• Please review our updated Terms and Rules here

VOTRAX Type-'N-Talk

I know this is a major necropost but the proper way to get a votrax type n talk working is quite simple: first of all be sure you have the right power supply: a votrax personal speech system uses a power supply quite different to the one used on the votrax type n talk, despite using the same din-5 connector!
Secondly, the baud rate needs to be set properly; the type n talk uses a rather hacky way to do this: there are 8 dipswitches, and each one connects the clock input of an m6850 ACIA to a different clock rate divdied by a 4040 divider. You must have only ONE dip switch set or it will NOT work! (see the schematic at http://www.kevtris.org/Projects/votraxtnt/tnt_schematic.jpg)
Most people will probably want to set just dipswitch 8, for 9600bps.
The serial port, as far as I recall, is set to be a 'DTE' device and so requires use of a null modem cable to connect to a computer (which is another 'DTE' device). Flow control uses RTS and CTS lines. serial words are 8 bits long with no parity, and one stop bit, I believe. so the standard (with dipswitch 8 set) is 9600,N,8,1

As for data sent to the device, it is an unlimited text to speech engine using the "NRL" text-to-phoneme algorithm and a small dictionary of override words (to replace a few common ones which NRL doesn't pronounce properly).
You can also make it use votrax phonemes by starting a phoneme string with a ~, using ascii characters whose lower 6 bits control the phoneme spoken, and ending with a ?. The 64 votrax phonemes are documented here: http://www.sscnet.ucla.edu/geog/gessler/borland/votrax-phoneme-chart.jpg

The user manual for the type n' talk can be found here, in html-ized form:
http://members.tripod.com/werdav/txtospm1.html

LN
 
Last edited:
Glad I'm not the one digging this up from the before time, but I am looking for a modern way to reproduce the sound from either one of these or the earlier SC-01-A. Preferably something that will run on mac. I've been tracking down alternative text to speech voices, but each one seems to be trying to sound more human than the last. I need the authentic retro sounds from the 80s. Any help is much appreciated.
 
Lallander,

Does it have to be a reproduction, or would you be satisfied with the real thing?

On the ground in an abandoned pile of junk at the end of the (rainy) Timmonium MD hamfest this spring I found two Micro Mint parallel "sweet talkers" that use SC-01-A chips. Are you up to the challenge of real hardware hacking?

Lou

P.S. Did you ever get the 11/23 system going?
 
The 11/23 needs some internal work that I am not quite up to. I am no hardware hacker as you put it. I need a way to record the files onto a computer (mac in my case). The easier this process the more likely I am to use it. I would love to own one of the chips, I can never pass up retro computing stuff. It doesn't have to be that specific chip, I just need a classic 80s robot sound. I found two possible leads on this. There was a Java based emulator that was on a now defunct site, and apparently there was a SC-01 emulator as part of a pinball emulator. Stripping out only the Votrax bits is going to be a royal pain tho.
 
Ok, I've studied the chip's user manual a bit, and checked out a demo of one a guy has running on his website. I've convinced myself I could get this running with the aide of my Arduino. I would need some help making sure I am supplying the right amount of power to things, resisters and all that, but the actual programming of it looks rather simple. This is not the ideal of course, but a project like this could be fun, and give me a reason to dust off the old Arduino.
 
Lallander,

Ok then. Send me a PM with an address with where to mail these to. No gaurantees, so invest as little effort as possible with them until you get signs of life. (Don't go designing a whole project around them yet.) Data sheets are around on the internet. I know I printed some out once.

Lou
 
Last edited:
Alrighty then, once I get the chips that are kindly being donated, I will start a new thread specifically for it. I have done some work with the Arduino so I'm not a complete noob there, but I don't have much experience with electrical engineering. I will need people to help look over wiring arrangements to help me make sure I don't fry anything. I have the Arduino Diecimila and a prototyping shield for it. This is a good thing considering I do not know how to solder. :eek:
 
I finally just got a USB-Serial adapter in the mail and am trying to get my Votrax to connect. So far I've tried it on both the PC and the Mac with no success. It looks like I'm getting a connection but sending keystrokes has no effect.

I had the serial adapter plugged directly into the Votrax first, but I also have a breakout box that we can try.

The terminal application is set to 96008N1, and the Votrax is set to 9600 per the user's manual.

