Jump to content
drac030

ICD Multi I/O

Recommended Posts

36 minutes ago, danwinslow said:

I'd like to see a PBI internet device. I would think it would be faster than the SIO pathway.

Keep in mind, it's not just about speed either.  With SIO storage and serial modems, you cannot be receiving data and saving it to disk like you can with a PBI solution.

  • Like 1

Share this post


Link to post
Share on other sites
3 hours ago, Stephen said:

Keep in mind, it's not just about speed either.  With SIO storage and serial modems, you cannot be receiving data and saving it to disk like you can with a PBI solution.

Good point. All in all, the PBI seems to have been really underutilized, and I'm not sure why.

  • Like 1

Share this post


Link to post
Share on other sites

Lobbying by the SIO consortium?

 

But seriously, is there any Atari that can't have a PBI?  We all know the 1200xl can be made to have one, and I'm nearly positive at least one of the 800 mods allows that capability... can anything be done for the XEGS and 400?

 

 

Share this post


Link to post
Share on other sites
2 minutes ago, Tillek said:

But seriously, is there any Atari that can't have a PBI?  We all know the 1200xl can be made to have one, and I'm nearly positive at least one of the 800 mods allows that capability... can anything be done for the XEGS and 400?

Incognito allows the connection of PBI devices to the 800 (an XL/XE machine architecture is required for PBI functionality). The XEGS can also be modified for PBI connectivity, but if all you want to put in the cartridge port is another cartridge, there are now a couple of cartridge-slot HDD adapters which are simultaneously capable of emulating almost every available cartridge type.

 

Share this post


Link to post
Share on other sites
4 hours ago, danwinslow said:

Good point. All in all, the PBI seems to have been really underutilized, and I'm not sure why.

Multiple amazing upgrade paths for the Atari have been under utilized for decades.  I don't know why - it's like the Atari methodology of shooting for the lowest common denominator (48kB) machine stuck.  PBI - nah, don't use it.  ECI - definitely not.  VBXE - why the hell would we use that mod.  Far easier to speculate about 20 different ways to get 80 columns on the machine, while we already have it and more!

 

I find it SO strange, that a machine clearly designed from day one in 1978 to be upgradable in both software and hardware, has more resistance to upgrades than any other platform I know of.  All I wanted to do in 1985 was upgrade it.  It's in 2020 and some people almost deem it sacrilegious to go above 16kB, CTIA and cassette.  Very strange.

  • Like 2

Share this post


Link to post
Share on other sites
11 hours ago, danwinslow said:

Good point. All in all, the PBI seems to have been really underutilized, and I'm not sure why.

Mainly, due to the 1090 never reaching the market.

 

  • Like 1

Share this post


Link to post
Share on other sites
4 hours ago, MrFish said:

Mainly, due to the 1090 never reaching the market.

 

Not following that completely. Do you mean that the PBI is too hard to develop for, and the 1090 would have taken care of those difficulties? Are PBI handlers a lot harder to write?

 

Anyway, seems to me that the amazingly capable folks around here could handle it. For that matter, if somebody decided to go ahead and build something like the 1090, some kind of board that exposed the signals and had maybe some RAM of it's own for handlers and some slots/pinouts, etc., that would be useful and I know I'd buy one. There's a chicken/egg problem there, since there's no general reason to buy one without a useful device on it, but if someone were to produce a development board maybe that would get things rolling.  Even just something with some handler Ram and pinouts designed to hook up to Arduino or Raspberry PIs would be cool. SIO is cool and you can do a lot with it, but it's not the fastest thing in the world. Using up the cart port is relatively easy but kind of a pain one way or the other. 

Edited by danwinslow
  • Like 1

Share this post


Link to post
Share on other sites
7 hours ago, danwinslow said:

Not following that completely. Do you mean that the PBI is too hard to develop for, and the 1090 would have taken care of those difficulties?

