plastic or metal case. Or, the meter case can be mounted separately on the instrument panel while the electronic circuitry can be fastened back of the instrument panel.

Transistorized Electronic Tachometer

by R. G. Muggins (EAA 4755) 4915 S. Detroit Tulsa, Oklahoma You can build your own inexpensive transistorized electronic tachometer for your Volkswagen experimentalaircraft engine. The circuit uses only readily available parts, and differs little from the familiar conventional monostable multivibrator. It operates off your magneto which supplies the voltage and triggering signal, so it requires no battery. The voltage required is so small that it does not interfere with the ignition system. HOW IT WORKS The tachometer circuit (Fig. 1) is nothing more than a simple monostable multivibrator (Ql and Q2) triggered by a shaped positive-going rectangular pulse produced by the opening and closing of the magneto points. Pulse shaping is accomplished by the C2-R6 combination. The average current in Ql's collector is monitored by an O-l mA fullscale milliammeter. Since the collector current will be directly proportional to the trigger frequency — determined by the engine rpm — the meter can be calibrated in terms of rpm. The accuracy of the reading is determined essentially by the accuracy of the meter used. CONSTRUCTION A convenient layout for the tachometer is given in Fig. 2. The parts are laid out on a 2V4 by 2V6 inch perforated phenolic board. Since the R3 potentiometer is used when calibrating the meter, be sure to locate it where it is accessible for this purpose. Although an inexpensive O-l mA d.c. meter was selected to keep cost low, a 3'/2 inch wide-view panel meter is preferable. The circuit board can be mounted on the back of the meter, and the entire unit can be housed in a 18

APRIL 1970

CALIBRATION

In order to test the circuit and have a means of calibration, the author built the test unit shown in Fig. 3 by mounting a small, light-weight Slick magneto on a board and connecting it up to a synchronous electric motor that turns 1,800 rpm, then connected four spark plugs so that the magneto would be operating in its normal manner. The electronic tachometer was installed on a small panel, and one lead connected to the P terminal on the magneto and the other to the ground. The electric motor was turned on and R3 adjusted until the meter read .18 or 1,800 rpm on the dial. An easy way to calibrate your electronic tachometer is with another tachometer. Locate an airplane with a four-cyclinder engine that has an accurate mechanical tachometer. Connect up your electronic tachometer to its magneto P terminal and to the ground. Run the engine to 1,000 rpm, then adjust R3 for a reading of 0.1 mA on the meter. With this adjustment, the meter is calibrated so that each 0.1 mA increment on the dial represents 1,000 rpm.

Fig. 1 PARTS LIST

Cl, C2 - O-.ln F, 200 Volt Copocitoi 01, Q2 - 2N414 Tran»ittor Rl, R 2 - 2200.0hm. 1/2 Wott R

R3 - 1000-Ohn,. Print.,) Cii

Miniotvr. Pot.ntiofllotvr

R4 - 6BOO.Ohm, 1/2 Wott R« °.5 - 1000.Ohm, 1/2 Wott R.

R6 - 1120.Ot.rn, 1/2 Wott R. Ml -0-1-m A d . c . M,11,0mm.

SI - Pon.l Mount Switch

Miic. - 2 Trontiitor Socfc.ti, 2.1/2"

2.1/2" Ph.nolle Circuit Beard,

Wir., Sold.r, Enclo>ur. (Opti nol)

Fig. 2 -0

R5 (

RtL.F.IIittc

SPARK PLUGS

Fig. 3 ELECTRIC MOTOR

LEADS FROM MAG TACHOMETER

PANEL

More On The Transistorized Electronic Tachometer By Lowell D. Hamilton, EAA 48677 10307 Collingswood Road La Porte, Texas

A SPORT AVIATION entitled "Transistorized Electronic Tachometer", I had mixed feelings. On the one hand, FTER READING the article in the April, 1970 issue of

my background as an electronic engineer told me the article was in error as to theory and as such the tachometer could give highly erroneous readings; on the other hand, while having the components with which to build the tachometer, I did not have a magneto or an oscilloscope with which to confirm my analysis. I did not wish publicly or otherwise to bring discredit on the described tachometer

or the author of the article without positive proof of incorrect information. However, after due consideration, and in the interest of safety, I have decided to present my analysis of the described tachometer, even though I lack the components and instrumentation with which to obtain positive proof. There are three main points I wish to present. These are: 1. The described operation does not agree with theory. 2. The described calibration technique does not check linearity nor accuracy of calibration at

end points of operation (i.e., low and high rpm).

I

3. The described tachometer is a potential safety hazard to the user since the readings are probably erroneous at operating rpm. The tachometer described by the author is a project worthy of undertaking. However, there is a serious discrepancy in both the circuit and the accompanying information which will present the builder and user of this circuit with a tachometer which will give erroneous readings, if it works at all. Since I do not have at my disposal an oscilloscope and magneto with which to test the recommended circuit, I cannot state whether the readings at 4000 rpm will be high or low, nor can I state how much error will be introduced. I can only state that the circuit will not perform as it is intended to perform. The theory on which the circuit is based is that the milliammeter, M,, will read the average current flowing in the collector circuit of Q,. The circuit described is a monostable multivibrator. Since a properly operating monostable multivibrator will produce a constant pulse width at the collector of Q, each time it is excited or triggered into operation, and since this pulse width can be expressed as the product of R., and C,, the percent of current flowing in Q, compared to the constant current flowing if