I'm attaching a picture because I have no idea how to use this box - can someone help?
 

Attachments

  • IMG_0597.jpg
    IMG_0597.jpg
    92.3 KB · Views: 1
Lallander said:
It doesn't have to be that specific chip, I just need a classic 80s robot sound.
Click to expand...
Have you considered vintage software speech synthesis? I am mainly thinking about programs like S.A.M, the Software Automatic Mouth for Commodore 64, Apple II and Atari 800. It does a pretty good job at speech synthesis, although it still doesn't sound very much like a real human. I am thinking you could obtain a copy of this program, enter some commands and make it say whatever you like. Then you record these sentences.

Hey, it may even be possible to do in emulation if you don't have access to the real thing. Please forgive me if I misunderstood your intention, how or for what you were going to use the synthetic sounds. I assumed you will use recordings and mix into something else.
 
The end goal for me is to have a simple way to recreate as authentically as possible the old robot sound. I then want to be able to save it to wav. On the mac I have the built in say command that I can output directly to a file. I doubt I will be able to find a mac speech compatible module, but that is the ideal. Whatever I can get to work is great, but the easier it is from the user standpoint the better. The hardware chip is not very likely to do what I need. It will however be a fun project, and might even become the basis for a prop.

I don't know how many people here are familiar with the tabletop RPG Paranoia. It is set in a dystopian underground society controlled by an insane AI. I am wanting to use this as His voice. The hardware version could very likely be used for local in person games. What a prop that would be.
 
I know this isn't vintage nor the hardware approach but have you just tried the native speech engine in XP since I see a few posters were using XP? http://support.microsoft.com/kb/306902 Get a copy of VB and embed that object and you can make a quick little sayit program (or add it to a picture of Michael Jackson and jack the pitch up a bit as my friend quickly did to my quick test app).
 
Three problems.
1. Not vintage. (this is the biggest)
2. Mac has a superior speech program.
3. It would require using windows. :crazy:
 
You may find an emulator a relatively way to go ahead, in particular if it lets you save the sound output directly to WAV. At least WinVICE (probably the Mac OSX port as well) has an option to record sound to WAV. Of course if you're dedicated you could strip selected portions of the emulator source code (e.g. CPU and SID emulation) and try to build a smaller program that runs the software speech synthesis. Or well, you may even be able to extract the phoenemes from S.A.M. and make your completely custom program that will work like the speech synthesis on modern computers, just using older algorithms and sound samples. This last task may be quite tricky, but I would assume interfacing a hardware chip is no easy task either, at least if you want to control it from the Mac and record the output.
 
I've found a few leads. I'm still looking into the hardware side of things, and will definitely try to do something semi related once I get the speech chips. On with the leads.

http://pinmame.retrogames.com/
This has a SC-01 emulator built in, the trouble is separating it out and getting it working cross platform.

http://www.nntpnews.net/f4625/no-one-can-escape-black-hol-umm-java-applet-7133495/
This site supposedly linked to a java emulator at one point, but it seems dead now.

http://www.youtube.com/watch?v=GTMow2Sn67A
This guy's video as well as his site linked in the info box is interesting. It has a few limitations that will keep me from going this route. It has a massive delay before it streams you the results, it can't parse text, and you can't modulate the voice at all.
 
Alrighty then. I got the chips in today. Where would be the best place for an Arduino project thread? Vintage hardware since I am trying to integrate some old ICs?
 
As a small side note... doesn't the Votrax Type & Talk allow one immense flexibility over the overall sound of the voice? (I heard that one Votrax synth would even let you speak using phonemes as opposed to just reading words, also said synth is capable of letting you inflect words to varying degrees as it is speaking them. (Apparently with good enough programming, one can make it sound quite exactly like a human to the casual listener.))

That's just my two cents.
 
While the sc-01-a on the type and talk CAN do 4 pitches, the inflection lines are not actually hooked up to anything. see Kevin Horton's schematic of it at http://www.kevtris.org/Projects/votraxtnt/index.html

The later Votrax PSS (personal speech system) is much more dynamic and allows very fine 16 step inflection control, etc by varying the sc-01's RC clock using a chopper and handling phoneme lengths via software control (i.e. ignoring the A/R pin).

LN
 
You must log in or register to reply here.
Back
Top