seem slow... I know there is some optimization that can be done, but surely the joystick port is a limiting factor.  A PBI XEP 80 would tick a number of boxes.

[Mathy]

Hello guys

 

What Atari did with the XEP80 is an abomination.  The software is horrible, and the hardware design ... The NS405 can handle 64kb, 16 bits wide, and as many bits for colour as you want.  As in for example 8 for red, 8 for green and another 8 for blue as is mentioned in the datasheets.  Which also mentions bank switching by the way.

 

Sincerely

 

Mathy

 

[+Stephen]

59 minutes ago, Mathy said:

Hello guys

 

What Atari did with the XEP80 is an abomination.  The software is horrible, and the hardware design ... The NS405 can handle 64kb, 16 bits wide, and as many bits for colour as you want.  As in for example 8 for red, 8 for green and another 8 for blue as is mentioned in the datasheets.  Which also mentions bank switching by the way.

 

Sincerely

 

Mathy

 

I know - when I read the datasheet on the controller, and realized what we actually got, it was a very sad moment indeed.

[Faicuai]

2 hours ago, Mathy said:

Hello guys

 

What Atari did with the XEP80 is an abomination.  The software is horrible, and the hardware design ... The NS405 can handle 64kb, 16 bits wide, and as many bits for colour as you want.  As in for example 8 for red, 8 for green and another 8 for blue as is mentioned in the datasheets.  Which also mentions bank switching by the way.

 

Sincerely

 

Mathy

 

No, it's not an abomination...

 

It is the result of selecting a HW component that met the project's original goal (e.g. driving a 80x25 chars. screen over PIA/Joystick ports) with and over-spec'ed chip available cheaply enough. Who the hell, anyways, could drive color output and screens through that PIA port? That was not certainly the primary aim of the product.

 

Thanks to Avery's hard work put into it, his latest ULTRA drivers (for which I developed a special wrapper that allow them to boot even from U1mB/Incognito BIOS !), you have a FAST, and PORTABLE 80-col solution, that can be deployed across your ENTIRE 8-bit collection, effectively supporting E: editor, per OS specs. (e.g. Atari Basic, Altirra Basic, Basic-XL, Assembler Editor, Atari Macro Assembler, FAST BASIC 80-col editor, an others.) 

 

Not only that, in its current form, both Atari host and XEP80 have the potential to go EVEN FASTER, provided that you use Moto's PIA ships, as seen under the scope, here:

 

 

With STOCK hardware, the XEP80 will dump FastBasic's latest manual (ENTIRE 45+ Kbytes) in 1m:17secs, via TYPE command on OSS dos, and in 1m:37secs under SDX.

 

I challenge ANYONE here to show faster times from ANY other 80-col solution (including VBXE or fastest ANTIC-based 80-cols dirivers) running the same test.... except, maybe, my Bit3 card on 800's Slot-3. That would be certainly interesting to see.

 

For those who want to get a glimpse (and enjoy the XEP80 at its very best), just launch Altirra's latest version, add it as a device, and let that 80-col. goodness and speed impress you with "console" (E:) work... I promise you will not be disappointed.

Cheers!

 

 

 

Edited January 11, 2021 by Faicuai

[kogden]

On 1/9/2021 at 10:41 PM, Gunstar said:

I think the PBI "device" we need most is an expansion box. I'm not talking about the 1090 or something similar, just a box (similar to a 1090 only as far as the case goes) that has multiple PBI slots and PCB edge connectors for multiple PBI devices.

 

I've been thinking this for a while and started thinking about giving it a shot to build a simple backplane with internal  SCSI-style 50-pin locking connectors for each slot.  I had started to come up with a simple IDEa-like but simpler IDE interface for CF cards that would fit as well.  Life unfortunately got in the way.

 

I *LOVE* my IDEPlus 2.0, until I had one I never knew how cool it would be to have a real PBI disk controller.  A RAM320XL-style RAM expansion and a card with a couple 16550 UARTs would make me happy as well.

