Home Page Forums MAME MAME TMS-09xx/1xxx thread (was New Dumps)

Could you make an audio recording of these?:

- race car (hold button)
- default synth
- F0A F0B F7C F0D F0E F0F
- F0A F0B F15C F0D F0E F0F
- F0A F0B F255C F0D F0E F0F

In MAME, the default synth sound doesn't oscillate pin 13. It just goes high and stays high, that's why volume doesn't decay there.
Does it oscillate on the real device? If so, it's an emulation bug with the non-analog part.

Command C with custom synth sounds does oscillate pin 13 on MAME.

At the moment I have no sound recordings (only captured the pin levels in audio mode by logic analyzer) but did some video including sound that is still in the digicam. (I will put it on a file hoster when I get it downloaded on the laptop.)

The default synth (press 'musical note' and any key except 'F') should sound like a guitar (fast attack, then decaying within about 0.7s). Enter now 'F' '0' 'C' to disable the envelope, which turns it into a steady toot at 100% volume (stays on so long the note key is held). Pin 10 apparently charges the volume capacitor to 100% at begin of any note.

The total volume and decay level also strongly changes with battery voltage. With full batteries it can be unpleasantly loud, and sounds decay slower and some less low (i.e. louder ADSR "sustain" phase). The "attack" and "release" phase of ADSR are not adjustable. So this should be an analogue parameter (like a channel volume control).

The "race car" volume decays to zero when holding the button (car drives away). Also "helicopter" gets quieter and louder again in a loop when holding its button. Also the "bee" grows quieter and louder with button held.

I have finally uploaded 5 videos about usage, synth mode and waveforms of the Sound FX Phasor.

https://www.file-upload.net/download-15071085/SoundFXPhasor_videos1.7z.html

Caution: This archive is 386.63MB huge (over half of a CD) and the filehoster may be quite slow and will delete it after a while. So only download this if you need to and save it.

The videos are uncut without further postprocessing. To conserve memory I already record in lowest video mode, but filming a scope screen during measurement and experimentation tends to become lengthy. The audio quality from the internal mic of my digicam (Samsung WB210) is not great, but should be sufficient to estimate envelope behaviour and how the synth mode works.

Thanks for the videos, it confirms the sound in MAME is not good.
Maybe the best solution in this case is netlist audio, but I've no experience with that.

*edit* if you have the time, can you capture pin 13 on the oscilloscope? With the default synth sound, and several command C parameters.

It's not rocket surgery, you'd pretty much just need a schematic and the luck to have all of the passives already handled by the netlist system. If you need a quick run-down on it, feel free to pop a friend request my way on Discord.

Doesn't the debugger give you pin13 digital behaviour anyway? (12-bit PIC is fully emulated - isn't it?)

I had recorded the pins in audio mode with a logic analyzer (Zaleae Logic 16) in PulseView (Linux Mint) after I dumped the rom. The signals are quite interesting to see how the phasing effects work (likely canceling out 2 waveform pins to produce the pauses of the square LFO, or does it saturate the transistor by shifting its operating point??).

Here is the 8.5MB archive of it. The .sr files can be opened with PulseView:
https://www.file-upload.net/download-15071500/SoundFXPhasor_soundpins.7z.html

The programmer of Mattel's Hockey posted an interesting note to Hacker News: https://news.ycombinator.com/item?id=34272079

I added netlist sound to sfxphasor, it does sound better, but some of the resistor/cap values were hard to see on the pcb photo.

Is R12 brown,black,red or brown,black,orange?
Is R7 red,violet,orange, or red,violet,brown (orange makes it sound bad)
What are the colors on C6? brown,black,orange? Or is that backwards?
Is C5 100nf like C3?
Are the diodes 1N4002? I don't think this one matters much.

netlist is here: https://github.com/mamedev/mame/blob/master/src/mame/handheld/nl_sfxphasor.cpp

There's another handheld that needs netlist sound, Coleco Galaxian. It has a sound effect when an enemy dives down. Kevtris made photos of the PCB but one capacitor is hiding in the photo, and some resistor color bands are hard to see.
It also has a weird "PUT" transistor that I don't think MAME supports.

Nice work on the netlist sound. Told you it was straightforward.

PUT is a so-called “programmable unijunction transistor”. It isn’t really a unijunction transistor at all, it’s more like a kind of thyristor (NPNP construction). I believe you’re correct, MAME lacks a suitable device right now.