The Smith-Kettlewell Technical File

A Quarterly Publication of
The Smith-Kettlewell Eye Research Institute’s
Rehabilitation Engineering Research Center

William Gerrey, Editor

Issue: [current-page:title]

Original support provided by:
The Smith-Kettlewell Eye Research Institute
and the National Institute on Disability and Rehabilitation Research

Note: This archive is provided as a historical resource. Details regarding products, suppliers, and other contact information are original and may be outdated.

Questions about this archive can be sent to


The "Ram-Talker": A Practical Speech Digitizer and Recorder

Modifying the TDI Electronic Carpenter's Level

"The Eggs Are Ready," or The Alexy Appliance-Alert System


By Susan Fowle, Tom Fowle, Al Alden, Bill Gerrey and Jay Williams


A method of digitally recording speech in computer memory, then playing it back, is described. This system demonstrates the principle of "delta modulation," as well as concepts involved in controlling and synchronizing access to digital memories. Projections as to how this circuit could become a 16-word addressable speech system are advanced here (in a section called "Projections").

[The editor is delighted with how easy it is to adapt this very circuit for a 16-word voice output system. More papers on this subject will be given, so if the details are not clear now, they will be. What this does for individual builders (without fancy microprocessor development systems) is give us a replacement for the TSI Mini Speech Board. Besides being multilingual, this frees us from worry about companies ceasing production of our favorite boards, and in so doing makes permanent the designs of talking instruments. I am ecstatic, ol' co-authors, and we haven't heard the last of this.]

We are all certainly aware of the wide variety of uses for digitally recorded human speech. We use "same" every day in our talking clocks, etc. As was written earlier ("Choosing Appropriate Speech Technology," SKTF, Spring 1985), digitizing speech can produce very high-quality results, but because of limited memory availability, we can usually have only a limited vocabulary. Nevertheless, this is ideal for applications where we want a single message or a limited group of messages to be repeated often, and where we would like to keep the technology simple and attain high-quality speech.

The system described here uses commonly available parts and easy-to-implement techniques of sampling to make a device which can record several seconds of speech (or any other audio-frequency information) in an inexpensive CMOS memory chip. The memory can be "battery-backed" so as to keep the message alive when the system is "powered down." The precise uses to which a device of this kind may be put are up to you; it is fun to play with, fun to build, and the authors sincerely hope that it will give you some ideas for projects of your own design.

[Along those project lines, the digital portion of this circuit was originally designed by Susan Fowle to record the pulsed infrared signals from the remote controller for a television and video recorder. This was done so as to allow the machines to be turned off via a timer whose output was an infrared control signal. (The alternative would have been to remove AC power with the timer, but this would result in losing user-programmed features of the machine.) Expanding the capacity for storage and retrieval of digital pulse trains so as to accommodate speech was a logical next step (having nothing to do with the original TV controller.)

Efficient Digitization of Speech

As was discussed in "Choosing Appropriate Speech Technology," the principle of digitizing an essentially analog signal, such as the output of a microphone, is as follows:

The voltage to be recorded is sampled at a rate which theory dictates should be at least twice the frequency of the highest desired frequency component of the source. Each "sample" is then turned into a binary number and recorded in memory. To play it back, the data from the memory is recovered ("read") at the same sampling rate, and turned back into an analog voltage whose value varies to replicate each recorded number. This analog voltage is put through a low-pass filter to remove high-frequency components inherent in the digital processing, then fed to an audio amplifier.

In the above discussion, it is probable that the digital number derived from the sample will be made up of several, usually at least eight, binary bits. Binary math says that an 8-bit binary number can contain the equivalent of the decimal numbers from zero to 255. This means that the digital number has a reproduction accuracy of 1/256th of the maximum excursion allowable. This is great for audio quality, but it means quick use of a lot of memory.

If, for example, we thought we wanted a high-frequency response to 20,000 Hz and we therefore chose a sampling rate of 40 kHz, we would find that we need 320,000 bits of memory per second. We could build a device of this sort, but at five bucks a memory chip, the project would take on a measure of extravagance (can you afford $25 a second?). Also, an 8-bit analog-to-digital converter of sufficient speed would cost at least another $50 again. We need another--perhaps less "hi-fi"--way to get usable results; for example, we can take a much less sophisticated sample as described below:

Delta Modulation

It turns out that we can continually compare the analog signal for each current sample to the signal at the previous sample, noting only whether the voltage has risen or fallen; we can then record just this information as a single binary bit. If comparison shows that the signal has risen, we record a binary one; if the signal has fallen, we record a binary zero.

This system is called "delta modulation" (coming from the common use of the Greek letter delta as an indicator of change). Our one binary bit is merely an indicator of the direction of change in the modulation. By using this method, we do not need to store 8-bit binary numbers for each sample. In fact, we can store eight of these "samples" in the same space needed to store just one digital representation of actual amplitude. Ah, that's better now.

It turns out that speech quality that is quite usable is attainable by taking our sample at a rate of, say, 10 or 12 kHz, thus giving us a high frequency cut-off of 5 or 6 kHz--more or less acceptable in communications systems.

We are using only one bit at a time; sampling at perhaps 12,000 times a second with a 64,000-bit memory, we get a little more than five seconds of recording time (not enough time for recording our latest symphonic composition--unless you are a fanatic minimalist--but plenty of time for saying, "Hello there."). The digital signal looks like a non-periodic rectangular wave form.

When this digital one-bit-at-a-time stream is retrieved from the memory, it is merely passed through an integrator which accumulates the "deltas" in the process, and gives a fair representation of our input waves. The signal is then passed through a low-pass filter with a cut-off frequency around our desired high-frequency component.

[For those of you who have heard of the Smith-Kettlewell "Talking Signs," you may be interested to know that this scheme is the one used in recording the messages in the infrared transmitters, except that in that system the speech data is impressed into erasable programmable read-only memories (EPROM's). EPROM's have the advantage of being permanent, but they have the disadvantage of requiring rather more fancy electronics in order to program ("record") them. Also, the memories available at the time of their design could store only sixteen kilobits, giving a recording time of around one second. In a future issue, we hope to describe the transmitter and receiver for the infrared Talking Signs system.]

Ram-Talker Circuitry

Power Supply

Most of the circuit runs off a regulated 5 volts; the output audio amplifier runs off the unregulated line--7.5 to 12 volts. In addition, the delta modulator integrators need an adjustable bias voltage (1/2 the 5-volt VCC, plus/minus 1/2 volt).

The negative terminal of the unregulated supply (probably a 9-volt battery) is grounded. The positive of this unregulated voltage goes to the input terminal of a Fairchild uA7805 5-volt regulator. This input terminal is bypassed to ground by 220uF in parallel with 0.1uF (negative of the electrolytic at ground). The "Common" terminal of the 7805 is grounded. The 7805 output is the regulated supply; it is bypassed to ground by 100uF in parallel with 0.1uF (negative of the electrolytic at ground).

[If the RAM supply is to be "battery-backed," a few changes are necessary. (The wire-wrap table at the end of this article does not reflect these changes. Instead of tying the RAM's pins 26 and 28 high, they must be dealt with as described here.) The Hitachi constraints on the RAM supply voltage, as well as the demand that other logic not present the RAM with input signals more than 0.5V above the positive supply, leads to the following:

An LM317 is arranged to supply 5.6V. The unregulated voltage goes to its "Input" terminal; this terminal is bypassed by 220uF in parallel with 0.1uF (negative of the electrolytic at ground). The "Output" pin of this regulator goes through 160 ohms, then through 510 ohms to ground; the junction of these resistors goes to the "Adjust" pin of the LM317. The "Output" pin goes to the anode of a 1N914 diode, with the cathode of this going to the plus 5V line of the project. This plus 5V line is bypassed by 0.1uF in parallel with 100uF (negative of the electrolytic at ground).

