New PCJR Speech board design, need Input !

[Chuckphd53]

Guys,
I am getting close to making time to designing the Speech board I mentioned a while back.
I’ve been giving it some thought and would like some input from the Jr experts we have on this board.

here is a list of items I am considering: and would value some input.

-I have two voice chips to choose from, One is the 2004Winbond chip that is an extreme High end voice output, I can program it to be Female Or Male voice but not both. the feedback we get from our RailYard customers that we do controllers for, is they like the Female voice. This is probably because the guys like to hear women talking to them (grin).

- the next chip I have is the Epson speech chip that has multiple voices both male and female, I beleive there are 7 software selectable voices. You can also switch to two types of spanish voices. This chip has a lower voice conversion and to be blunt, sounds similiar to the voice Steven Hawkings preferred to use.

- Both chips have the ability to adjust the volume in software, so the user can adjust it as needed.

- I am using the Modem slot for the board.
- I will am thinking of making the IO address jumper selectable.
- On board will be a small 8ohm speaker that you should be able to hear thru the front case grills slots.
- there will be an RCA jack that allows you to bring out the Speaker or a standard audio level output that goes to your audio IN amplifier. Jumper selectable.
- on board there will be an trim pot to adjust the main volume out for the speaker, this does not interfere with the software setting of the volume, it just sets the
baseline max volume to the speaker. Considering a jumper to allow the user to add and external volume control, by just deselect the trim pot, via jumpers and
add external POT with some exteneder wires that will plug into a 3 pin header.
- there will be an AT type command set to allow user to adjust the chip voices/language(epsonchip) / and volume(both chips)
- maybe an on board led that turns on when the voice is executing, with jumper to bring the Led to front panel for user viewing.
- simple use, same as sending strings to a modem port, which is typical ‘text to voice’ conversion.
- I have the option to include some controllable I/O maybe 8 inputs, and 8 outputs for user use. These can be optically isolated to control whatever.
- is there any need for any A/D channel I/O .
- Also consider LED activity for any optional items we are talking about.
- I can even add an ethernet port here, my son is the expert on this implementation, but it would use standard ether to serial type communication.
( I will review this more with him)

This will use a standard Arm processor as the main brain here.

I have thought about using the Jr to control my train setup ( when I get time for that again), but it can easily be interface to any type of control/monitoring system via the
I/O I spoke of.

Using the Jr for a control system is overkill of course! But having the voice now gives a wide range of use for the system again, I you are willing to program your favorite activity into this system..... brush up on your Turbo Pascal...

Anyway I have been working this in my head for a couple months now and wanted to give the Jr experts a chance to’voice’ their opinion. (pun intended)

Chuck

[alanh]

- I am using the Modem slot for the board.
- I will am thinking of making the IO address jumper selectable.

The modem slot has fixed address decoding. The only addresses available on it are 0x3f8 -> 0x3ff

[Chuckphd53]

good point, but what about the A9 line, it is usable, to give address 1F8-1FF... But I have not looked for another device using this address yet...
thanks Alan noted!

[alanh]

good point, but what about the A9 line, it is usable, to give address 1F8-1FF... But I have not looked for another device using this address yet...
thanks Alan noted!

It's not that simple. The modem/floppy CS line only goes active for addresses 0x0f0 -> 0x0ff and 0x378 -> 0x37f. The floppy controller is active on the lower range. Leaving only the upper as intended by IBM engineers.

[Chuckphd53]

Ah , ok thanks,

a side note, has anyone documented the decoders on the main board, stuck a logic analyzer and took notes?

hard to think it hasn’t been done

[alanh]

Most of the logic decode is done through two ROMs.

ZM52:

bit 0 - along with ZA3 for /b845
-----
I/O 0x3d0
I/O 0x3d8

bit 1 - /NMI CS
-----
I/O 0x0a0

bit 2 - /TIMER CS
-----
I/O 0x040

bit 3 - /SERIAL, SOUND CS
-----
I/O 0x0c0
I/O 0x2f8

bit 4 - /GAME, INTR CS
-----
I/O 0x020
I/O 0x200

bit 5 - /8255 CS
-----
I/O 0x060

bit 6 - /MODEM, FLOPPY CS (also feeds /CARD SLCTD)
-----
I/O 0x0f0
I/O 0x0f8
I/O 0x3f8

bit 7 - /I/O ON CARD (also feeds /CARD SLCTD)
-----
I/O 0x020
I/O 0x040
I/O 0x060
I/O 0x0a0
I/O 0x0c0
I/O 0x200
I/O 0x2f8
I/O 0x3d0
I/O 0x3d8




ZM56:

bit 0 - /VIDEO RAM CS (gate enable for video subsystem back to CPU)
-----
MEM 0x00000
MEM 0x08000
MEM 0x10000
MEM 0x18000
MEM 0xb8000

bit 1 - Mux control - CPU PG0-2 to CPU_A14-16
-----
MEM 0xb8000

bit 2 - /ROM_SEL
-----
MEM 0xd0000
MEM 0xd8000
MEM 0xe0000
MEM 0xe8000
MEM 0xf0000
MEM 0xf8000

bit 3 - /MEM ON CARD
-----
MEM 0x00000
MEM 0x08000
MEM 0x10000
MEM 0x18000
MEM 0xb8000
MEM 0xd0000
MEM 0xd8000
MEM 0xe0000
MEM 0xe8000
MEM 0xf0000
MEM 0xf8000

[Chuckphd53]

Great stuff Alan...
Looks like you get little sleep also..
Edit:——

well after 10+ years of abscence from thr jr, I had to tear back thru the docs, and it was baffeling that the Disk controller
used the Floppy CS to work on the Floppy controller board. The logic just did not work, and I hate the ‘mix’ of old logic symbols with later versions ugh...
It was not until I went back and read the disk controller chapter that it talks about the Diskette CS going LOW, THIS made sense and fit the proper logic
operation. In the rush to get the HW Docs out the door, I guess attention to detail was low priority. So now considering the FLOPPY CS is really /FLOPPY CS.
I see where the A9 line now plays a significant roll in the decoding....

thanks to Alan for the Decoder Map...
Chuck

[Brutman]

I'd consider finding a trace or a chip pin that has the audio signal from the bus. You can then use a micro-hook type wire to make a non-permanent attachment, allowing the speech go be mixed by the internal audio electronics, which will allow it to play through the audio port on the back or the monitor.

[alanh]

Audio input mix is available on the side-car connector

[Brutman]

I'm thinking that it has to be available closer than the sidecar bus .. that's a long way for a jumper wire.

(The optimal solution is left as an exercise to the interested student ...)