Pinouts for TI speech ICs

In general, you can't. TI uses the "TMC" and "CD" prefixes for custom chips, for which documentation is not publicly available.

Could be. I finally got through to Texas Instruments and got a reply where they told me they couldn't be of any help as they no longer had any access to information regarding those (discontinued) products. But I've picked up a little information here and there, enough to know that the TMC-0281 is a speech-synthesis chip. One of the two other ICs is indeed a ROM chip, though I'm not sure which one.

Any ideas where I can go to get information and pinouts?

Even TI can't help me. Are pinouts/information for customs like this kept "secret", or is it that TI simply didn't put any effort into publicizing it? If the latter is the case I'm sure someone somewhere must have a datasheet, or some sort of information, but where?

what is the package and number of pins of the TMC-0281? do you think it could seem to chip speech processor 28 pins as the SP0256 ?

On 29 Jun 2006 13:25:31 -0700, "Arild P." put finger to keyboard and composed:

The 23xx number looks ROM-ish. For example, the 2316, 2332, and 2364 were Commodore 24-pin, 2K x 8, 4K x 8, and 8K x 8 ROMs. Let me know if you need the pinouts.

- Franc Zabkar

--
Please remove one 'i' from my address when replying by email.

On 29 Jun 2006 13:25:31 -0700, "Arild P." put finger to keyboard and composed:

Try to locate the data and address buses. These should be relatively obvious. Then look for a crystal, a reset pin, supply rail, ground. With any luck the pinouts will be similar to a standard product.

FWIW, searching for 8011 at datasheetarchive brings up numbers such as "7801102RA" whose datasheet is that of a 20-pin 74373 or 74374 octal D-type latch/flip-flop.

- Franc Zabkar

--
Please remove one 'i' from my address when replying by email.

On 29 Jun 2006 13:25:31 -0700, "Arild P." put finger to keyboard and composed:

AFAICT the CD2354 is a TMC0350 256kbit ROM (Voice Synthesis Memory, VSM) that was used in "Speak & Read" and "A Dog on a Log". ???

- Franc Zabkar

--
Please remove one 'i' from my address when replying by email.

Decades ago, there was an article in Byte magazine about the internals. Note it was the second article, because there had been one where the author merely speculated on what he thought was inside. The second article gave more insight, and if I recall he did discuss the type of speech synthesis the Speak and Spell used. I don't recall dates, or where you might find the articles now.

Michael

Hello Franc : I am in need to identify some ROM chips + pinouts etc that are on microcomputer cartridges , P/n are : CD2155 CD2156NL CD2157 (those three previous p/n are in the micromputer unit (inside)) CD2215A CD2204A CD2293 etc in the cartridges those little memories are dip 2x9pins and the microcomputer unit is a Ti-99/4a. thank you for your help, @+ Matthieu

(snip)

Try the TI-99/4A home computer page

There's a discussion forum there as well, so you should be able to get in touch with people using the same systems.

I wrote:

Yes, the information was considered to be a trade secret.

No, it's a TI design, predating the SP0256. It was custom-designed specifically for the Speak-n-Spell, though it is probably similar to some of their later, more general chips.

Franc Zabkar wrote:

No, the TI CDxxxx numbers are for custom parts that are not generally related to any standard parts like the 23xx ROMs.

As it happens, the two parts in question probably are both ROMs, but they have a very non-standard interface.

On Fri, 30 Jun 2006 06:34:19 +0200, Matthieu Benoit put finger to keyboard and composed:

I think Eric is correct in saying that the correlation between TI CD23xx ROM numbers and Commodore 23xx numbers is purely coincidence. After further research I've discovered that "CD" means "Custom Data".

See

The abovementioned website lists many ROM chips which are marked with the base part number (eg TMC0350) as well as the specific ROM code (eg

23xx).

See

In the OP's case, the CD2354 appears to be a TMC0350 ROM chip programmed with ROM code 4 or 54. I suspect you have the same scenario although your base part number would be different (the TMC0350 is a

28-pin IC).

- Franc Zabkar

--
Please remove one 'i' from my address when replying by email.

On 25 Jun 2006 14:21:47 -0700, "Arild P." put finger to keyboard and composed:

AFAICT, the CD8011A must be a CPU, possibly similar to a TMS1000, with a custom ROM code (11?).

- Franc Zabkar

--
Please remove one 'i' from my address when replying by email.

Could be. Since the TMC-0281 is a speech-synthesis chip, the other two are probably a CPU and a ROM chip.

Looking at the following website which lists many (but not all) chips used in the TI speech products:

.... I gather that the TMC-0351 is the ROM chip used in the regular speak & Spell. I also have this confirmed from a (very badly scanned) schematic of the Speak & Spell I was overwhelmed to find the other day. In that same category at the above website I find a chip called the CD-2353 and CD-2355, but no mention of the CD-2354. Still, my guess is that the CD-2354 is the ROM chip. This leaves the CD-8011 which there also is no mention of, but could be the CPU as you suggest.

Just guessing, but estimated guesses nonetheless ;-)

.... but getting back to the TMC-0281 (speech synthesis chip). I believe this is the one I need more information about as I'm trying to create an audio line-out from the Speak & Spell compact (which uses a piezo-speaker). According to the schematic I found on some forum (since that forum was closed to non-members I've uploaded it here so you can follow what I'm talking about:

pins 11 and 12 handle the audio output. But there's some inconsistency as the quality doesn't improve even after cutting the traces to the rest of the circuitry and feeding pins 11 and 12 directly into my mixer/amp.

Since the regular Speak & Spell uses the same speech chip I thought I'd see what would happen if I did the same thing with that toy, so I cut its traces from pins 11 and 12 and fed the audio directly into my mixer/amplifier. Same sound as I've always heard it (and so much cleaner and better than the Speak & Spell Compact). Very strange. Could it be that the "shaping" (filtering/adapting etc. for the piezo-speaker's limited frequency characteristics) could be done

*before* the output stage? That would explain things.

Another thing... I just had a look at both circuit boards and I have to correct something I said above. The TMC-0281 are *seemingly* the same in both toys, but when I look closely I find the following markings on the ICs:

First, the "Speak & Spell":

TMC0281NL WA 8114 Philippines

Then the "Speak & Spell Compact":

TMC0281D-NL ME 8149 Philippines

I assume that the numbers/lettering in the second line are just production numbers, but in the first line there's a "D" suffix in the Speak & Spell Compact's chip. Could this denote a different version of the chip, or just a production run or something?

If the chips were socketed I could simply exchange them and see if there would be any difference in audio, but I'll have to de-solder them (28 pins) if it should come to that, so I'd like to hear what you all have to say about it first.

On 30 Jun 2006 05:13:36 -0700, "Arild P." put finger to keyboard and composed:

There is a "CD2534" but I suspect that is a typo, and that it should really be "CD2354".

- Franc Zabkar

--
Please remove one 'i' from my address when replying by email.