The "Output" of the regulator also goes to the anode of another 1N914 diode; the cathode of this diode goes to pin 28 of the Hitachi HM6264LP-15 RAM. Pin 28 of the RAM also goes to the cathode of a third 1N914; the anode of this diode goes to the positive of a 3V battery, the negative of this battery being grounded. When pin 26 of the RAM is brought low, it throws the chip into "standby mode"; thus, pin 26 goes to the plus 5V line of the rest of the logic.]

The adjustable bias circuit is constructed near the TL074 op-amp chip. The 5-volt line goes through 10K to one end of a 5K trim pot, with the other end of this pot going through 10K to ground. The arm of this pot is the adjustable bias point, and it is bypassed to ground by 10uF in parallel with 0.1uF (negative of the electrolytic at ground).

Delta Modulator Circuit

A TL074 quad op-amp chip has its pin 11 grounded, while pin 4 goes to plus 5V. All the non-inverting inputs go to the bias point; pins 3, 5, 10 and 12 are tied together and go to the arm of the 5K pot of this bias circuit.

A closed-circuit mini phone jack, wired as a "shorting jack," is used for the microphone; thus, the switch contact is jumpered to the sleeve. The sleeve of this jack is grounded. The tip contact goes through 0.1uF, then through 11K to an inverting input of an op-amp (pin 2) in a TL074 quad op-amp chip. A low-pass feedback network, 220K in parallel with 100pF, goes between pins 1 and 2 of the TL074. Pin 1, the output of the first stage, goes through 11K to pin 13, another inverting input. Similarly, between pins 14 and 13 is the parallel combination of 220K and 100pF.

An LM311 comparator has its pins 1 and 4 grounded. Pin 8 of the 311 goes to plus 5V. (Pin 1 is the emitter of the 311's output transistor, thus giving the option of referencing its output somewhere other than minus V. It is referenced to ground here.) Pin 7, the open-collector output of the 311, goes through a 4.7K pull-up resistor to plus 5V. Pin 3, the 311's inverting input, goes to the adjustable bias point. (Unlike op-amps, the LM311 comparator's inputs are the wrong-way-round; pin 2 is non-inverting, while pin 3 is inverting.)

Pin 2, the 311's non-inverting input, goes through 47K to the output of the mike preamplifier (pin 14 of the TL074). The output of the comparator, pin 7, goes to the "D" input, pin 9, of a CD4013 D dual flip-flop. The "clock" for this flip-flop is derived elsewhere in the digital Ram-Talker circuit to be described later. (For testing, however, the clock terminal, pin 11, can be fed with any signal--perhaps a variable one from 10kHz to 100kHz. However, before applying power, temporarily tie all unused inputs of the other half of the 4013 to ground.)

On the 4013, pin 14 goes to plus 5V; pin 7, along with "Set" and "Reset" pins (8 and 10) are grounded.

The "Q" output of this flip-flop, pin 13, goes through 68K to the inverting input (pin 9) of an op-amp (used as an integrator) in the TL074. Between pins 8 and 9 of the TL074 is the parallel combination of 0.047uF and 100K. The output of this integrator, pin 8, then goes to the free end of another 47K resistor off the non-inverting input of the comparator (pin 2 of the 311), thus summing the integrated signal with the microphone input.

The "Q" output of the flip-flop, pin 13 of the 4013, is the digital signal output of the delta modulation encoder, and goes to the serial input of a CD4034 shift register--used to create 8-bit bytes for the memory.

Calibration of the Delta Modulator

If the bias for the integrator is not just right (exactly at the midpoint of the swing of the flip-flop's output, also accounting for offset in the integrator op-amp), the output of the flip-flop will not go "one up, one down, one up ..." as desired for no input signal. Instead, the flip-flop could easily go "three up, two down," etc. An oscilloscope could easily display this effect with no signal applied (the input shorted to assure that there was no signal input).

As it happens, the effect of misadjustment is quite apparent with the input shorted. While listening to the modulator--with the demodulator--directly, a noisy tone will be heard which can be brought to 0Hz (tuned for "zero-beat") with the 5K pot near midposition. Once this is done, the 5K rheostat in the clock circuit should be scanned throughout its range to see if noise occurs at certain frequencies; if noise does occur, readjustment of the bias is warranted.

Delta Demodulator

A single-pole double-throw toggle switch is provided so that the demodulator can either listen to the recorded memory or to the digitized input directly. The arm of this switch goes through 68K to the inverting input (pin 6) of an integrator in the TL074. The feedback network between pins 6 and 7 of the TL074 consists of 0.047uF in parallel with 100K. The output of this integrator goes into the low-pass filter:

Pin 7 of the TL074 goes through two 22K resistors in series to the top of a 10K volume control; the bottom of this control is grounded. The junction of the fixed resistors goes through 0.047uF to ground. The volume control is shunted by another 0.047uF capacitor. The arm of the volume control goes through 0.1uF to pin 3 of an LM386 audio amplifier. To suppress oscillations in the amplifier, pin 3 also goes through 0.01uF to pin 2.

Pins 2 and 4 of the LM386 are grounded. Pin 6 goes to the unregulated positive supply line. To suppress oscillations in the amplifier, it will probably be necessary to bypass pin 6 close to the chip, perhaps with 100uF (negative at ground). Pin 7 of the 386 is bypassed to ground by 22uF (negative at ground). Between pins 5 and 4, located close to the chip, is 0.1uF. Pin 5 goes to the positive end of a 100uF capacitor, with the negative end of this going through the speaker to ground.

Digital Circuit Description

This circuit takes serial clock synchronized data (such as speech from a delta modulator) and inputs it to an HM6264LP-15 64-kilobit RAM. Data may then be retrieved serially--played back through the above low-pass filter and amplifier. This circuit description, intended to give a theoretical overview of the digital circuit, is for understanding only. It is recommended that the builder read this section for basic understanding, then build the device by following the wire-wrap table listed later.

The system clock is built around one of the inverters in a hex Schmitt trigger chip (CD40106). An input (pin 1) goes through 0.0047uF to ground; pin 1 also goes through a 1k resistor in series with a 5k rheostat to the inverter's output (pin 2). The approximate values for the system clock may be had by using the formula:

R equals 1 over the product of 1.7 times f times C, where "f" is twice the desired sampling frequency. (This clock runs at twice the sampling frequency so as to derive the necessary timing signals which transfer parallel data to and from the 6264 RAM between serial input samples of the 4034 shift register.)

The system clock goes to one input (pin 13) of a CD4073 triple 3-input AND gate. The second and third inputs of this gate are controlled by an RS flip-flop which directs the "start-stop" activities. The output of the AND gate, pin 10, goes to the clock input (pin 1) of a CD4520 dual 4-bit "up counter." The Q1 output of the CD4520 bears the clock signal at the desired rate of sampling (half the frequency of the "system clock").

The Q1 output of the 4520 (pin 3) goes to the clock input of the CD4034 (pin 15). It then goes through one of the inverters in the CD40106 to the clock input of the delta modulation circuit's flip-flop (the CD4013's pin 11). (This inversion is necessary because both the delta modulator and the shift register have clock terminals which trigger on the positive edge; if they were clocked by the same signal, we would send data into the shift register while the modulator was changing state. This would be a heck of a way to make a random-noise generator.

