SIO connector alternative

[raster/c.p.u.]

Some time ago somebody started the discussion about alternative connector for Atari SIO port. Because I have a few devices which are connected to Atari SIO and I plan to make others, I like this idea for some new _standard_ for Atari SIO port.

I don't know if discussion about it has ended and if any new standard was choosen already?

 

(I want to create adapter cable between Atari SIO and new standard connector and I want to be conformable to the others. )

[Rybags]

From memory, the popular suggestion seemed to be the DB-15 connector, as already used in VGA monitor plugs.

 

I remember there being a thread here about it, that linked to another site where the "standard" was being discussed.

 

No idea though if any further progress was made or a standard agreed upon.

[krupkaj]

I think you mean DB 15 as used in sound cards for joystick and MIDI, not high density VGA connectors with three rows of pins.

[raster/c.p.u.]

From memory, the popular suggestion seemed to be the DB-15 connector...

Yes, standard 15 pin cannon DB-15 connector was mentioned there I think, but there is a lot of variations: male/female, pinouts assignment (1=1,..,13=13, 14+15unused / or Atari SIO assignment style / or ? ).

btw - For a long time I have decision there (for DB-15) would be best the Atari SIO style pinouts assignment and male gender, but now I discover the problem with connector bottom-top orientation when DB-15 for PCB is used (PCB would have to be bottom up and all the other electronic parts too).

[Sikor]

Hi, Raster!!! Look here: http://atariarea.krap.pl/forum/viewtopic.php?id=3352 and here: http://atariarea.krap.pl/forum/viewtopic.php?id=2564&p=2 - some times ago I proposed on atariarea forum to make ESIO port (Enchanced SIO) on DB15 or D-SUB 15 connector.

VGA connector is smaller, than - You can use it for example between standard SIO and cartridge port. I propose to connect new signal to pins 14 and 15 - for example HD Floppy detector or simple Audio IN (Left/right, analog) or others.

Edited September 26, 2006 by Sikor

[raster/c.p.u.]

...to make ESIO port (Enchanced SIO)

I'm sorry. I don't care to design any "enhanced" SIO port, I'm looking for substitution of rare Atari SIO connectors (male/female).

No building into Atari will be necessary, but one easy adapter cable will be sufficient for all sio devices.

Ok, here is my suggestion on the picture in attachment.

Advantages of this idea:

- pinouts placement is similar to Atari SIO (no wires crossing)

- d-sub M15 is male as well as Atari SIO connector

- d-sub M15 has proper top-bottom orientation for PCB mounting

Your opinion?

[Almost Rice]

There is already a SIO alternative made by a8maestro.com.

 

It uses a DB25 connector. I have not used it so I do not know if its what you are looking for. Also DB25 switches and parts are probably going to be easier to find than DB15(in the US for sure). I do not know about your parts availability in the Czech republic.

Edited September 26, 2006 by Almost Rice

[raster/c.p.u.]

There is already a SIO alternative made by a8maestro.com.

It uses a DB25 connector. I have not used it so I do not know if its what you are looking for. Also DB25 switches and parts are probably going to be easier to find than DB15(in the US for sure). I do not know about your parts availability in the Czech republic.

Thanks. I wanted explore it, but I can't find scheme of "EASIO-S SIO adapter cable" wires connection on a8maestro's website.

btw - cannon d-sub15 connectors availability is good in the Czech Republic. It's a "normal" connector as well as d-sub9 and/or d-sub25.

[HiassofT]

Bernhard Engl and I are currently working on a SIO connector replacement that uses standard RJ45 (network) cables to connect peripherals to the Atari. The first prototypes are an adapter SIO to RJ45 and an SIO2PC interface with a RJ45 connector. New peripheral devices could then just provide 2 RJ45 jacks and can be daisy-chained with network cables just like standard SIO devices. To connect a new RJ45 devices to standard SIO devices, you just need a single adapter. You could even add a RJ45 jack to your Atari and connect devices directly.

