The History of the Dynamometer

Send This Page To A Friend
Fade To White
Before the dynamometer there were a number of independent testing organizations that selected cars, took them out, gave them the works, and then wrote up the results. Accurate � but a little clinical. How much bias went into such tests was a matter of business ethics, a subject that's far from the scope of this article. Suffice to say those that could afford to have their cars tested this way were usually the manufacturers, and there lay the problem � enthusiasts wanted an unbiased result, tests conducted by people not being paid for by the company that made the car.

To make the accurate testing of the car more accessible, free from the men in white coats, and available to road testers and the lay-person alike, a cheap and simple device was needed. The first attempt to fill the void was the gravity meter - a device that is best described as a dynamometer that went right along with your car. Instead of Mohammed travelling to the mountain, the mountain came to your dashboard.

Old Man Gravity



Without attempting to go into the remote details of the theory behind such meters, we can journey into physics far enough to set down that the force of gravity inexorably holds us pinned to one spot on this whirling globe. Every time we flick an eyelid, we are jousting with Old Man Gravity. Engineers have become so intimate with the resisting old codger that they simply call him "G." It is Gravity which holds back your car, your boat, even the pen you lift from the desk. If "Gravity" was not eternally pinning us down, we would float around in a world of infinite speed. Of course, Gravity has some good points. It would, for instance, be embarrassing to have the soup refuse to stay in the tureen.

As the villain is Gravity, it follows that we can measure the gravitational perversity and thus find out how much it holds us back. Knowing that, we can then evaluate the amount of effort we must exert to overcome it. The simplest method of illustrating the point is with a pendulum. It does not have to be a fancy gimmick. Take a piece of string, attach a weight to one end, and hang on to the other. The weight will hang vertically and relatively immobile. Move your hand horizontally. The weight - grasped by the invisible hand of Gravity - will lag behind the movement. If you were to rig up a method of measuring that lag, you would have created a pocket-size dynamometer. It would be enough of an instrument to tell you more about the performance of your car than you would be able to get in any other convenient an inexpensive way. The gravity meter is just such an instrument.

The Tapley Meter and Perfometer



Before World War 2, and for a time after, it was the Tapley meter that was used. In the early 1950s an improvement on the Tapley meter come on the market - aptly called the Perfometer. The Perfometer was manufactured by the Autosphere Corporation in New York City. About the size of a speedometer, it could be easily installed on any car. At the time there was a common misconception that the Perfometer was a simple vacuum gauge - mainly because it required no connection to the engine by means of wires, pipes, tubes, nor anything else. Beautifully simple, it was attached to the car at any spot that was convenient. You could take it off and switch it to another car in a matter of minutes. You could stick it on a lawn mower, wheelbarrow, go-kart. A brilliant bit of kit for the road tester and motoring journalist - you can take it along in your suitcase on a plane, train, or boat. Whenever and wherever there was motion, the instrument would read the force being exerted to overcome Gravity.

The dynamometer is a pretty handy device. You can not only determine the efficiency of your engine, but you can do a host of other things such as measuring braking efficiency, evaluating the results of changing plugs, using a different fuel, trying a different grade of oil, and, in the case of roadster or convertible, finding out whether the car is faster with top up or down. We doubt many of you have ever taken readings of your car today, with the intention of comparing it five years from now. For most of us, seeing the Top Gear boys get about in a clapped out high mileage car, and test the output against what "should" have been available when new, is closest we get to a dynamometer.

Using a Perfometer



How do you use the Perfometer? It is impossible to describe all the uses within the scope of this article. Obviously you need to fit it to the dash, or the steering column, or anywhere else where it can be seen in plain sight. When the devices first came into use it was essential to have a helper riding alongside, equipped with pencil and pad. You needed to adjust the Perfometer so that the needle read zero when the car was perfectly level. If your garage floor was truly horizontal, you loosened a bolt, tilted the device until the needle read zero, and tightened up. It was that simple. Using a spirit-level was help, but for the die-hard motoring enthusiasts you could double check the surface more accurately by driving in head first and setting the Perfometer at zero, then reversing into the same spot to see how the needle read. If slightly out, you needed to halve the difference and tighten the bolt. Thankfully this procedure usually only needed to be completed once.