As mentioned earlier, the system clock signal is gated on and off by one of the gates in the CD4073. Controlling this gate is a "start-stop" flip-flop. (The needed RS flip-flop is made using a spare D flip-flop in the CD4013; grounding both the "D" and "Clock" inputs, the "Set" and "Reset" terminals can be used in the manner of an ordinary RS flip-flop.) Pins 11 and 12 of the 4073 AND gate go to the "Q" output of the CD4013 (pin 1).

The "Set," pin 6, of this start-stop flip-flop goes through a resistor to ground, as well as going through an SPST pushbutton "start switch" (S2) to plus 5V. This point is also connected to 4520 resets (pin 7 and 15), and to the reset of the CD4040 (pin 11). This assures that when the big "go" button is pushed, the cascaded counter--and thus the memory addressing--is started from zero. The "Reset" terminal of the RS flip-flop, pin 4, gets its "clear" signal from the 14th bit in the counter which addresses the RAM. (This 14th bit turns out to be pin 12 of the CD4520.)

The CD4040 12-bit counter and the second half of the CD4520 are connected in cascade to make a counter with enough bits to fully address the RAM. (In order to follow the 4040 with the 4520, the latter must be wired for negative-edge triggering.) The 4520 "Clock" (pin 9) is grounded and the "Enable," pin 10, is used as the clock; it goes to pin 1 (Q12) of the 4040.

This "address counter" is clocked from the inversion of Q4 off the first half of the 4520. Thus, the 4040's clock input, pin 10, goes to pin 4 of the 40106, with pin 3 of this Schmitt trigger going to pin 6 of the 4520.

The first half of the CD4520, in conjunction with a 5-input AND gate, marks the 8-bit strings of clock signals that define the "samples" for the memory and the shift register. In the "record" mode, the system sends a "Write-Enable" signal (which must be inverted) to the RAM. In the "play" mode, it sets the shift register between serial and parallel modes as necessary.

The needed 5-input AND gate is made using two 3-input ones in the CD4073. On this 4073, pin 6, the output of one gate, goes to pin 8, the input of a gate whose other inputs are pins 1 and 2. Thus, the output of the resultant gate is pin 9, with its five inputs being pins 1, 2, 3, 4, and 5. The first four of these inputs go to Q1 through Q4 on the 4520 (pins 3, 4, 5 and 6).

The final input of this gate, pin 5 of the 4073, gets one of two versions of the system clock via a switching arrangement which will be described later.

[For the technical masochist: During the record process, it is necessary to write data to the memory--from the shift register--after the eighth bit has been entered, but doing so before the next shift. The inverted system-clock signal (at twice the frequency of the sampling signal) generates a signal during the middle of the "sampling-frequency" clock. In the play mode, however, the system clock is not inverted so that the data accepted from the memory is not mixed with a bit of serial data which would otherwise be shifted into the 4034 as the new byte is being read.]

Pin 9 of the 4073, the output of the combined AND gate, is inverted; it goes to pin 5 of the 40106. The inverter's output, pin 6, bears the "Write-Enable" signal whose associated switching arrangement will be dealt with shortly.

A four-pole double-throw switch (S1) selects the "record" and "play" modes. (The four poles of this switch will be termed S1A, S1B, S1C and S1D.)

S1A determines the direction of data in the 4034 shift register--from the A to the B register in "play," and from B to A in "record." It also enables the tristate outputs of the RAM in the "play" position. To do this, the arm of S1A is grounded. The A/B pin of the CD4034, pin 11, goes through 12K to plus 5V; pin 11 also goes to the "record" side of S1A. Pin 22 of the 6264 RAM (the output "NOT-Enable") goes through 12K to plus 5V; pin 22 also goes to the "play" side of S1A.

S1B is associated with the "Write-Enable" pin of the RAM (pin 27). This pin 27 is taken through 12k to plus 5V; it also goes to the "record" contact of S1B. The "play" side of S1B has no connection. The arm of S1B goes to the output of Schmitt trigger inverter pin 6; the input of this inverter goes to the aforementioned 5-input AND gate with which the "Write-Enable" signal is derived.

The 4034 shift register is put in the serial synchronous mode by taking both its "Parallel/Serial" terminal (pin 13) and its "Asynchronous/Synchronous" terminal (pin 14) to ground: it is put in the parallel asynchronous mode when these pins are taken to plus 5V. Therefore, both pins 13 and 14 go to the arm of S1C. The "record" side of this switch is grounded. The "Play" contact goes to pin 9 of the 4073, the output of the 5-input AND gate. Thus, after eight clock pulses, new information is "read" from the memory in parallel format, then discharged serially.

S1D selects the appropriate system clock--either direct (in "play") or inverted (in "record"). Thus, the final input of the 5-input AND gate, pin 5 of the 4073, goes to the arm of S1D. The "play" side of this pole goes directly to the clock input of the 4520 (pin 1, which comes from pin 10 of the 4073 "starting" gate). The "record" side goes to pin 12 on the 40106 inverter chip, with pin 11 going to the 4520's pin 1.

On the 4034 shift register, the "A" bus lines, A1 through A8 (pins 16 through 23), are connected to the 6264's I/O pins (pins 11 through 13 and 15 through 19). Also on the 4034, the "A-Enable" (pin 9) is tied to plus 5V. The 6264 records when its "NOT-Chip-Select" (pin 20) is pulled to ground by the 4013 start-stop flip-flop's Q-NOT output (pin 2).

As mentioned above, the 6264 RAM is addressed by the combination of a 4040 and the second half of the 4520. The outputs of the 4040 are connected to the first twelve address lines of the RAM, while the 4520 Q1 (pin 11) is connected to the RAM's thirteenth address line (pin 2). (Rather than just duplicating their stated connections here, we refer you to the wire-wrap table that follows this section.)

Delta modulation is accepted on the 4034's serial data input (pin 10). Digital (undemodulated) playback information is taken from the 4034's B8 line (pin 1).


A detailed look at the above circuit suggests that, if you shortened the address counter (dropping four bits), you could then hand the four highest address lines of the RAM over to external logic. This would give you sixteen addressable speech segments--each almost half a second in the slow speed--which could speak any language you choose to record. A handshaking signal could be derived from the "tenth bit" in the shortened counter (thus alerting other circuitry that the word had finished). A handshaking signal already exists which can indicate to other logic that a word is being spoken; the Q output of the start-stop flip-flop (pin 1 of the 4013) will go high during speech, and the NOT-Q (pin 2) will go low during speech. External logic can start speech by pulsing the normally open contact of the start button to logic high. (Don't forget that there are two spare inverters available on the 40106.)

Of course, storing this speech in a nonvolatile PROM would be the logical next step in designing a speech board for talking instruments. While "burning" PROM's is usually done with the assistance of a computer, this project lends itself to mere copying of the RAM into the PROM (which, for now, must be done at a much lower data rate). Given strange programming voltages required for burning PROM's and EPROM's, this would be quite an electronics project all right, but it's absolutely feasible to build your own home recorder for speech in programmable read-only memories.

Ram-Talker Wire-Wrap Table

[Thank you, Susan Fowle, for the basic design and for this table. What a stormer!]

Note: This table only applies to the purely digital portion of the circuit, since wire-wrapping all but a few connections in the delta modulator is not possible (given that resistors and capacitors do not have square leads yet).

