Jump to content
Sign in to follow this  
mytek

POKEY (AUDIO) and GTIA (BELL) Output Specifications

Recommended Posts

I'm away from a real Atari 8-bit computer and a scope, so I was hoping someone in the know can help me with this.

 

What I am seeking is the typical audio output (at maximum volume, peak-to-peak voltage) straight out of POKEY's 'AUDIO' pin-37 with a typical 1K pull-up resistor, and the same for GTIA's 'BELL' pin-15.

 

Thank you in advance,

 

- Michael

 

Share this post


Link to post
Share on other sites

I'm away from a real Atari 8-bit computer and a scope, so I was hoping someone in the know can help me with this.

 

What I am seeking is the typical audio output (at maximum volume, peak-to-peak voltage) straight out of POKEY's 'AUDIO' pin-37 with a typical 1K pull-up resistor, and the same for GTIA's 'BELL' pin-15.

 

Thank you in advance,

 

- Michael

 

I can't tell you the POKEY audio output easily, but GTIA *should* be a TTL level signal. At least it should look close to that in-circuit.

  • Like 1

Share this post


Link to post
Share on other sites

I can't tell you the POKEY audio output easily, but GTIA *should* be a TTL level signal. At least it should look close to that in-circuit.

 

Yes that makes perfect sense, because the audio requirements are probably very simple for that sound aspect, as in make some different frequency square wave clicks and beeps. POKEY on the other hand is much more sophisticated, and was designed with the intention of producing audio of a better quality. So I'm leaning towards a standard consumer line level audio output of nearly +/- 1 Volt Peak-to-Peak output being what Atari shot for in the chip design.

 

Here's an interesting POKEY audio output circuit I found online. Anyone care to speculate what the RC network tied to ground is all about? Some kind of filter?

 

Atari%20POKEY.jpg

So what I'm considering doing, is to make a TK-II-STEREO board (combined PS/2 keyboard adapter, dual POKEY, and built-in U-Switch). And I would like to keep the audio circuit and parts count minimal, while still producing good audio quality. I looked at Lotharek's Stereo Board schematic, and the audio section uses a lot of components. Is this really necessary?

 

P.S. Not knocking Lotharek's design. Just trying to reduce parts count.

 

- Michael

Edited by mytekcontrols
  • Like 2

Share this post


Link to post
Share on other sites

Nice! I've been wanting a combo board to save space.

 

Yep I've finally started working on the list of things Steve Cardin suggested months ago.

  1. Dual keyboard support -- Done
  2. Combine 3 things associated with POKEY in one board,
    • Dual POKEY
    • TK-II instead of AKI (Candle Stereo Board)
    • U-Switch for electronically controlled Mono/Stereo switching

Should I also add SIO2PC-USB?

 

- Michael

  • Like 1

Share this post


Link to post
Share on other sites

Ok, measured in 130xe with the BIST sound test, and the gtia bell with ctrl 2 buzzer. Can't check maximum volume, because I can't type the sound statement in basic (couple columns of keys out).

post-27376-0-43036900-1468125567_thumb.jpg

post-27376-0-92035400-1468125726_thumb.jpg

Edited by Joey Z
  • Like 1

Share this post


Link to post
Share on other sites

Why you get at the audio out is different to what you get at the pins.

 

I put Pokey's audio out onto a 'scope, you get a fairly uniform square wave as opposed to the audio port which sees the waveform squashed as if it's going through an ADSR sequence.

 

GTIA's audio output is just a general-purpose IO, note that the 5200 uses the pins normally assigned to console keys to select which controller to read.

I would think that the GTIA output at the pin is probably close to what you'd get from the PIA, ie 0 and +5V levels, though external pullup would mean there's likely an "attack" phase where a 0 -> 1 transition takes a few cycles to ramp up.

 

Pokey of course, the audio out pin is an analog quantity though there would be some finite number of levels.

  • Like 1

Share this post


Link to post
Share on other sites

Ok, measured in 130xe with the BIST sound test, and the gtia bell with ctrl 2 buzzer. Can't check maximum volume, because I can't type the sound statement in basic (couple columns of keys out).

 

Thanks Joey, that gives me a very good idea of what I'll be dealing with. The GTIA BELL is as expected, and the POKEY although not at full volume is indicative of being in the range of consumer line level outputs (CD, Tape, ect.). So I think a fairly simple passive mixing circuit with attenuation on the BELL output should work reasonably well.

 

 

Why you get at the audio out is different to what you get at the pins.

 

I put Pokey's audio out onto a 'scope, you get a fairly uniform square wave as opposed to the audio port which sees the waveform squashed as if it's going through an ADSR sequence.

 

GTIA's audio output is just a general-purpose IO, note that the 5200 uses the pins normally assigned to console keys to select which controller to read.

I would think that the GTIA output at the pin is probably close to what you'd get from the PIA, ie 0 and +5V levels, though external pullup would mean there's likely an "attack" phase where a 0 -> 1 transition takes a few cycles to ramp up.