 

Currently the signals ClockIn, ClockOut, Proceed, Interrupt are missing (since the RJ45 cable has only 8 wires), but we are thinking about building an "active" version of the interface that multiplexes Proceed and Interrupt over a single, additional wire.

 

It would be interesting to know which extensions really use the Proceed and Interrupt lines and which timing constraints must be met. None of the standard peripherals (disk drives, tape recorders, printers) use these signals, so the current prototype already works fine for these applications.

 

The nice thing about the RJ45 jacks is that they are quite small and easyly available, as well as RJ45 network cables. The SIO end of the interface is built from pins out of a female DB25 connector and just soldered to the PCB.

 

Here's an image of the SIO-to-RJ45 adapter and of the RJ45 SIO2PC interface:

 

Here's an image of the SIO-to-RJ45 adapter plugged into an 800XL (the same adapter can be plugged into a 1050 etc.):

 

Please note: this is just a first proof-of-concept prototype. If you have some more ideas/comments/... please let us know.

 

so long,

 

Hias

[ataridano]

I, for one, never liked the idea of trying to mix and match the existing SIO connectors with another type of connector, so I've been working on a small board that has holes to mount pins to make a male SIO connector, and this could be connected to any new hardware, and mounted on stand-offs inside the case. The origional plan was so that I could use a plain old SIO cable with an SIO2IDE drive, but thought it would be useful for other things as well. I know it doesn't solve the problem of creating more SIO cables, but at least I don't have to hack a cable to attach new hardware. I haven't ordered boards yet but hope to before too long, If anyone is interested let me know.

 

-Dan

[MEtalGuy66]

I, for one, never liked the idea of trying to mix and match the existing SIO connectors with another type of connector, so I've been working on a small board that has holes to mount pins to make a male SIO connector, and this could be connected to any new hardware, and mounted on stand-offs inside the case. The origional plan was so that I could use a plain old SIO cable with an SIO2IDE drive, but thought it would be useful for other things as well. I know it doesn't solve the problem of creating more SIO cables, but at least I don't have to hack a cable to attach new hardware. I haven't ordered boards yet but hope to before too long, If anyone is interested let me know.

 

-Dan

 

Dan, Id love to see your idea. I could possibly incorporate it in devices I may build in the future.

[MEtalGuy66]

Bernhard Engl and I are currently working on a SIO connector replacement that uses standard RJ45 (network) cables to connect peripherals to the Atari. The first prototypes are an adapter SIO to RJ45 and an SIO2PC interface with a RJ45 connector. New peripheral devices could then just provide 2 RJ45 jacks and can be daisy-chained with network cables just like standard SIO devices. To connect a new RJ45 devices to standard SIO devices, you just need a single adapter. You could even add a RJ45 jack to your Atari and connect devices directly.

 

Currently the signals ClockIn, ClockOut, Proceed, Interrupt are missing (since the RJ45 cable has only 8 wires), but we are thinking about building an "active" version of the interface that multiplexes Proceed and Interrupt over a single, additional wire.

 

It would be interesting to know which extensions really use the Proceed and Interrupt lines and which timing constraints must be met. None of the standard peripherals (disk drives, tape recorders, printers) use these signals, so the current prototype already works fine for these applications.

 

The nice thing about the RJ45 jacks is that they are quite small and easyly available, as well as RJ45 network cables. The SIO end of the interface is built from pins out of a female DB25 connector and just soldered to the PCB.

 

 

Hmm. If we go to a new standard, I think it should be an improvement from the standpoints of noise reduction. I agree that cat-5 cable is obviously cheap & available.

 

However, if we are going to create a new standard, Id improve it. I have a stack of 8 XF551s.. By the time the SIO chain gets to the last drive, its very noise-prone.