 

I'm saying that because the 1090, which was intended by Atari to be the primary means of utilizing the PBI bus, never came to market, people/companies generally sought other means of providing upgrades for the XL/XE machines. If the 1090 had existed, companies making upgrades would only need to produce cards that would fit into and work with it. Without the 1090, it was then a companies task to not only bring the upgrade, but also an enclosure, and other necessary electronics and hardware. There's a lot less investment risk in a single board that fits into a 1090 compared with a stand-alone upgrade product. Even today, most upgrades that utilize the PBI/ECI are missing an enclosure; and many people aren't interested in having a bare board laying around on their desktop.

 

  • Like 1

Share this post


Link to post
Share on other sites
1 hour ago, MrFish said:

 

I'm saying that because the 1090, which was intended by Atari to be the primary means of utilizing the PBI bus, never came to market, people/companies generally sought other means of providing upgrades for the XL/XE machines. If the 1090 had existed, companies making upgrades would only need to produce cards that would fit into and work with it. Without the 1090, it was then a companies task to not only bring the upgrade, but also an enclosure, and other necessary electronics and hardware. There's a lot less investment risk in a single board that fits into a 1090 compared with a stand-alone upgrade product. Even today, most upgrades that utilize the PBI/ECI are missing an enclosure; and many people aren't interested in having a bare board laying around on their desktop.

 

Ah. Yeah, I was mostly talking about underutilization over the recent past. We've seen a lot of amazing stuff done, but very little (that I know of) utilizing the PBI/ECI.

Share this post


Link to post
Share on other sites
7 minutes ago, danwinslow said:

Ah. Yeah, I was mostly talking about underutilization over the recent past. We've seen a lot of amazing stuff done, but very little (that I know of) utilizing the PBI/ECI.

I'm talking about the recent past too. If the 1090 had been sold, things being done in recent years would have been drastically different.

 

Yes, not a lot of PBI/ECI devices, but we did have the updated MIO, IDE + 2.0, Turbo Freezer 2011, and SysCheck, for instance.

 

Share this post


Link to post
Share on other sites

sadly our project is over with the passing of Curt V.... I haven't heard from anyone about it since then... don't remember who else was working with on the project as it was a while back now and only in mentions of this or that, last conversation was about the buffer cable and the expense of a proper cable... perhaps cannibalizing ti- expansions box cables and adding the buffers to them would have worked out better...

as for cart port use, not and issue on 800's XL's etc... for the XE's the cart port were added to the device or adapter card

Edited by _The Doctor__

Share this post


Link to post
Share on other sites

The lack of a proper buffered expansion bus on the atari 8bit platform means that anything connected to the data bus skews the bus capacitance/timing all to hell and creates flakiness & instability.  That is the problem with developing for PBI. The PBI standard, itself is very straightforward. 

 

This is not a unique problem n the 8-bit world. The C=64 is in no better shape in this respect.  It was all about cost reduction..  

 

Share this post


Link to post
Share on other sites
21 hours ago, Stephen said:

I find it SO strange, that a machine clearly designed from day one in 1978 to be upgradable in both software and hardware, has more resistance to upgrades than any other platform I know of.

Well when you have no less than 8 different machines essentially all using the same chip set, but all built with no standardization across those machines for expansion other than SIO or re-utilization of joy stick ports.

 

Here's where the C-64 wins. Because you can pretty much count on stuff being in the same relative location, and having the same expansion port from a physical and pin out aspect. Yes we saw a lot of different models of our A8 line, but with the way that was executed it simply promoted mass non-standardization. Expansion port aside, we can't even count on the main IC chips being in the same places or orientation on the individual motherboards, which makes it very difficult to design a drop-in piggyback upgrade board to tie into those chips. Case in point, I had an Osborne 1 which normally had a built-in display that allowed scrolling around within a 52 character wide window to view an 80 column screen. A 3rd party produced an 80 column drop-in (no soldering) piggy-back board that was a fairly simple install that allowed for an external 80 column text monitor to be used. It worked with all Osborne 1 systems because they all had an identical motherboard layout.