While each chip has its own mini table within this section, this "mini table" is not all-inclusive; i.e., if a string of connections has previously involved a pin of the chip being worked on, this connection will not be relisted in that chip's own mini table. That's a good thing, since if you just follow every step in due order, no checking need be done to see if some connection has already been made. This is being mentioned because the meddling editor has made exceptions to this rule--the switches. The switches will have their own tables that list all their connections. (However, so that you can kind of tell where they go as you read these tables for pleasure, they will be shown in parentheses within the other mini tables.)

We decided to be kind and explain how to read the following wrap table.

As an example:

  • 4520-1 to 4073-10; 40106-13; (S1D play)
  • 2,16 to plus 5V

should be read as "the 4520's pin 1 goes to the 4073's pin 10 and to the 40106's pin 13 and to S1D's play terminal. The 4520's pins 2 and 16 go to plus 5V." Good luck!

  • CD4520 Dual 4-bit "Up Counter":
    • 4520-1 to 4073-10; 40106-13; (S1D play)
    • 2,16 to plus 5V
    • 3 to 4073-1; 4034-15; 4013-11
    • 4 to 4073-2
    • 5 to 4073-3
    • 6 to 4073-4; 40106-3
    • 7,15 to 4040-11; 4013-6; 12K to ground; (S2 pushbutton to plus 5V)
    • 8,9 to ground
    • 11 to 6264-2
    • 12 to 4013-4
    • 13,14 to NC (no connection)

CD4034 Shift Register:

  • 4034-1 to Undemodulated Digital Output
  • 2 through 8 to NC
  • 9,24 to plus 5V
  • 10 to 4013-13
  • 11 to 12K to plus 5V; (S1A record)
  • 12 to ground
  • 13,14 to (S1C arm)
  • 16 to 6264-11
  • 17 to 6264-12
  • 18 to 6264-13
  • 19 to 6264-15
  • 20 to 6264-16
  • 21 to 6264-17
  • 22 to 6264-18
  • 23 to 6264-19

CD4040- 12-Bit Counter:

  • 4040-1 to 6264-23; 4520-10
  • 2 to 6264-5
  • 3 to 6264-6
  • 4 to 6264-4
  • 5 to 6264-7
  • 6 to 6264-8
  • 7 to 6264-9
  • 8 to Ground
  • 9 to 6264-10
  • 10 to 40106-4
  • 12 to 6264-25
  • 13 to 6264-3
  • 14 to 6264-24
  • 15 to 6264-21
  • 16 to Plus 5V

HM6264LP-15 RAM:

  • 6264-1 to NC
  • 14 to Ground 20 to 4013-2
  • 22 to 12K to plus 5V; (S1A play)
  • 26,28 to Plus 5V
  • 27 to 12K to plus 5V; (S1B record)

CD40106 Hex Inverting Schmitt Trigger:

  • 40106-1 to Through 0.0047uF to ground
  • 2 to 4073-13; Through 1K in series with 5K rheostat to 40106-1
  • 5 to 4073-9; (S1C play)
  • 6 to (S1B arm)
  • 7,9,11 to Ground
  • 8,10 to NC
  • 12 to (S1D record)
  • 14 to Plus 5V

CD4073 Triple 3-Input AND:

  • 4073-5 to (S1D arm)
  • 6 to 4073-8
  • 7 to Ground
  • 11,12 to 4013-1
  • 14 to Plus 5V

CD4013 Dual D Flip-Flop:

  • 4013-3,5,7,8,10 to Ground
  • 9 to LM311-7
  • 12 to NC

S1 Four-Pole Double-Throw Record/Play Switch:

  • S1A arm to Ground
  • S1A record to 4034-11; through 12K to plus 5V
  • S1A play to 6264-22; through 12K to plus 5V
  • S1B arm to 40106-6
  • S1B record to 6264-27; through 12K to plus 5V
  • S1B play to NC
  • S1C arm to 4034-13,14
  • S1C record to Ground
  • S1C play to 40106-5; 4073-9
  • S1D arm to 4073-5
  • S1D record to 40106-12
  • S1D play to 4520-1; 4073-10; 40106-13

Pin Identifications

Texas Instruments TL074 Quad FET Op-Amp:

  • 11--VCC
  • 4--VSS
  • 1--Output A1
  • 2--Inverting Input A1
  • 3--Non-Inverting Input A1
  • 7--Output A2
  • 6--Inverting Input A2
  • 5--Non-Inverting Input A2
  • 8--Output A3
  • 9--Inverting Input A3
  • 10--Non-Inverting Input A3
  • 14--Output A4
  • 13--Inverting Input A4
  • 12--Non-Inverting Input A4

National LM311 Comparator:

  • 4--Minus V
  • 8--VCC
  • 1--Minus V Reference for the Output Swing
  • 7--Output (open collector)
  • 2--Non-Inverting Input
  • 3--Inverting Input
  • 5 and 6--Offset Correction Pins (not used)

National LM386 Audio Power Amplifier:

  • 4--Ground
  • 6--VCC
  • 7--Preamp Bypassing
  • 2--Inverting Input
  • 3--Non-Inverting Input
  • 1 and 8--Gain Boost Pins (not used)

RCA CD4013 Dual D Flip-Flop:

  • 7--VSS
  • 14--VDD
  • 1--Q1
  • 2--NOT-Q1
  • 3--Clock1
  • 4--Reset1
  • 5--D1
  • 6--Set1
  • 13--Q2
  • 12--NOT-Q2
  • 11--Clock2
  • 10--Reset2
  • 9--D2
  • 8--Set2

RCA CD4040 12-Bit Counter:

  • 8--VSS
  • 16--VDD
  • 11--Reset
  • 10--Clock
  • 9--Q1
  • 7--Q2
  • 6--Q3
  • 5--Q4
  • 3--Q5
  • 2--Q6
  • 4--Q7
  • 13--Q8
  • 12--Q9
  • 14--Q10
  • 15--Q11
  • 1--Q12

>RCA CD4073 Triple 3-Input AND:

  • 7--VSS
  • 14--VDD
  • 1, 2, 8--In1
  • 9--Out1
  • 3, 4, 5--In2
  • 6--Out2
  • 13, 12, 11--In3
  • 10--Out3

RCA CD4034 Shift Register:

  • 12--VSS
  • 24--VDD
  • 16 through 23--A Bus, A1 through A8, respectively
  • 8 through 1--B Bus, B1 through B8, respectively
  • 9--A-Enable
  • 10--D Serial Input
  • 11--A/B (When high, A pins are inputs and B pins are outputs.)
  • 13--P/S (high for parallel.)
  • 14--A/S (high for asynchronous.)
  • 15--Clock

RCA CD4520 Dual 4-Bit Counter:

  • 8--VSS
  • 16--VDD

First Counter:

  • 1--Clock
  • 2--Enable
  • 3--Q1
  • 4--Q2
  • 5--Q3
  • 6--Q4
  • 7--Reset

Second Counter:

  • 9--Clock
  • 10--Enable
  • 11--Q1
  • 12--Q2
  • 13--Q3
  • 14--Q4
  • 15--Reset

RCA CD40106 Hex Inverting Schmitt Triggers:

  • 7--VSS
  • 14--VDD
  • 1--In1
  • 2--Out1
  • 3--In2
  • 4--Out2
  • 5--In3
  • 6--Out3
  • 9--In4
  • 8--Out4
  • 11--In5
  • 10--Out5
  • 13--In6
  • 12--Out6

Hitachi HM6264LP-15 8K by 8-Bit Static RAM:

  • 14--VSS
  • 28--VDD
  • 1--NC (no connection)
  • 10, 9, 8, 7, 6, 5, 4, 3, 25, 24, 21, 23, and 2--Address Lines, A1 through A13, respectively
  • 11, 12, 13, 15, 16, 17, 18, and 19--Input/Output Lines, 1 through 8, respectively
  • 20--NOT-Chip-Select No. 1
  • 26--Chip-Select No 2
  • 22--NOT-Output-Enable
  • 27--NOT-Write-Enable

Fairchild uA7805, National LM340T-5 5-Volt Regulator:

With the mounting surface toward you and with the leads pointing upward, the three leads are, from left to right: Input, Ground, Output.

