Voice Module
(Sound Recorder/Playback Module)





Introduction

Functional Description

Circuit Design and Components

Board Layout

Final Assembly




Introduction

Flipping through Jameco's catalog, I came across the Single-Chip Voice Record/Playback Device and decided Conga needed to talk back. Looking into the device, I realized that I could interface this device to a CPLD and come up with a simple design to help me quickly get started on Conga. I was also eager to start playing with my schematic capture tool and start learning how to lay out boards.

The design was pretty much there; all I had to do was find the parts, put together the schematic in OrCAD Capture, and then lay out my first board. Most of the time was spent on learning OrCAD Layout, as I had no prior experience with it. I don't claim to have learned the tool in the four months time (an hour each night) but I learned enough to get started. Now, let's see how Conga gets her voice....


Functional Description

The Sound Recorder/Playback Module structure simply looks like this:



16 Button Switches allow recording or playback of 16 different sounds. Alternatively, there is an External Input Connector which an external module (such as the processor board) can also access the 16 sounds.

The Mode Switch has two positions. When it is in the REC position, pressing a Button Switch will turn on the microphone and the controller will record until the Button Switch is released. When the Mode Switch is in the PLAY position, pressing the same Button Switch will play back what was previously recorded.

Eventually, this module will be installed within Conga, and the External Input Connector will allow an external processor board to control the Sound Recorder Playback Module. When the processor board senses a touch, it will play the giggling stream; or when it senses that Conga is hungry, it will signal the module to play the "I'm hungry" stream.


Circuit Design and Components

The schematic looks like this:


(Click on the schematic to download a .PDF version of the schematic. You will need Acrobat Reader to view the file.)

ISD25120 - SINGLE-CHIP VOICE RECORD/PLAYBACK DEVICE (U2)
Winbond Electronics' ISD2560 is the main device used to record and play back Conga's voice. Take a look at the datasheets for a more detailed description of this device.

There are four variations of this device: 60-, 70-, 90-, and the 120-second version. They all come in the same packages, but the sampling rate differs. The ISD2560, capable of holding a 60-second message, has a sampling rate of 8KHz. The ISD25120, capable of holding a 120-second message, only has a sampling rate of 4KHz. The human voice ranges from 100Hz to over 7000Hz. Since Gorillas have a deep voice, the 4KHz device should be sufficient. If not, the ISD2560 is pin-for-pin compatible with the ISD25120; thus, I can swap out the device if I need to.

ISPLSI1016E - COMPLEX PROGRAMMABLE LOGIC DEVICE (U3)
OK...A little overkill for what I'm using it for, but it's what I was able to quickly dig up. This CPLD multiplexes each Switch Button, or equivalently each single input from the External Input Connector (J1), to four address lines on the ISD25120. The CPLD also sequences the ISD25120's control signals, /PD and /CE.

If you're curious to see what the code looks like, feel free to take a peek of the code to the left... (Any text editor should be able to read this ABEL file.)



Board Layout

To the left, you'll see the layout design from OrCAD Layout (click on the image to enlarge). Fat traces were used for the power traces. Only two layers were required, thus making the bare PCB cheaper.



To the right, you'll see the board after if came back from the board house. Pretty, ain't it?



Final Assembly

After installing all the parts, I hooked the assembled unit to my trusty power supply, and vwalla! It works!!!

OK...not exactly. You could hear a click when a button was pressed, but nothing could be heard after supposedly recording a sound.

Turns out there was one minor error was discovered on the schematic level: I've modified the schematic in the Circuit Design and Components section above Thus, I had to make room for the capacitor by moving some resistors. The picture and schematic piece below shows the rework.