 

Share this post


Link to post
Share on other sites
35 minutes ago, mytek said:
22 hours ago, Stephen said:

I find it SO strange, that a machine clearly designed from day one in 1978 to be upgradable in both software and hardware, has more resistance to upgrades than any other platform I know of.

Well when you have no less than 8 different machines essentially all using the same chip set, but all built with no standardization across those machines for expansion other than SIO or re-utilization of joy stick ports.

 

Here's where the C-64 wins. Because you can pretty much count on stuff being in the same relative location, and having the same expansion port from a physical and pin out aspect. Yes we saw a lot of different models of our A8 line, but with the way that was executed it simply promoted mass non-standardization. Expansion port aside, we can't even count on the main IC chips being in the same places or orientation on the individual motherboards, which makes it very difficult to design a drop-in piggyback upgrade board to tie into those chips. Case in point, I had an Osborne 1 which normally had a built-in display that allowed scrolling around within a 52 character wide window to view an 80 column screen. A 3rd party produced an 80 column drop-in (no soldering) piggy-back board that was a fairly simple install that allowed for an external 80 column text monitor to be used. It worked with all Osborne 1 systems because they all had an identical motherboard layout.

I think you misunderstood what I meant.  Regardless of how many models of the A8 we have, with the exception of the 400/800 line, every mod works in every machine and fits in one way or another.  This greatly expands the available userbase for all mods.  It's the community / public - not the machine I mean is resistant to upgrades.  I've never seen a more capable upgrade with such vitriol (or at least complete irrelevance / avoidance) directed towards it than the VBXE.  I wanted 80 columns on my machine back in 82 because even the shitty Apple 2 had it.  Even today, there's no less than 8 topics I can think of discussing 80 column mods or upgrades for the machine.  Most of them are pie in the sky threads about creating more new hardware.  Well we've had hardware than can do 80 columns and a shit ton more for over a DECADE now.  People say "ohh - it's making the machine too modern, it's not original, etc.".  But 1MB even 4MB of RAM is seen as awesome.  65816 processors - oh yeah, that's also cool.  Stereo, quad pokey, great.  SID, MIDI, S2 WaveBlaster for sound, check.  Mention VBXE though and it's like you have leprosy.  I just wish I understood why this one mod elicits this response, and every other single mod, even ones that are just dreams or people asking someone to "just make it" all seem to be super awesome, yeah, let's do it.

Share this post


Link to post
Share on other sites
2 hours ago, Stephen said:

I've never seen a more capable upgrade with such vitriol (or at least complete irrelevance / avoidance) directed towards it than the VBXE.

I can't speak for everyone, but my reason for not having a VBXE is first of all the expense, and 2nd of all the lack of RGB monitors in my country. The cost wouldn't be an issue if I could easily move it from one machine to another, but when you have many systems as I do, it would be costly to outfit everyone. Of course I could dedicate one machine to the cause, but which one? And then there is the chicken and the egg problem when it comes to VBXE support for all the magic it can do, something which has hardly changed in the entire time its been available. However if you want very high quality and very compatible 80 column text, the VBXE looks like the ticket.

 

Lately I find myself going back to S-Video and setting my Sophia DVI and RGB boards aside. The difference for me was to focus on producing a very high-quality S-Video output and going with a high-quality S-Video monitor. For standard Atari video I'm seeing excellent results, and not that much different than DVI to warrant the extra cost. Yes DVI is very crisp, but sometimes almost too much so when viewing the Atari's relatively low rez text. And UAVs are cheap enough I can put one in every machine. For big monitor use, I only need one retrotink connected to my 55" HDTV.

 

Edit: sorry for the off topic posts, this will be my last :) .

 

  • Like 1

Share this post


Link to post
Share on other sites
13 hours ago, MrFish said:

I'm talking about the recent past too. If the 1090 had been sold, things being done in recent years would have been drastically different.