National LM317T "Adjustable" Voltage Regulator:

With the mounting surface toward you and the leads pointing upward, the three leads are, from left to right: Adjust, Output, Input.



By Jay Williams


This paper describes the modification of a commercial electronic level. Called "Levelite," the device has five LED's which indicate when a desired degree of incline has been established. Besides horizontal and vertical planes, a rotatable pointer can be used to set the sensor for indications at 5-degree increments over a range of 180 degrees. As purchased, the Levelite instrument emits a beeping tone when the desired incline has been reached; however, this position is hard to find without feedback from the LED's, and our modification has been added to provide tones for feedback to the blind user.

[Smith-Kettlewell wishes to profoundly thank the company, Fourth Corner International, for their landmark cooperation in providing us the needed information. In particular, the open-minded sharing of technical data by the Levelite's engineer, Galen Biery, has been without parallel in the editor's experience, and the officer in charge of marketing the instrument, Doug Starcher, is himself on the level. Thank you.]

Introduction and Description

No matter how useful a device is, we may not appreciate its usefulness if a good one is not available. A good example is a user-friendly carpenter's level for the blind. Dangling plumb lines and rolling marbles may seem to work, but having an accurate instrument present you with unambiguous information makes you wonder how you got along without one. (Such products of recent vintage have either been insensitive or grossly underdamped, making the user exercise interpolation and wait a long time for a valid reading.) A significant find is the commercial Levelite instrument, which not only works well as a visual device but which is easily modified.

The Levelite instrument has two components: A long flat module contains the electronics, the sensor (which can be set in steps of 5 degrees), and the read-out system (five LED's and a beeper). The other component is a piece of aluminum channel into which this module is inserted. (The company offers four lengths of this channel: 2, 3, 4, and 6 feet. These "bars" are called "rails" by the factory, and more than one length can be gotten; see ordering information at the end of this article.)

The channel looks like an aluminum trough with straight sides, and measurements are made with this "trough" lying on one edge. The sensor module is fitted into the trough and forced back as far as it can go, so that it lies against the back of the trough. The module is held securely in place by two springy metal projections on its upper edge.

If, for example, you want to be sure that a horizontal surface is "level" (zero degrees), lay the bar on its side with the front panel of the sensor module facing you. Orient the module so that the sensor dial is at the right-hand end. (The sensor module can be installed anywhere along the channel.) The sensor, which has a pointer knob, must now be set to zero (to indicate levelness of the horizontal surface in our example). The zero-degree position will be found with the pointer pointing straight to the left. Pressing the "on" switch will bring the electronics to life. If the surface is level, a high-pitched beeping will be heard, and a green LED will be visible. If the surface is not level, no beeping will be heard, and one or both red LED's (off to one side of the green) will be visible.

As purchased, only the presence of the sensor dial is tactually apparent because everything else is hidden behind a translucent front-panel membrane. Beginning just beyond the tip of the pointer, there is a long slender window. Behind this window are the Levelite's five indicator LED's. The on-off switch is hard to find; it is located about a finger's width above the left end of the long window, and takes the form of a pushbutton which is actuated by pressing in on the membrane of the front panel. (If you rub the area with a fingernail or similar blunt instrument, you can hear the hollowness where the housing has been drilled out to accept the pushbutton underneath. To add to the confusion, a similar void under the membrane--below the long window and opposite the power switch--was made for a pilot light which does not exist in the latest Levelite design.) There is no "off" position; turning it on actuates a 2-minute timer which turns it off automatically.

Ironically, although you cannot see it, the sensor is an optical device. The heart of the sensor is the conventional "bubble level." Along the glass tube (and encapsulated where nothing can be seen) are sets of matched photo detectors, also provided with matched LED's that constitute their light source. The outputs of these matched sets are summed in an amplifier (one op-amp within an LM358 dual). (The second half of the LM358 is a follower which establishes a solid reference of 1/2 VCC--its input looking at the center of a resistor string.) The output of the summing op-amp is an analog signal which is decoded by four comparators (contained in an LM339 quad comparator); the free inputs of these comparators are attached to junctions in a string of resistors across the power supply. As you might expect, these four comparators' open-collector outputs directly drive (pull down on) the four red LED's--the ones we will be sensing with the electronics in our modifications.


In a sense, with patience, the Levelite instrument is usable as purchased. However, confusion frequently arises. The beep indication happens only at the condition of "levelness," a position which is extremely hard to find; you're frequently not sure if you aren't just missing this position, or if the instrument has "timed out"--turned itself off. Our additional circuitry tells you which way to move to attain levelness, and chirps a "good-bye" when the instrument turns off.

In our modification, an oscillator indicates the condition of the LED's either side of center. This oscillator generates four different pitches; one pair of pitches is distinctly lower than the other. For the LED's left of center, the low-pitched pair is used; a low tone tells you that only one LED is on, while a higher pitch tells you that both are on. Likewise, to the right of center, a high tone tells you that both LED's are lit, and a lower tone tells you that only one is on.

This oscillator system turns on and off simultaneously with the other circuitry in the module. Thus, you know when the "on" button has been pressed and the circuit is active. (We did, however, include a switch to defeat our oscillator in the event that the leveling job takes less than the two minutes provided by the Levelite's timer; turning it off can certainly reduce the "annoyance quotient.")

The housing for our modification is mounted on the left end of the sensor module. This added oscillator circuitry is contained in a standard project box. It is positioned so that its right apron guides your finger in finding the power switch.

Two other minor additions are recommended. Since the front-panel membrane rather successfully hides the "on" button, a dot of cement or Dymo tape should be installed to mark the spot; this spot can be easily located for marking when the board is out. Some way of marking the dial is also desirable; a single mark at "zero degrees" is sufficient, although a Braille dial could be generated for this. For other agencies (primarily the Washington State Commission for the Blind), the Levelite company has sold units without the membrane attached--sending the membrane separately. In a real production setting, this would be a boon, since the membrane material is ideal for holding Braille markings directly impressed on it.

Circuit Operation

In the Levelite circuit, a 14541 IC serves as a timer; its output operates the base of a 2N2222 whose collector goes to the positive of the battery, and whose emitter powers the device. Following this 2N2222, another 2222 regulates the main VCC line; its base goes to a zener diode, its emitter goes to the VCC line, and its collector goes to the emitter of the first transistor.

A series string of six resistors goes across this regulated supply; the junctions along this string (excluding the center point) go to appropriate inputs of four comparators which operate the red LED's.

The outputs of these comparators are open collector, and an LED is turned on when the output of its comparator is brought low. To produce the four pitches which represent the four off-center LED combinations, we use an NE555 timer chip as the audio oscillator.

