The "Fowle Gimmique"
Also, "The Dynamic Meter Reader"
History and Introduction by Tom Fowle, WA6IVG
As has been said many times, there are any number of techniques for audibly reading analog voltages, the appropriateness of which can be chosen to meet the circumstances. Some are mere "relative indicators" which facilitate adjustments -- "tweaking" for maximum and/or minimum, for example. Many systems whose indications can be "quantified" (where specific values are to be noted) use a graduated standard for comparison; this "standard" is often a linear potentiometer which is fitted with an appropriate Braille scale.
An early example of a quantifiable readout is a "chopper/comparator type". Such a readout emits a tone of constant pitch whose amplitude decreases to a "null" as the measured voltage approaches the setting of the calibrated standard. If an unknown is to be measured, the Brailled potentiometer is adjusted for a null in the tone. Used as a monitoring system, if a voltage is to be monitored for an extended period, and only a change is of interest, then a "chopper" type reader may be best, since the readout will be silent until a change takes place. In this latter service, one can go about his/her business until the tone becomes audible, thus signifying that the input signal no longer matches the setting of the standard.
More often, it is desired to know rapidly the relative state of the voltage under measure, and also to know when it crosses or achieves a desired level. Before the development of this circuit, "VCO" (voltage controlled oscillator) type readers were common, but most circuits required the input to be switched manually between the input under test and the standard. This unnecessarily used an extra hand to operate the switch.
The concept for this circuit was developed by Mr. Howard Moscovits, KB3ZX, (then KB6EY). He used an instrumentation opamp on the input to allow for removal of any offset voltages in the circuit, and a CMos NE555 dual timer. We were able to extend the frequency range of the VCO with other driving schemes (a voltage-controlled current source), and by using a bipolar 555, our circuit drives a permanent-magnet loudspeaker at comfortable listening levels.
Recently, the "pulsating VCO" concept has been implemented in computerized systems -- various microcontroller based devices allowing for some interesting improvements.
One modernized device is a meter reader designed by Mr. Rod Kreuter, WA3ENK; this can simulate the "gimmique" and also send raw data readings in Morse code, which gives it an alphanumeric readout as well. Thus, this reader can also be made to present directional information from the Ham-M style of antenna rotator by connecting it across the control box's meter terminals.
As a "meter reader" for general applications, it will be available with the ability to switch the tone's pitch from that indicating the input signal, to that which indicates the relative position of a calibrated pot. This switched tone is left on the "input" for 3/4 second and moves to indicate the "dial" for 1/4 second. It appears from initial testing that this system makes for even more rapid adjustment to match the input to the dial's position.
This fancy meter reader may be available from its designer as a kit, contact Rod Kreuter at: rak@chem.psu.edu.
The "Fowle Gimmique"--A General Discussion of This Brilliant Readout System
General Comments by Bill Gerrey
The fanciest successor to "The Transistorized Auditory Gimmick" was invented by the Smith-Kettlewell colleague Tom Fowle; this was published as "The Fowle Gimmique", Smith-Kettlewell Technical File, Volume 2, No. 3, Summer, 1982. The sensitivity of this VCO is adjustable over a wide range, and its output can be quantified in a couple of different ways. ("Calibration" of this circuit is completely independent of VCO sensitivity.) We have used the gimmique system in various applications over the years: our audible carpenter's level, analog multimeters, machinist's gauges, etc.
The gimmique has two adjustable features: First, it is a VCO whose pitch over a given input voltage range can be set to please the user. Second, a comparator is arranged to cause pulsation of the VCO tone when an input voltage exceeds a selectable setting.
In many applications, a linear potentiometer, supplied from a stable "reference", is fitted with a Braille scale and a pointer knob. "Calibration" means that the rotation of this Brailled pot covers a desired range of input voltage; often, this means that the full rotation of the pot corresponds to the full-scale deflection of a visual meter to which the gimmique is connected.
With a linear Braille scale, say with 10 major graduations and perhaps minor ones in between, this system would work for a voltmeter or a machinist's "dial gauge". As an adapter for a VSWR meter, a logarithmic scale is appropriate.
For VSWR measurement, the "scale" can be quite elemental; we typically place three-dot marks at 0 and 3-to-1, single-dot marks at 1.5- and 2.5-to-1, and a widely spaced 2-dot mark at 2-to-1.