 

Well, true enough of course. The recent past is always different if the less recent is. But even without the 1090 I'm surprised is what I'm saying, not that that the 1090 is/was useless, or wouldn't have made a difference. I agree that if the 1090 had been made there could have been a bunch more stuff. But even as things are, I mean, it seems like we could have made effectively our own 1090 or something similar.

12 hours ago, MEtalGuy66 said:

The lack of a proper buffered expansion bus on the atari 8bit platform means that anything connected to the data bus skews the bus capacitance/timing all to hell and creates flakiness & instability.  That is the problem with developing for PBI. The PBI standard, itself is very straightforward. 

 

This is not a unique problem n the 8-bit world. The C=64 is in no better shape in this respect.  It was all about cost reduction..  

 

Hmm. That's interesting. Is that why they dropped the 1090? I doubt it, they usually had lots of other dumber reasons than that as to why they did things.

Share this post


Link to post
Share on other sites

The 1090 was dropped before the modern XL/XE existed.. By the time the 800XL and 600XL were finalized, ATARI's current product development dept. had resigned themselves to a)producing the XL line cheap enough to compete with the C= 64 price point and b)only using the PBI port for a memory expansion for the 600XL.  Any 3rd party developers wanting to use the PBI bus for complex expansion devices basically had to deal with any problems incurred on their own.

 

The last machines that had a buffered expansion bus were the 14xxXL line. And even some of that was never standardized. The 1090 has provisions on the board for full buffering, but it can be assembled to bypass it. I have seen 1090 prototypes built multiple ways, in this respect. I'm not sure of the level of functionality/stability any given spec had in this regard. They were ALL still prototypes and limited promo/beta/show units. Anywayze, the main point is that Atari originally envisioned the PBI as a workable expansion system, but it got bastardized before it ever saw consumers' hands.. Sad but true.. 

 

When the XE line came along, there were a whole new crop of people at ATARI (both engineering and marketing).  The ECI bus was basically just a more centralized/complete connection standard for what had been the "XL PBI PORT" (and a few additional choice signals) so as not to "close the door" on the possibility of more 3rd party local bussed expansion hardware, but also to reduce costs even further by making the cart port and "ECI" physically overlap..

 

Even in the XE era, ATARI, itself had no intention whatsoever of selling a complex parallel-bus connected device.. They knew the stability problems (and related support nightmare) that would cause. Hence the ridiculous nature of the XEP-80 80 column "solution"..

 

Both ICD and CSS had "custom tuning" departments that required you to ship them your ATARI so they could hook it up in conjunction with the device (MIO or Black Box) and make necessary timing adjustments (add caps.. Swap ICs.. etc.) to various circuits in the event of stability issues that could not be resolved by normal user operator means.  Most cases of this involved heavily expanded machines, but not all.  You can read about a lot of these cases on the old Compuserve, Genie, and usenet archives from the late 80s to mid 90s..  

 

The last thing that's worth noting here is that in recent decades, a lot of people (Candle, HiasSoft, Mega-Hz, Simius, Warerat, myself, and others) have all tried to come up with a somewhat universal "PHI-2 timing fix" that would address these stability issues in a way that would negate (or at least reduce) the amount of "custom tuning" needed to achieve/maintain stability on an ATARI with a heavily expanded parallel bus.  To my knowledge, there is no universal system-wide fix.. There are device specific solutions that can be added to hardware designs which will augment the PHI2 timing in relation to the onboard chips on that device.. But there is still the possibility (Heh. Probability) of the total bus load/capacitance effecting the PHI2 to bus timing with respect to the Atari's internal chips and causing flakiness in many cases. The more bus capacitance (the more crap you've got connected to the system's physical bus), the more likely/prominent this is.. 

 

So.. This, I'm afraid, is simply the "nature of the beast" with these systems...  And as I said, it's all about cost reduction back in the 80s.   

 

For contrast, Apple II is a good example of a 6502 platform on which a relatively stable expansion bus standard was developed.. 