The connection of the 555 is rather unique. In conventional arrangements, different pitches are gotten by pulling up on different charge resistors. However, because the open collector comparators in the sensor module pull down to actuate their LED's, some arrangement was needed to influence the frequency of the 555 by pulling down on resistors instead of up.

Basically, the 555 is wired so that its output drives pins 2 and 6 through a charging resistor. However, a diode between this charging resistor and the output makes the circuit monostable; i.e., for half of the cycle, the output pulls pins 2 and 6 high (charging the capacitor). The circuit is content to remain in this state until pins 2 and 6 are brought below 1/3 VCC (discharging the capacitor). The discharge phase is controlled by various combinations of four resistors at the junction of the diode and the resistor going to pins 2 and 6. The frequency of the oscillator is dependent on this discharge function, and increases as the value of pull-down resistance decreases.

For pull-down resistors, there are two 47K for one pair of off-center LED's, and two 130K for the other pair of off-center LED's. In the Levelite circuit, when an open-collector comparator output is brought low, its LED turns on and the oscillator emits a pitch. As the level is tipped more and more to one side, both off-center LED's become lit; this puts two pull-down resistors of the 555 in parallel. Hence, the pitch is higher than it is when only one LED is lit. Further, one pair of LED's employs 47K's, and the other, 130K's. The result is four different pitches which are not likely to be confused by the user. Because the oscillator is monostable, it is silent when the level is centered, since there is no resistor being pulled down by the comparators.

[Operation of the green LED and the beeper is not directed by a comparator. Rather, an input of a Schmitt trigger 2-input NAND gate (one of four gates in a 4093 quad) looks at a common pull-up resistor shared by the two innermost red LED's; when neither of these LED's are being actuated, pulsing of the green LED and operation of the beeper are taken care of by this 4093.]

We wanted to use the Levelite's battery to power our modification. Furthermore, we thought it would be better not to have a separate switch on the modification which could be left on to run the battery down. Therefore, we used the Levelite's VCC line to control a transistor switch in the positive lead of the 555.

Circuit for the Smith-Kettlewell Modification

The minus 9-volt battery lead is grounded to both the oscillator circuit and the Levelite circuit. (This battery is already contained in the housing of the Levelite.) For the modification circuit, the plus 9-volt lead of the battery goes to the emitter of a 2N4036 transistor. The collector of the 4036 goes to the VCC line of the oscillator. The 4036 base goes through 47K to its emitter; this base also goes through 1K to the collector of a 2N2222 whose emitter is grounded. The 2222's base goes through 47K to pin 3 of the comparator chip (National LM339), this pin 3 going to the VCC line in the level circuit.

Pin 1 of the 555 is grounded; pins 4 and 8 are tied together and go to our oscillator's VCC line (the collector of the 4036). This VCC is bypassed to ground by 100uF, positive lead at VCC.

Pins 2 and 6 are tied together and go through 0.01uF to ground. Pin 3 goes to the anode of a 1N914 diode whose cathode goes through 22K to pins 2 and 6. The junction of the resistor and diode goes to one end each of four resistors, two 47K and two 130K. Each of these four resistors goes through a diode to the appropriate comparator output in the Levelite. Our order for doing this is as follows:

Pin 13 of the LM339 in the Levelite goes to the cathode of a diode whose anode goes to one of the 47K resistors. Pin 1 goes to the cathode of a diode whose anode goes to the other 47K resistor. Pin 2 of the 339 goes to the cathode of a diode whose anode goes to one of the 130K resistors. Pin 14 goes to the cathode of a diode whose anode goes to the other 130K. (You may reverse the high and low tone pairs by associating pins 2 and 14 with the 47K resistors and 1 and 13 with the 130K resistors.)

The output, pin 3, goes through 47 ohms (1/2 watt) to the positive end of a 10uF electrolytic capacitor. The negative end of this capacitor goes to one side of the "silence" switch (SPST toggle); the other side of this switch goes through the speaker to ground.

Construction Tips

A piece of perforated board, 1.5 by 2.5 inches, will easily accommodate the switching transistors and oscillator circuit. To contain this circuit, we use a box 3 inches long, 2-1/8 inches wide, and 1-1/4 inches high (Radio Shack 270-230). The speaker, available from Mouser as 25SP016, has a square mounting flange with holes at its corners. From the square flange, which is 1-5/8 inches on a side, remove material from two adjacent corners so the speaker will fit between the two screw posts at one end of the box. Plan your board layout so that some unused board remains at one end; that way, the components will have room to protrude, and blank board will be sandwiched between the speaker magnet and the surface of the sensor module. (The lid of the box is not used, but it makes a fine template for drilling holes in the Levelite cabinet.)

Free motion of the board may be reduced by adding foam rubber or other insulating material under the component side of the board (right beside the speaker). I mounted the "silence" switch on the end farthest from the speaker--the right-hand apron near the power switch of the completed unit.

Disassembly of the Levelite

Four screws hold the bottom cover in place; four screws hold the circuit board in. These eight screws sport splined heads for which you probably haven't the right tool. Somewhere in your collection of jewelers' screwdrivers is one which will fit snugly. Remove the battery door and get your parts bin ready.

Remove the cover carefully! There is a large spring between the cover and the rotatable sensor. This sensor dial will now lift or fall out if left alone, and along with it will come a thin felt washer that lines the bottom of the hole in which its housing rides. It would be well not to remove this spring, but to stretch a piece of tape across the sensor assembly to hold it in place (anchoring the ends of this tape on the sides of the Levelite's cabinet).

On the solder side of the board is a large piece of cloth tape which protects the sensor wires as they move over this jagged surface. Remove this tape and save it for reinstallation.

Finally, remove the board. (Originally, four screws were used, one in each corner. However, in our latest acquisition, only two screws were used. Take note of which corners are secured so that reinstalling these screws will be easier.)

Mounting the Oscillator Box

Luck is with us here. The distance between the two screws at the end of both containers is identical. Drill completely through the two end posts next to the battery compartment on the Levelite module. Long screws will be used to fit through these posts and into the oscillator box. Next, use the aluminum cover of the oscillator box as a template to drill the remaining two holes (these being used to secure the right hand of the oscillator box). Thus, 1-1/2 inch long self-tapping screws are used to hold both boxes together at the battery end, and the original screws for the oscillator box are fed through from the inside of the Levelite cabinet to hold the other end. The aluminum cover plate that came with the Radio Shack box is not used, and can be discarded.

This scheme of mounting, although convenient, is slightly flawed in the following two ways. First of all, the positioning of the oscillator box enforced by making the screw holes coincide causes it to present a slight overhang at the left end of the Levelite module. Second, 1-1/2 inch self-tapping screws of the proper sort are hard to find. (We used screws intended for mounting wallboard to aluminum studs; their heads are not exactly right, and the pitch of their threads is extremely low.)

An alternative would be to use a piece of aluminum angle as a bracket. (A piece 1/2 inch on a side and 1-7/8 inches long will do.) Tapping the angle bracket appropriately, a single 3/8-inch long 4-40 machine screw can hold the angle to the Levelite's cabinet (the head of this screw will protrude into the battery compartment). Two holes can then be drilled in the left apron of the oscillator box to meet matching tapped holes in the angle bracket. (The aluminum angle should be contained inside the oscillator box, not "lipping" over the outside of the left end. This requires some careful measurement, but it looks elegant.)

Next, drill a hole through the cabinet of the Levelite, somewhere under the oscillator box, through which leads can pass from the Levelite circuit to the modification. There is a ridge in the Levelite case between the battery compartment and the board compartment. Put this hole close to the ridge in the "board compartment." When assembled, these leads should have clearance between the board and this ridge.