A simple way to generate the scale is to temporarily mount the "scale blank" to the pot shaft, then supply current to the visual meter so as to make it deflect to these major "readings", and with a Braille slate sandwiching the blank, make the desired marks. Positions o the blank can be calculated arithmetically with the expression:
Fraction of full rotation equals the quantity VSWR minus one, that quantity divided by the quantity of VSWR plus one. For example, for the VSWR mark representing three-to-one, the equation is solved with the fraction "three minus one over three plus one"; which is two over four or half the rotation of the pot.
While it is easy enough to make customized scales for each application, if a master on heavy paper is made, vacuum-formed plastic copies can be generated on a "Thermoform Machine", a Braille copy machine residing in virtually every agency for the blind.
{The term "Braille" here is used loosely. Braille is an alphanumeric code of writing, equivalent to print. On most "Braille dials" of watches and instruments, one-dot, two-dot and three-dot patterns have no alphanumeric meaning, but serve as "tick marks" of appropriate prominence. Braille numbers cannot be rotated; 4, 6, 8 and 0 are all three-dot similar right-angle figures whose rotation determines which is which.}
In operation of the VSWR meter reader, you would first set the pointer knob fully clockwise to its "top of scale". Next, with the meter being "read" set to the "forward-power" position, its sensitivity knob would be turned until the VCO output pulsates (a "chopper" turns the VCO on and off about 4 times per second). Next, after switching the VSWR meter to "reflected power", the pointer knob on the gimmique is turned counterclockwise until the VCO is made to pulsate again, at which point, the pointer's position is read against the Braille scale. As equipment/antenna adjustments are made, improvement of the VSWR is indicated by decreasing pitch of the VCO.
On our carpenter's level, the Fowle Gimmique has no pointer knob; the threshold -- the "boundary" -- at which pulsating starts to interrupt the tone of the VCO corresponds to the 0-degree position of a tilt sensor. The pulsating pitch of the VCO increases as the beam is tilted to the right, and decreases in smooth pitch as the beam is tilted to the left. Everyone would want one of these levels; you can "level" a shelf or a long fence rail, look at your thumb in relation to the hammer, and just listen, without staring at the level's "bubble", for the sound to change from smooth to pulsating. Moreover, with help from the changing pitch, you always know which way, and how far, to move to find that critical position -- between pulsation and continuousness of the VCO tone.
The circuit here may vary slightly from versions in other published documentation. For example, the print diagrams may show a third trim pot (as I remember, a high-resistance one) with which the VCO can be guaranteed to sound at very near 0 volts input. On the ones I build, I provide that there always be a slight amount of charging current to the 556; you will note my 22meg tenth-watt resistor running from the top end of the 22K charging resistor to the VCC line. For my applications, I don't mind a few millivolts of error right near 0 volts, and the 22megs further provides that the unit "ticks" when not in use, thus reminding me to turn it off between uses.
Another omission on my units is a diode in the ground line, needed when other op-amp chips fail to work at the ground rail. Where as the CA3240 I use is a little current-hungry, its FET inputs solve this problem.
You will note the uncommon arrangement of transistors; the 2N2907 PNP units are wired as a "current mirror" which inverts the current from the 2N2222 so as to source current in the opposite sense to the 556 charging circuit. Clever, what? Somewhere, one of our print circuits shows a way around the current mirror, but the price paid for that is that the VCO collapses and falls silent for high input voltages. I prefer the version with the current mirror, since the VCO will always emit a tone, even if its upper limit is saturated.
The trim pot which is connected as a rheostat in the emitter of the 2N2222 adjusts the "gain" of the VCO; we typically adjust it so the pitch is about 2.5 kHz at the top of expected input range. In some implementations, we pick a fixed value of emitter resistor, after we have decided what we want.
In service as a "meter reader", the rheostat in series with the Brailled pot is set so that rotation of the pointer knob against the Braille scale covers the precise range of the expected input; full scale on a meter being "read" would correspond to the clockwise setting of the control, for example. On a "fixed" device -- our carpenter's level based on the Midori tilt sensor being an example -- a trim pot in place of a Brailled pot, and a fixed resistor goes from its clockwise end to the voltage reference (the counterclockwise end of this trimmer being qrounded).