 

Cat-5 is a twisted-pair cable. It would be excellent if there were enough pairs to dedicate a pair of wires to each signal (in otherwords, each signal could be coupled with a ground).. But since there are only 4 pairs (8 wires total), you are talking about using the wires individually for signals, and you gain nothing from the standpoint of reduction of noise/crosstalk.

 

There are other cable standards which have 8, 16, or more twisted pairs, and even have double shield layers under the outer sheathing.. I plan to use this type of cable to "fix" all my existing SIO cables at some point, but as far as creating a new connection standard, this really doesn't help us.

 

Im just saying that if we are going to create a new standard for SIO cables in their entirety, I'd go towards something that is going to increase overall reliability/stability.

Edited September 29, 2006 by MEtalGuy66

[ataridano]

 

Dan, Id love to see your idea. I could possibly incorporate it in devices I may build in the future.

 

 

My idea is really simple, it's just a PCB with pads or holes on it spaced such that they line up with the pins of a male SIO connector. My plan was to find pins of the correct diameter and solder them in place, poking out of the board. Then mount the board such that I'd have an opening cut in my case and the pins exposed behind the opening. At the rate I get things done (not nearly as fast at the MIO) it might take me a while to get them ordered, but it is on my to-do list, and I'd gladly send you one.

[ataridano]

 

Cat-5 is a twisted-pair cable. It would be excellent if there were enough pairs to dedicate a pair of wires to each signal (in otherwords, each signal could be coupled with a ground).. But since there are only 4 pairs (8 wires total), you are talking about using the wires individually for signals, and you gain nothing from the standpoint of reduction of noise/crosstalk.

 

While it's kinda "cheating" the power pins could be treated as a virtual ground, while not at 0V they should be a relitavely constant voltage and could probably be used as a ground. This may help some as long as there aren't voltage spikes. Then again, the SIO doesn't run at that great a speed, so there might not be much advantage in even the best cable, as you'd still have all the unshielded areas inside the drive/periphal inclosures.

[ijor]

Currently the signals ClockIn, ClockOut, Proceed, Interrupt are missing (since the RJ45 cable has only 8 wires), but we are thinking about building an "active" version of the interface that multiplexes Proceed and Interrupt over a single, additional wire.

 

It would be interesting to know which extensions really use the Proceed and Interrupt lines and which timing constraints must be met.

 

I'm not aware about any usage of those lines. But I understand that ClockOut is used by Indus drives. At least I remeber somebody mentioned this here some time ago. Even if it doesn't, it might be useful for autodetection the SIO rate with SIO2PC applications.

 

I have a stack of 8 XF551s.. By the time the SIO chain gets to the last drive, its very noise-prone.

... I plan to use this type of cable to "fix" all my existing SIO cables at some point

 

You should consider that by using much better insulation you would lose the SIO sound when SOUNDR variable ($41) is off. Because this is precisely a product of the crosstalk in the signals.

 

Some might not care, some might even say this is a good thing. But this is something that should be considered.

[kenfused]

 

 

You should consider that by using much better insulation you would lose the SIO sound when SOUNDR variable ($41) is off. Because this is precisely a product of the crosstalk in the signals.

 

Are you sure about that. I know with SOUNDR off, you can sometimes still here the sound a little which is probably due to crosstalk or leakage somewhere, but when its on, I thought that was a function of pokey?

[Rybags]

With 8 wires, shouldn't you have:

+5 V

GND

Command

Data In

Data Out

Audio In

Cassette Motor

 

and one extra. I would assume you're using it for +12 Volts. But, isn't it unused by anything out there? And not even connected on the 1200XLs?

 

AFAIK, nothing uses the Interrupt and Proceed lines. But having one of them would be handy for anything that might need them in future. As it is, it's really the only reliable way for an otherwise idle peripheral to get the computers attention.

 

A simple way of multiplexing might utilize the data in line to differentiate between Proceed and Interrupt. POKEY allows sampling directly the input on data in, although using that method would possibly cause the serial input shift register to start receiving non-existant data.