Orientation to the Levelite Board

Familiarity with the Levelite board layout will help in determining where to solder the seven leads which go to the oscillator box. Hold the Levelite board with the component side up and the battery connector at the left end.

A large round capsule will be apparent at the bottom-left corner; this is the beeper. Along the bottom edge and immediately to the beeper's right is the LM358 dual op-amp (used to compile the various sensor signals and present them as an analog output). At the bottom-right corner are two trim pots which are for calibration of the instrument. The 4093 Schmitt triggers are at the upper-right corner; this chip is oriented horizontally.

Although you needn't tamper with them, there are a total of three trim pots on the board. The 4093 has one of its gates wired as an oscillator which drives the beeper (actually controlling the frequency of the beeper's audio tone). This adjustment is a square pot on the solder side of the board just behind the beeper.

For calibrating the Levelite, currents to the illumination LED's on the encapsulated vial can be adjusted. The anodes of these two LED's go to VCC. The cathodes go to the ends of a 1K pot. The arm of this pot goes through 2.7K, then through a 20K rheostat to ground. The trim pot closest to the bottom right corner is the 1K "balance" pot, and the one immediately to its left is the 20K rheostat. Lowering the resistance of the rheostat "broadens the window" (by swamping the vial with more illumination) over which the level detects "0".

Immediately above the beeper is another horizontal chip, the 14541 timer. In the upper-left corner are the two transistors, one for switching off the power and one for regulating the Levelite's VCC.

As you might expect, the horizontal line of upright flat devices on the right half of the board contains the five LED's. Above the left two is the pushbutton power switch.

The chip of interest is the LM339. A 14-pin package, this is vertically oriented, and is placed to the left of the five LED's. As we view the board from the component side, pin 1 is at the 339's upper-left corner.

Pins 1, 2, 14 and 13 of the LM339 go to charging resistors on the 555 (see the above circuit for specific connections). Pin 3 of the 339 goes to the 47K resistor which controls the 2N2222 in the above modification circuit. These five connections are made at the ends of the chip pins, on the solder side of the board.

Power connections are made directly to the points where the Levelite's battery connections come through on the solder side. The lead closest to the beeper is negative (ground). The lead closest to the two transistors is plus 9V; this goes to the emitter of the 2N4036.

Don't forget to pass all leads through the hole you drilled in the Levelite's housing before you solder them. The leads should be long enough to lie along the Levelite board some distance before passing through to the oscillator box; I like them to be 5 inches long.


Making sure that some of the extra lead lengths to the modification are fed through the hole behind its board, mount the oscillator box (at least, install the screws which can only be reached from inside the board compartment of the Levelite). Secure the Levelite board with its four screws (unless only two were provided). Lay the large piece of cloth tape down against the solder connections of the Levelite board; be sure not to entrap any of the rotatable sensor's wires under it. Finally, loop any extra lead lengths to the modification over the tape as well, perhaps even taping these down with an added piece.

With the front panel of the Levelite facing the table, remove the tape from the rotatable sensor. Be careful with its leads. See that the wires move freely when the pointer knob is rotated. Make sure that the large coil spring is in place; then attach the metal backplate.

Summarized Operating Instructions

Operation of the modified instrument is simple. First, with the aluminum "extrusion" lying on its edge (the open side toward you), insert the electronic assembly into it with the sensor dial at the right end and the added housing to the left. The edge now on the table is the one to be used in testing surfaces. Be sure that the "silence" switch of the modification is switched on; then push the Levelite's power switch. Tones will be heard as the level is moved around.

A surface at any incline can be tested, and its orientation can be corrected to conform to any of the 5-degree steps provided for by the Levelite's rotatable sensor. The important thing to note is that, in order for an indication to be had, the dial should be turned so that its pointer is straight to the left; this puts the sensing device in its active region.

As the device is tilted through the active region of the sensor, the various tones will be heard; you will have to do this slowly, due to its high sensitivity. At the zero-degree position, the Levelite's beeper will be heard and the modification (the oscillator box) will be silent. As you slowly tip the instrument one way or the other, a pair of tones will result, the higher one meaning that you are far away from level (about 3 degrees). Train yourself to recognize that the tones to one side are higher than those at the other side of level.

Once on, the circuitry will stay active for two minutes. No "off" switch was provided for on the original instrument. The "silence" switch on our unit may be turned off after the job is done. If left alone, our modification will go off with the rest of the circuitry. If your time runs out in the middle of a job, our circuit will let you know, either by the absence of sound, or by a little chirp it makes in the throes of death.

Parts List

Resistors (1/4-watt 5% unless otherwise stated):

  • 1--47 ohm 1/2-watt
  • 1--1K
  • 1--22K
  • 4--47K
  • 2--130K


  • 1--0.01uF ceramic or Mylar
  • 1--10uF 10V electrolytic
  • 1-100uF 10V electrolytic


  • 1--555 timer (A CMOS version is not necessary, since the driving currents for the LED's in both the sensor and the display are quite high.)
  • 1--2N2222
  • 1--2N4036
  • 5--1N914


  • 1--SPST toggle (for "silence")
  • 1--Small speaker, such as Mouser 25SP016.
  • 1--Small project box, such as Radio Shack 270-230.
  • 1--Small piece of perforated board, perhaps 1-1/2 by 2-1/2 inches.

The Levelite must be purchased in two or more units. The electronics section costs just over $50. The aluminum "rails" (extrusions) come in
2-, 3-, 4-, and 6-foot lengths, priced at about $5 a foot.

Address your correspondence and telephone inquiries to Mr. Doug Starcher, Fourth Corner International, 1805-B West Bakerview Road, Bellingham, WA 98226. Telephone (206) 734-0602.


By Albert Alexy, Jr.


This device audibly informs the user when an appliance is drawing current. The original use of this system was on my father's electric egg cooker. It solved the problem of knowing when the egg cooker was finished and the eggs were ready to serve. In this application, an audible beep sounds every few seconds until the cooker turns off, alerting you that the eggs are done. An alternative use is to alert the homemaker that other silent appliances (such as toaster ovens, electric fryers, or curling irons) have been left on.

Description and Circuit Operation

Built into a project box, the instrument has an electrical outlet for the appliance and a power cord of its own. In operation, it is plugged into any wall outlet; then the appliance to be monitored is plugged into this device. Whenever the appliance is consuming more than 20 watts of power, an audible beep is heard. The repetition rate of the beep can be adjusted--by a front panel control--from a continuous tone to a one-second beep every 5 minutes. Also located on the front panel are the speaker and volume control.

An interesting effect is noticed when this instrument is used with appliances containing heating elements, such as toaster ovens, electric fryers, and curling irons. When used on these appliances, the Egg Timer will beep slower as they approach operating temperature. On the other hand, it will not operate properly with appliances that do not shut off completely, such as coffee pots that switch from brew to warm.

Tone Oscillator

The "beep" is generated by the second half of the IC (a 556 dual timer chip). Its pitch can be adjusted by changing the 50K pot labeled "pitch." A value of approximately 29K causes a beep frequency of 500Hz, and a value of 7K causes the beep to be at 2000Hz. The beep is kept turned off by the reset pin (pin 10) which is controlled by the first half of the chip.

Beep Controller

The first half of the 556 determines how long the beep stays on and off. The 1K resistor labeled "on time" can be adjusted for the amount of time the beep will sound. A value of 1K produces about a one-second beep. The 100K pot labeled "rep rate" (mounted on the front panel) determines the length of time the beep will stay off. With no current flowing through sensor T1, this section will be held in reset by pin 4.