0.3 VOLTS RM| = 1000

2 6

R| = 2200 4 6 INDICATED RPM 11000

FIG. 1 JANUARY

1971

FIG. 2

FIG. 3

10

QI is forced to conduct continuously can be stated mathematically as follows: Percent of average current flowing in Q, collector — product of following: rpm of engine number of cylinders Vi (for four-cycle engines); 1 (for two-cycle gines) 1/60 (to convert minutes to seconds) value of R:J in ohms value of C, in farads 100 (to convert to percent) 0.65 (explained in next paragraph) Assembling the above, the mathematical equation becomes: Percent of average current = rpm (number of cylinders) (R..) (C,) (100) (0.65) (60) (2) The factor 0.65 is a correction factor that partially compensates for differences in transistor biasing points and the reverse current flowing in R, while C, is discharging itself through the series combination of R:, and R,, during the period that Q, is conducting fully and Q^ is in a non-conducting state. However, for purposes of making my point that the article has badly misled the reader, it will suffice. Now, substituting the values given in the article and using a four-cylinder four-cycle Volkswagen engine at 1800 rpm we have: Percent of average current = (1800) (Vfe) (4 cyl.) (0.65) (1000) (0.1x10-6) (100) (1/60) = 39x10-2 = 0.39 percent If now, as the article states, we have adjusted the engine for 1800 rpm and the meter M, to read 0.18 ma, and since this reading of 0.18 ma corresponds to 0.39 percent of steady state saturation current, the steady state current would be: 0.18 ma = s.s. current O 3 9 1 0 0 s.s. current = 46 ma Now, assuming that the resistance of M; is 1100 ohms (and this is not an unusual value for a 1 ma meter), and assuming the Collector-to-Emitter voltage is 0.3 volts when Q, is saturated, then we have the circuit shown in Fig. 1. Reducing this to an equivalent circuit we have the circuit shown in Fig. 2. Now, writing a simple voltage loop applying Ohm's law of E = IR we have: V, = 0.3 v. + R M (IM) + R, (I M ) i Collecting terms and solving for V8 (the voltage source for the transistors) we have: V. = IM (R, + RM ) + 0.3 1

Substituting values we have: V8 = I M (2200 + 1100) + 0.3 = I M (3300) + 0.3

= (6.046) (3300) + 0.3= 152 + 0.3 = 152.3 volts

Since the manufacturer gives a Vceo (Collector to Emitter voltage, open circuit) of 15 volts maximum for the 2N414 transistors, clearly the 152.3 volts would destroy the transistors. Also, the peak current of 46 ma, while capable of being handled by the transistor, would probably ruin the meter rated for 1 ma. But wait, you say, where is the voltage source for the emitters of the

transistors? Where is it indeed? The circuit in the au-

thor's diagram shows no such connection, and without a voltage source the transistors cannot conduct current properly and operate as expected to. This then, is a prime discrepancy. There are more, although the lack of a voltage source overshadows all others. Other discrepancies arc as follows: 1. The transistors are germanium with high-leakage currents, they arc temperature sensitive, and have much higher failure rates than silicon types. 2. There is no back bias supplied to the bases of the transistors, a serious oversight since the transistors arc germanium and subject to base leakage which can result in both transistors conducting simultaneously which is most undesirable since the meter readings will be highly inaccurate. 3. There is no protection on the triggering circuit to prevent damage to the transistors due to spikes at the trigger source. 4. The pulse width of the monostable multivibrator would be too small to be effective due to the RC time constant of R3 and C, T = R3 C, = (1000) (0.1 x 10-6) = 1 x 1(H = 100 Us This is less than four percent of the time between trigger pulses (3000 us) at the maximum rpm of 10,000. 5. Calibration of the tachometer at only one point does not prove the tracking or accuracy of the tachometer over the range to be used. On the subject of calibration, reference to Fig. 3 shows that while the transistorized tachometer may be accurately calibrated for 2000 rpm, it could be inaccurate at the desired rpm (say 5000). Two possibilities exist: 1. The tachometer registers more rpm than the engine is developing, thereby misleading the user into thinking he is developing more power than he actually is. 2. The tachometer registers less rpm than the engine is developing, thereby resulting in overspeed of the engine, with subsequent early failure of the engine plus less power possibly being developed should this cause the user to operate on the backside of the power curve.

In closing then, it can be stated that in order for the tachometer to be accurate, the first requirement is for a stable voltage source to drive the current through the meter. To fulfill the objective of an accurate tachometer the following requirements must be fulfilled: 1. Provide a regulated voltage source. 2. Use silicon transistors (these have low leakage and can usually be operated over a fairly wide range without back bias. Also, they are much less prone to effects of temperature changes). 3. Provide protection against undesirable characteristics of the triggering source. 4. Use an RC time constant that will give a duty cycle of approximately 0.35 to 0.40 at the maximum operating point of the tachometer (i.e., if the maximum operating point is 5000 rpm, corresponding to a pulse repetition frequency of 10,000 for a four-cylinder engine or a time between pulses of 6000 microseconds,

then design the pulse width of the monostable multivibrator to be 2100 to 2400 microseconds. 5. Use at least a three-point calibration check to ascertain that the tachometer is accurate at low, medium, and high rpm readings. (A five-point or greater calibration check is even more desirable). SPORT AVIATION

45