 

Pokey of course, the audio out pin is an analog quantity though there would be some finite number of levels.

 

I'm sure there is a bit of filtering going on after the audio signal exits POKEY, thereby accounting for that squashed waveform you were seeing. Not sure if that's intentional, but if so it might have been an attempt to convert the square wave into something more like a sine wave.

 

- Michael

 

P.S. I was just kidding about adding the SIO2PC-USB to my project. There are plenty of good solutions for that already.

Edited by mytekcontrols

Share this post


Link to post
Share on other sites

*Should say "What you get at the audio out..."

 

Yep, plenty of arcade games of the time sound different to what you hear in emulation too. I think the reasoning is to soften the audio a bit, makes it sound more natural if the waveforms are rounded out a bit. Though of course the type and quality of speakers used makes a huge difference as well.

Share this post


Link to post
Share on other sites

*Should say "What you get at the audio out..."

 

Yep, plenty of arcade games of the time sound different to what you hear in emulation too. I think the reasoning is to soften the audio a bit, makes it sound more natural if the waveforms are rounded out a bit. Though of course the type and quality of speakers used makes a huge difference as well.

 

No problem, I understood exactly what you meant.

 

When I get back to my shop on Monday I'll be doing a little bit of 'scoping' things out myself, and also listen to what is coming directly out of POKEY vs the transistor buffering/conditioning circuitry that Atari designed. Also will try that other passive circuit I posted a bit earlier to see how that compares. Should be quite interesting.

 

Thanks for the feedback :)

 

- Michael

Share this post


Link to post
Share on other sites

What you should see at the audio pin is it pull up to a little below the the voltage at the pullup resistor. The output will be a square wave and the low voltage level will vary with volume setting. The DC offset needs to be filtered out to get a usable audio signal.

  • Like 1

Share this post


Link to post
Share on other sites

GTIA output is, of course, digital, very close to TTL levels.

 

The analog specifications for POKEY are in the datasheet: http://visual6502.org/images/C012294_Pokey/pokey.pdf

 

But I agree with others that what matters is the output of the analog on-board circuit, not just the output at the chip pin.

 

Actually when I posted my initial questions I wasn't looking for the output level coming out of the on-board audio circuit, since I would be creating my own hopefully much simpler audio conditioning circuit from scratch. I needed to know specifically what was coming directly out of pin-37 on POKEY. I have looked at that spec you linked to, but unless I missed something it wasn't perfectly clear as to what signal levels were coming out of the chip, although it did describe the wave forms being generated.

 

 

What you should see at the audio pin is it pull up to a little below the the voltage at the pullup resistor. The output will be a square wave and the low voltage level will vary with volume setting. The DC offset needs to be filtered out to get a usable audio signal.

 

Yes I can see the offset in the waveform that Joey submitted, where the top is at 5 VDC, and the bottom is about 1 volt lower. So in that example, we are seeing approximately a +/- 0.5V modulated signal with a DC offset of nearly 4.5V (to zero crossing of square wave). To get rid of the DC offset, a series connected capacitor makes for a good DC blocker, only allowing the AC (modulated) signal to pass through.

 

I know you have been working on the XM7800 project. Just curious as to what kind of circuit are you guys using for the audio output from the on-board POKEY in that design?

 

- Michael

Edited by mytekcontrols

Share this post


Link to post
Share on other sites

 

Yep I've finally started working on the list of things Steve Cardin suggested months ago.

  1. Dual keyboard support -- Done
  2. Combine 3 things associated with POKEY in one board,
    • Dual POKEY
    • TK-II instead of AKI (Candle Stereo Board)
    • U-Switch for electronically controlled Mono/Stereo switching

Should I also add SIO2PC-USB?

 

- Michael

 

Quad POKEY. :)

Share this post


Link to post
Share on other sites

 

I know you have been working on the XM7800 project. Just curious as to what kind of circuit are you guys using for the audio output from the on-board POKEY in that design?

 

- Michael

It's not my design and I'm not sure if the schematic I have shows the latest, but it appears to be similar to the Ballblazer carts. A 1k pullup with a 12k resistor in series. It goes thru a 2.2uF cap prior to being mixed with the other audio sources.

Share this post


Link to post
Share on other sites

 

Quad POKEY. :)

 

Maybe I can piggy-back 4 Pokey chips :P Seriously unless someone finds a good source for Quad Pokey's (doubtful), or develops a core for a quad Hokey, it ain't going to happen.

 

 

It's not my design and I'm not sure if the schematic I have shows the latest, but it appears to be similar to the Ballblazer carts. A 1k pullup with a 12k resistor in series. It goes thru a 2.2uF cap prior to being mixed with the other audio sources.

 

Thanks tep. So it's really starting to sound to me like a purely passive approach to the audio circuit would work fine.

 

- Michael

Share this post


Link to post
Share on other sites

Well here is my first stab at creating a new Stereo Board with TK-II, U-Switch, and Dual Pokey's incorporated in a single board.