Current Detector

The current detector consists of a modified transformer for current sensing. The modification consists of carefully removing the secondary winding and replacing it with one turn of heavier gauge wire.

The unit chosen is a modified 110-volt/12-volt, 2-amp transformer. The one used must have the 110V winding on the center core and the 12V winding around the outside of this 110V winding. Due to the thickness of the 12-gauge wire being installed, a 2-amp transformer was used for reasons of its physically large core (thus leaving plenty of room for the 12-gauge wire). (To help in identifying which winding is which, it should be noted that the 110V one is of fine wire, and the 12V winding is much heavier.)

Proceed to remove the outer (12V) winding. When you get to the last few wraps, be careful not to damage the finer 110V winding underneath. Discard the wire removed. Obtain a piece of insulated 12-gauge stranded wire about 12 inches long. (Note: This wire will carry all of the current used by any appliance connected. Do not use a smaller gauge wire.)

With insulating tape, wrap the 110V winding remaining on the core. Wrap one turn of the number 12 wire around the transformer core and secure it in place with "Ty-Wraps" or insulating tape. Be sure not to chafe the insulation while wrapping it onto the core, and ensure that it is well insulated from the transformer core.

The output of the current-sensing transformer is detected by a bridge rectifier, then limited by a zener diode. With no current flowing through T1, no voltage will be produced by BR1. Consequently, the circuit will be held in reset. With a load of 20 watts, sufficient voltage will be produced by BR1 to turn on this half of the 556, allowing the beep to cycle. As the load increases, the output of BR1 will increase until it is clamped to 5V by the zener diode.

Power Supply

The power supply is composed of transformer T2 which lowers the 110V line power to 6 volts. BR2 rectifies this, which is then filtered and regulated by IC-2 to 5 volts DC. These 5 volts are designated VCC and used by the rest of the circuit.


The prototype Egg Timer was built in a 3 by 4 by 5-inch Bud box. A speaker, volume control, and rep-rate control were mounted on one end. A standard duplex outlet (i.e., the same as the one used in house wiring) was mounted on the other end along with a strain relief type grommet for holding the input power cord. Be sure to use either 12- or 14-gauge wire for the cord--it will carry the same amount of current as the appliance plugged into it. A slotted hole in the back of the box makes it easy to hang the box on the wall, and the cords will hang out of the way.

Circuit for the Alexy Appliance-Alert System

A 3-wire 12- or 14-gauge line cord is used. The "third" prong (the green wire) of the cord goes to the corresponding "U-Ground" screw on the receptacle. The white wire of the cord (often called "neutral") goes to the wider of the two slots on the receptacle. [This "neutral" side of the plug sometimes is given a prong which is wider than the other "hot" side. Orient the plug so that the round U-Ground pin is at the bottom, and point the prongs away from you (as if you were plugging it into the socket). The round prong is ground (green), the left-hand flat prong (which can be a wide one) is "neutral" (white), and the right-hand prong is "hot" (black).]

The hot prong goes through a fuse, then through the 12-gauge one-turn loop on the current-sensing transformer to the hot side of the receptacle. The junction of the fuse and the current sensor goes to one side of the primary winding on the power transformer T2. The other side of the T2 primary goes to the "neutral" prong on the receptacle.

The secondary of T2, 6 volts AC, feeds a 50-PIV 1-amp bridge rectifier, BR2. The negative output terminal of BR2 goes to circuit common (the "ground" line for the IC circuitry). The positive output of BR2 goes through 10 ohms to the positive side of a 500uF electrolytic; the negative end of the capacitor goes to circuit common.

The junction of the 10-ohm resistor and the filter capacitor goes to the input terminal of IC-2, a 3-terminal 5-volt regulator. The "Common" terminal of this regulator goes to circuit common; its output goes to the VCC line of the IC circuitry.

Pin 7 of the 556 goes to circuit common. Pin 14 goes to VCC. The "Control" terminals of the 556 (pins 3 and 11) are each bypassed to circuit common by 0.01uF.

The 110V winding of the modified transformer (current sensor) feeds a 200-PIV 1-amp bridge rectifier, BR1. The negative output of BR1 goes to circuit common. The positive output of BR1 goes through two 4.7K resistors in series to circuit common. Across the second of these resistors (the one going to circuit common), are two components: One is a 3.3uF electrolytic capacitor. Also across the resistor is a 5.1V zener diode (anode at circuit common).

The cathode of this zener diode (which is also the junction of the resistors) goes to pin 4 (reset) of the 556. Pins 2 and 6 are tied together and go through 100uF to ground (negative of the electrolytic at ground). Pin 1 goes through a 1K resistor to VCC; if this is made a 1K rheostat, the duration of the beep will be adjustable. Pin 1 also goes to the anode of a diode, the cathode of which goes to pins 2 and 6. Pins 2 and 6 go to the anode of another diode; the cathode of this diode goes through a 100K panel-mount rheostat ("rep rate") to pin 7.

Pin 5, the output of the first half, goes to pin 10 (reset) of the beeper half. Pins 8 and 12 are tied together and go through 0.1uF to ground. Pin 12 goes through a 50K rheostat to pin 13; this is the pitch control. Pin 13 goes through 1K to VCC (which is at pin 14).

Pin 9 goes through a 500-ohm panel-mount rheostat (volume) to the positive end of a 3.3uF electrolytic capacitor. The negative side of this capacitor goes through the speaker to circuit common.

Parts List


  • 1--500 ohm panel-mount linear pot wired as a rheostat
  • 1--1K (optional replacement for the resistor off pin 1) PC-mount trim pot wired as a rheostat
  • 1--50K PC-mount trim pot (pitch control)
  • 1--100K panel-mount linear pot wired as a rheostat

Resistors (1/4-watt 5% unless otherwise specified):

  • 2--1K
  • 1--4.7K (1/4-watt okay as the bottom resistor--across the zener)
  • 1--4.7K 1-watt (off the output of BR1)


  • 2--0.01uF disc ceramic
  • 1--0.1uF ceramic or Mylar
  • 2--3.3uF 10V electrolytic 1--100uF 10V electrolytic


  • 2--1N914 Diodes
  • 1--5.1V Zener Diode
  • 1--BR1 200V 1-amp bridge rectifier 1--BR2 50V 1-amp bridge rectifier 1--5V 300mA regulator (such as 7805) 1--556 dual timer chip


  • 1--110 to 6 volt 300mA transformer
  • 1--transformer to be modified (such as 110 to 12 volts at 2 amps)
  • 1--box (such as 3 by 4 by 5 Bud)
  • 1--duplex 3-wire outlet
  • 1--12- or 14-gauge 3-wire power cord
  • 1--20amp 125V slow-blow fuse

Pin Connections

uA7805T Voltage Regulator

With the mounting surface toward you and the leads pointing up, the three connections are, from left to right: Input, Common, Output.

  • 556:
    • Pin 7--Ground
    • Pin 14--VCC
    • Pin 1--Discharge 1
    • Pin 2--Threshold 1
    • Pin 3--Control 1
    • Pin 4--Enable (Reset) 1
    • Pin 5--Output 1
    • Pin 6--Trigger 1
    • Pin 13--Discharge 2
    • Pin 12--Threshold 2
    • Pin 11--Control 2
    • Pin 10--Enable (Reset) 2
    • Pin 9--Output 2
    • Pin 8--Trigger 2