[HiassofT]

Concerning noise immunity: DataIn and DataOut will be on a wire pair that is twisted with GND (or VCC, respectively), so interference shouldn't be a big deal. Additionally, if we use a shielded jack and STP cables, the cable shield will be also on GND.

 

Currently we also run the SIO pin 12 ('NC') through the RJ45 cable and though about using it for future expansions. Some kind of 1-wire-bus that is connected to a small µC inside the Atari which can then be used for external keyboards or other stuff.

 

The missing ClockOut seems to be a problem that I wasn't aware of yet. Another new interface (SIO2USB) also seems to use it.

 

Yet another idea to get a free pin was to throw out the AudioIn pin and add a Cynch/RCA jack to the adapter so that the AudioIn signal is run through a separate cable. This will also eliminate crosstalk with the AudioIn signal and IIRC this signal is only used by the tape recorder. We also thought about building a tape recorder interface so that one could connect a standard tape deck to the Atari, but I'm not sure if anyone would still want to use tapes these days...

 

If one wants to use long cables (or a lot of devices), an active adapter with additional line transceivers is needed. We are also thinking about that.

 

so long,

 

Hias

[drac030]

This will also eliminate crosstalk with the AudioIn signal and IIRC this signal is only used by the tape recorder.

 

... and floppy drives.

 

PS. Honestly, I don't think that an "alternative SIO connector" with any missing signals (i.e. a crippled SIO connector, in fact) will get a wider warm acceptance.

Edited October 2, 2006 by drac030

[HiassofT]

This will also eliminate crosstalk with the AudioIn signal and IIRC this signal is only used by the tape recorder.

 

... and floppy drives.

Which floppy drives have the AudioIn pin connected? My 1050 schematics only list DataIn, DataOut, Command, VCC/Ready and GND. If one really wants to have/simulate crosstalk, he/she can easily do that by using a few resistors to connect DataIn/Out to AudioIn.

 

PS. Honestly, I don't think that an "alternative SIO connector" with any missing signals (i.e. a crippled SIO connector, in fact) will get a wider warm acceptance.

ACK, and that's also one of the reasons for my postings. I'd like to know which of the signals are really needed, which electrical and timing constraints must be met and then decide which way to go.

 

So far, all other proposed alternatives (except re-building compatible SIO connectors) are quite unsatisfactory. DB25 (which could carry all signals and also has the pro that cables are easily available and cheap) is way too big, DB15 (PC-joystick) cables aren't that common and easy to get, HD15 is expensive and you can't be 100% sure that really all pins are connected 1:1, and DB9 has too few pins.

 

Personally I think that Bernhard's idea (it wasn't mine, honestly) of using RJ45 cables is quite nice and worth thinking about, but in no way I want to enforce a new (pseudo) standard. IMO we all should stay with the old standard SIO connectors until a wide majority of the community has decided upon a new standard - IMO noone needs several new, incompatible connectors.

 

so long,

 

Hias

[drac030]

Which floppy drives have the AudioIn pin connected?

 

At least the LDW 2000 and CA-2001. I am not sure about the 1050 Top Drive. I don't know what you mean by "crosstalk", the drives use the AudioIn signal to make beeps at certain error conditions.

 

ACK, and that's also one of the reasons for my postings. I'd like to know which of the signals are really needed,

 

This is what I am talking about: any missing signals = crippled SIO. Which implies, that any alternative connector has to provide all existing signals (IMHO), as even if something has not been used so far, you never know, what future devices you kill cutting the wire. There's for example the CLOCK IN (aka BIDIRECTIONAL CLOCK) apart of the CLOCK OUT discussed here, and I know that at least one guy is building an experimental device using that signal. So you can imagine that he'll really love any new connector without clock signals connected.

 

Atari designed the SIO to be an universal connector and implemented several lines in it to accomplish that. That universality is one of SIO features, and so limiting it to signals "only really needed" (for whom?) would severely limit facilities the SIO provides.