 

CTNZr9q.png

 

All of the audio circuits have been tested (results shown as pseudo O-Scope trace in diagram). The only uncertainty I have at the moment is whether my POKEY SELECT A4 inversion will actually be fast enough. In order to limit chip count I am utilizing two of the 74HCT4066's switches as inverters. The one associated with Mono/Stereo audio switching works fine as expected, since it is a slow speed implementation. However I've never tried to use this method for high speed logic before, so it might not work when it comes to the POKEY SELECT line.

 

I decided to use transistors on the audio outputs for buffering and gain adjustment. Still a relatively simple circuit.

 

- Michael

  • Like 2

Share this post


Link to post
Share on other sites

Minor update.

 

daZzwdz.png

 

 

Added BELL audio mix circuit to right channel so that it'll appear at both audio outputs. Also eliminated 1K pull-up on original (left) Pokey audio out since it already exists on the Atari motherboard.

 

- Michael

  • Like 1

Share this post


Link to post
Share on other sites

 

Maybe I can piggy-back 4 Pokey chips :P Seriously unless someone finds a good source for Quad Pokey's (doubtful), or develops a core for a quad Hokey, it ain't going to happen.

 

- Michael

Let me know if you need help with a Quad Hokey. Not much to do:)

 

Mark

Share this post


Link to post
Share on other sites

Let me know if you need help with a Quad Hokey. Not much to do:)

Mark

Hi Mark,

 

Intriguing idea, but I must confess I don't completely understand how this would be best implemented. So for starters I doubt that we are talking about quadrophonic sound as in 4 independent audio outputs. That leaves it as the possible doubling up of voices available on each of the two stereo channels. So with that in mind, idealy it would be a Hokey that completely took the place of the original Pokey, had 2 additional address inputs (A4-A5), and also was inclusive of the SIO, keyboard scan, and paddle inputs. Is this possible or practical at this time? Or are we still talking about a lot of development left to do?

 

- Michael

 

P.S. This would be a fantastic accomplishment, and certainly warranted due to the seeming scarcity and increasing prices of Pokeys in the wild.

Edited by mytekcontrols

Share this post


Link to post
Share on other sites

Keyboard scan, sio, paddles, etc are all there. Will need to cut and paste to make 4, mix down to stereo and check it runs with a real clock on the real bus. Also some fixes on sio clock in/clock out.

 

Circuit wise, need a suitably low cost fpga and level shifters for 5v/3.3v problems. I think max10 simplifies power supply side and has schmidt trigger inputs.

 

So yeah, not for free but an option with some work.

 

Mark

Edited by foft

Share this post


Link to post
Share on other sites

Keyboard scan, sio, paddles, etc are all there. Will need to cut and paste to make 4, mix down to stereo and check it runs with a real clock on the real bus. Also some fixes on sio clock in/clock out.

 

Circuit wise, need a suitably low cost fpga and level shifters for 5v/3.3v problems. I think max10 simplifies power supply side and has schmidt trigger inputs.

 

So yeah, not for free but an option with some work.

 

Mark

 

Hi mark,

 

If you decide to pursue this, I would love to use it and of course layout the circuit board to suit. However I will need some recommendations on what I'll need (or anyone that produces this) to burn the FPGA's and specifically what FPGA is needed, as well as more specifics on the 5v/3.3v interface issues. As with previous designs I've been involved with, this one would also be a full disclosure project for either DIY, or for someone to pick it up and run with the production side of things.

 

Let me know if you are game for this, and if so I think a new Topic needs to be created around it.

 

- Michael

Share this post


Link to post
Share on other sites

I think You can add Covox (DAC/DACs - for stereo and address decoding/setting circuit).

FTDI for SIO2USB of course could be nice.

Edited by lemiel

Share this post


Link to post
Share on other sites

I think You can add Covox (DAC/DACs - for stereo and address decoding/setting circuit).

FTDI for SIO2USB of course could be nice.

 

Although all of that would be pretty cool, I've got to draw the line with what I am willing to do and what truly makes sense to put into this particular project. SIO2PC-USB has been well covered by all sorts of great solutions both internal and external, and it just doesn't seem imperative that this new Stereo Board include that. Actually the driving force behind this project for me, was to allow TK-II to be installed inside the computer without creating the mini skyscraper that results when also wanting a Stereo Pokey as well. And then it just made sense to also include the U-Switch along with that. Another goal of this new iteration, is that it be easy to build by the DIY person, so no surface mount parts. Of course this means reducing the parts count as much as reasonably possible to still keep the board small. If Mark can come up with a Quad Hokey FPGA design, then this size reduction can be fully realized, and a better fit will be possible across all platforms.

 

And with 8 voices per channel (16 total), do we really need covox? I'm not knocking the suggestion, but once again I am leaning towards the KISS principle.

 

- Michael

  • Like 1

Share this post


Link to post
Share on other sites

Agree we need to avoid too much scope creep. I've created a new topic for this with a few initial notes.

  • Like 1

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.

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...
Sign in to follow this  

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...