 

 

  

 

 

Edited by MEtalGuy66
  • Like 1

Share this post


Link to post
Share on other sites
On 1/5/2021 at 12:21 PM, MEtalGuy66 said:

Let's see.. The original ICD design has pitifully inadequate heat dissipation on the power circuits..  The rectifier diodes cook themselves, along with the PCB..  ESPECIALLY on 1meg models

Hi, MEtal, I have a mio and the 4 diodes, one end of them are up on the board.. I don't remember if I did it or it was given to me back in the 1988 or so.. 

My question or two are when the diodes / IC reg. went, did it do and of the other IC and things bad. did it over voltage them or ???

Did you see any other problems when the reg/diodes went?? What was the input voltage for the unit??

Sorry to ask but I just found all my Atari stuff and I am trying to remember all the parts and put my old BBS back on the air...

 

Peter

 

 

Share this post


Link to post
Share on other sites
10 minutes ago, Peter Rabitt said:

Hi, MEtal, I have a mio and the 4 diodes, one end of them are up on the board.. I don't remember if I did it or it was given to me back in the 1988 or so.. 

My question or two are when the diodes / IC reg. went, did it do and of the other IC and things bad. did it over voltage them or ???

Did you see any other problems when the reg/diodes went?? What was the input voltage for the unit??

Sorry to ask but I just found all my Atari stuff and I am trying to remember all the parts and put my old BBS back on the air...

 

Peter

 

 

Usually, the diodes get so hot that they delaminate the copper traces from the PCB, burn off the green soldermask, and even turn the fiberglass of the board dark brown/black..  Before finally prematurely failing due to obvious over-heat (over-current and lack of cooling) conditions.  

The large aluminum plate is adequate heat-sink for the 7805 regulator, but I have seen those fail too. 

 

If you want to fix the power circuit back to factory functionality, just replace the 7805, diodes (Id go much bigger or use a 2A+ rated bridge rectifier unit instead,) and the 3 largest electrolytic caps (these are probably dried up and shorted with age).

 

If you send me a pic of your board in PM, I can make further suggestions. 

 

 

Edited by MEtalGuy66

Share this post


Link to post
Share on other sites
14 hours ago, danwinslow said:

Well, true enough of course. The recent past is always different if the less recent is. But even without the 1090 I'm surprised is what I'm saying, not that that the 1090 is/was useless, or wouldn't have made a difference. I agree that if the 1090 had been made there could have been a bunch more stuff. But even as things are, I mean, it seems like we could have made effectively our own 1090 or something similar.

There's been a fair amount of talk over the years about building some sort of 1090-type device. @Dropcheck apparently did a fair amount of assessment about recreating something close to the 1090 itself (not sure at what point or why she abandoned the idea), and @Spaced Cowboy actually did a lot of work on his own idea for an Atari 8-bit upgrade box (apparently efforts were halted mainly due to life adjustments as a result of COVID-19).

 

So, there's really no lack of interest or effort concerning such a device, as of late; but ultimately, there's a lot more to tackle when a box alone would just be setting the table. From what I recall, the device @Spaced Cowboy is planning isn't for everyone too (it has some amount of functionality built into it, and has a fairly high price tag as a result -- something like $300, IIRC), which means there could be some shortage of hardware developer support, as a result of a small user base.

 

 

Edited by MrFish

Share this post


Link to post
Share on other sites
On 1/8/2021 at 6:55 PM, MrFish said:

From what I recall, the device @Spaced Cowboy is planning isn't for everyone too (it has some amount of functionality built into it, and has a fairly high price tag as a result -- something like $300, IIRC), which means there could be some shortage of hardware developer support, as a result of a small user base.

Yes it was going to be a bit pricey, but extremely powerful - bringing much more to the table than a simple buffered parallel expansion box. If I recall, it was also going to feature a high speed serial I/O link between the A8 and itself, thereby allowing for a small diameter cable connection between the two, thus more flexibility in expansion box placement. I believe he created a webpage describing all the features, which there is probably a link in the original AA topic he started.

 

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...