[Rybags]

I actually like the idea of an RJ45 plug.

 

The whole idea is to provide an alternative - existing peripherals need not be modified if you don't want them to be.

 

In theory anyway, you could probably have 2 daisy chains going to a certain extent. One with legacy SIO and the other with RJ45. I assume electrical limitations would probably come into play.

 

The biggest pain with SIO, aside from it's "non-standardness" is the cables themselves. They are inflexible, too short, and too bulky.

 

Maybe the ideal setup would be a small external "hub". A standard SIO cable plugs into it, then it has a pass-through SIO as well as the reduced capability RJ45 variant.

[HiassofT]

At least the LDW 2000 and CA-2001. I am not sure about the 1050 Top Drive. I don't know what you mean by "crosstalk", the drives use the AudioIn signal to make beeps at certain error conditions.

Ah, OK. Thanks for the info, I didn't know this.

 

This is what I am talking about: any missing signals = crippled SIO. Which implies, that any alternative connector has to provide all existing signals (IMHO), as even if something has not been used so far, you never know, what future devices you kill cutting the wire. There's for example the CLOCK IN (aka BIDIRECTIONAL CLOCK) apart of the CLOCK OUT discussed here, and I know that at least one guy is building an experimental device using that signal. So you can imagine that he'll really love any new connector without clock signals connected.

 

Atari designed the SIO to be an universal connector and implemented several lines in it to accomplish that. That universality is one of SIO features, and so limiting it to signals "only really needed" (for whom?) would severely limit facilities the SIO provides.

IMHO it is very important to re-evaluate the old design before starting to work on a new design. The original SIO was designed some 30 years ago and the world of electronics has changed a little bit since then. So, my question is/was: which signals are really needed for new devices (old devices can still use the standard SIO connectors, there's no need to re-build your existing SIO chain). Interrupt and Proceed seem quite useful to me, but so far I can't see why anyone would like to use synchronous communications via ClockIn/Out these days. You can do anything you'd like asynchronously, even adapting to varying bitrates without loosing a single byte.

 

Please note: I really don't want to offend any one, I'm just asking for arguments pro/con keeping a signal ("it's always been there, so let's just keep it though it's never been used" doesn't count, obviously).

 

so long,

 

Hias

[drac030]

so far I can't see why anyone would like to use synchronous communications via ClockIn/Out these days. You can do anything you'd like asynchronously, even adapting to varying bitrates without loosing a single byte.

 

This is rather easy to argue:

 

* the CLOCK OUT is needed for disk drives (and devices alike), if they want to determine _reliably_ the baudrate requested by the computer. Accommodating to the baudrate can be done without that, by trial and error, but the "error" part of this method is a bit risky: trial and error means that the drive is quite often receiving garbage instead of a valid command, and may act unexpectedly. Actually I know a type of a disk drive, which, when received certain command and wrong baudrate, interprets it as FORMAT DISK, and - due to an independent bug in the ROM, proceeds with execution (_not_ funny). Determining the baudrate according to the CLOCK OUT signal is much more elegant and safe.

 

* the CLOCK IN signal can be used in a disk drive (and devices alike) to change the baud rate of the computer's Pokey without exchanging any commands between the computer and the drive. For example, the computer sends the normal READ SECTOR command at 19200, and the drive can, just by supplying an appropriate CLOCK IN signal, change the baudrate and send responses and data at the speed it likes.

 

That last possibility of course requires the XL OS SIO code to be slightly changed so that one could enable the external sync mode when needed. But this is not any problem these days.

Edited October 2, 2006 by drac030

[Rybags]

Another alternative. Two sockets.

 

Carry the most common signals over RJ45. Have a RJ11 or RJ12 (phone type) for the remainder.

 

The big advantage of this is that both network and phone cables can be used, unmodified, and at very low cost.