 

It would really be sweet to be able to use PBI as intended with multiple devices.  Lots of fun could be had.  A 1450XLD-style voice synth, COVOX, simple Z80 card for running CP/M based on an eZ80 or something.

 

A standard backplane supported by the community with cheap proto cards available could really open up a lot of possibilities for us.  Rather than a handful of really awesome devices that aren't interoperable or compatible, we could really have something managable.  It would make hardware hacking more newbie friendly as well.  

[MEtalGuy66]

On 1/9/2021 at 12:40 AM, BillC said:

So an MIO/Black Box may be capable of 115,200 baud by adding a 1.8432MHz external clock connected to the 6551 ACIA(may also require a W65C51N), and modified firmware.

 

 

It's already running on a 1.8432Mhz external clock.. There's a big assed 1.8432Mhz crystal right there next to the chip.. On both devices.. 

 

And yeah, it can be done in software. ADTpro on the Apple II does this using the 6551 based serial ports.. 

 

[Irgendwer]

Sounds like I have to pull out my plans for an 80 column solution out of the drawer... ?

 

[Wrathchild]

hmm, you could even use the UNO Cart to manage an 80 column gr.8 screen for you

[Mathy]

Hello Irgendwer

 

6 minutes ago, Irgendwer said:

Sounds like I have to pull out my plans for an 80 column solution out of the drawer... ?

 

Real 80 columns as in 8x8 bit characters?  Color?

 

Sincerely

 

Mathy

 

[Irgendwer]

1 minute ago, Mathy said:

Real 80 columns as in 8x8 bit characters?  Color?

Yes, real ones. Color comes from GTIA as the 80 column text display just monochromatic overlays the normal output (e.g use the 9 colours / 80 columns from graphic 10 and overlay that).

[Mathy]

Hello Irgendwer

 

So it is an internal upgrade?

 

Sincerely

 

Mathy

 

 

[+Stephen]

14 minutes ago, Mathy said:

Hello Irgendwer

 

 

Real 80 columns as in 8x8 bit characters?  Color?

 

Sincerely

 

Mathy

 

VBXE2   Sorry - couldn't help myself.

[Irgendwer]

2 minutes ago, Mathy said:

So it is an internal upgrade?

No, just a cart and monitor adapter, no soldering.

 

Principle is similar to what I once showed with the crosshair of the "CLM17" prototype:

[Mathy]

Hello Irgendwer

 

I still have some empty Atari-Max cartridges shells, red and light blue.

 

Would love to see that 80 column card turn into reality.

 

Sincerely

 

Mathy

 

 

Edited January 11, 2021 by Mathy

[Irgendwer]

1 minute ago, Mathy said:

Would love to see that 80 column card turn into reality.

Me too, as it would be a low cost solution and a new play field for some tricks.

I have all parts since 2+ years already here, but the usual suspects prevent the realization:

• Time
• Motivation 1: When I solved all issues in theory, I often loose interest. ?
• Motivation 2: Esp. for hardware projects, I made the experience, that a realization is not worth the hassle. I'm not talking about making "big money" but if I have even problems to get on-par with the development costs ((dev-)tools, parts, prototypes), motivation is reduced significantly. Compared to that, a software contest participation is a picnic in the park. ?  (The conditions of the hardware contest are very unattractive in this regard.)
• Motivation 3: Production problems (who & how)

Time will tell...

 

[Mathy]

Hello Irgendwer

 

When it comes to cost, ABBUC might be able to help.

 

Which part of the conditions of the HW contest should be changed in your opinion and why?

 

Sincerely

 

Mathy

 

[BillC]

3 hours ago, MEtalGuy66 said:

It's already running on a 1.8432Mhz external clock.. There's a big assed 1.8432Mhz crystal right there next to the chip.. On both devices.. 

I wasn't aware the MIO/Black Box already had a 1.8432MHz crystal connected to the 6551 external clock pins, so 115,200 baud should be just a matter of writing the correct DATA to the 6551 control register.