Dual Pokey... so much more could be done.

[Bryan]

When you put 2 Pokeys in a system, you get a lot more than 4 more voices. Someday I'd like to look into designing an upgrade that does more.

 

Here are the possibilities I've thought of:

 

1. 2nd SIO port, or a dedicated RS-232 port with a little extra interfacing.

One shortcoming of the SIO scheme is that you can't easily read or save data while using a serial connection. That problem would go away if you could open a 2nd channel. Also, a 2nd serial port gives you a way to talk to other microcontrollers.

 

2. 8 more Pot ports.

There's not much call for 8 player paddle games, but they can also be pressed into service as digital inputs allowing 2 more joysticks.

 

3. Keyboard IO

You could support a larger keyboard, or other alternative controllers. At the very least you'd have triggers for the above joystick ports.

 

 

[+mytek]

I would definitely vote for a 2nd serial port inclusive of RS232 conditioning circuits. Then route that out to a standard com port via a 9-pin female D-Sub connector. Ultimately that would make it much more useful, and could then directly connect with an older PC or a newer one utilizing the very common USB serial adapters. This would also provide the bridge for use with the multitude of RS232 instrumentation devices including ones utilizing ModBus RTU protocol. And with the standard SIO untouched, it would make for a great DAQ system with simultaneous data logging to disk.

 

- Michael

Edited November 8, 2016 by mytekcontrols

[Marius]

You are SO RIGHT!

 

And great of all, all these things can be there at the same time, since all the single things are there now too at the same time!

[UNIXcoffee928]

What I would like to have would be a Quad POKEY board that sent all of the port lines to a 1" to 2" high breakout-box with front & rear ports; a device which could be placed below the 800 computer unobtrusively. This would allow for the ease of use of the front joystick ports, like on the 400/800, and have separate audio lines & SIO in the rear of the device, to allow ports for each sound channel to be fed into an external mixing board, while hiding the mess of wires behind the computer.

 

Unfortunately, this would result in a ridiculous amount of work to achieve, due to the high number of signals that would each need it's own wire. While it could be done, it would end up being prohibitively expensive to produce. It would be a cool one-off design, or a bill of materials for a kit though.

 

Has anyone even made a Quad POKEY for any of the Atari 8-bits? I know that they were used in the arcade machines, so it is possible. It would probably need a CPU accelerator to get the most out of the quad configuration.

[+mytek]

I would definitely vote for a 2nd serial port inclusive of RS232 conditioning circuits. Then route that out to a standard com port via a 9-pin female MALE D-Sub connector.

 

 

 

Also it would be a good time to integrate SIO2PC-USB parallel connected to Standard SIO port.

 

- Michael

[mono]

3 extra TIMER IRQs...

[emkay]

Referring to the design of POKEY, it is created to take use of one 6502 CPU. So, if you want to have every added POKEY used correctly, you have to add an extra 6502 CPU. Anything else is dreaming on a pink cloud...

[UNIXcoffee928]

Referring to the design of POKEY, it is created to take use of one 6502 CPU. So, if you want to have every added POKEY used correctly, you have to add an extra 6502 CPU. Anything else is dreaming on a pink cloud...

A 65816 would probably be needed to have acceptable results with quad POKEY audio. You're probably right about many of the other POKEY functions needing a dedicated CPU... otherwise, some kind of buffering circuit would be necessary, and that would cause lag.

 

The Quad POKEY design was used in the "Star Wars" arcade machine, via two 6809s.

 

Specs for the Star Wars Arcade Cabinet:

 

Main CPU : Motorola M6809 (@ 1.512 Mhz)

Sound CPU : Motorola M6809 (@ 1.512 Mhz)

Sound Chips : (4x) POKEY (@ 1.512 Mhz), Texas Instruments TMS5220 (@ 640 Khz)

[Timothy Kline]

Sounds like a great idea for a drop-in card for Dropcheck's 1090 follow-up! A dedicated sound card for the Atari 8-bit via the PBI. ^_^

[shoestring]

You would definitely need another CPU. Most arcades had at least a second CPU for sound.

 

With just a single CPU and an extra POKEY you're not going to have enough cycles left for anything useful.

[Timothy Kline]

You would definitely need another CPU. Most arcades had at least a second CPU for sound.

 

With just a single CPU and an extra POKEY you're not going to have enough cycles left for anything useful.

 

Is it technically possible to set up that second CPU on a drop-in card that would install into Dropcheck's 1090 clone and onto the PBI?

 

Can two 6502s be on the PBI at the same time yet operate in their own 6502 ecosystem: the primary 6502 with the computer; the second 6502 strictly committed to the POKEYx4 "sound card" for processing?

 

And is the Atari 8-bit limited to POKEY for sound, anyhow? We have the VBXE, which takes the video to a whole new playing field, supplanting dependency on the limitations of the stock 8-bit's video.

 

Most computers have on-board sound, so the tech has reached that level of micro that it seems possible to do a VBXE-styled sound card that could bring, through programming API, as much as 5.1 audio, if the card supported it. Having this in an edge card in that 1090 clone means that the rear of the card could provide outputs, avoiding having to physically alter any actual casing.

 

What would such a card require in components? In PCB size? What about the DSP and the ecosystem it would require? Old sound cards would be a great and proven template for any project of this type if the endgame was an edge-based PCB.

 

Anyhow, just thinking out loud. I certainly don't have any of the skills any of this would require. But I would surely buy one to support the developer's efforts.

 

--Tim

[Rybags]

There's been plenty of 8-voice music around for years and it copes with a single CPU just fine.

 

Star Wars needed the second CPU since Star Wars uses sampled speech and 4 Pokey chips. Similar case with Gauntlet, Pokey doing speech + 2 other sound chips doing most of the remainder.

[+tf_hh]

Referring to the design of POKEY, it is created to take use of one 6502 CPU. So, if you want to have every added POKEY used correctly, you have to add an extra 6502 CPU. Anything else is dreaming on a pink cloud...

 

IMHO this is a failed assumption. The POKEY chip is the one and only in the Atari 8-Bit chipset which has two (!) chip-select signals. Atari would have advisedly the goal to enable more than one POKEY in a system or cabinet. And CPU power isn´t the issue here.

[emkay]

IMHO this is a failed assumption. The POKEY chip is the one and only in the Atari 8-Bit chipset which has two (!) chip-select signals. Atari would have advisedly the goal to enable more than one POKEY in a system or cabinet. And CPU power isn´t the issue here.

???

 

Even the combination inside the Atari isn't able to feed POKEY completely.

[+mytek]

Even the combination inside the Atari isn't able to feed POKEY completely.

I guess it all depends upon what you want to do, because the same could be said for the CPU/Antic/GTIA combo when trying to achieve greater detail and/or more sprites than what was originally envisioned. But when used within the confines and likely intentions of the chip design, two POKEY's can be utilized without over taxing the CPU. And this has been shown to be the case time after time (there are many examples of such to be found). However I'll concede that for a musician or the graphic artist, CPU cycle time rapidly becomes an issue when trying to obtain perfection.

 

- Michael

Edited November 9, 2016 by mytekcontrols

[ijor]

IMHO this is a failed assumption. The POKEY chip is the one and only in the Atari 8-Bit chipset which has two (!) chip-select signals. Atari would have advisedly the goal to enable more than one POKEY in a system or cabinet. And CPU power isn´t the issue here.

 

Not only is has two chip select signals, the schematics show that the polarity of those signals are programmable. But not there are no actual traces of that feature on the chip layout.

 

Not sure this was intended for multiple chips in the same system. May be more because Pokey was designed as a generic audio chip to be used in arcade systems?

[emkay]

Not sure this was intended for multiple chips in the same system. May be more because Pokey was designed as a generic audio chip to be used in arcade systems?

That's it. Also (the) POKEY(s) always had a dedicated CPU in the Arcade Machines.

If the 1090 System had found it's way to the light, things would have been more obvious.

[Timothy Kline]

That's it. Also (the) POKEY(s) always had a dedicated CPU in the Arcade Machines.

If the 1090 System had found it's way to the light, things would have been more obvious.

 

Correction:

When If the 1090 System has had found it's way to the light, things will be would have been more obvious.

 

 

Bring it on, Dropcheck! ^_^

 

--Tim

[Bryan]

That's it. Also (the) POKEY(s) always had a dedicated CPU in the Arcade Machines.

If the 1090 System had found it's way to the light, things would have been more obvious.

 

Not so.

 

Here's a Centipede PCB. The 40-pin ICs are a 6502 and a Pokey.

 

http://www.tvspels-nostalgi.com/Bilder/PCB/Atari/centipede_pcb.jpg

 

Tempest has 2 Pokeys, but only one 6502.

 

http://www.scvgames.com/images/items/tempest-pcb_02.jpg

 

I could list a lot more examples.

[emkay]

Not so.

 

Here's a Centipede PCB. The 40-pin ICs are a 6502 and a Pokey.

 

The SIO usage is tremendous and the music is well done.

 

http://www.tvspels-nostalgi.com/Bilder/PCB/Atari/centipede_pcb.jpg

 

Tempest has 2 Pokeys, but only one 6502.

 

http://www.scvgames.com/images/items/tempest-pcb_02.jpg

 

I could list a lot more examples.

I wasn't aware that a 6502 can do vectors that fast.

[emkay]

To get more serious.

2 POKEYs can already break a fluent gameplay, when using a 60Hz machine. SIO needs the CPU to pull the Data....

 

That's the point why every "use 2 or more chips of the same type in the A8" isn't really useful. With restrictions, you get a stereo , with restrictions, you get double interfaces for communications, but you cannot do that at a the fastest speed.... Well, you'd get the double of paddle registers, which seems useful in some way.

But, wasn't it better to look at different chip solutions?

IF POKEY was able to read and write to the memory, things would make more sense.

[emkay]

Correction:

When If the 1090 System has had found it's way to the light, things will be would have been more obvious.

 

 

Bring it on, Dropcheck! ^_^

 

--Tim

So , you're working on it?

[Timothy Kline]

So , you're working on it?

 

If I had the technical wherewithal, I definitely would be working on it.

 

Yet if Dropcheck is able to pull off her 1090 "reboot," it seems an ideal environment for the people who can do these types of projects to test their mad skillz on the new platform: the PBI bus.

 

Which goes to my post earlier, where it seems like we could have a soundcard on the level of a VBXE, providing extensibility to sound on the 8-bit in an edge-card PCB that could even provide input/output jacks similar to today's PC sound cards. If we can do it on a video level, surely we can do it on an audio level, too, right?

 

--Tim

[Bryan]

I wasn't aware that a 6502 can do vectors that fast.

 

It can't. Many Atari vector games have AMD2901 bit-slice chips to help with the math.

[emkay]

It can't. Many Atari vector games have AMD2901 bit-slice chips to help with the math.

That's the point.

There are no miracles.