Now for the road test. The first stop needed to be a weighbridge, where you could get the car weighed, including passengers. In the interests of accuracy, many road testers also filled the tank with petrol. The next step was to get the car well warmed up - and then head for the nearest smooth road. If you could find one that was perfectly level and free from curves, you were fortunate, but such an ideal test course actually not necessary. In fact, a smooth, straight road with a slight upward grade was made to order.

Forward Performance



The first test was usually the "ahead" performance of the car, making it necessary to drive with the car facing the slight grade. Drivers would start at the bottom in top gear and stay in that gear. Later on, again in the interests of accuracy, they would plot performance for all the gears. The initial procedure was to run the car as slowly as it would go as you approached the course. Back in the 1950s that was somewhere below 20 miles an hour. With your "test clerk" set with poised pencil, you needed to hit the accelerator smack down to the floor boards and leave it there. Remember that you needed to stay in top gear all the time. As the car jumped ahead, keep an eye on the speedometer while your helper was watching the instrument needle, which would be reading on the outer row of figures on the right-hand side of the dial. As the car hit 20 mph, you needed to let your helper know you had reached the speed, by yelling out "20." Your assistant needed to instantly scribble down the needle reading. Pretty amateurish - but it worked.

The driver then needed to keep their foot on the floor and sing out again at 30 mph and at 40, 50, and so on. At each of these speeds, the instrument reading had be taken. Check out the performance of the VF SSV Redline and you can see such methods would never work today. But back in the early 1950s, even if your car was quick for the time, it was possible to get readings as you accelerated. For the handful of high performance cars that made 10 mph readings impossible, you could take readings at 20, 40, and 60 mph and then go over the course again for the 30, 50, and - if you could make it - higher speeds. For even better accuracy, you could make a run and take readings at 20, 30, 40, etc., and then make a second run for 25, 35, 45, etc.

Rolling Resistance



All of the above based on the assumption that you were using a manual gear box. For automatic transmissions, the procedure was slightly more difficult - because the auto would, of course, want to shift down when you planted the foot. For many of the early road testers, the performance of a car was taken on the manual, then best guessed for the automatic, based on acceleration and quarter-mile times. But the testing was not over yet. The next step was to measure the resistance of the car to rolling without power. In other words, the amount the car was held back by wind resistance and rolling friction. This was even easier than the power test. If the road was perfectly level, then you could have used it, but if there was a slight grade, you needed to run down the incline. You took the power readings running up a slight grade, the decelerating readings needed to be run the reverse way. The grades were not so important, as the instrument automatically compensated for the incline or decline.

This time, you needed to start at your highest speed - or at least the highest speed in which you were interested. Back then, that speed was usuallly the maximum speed limit in Australia - 60 miles per hour. When the speedometer dropped to 60, you needed to push out the clutch and leave it out. Your helper would note down the needle reading for 60. This time, the readings would be rather low and would be on the left half of the dial. As the car decelerated, the driver needed to call out at 50, 40, 30, and 20 mph. For each, you would get a reading that measured the amount of power Gravity was exerting to stop your rolling. To be sure of the figures, you needed to make several runs in each direction and then strike an average. If you have read this far, and are thinking the procedure easy - have a thought for our motoring forefathers, that had speedometers that would jump about with the slightest vibration, or wobble with a slightly stretched speedo cable.

Unless you want to make brake-efficiency and other possible tests, you are finished as far as driving is concerned. Go on home, clear off a table, get plenty of scratch paper, and get to work. You are now an automotive engineer. Lay out a sheet like the one atop page 30. In Column A. insert your test speeds�20, 80, 40, etc. In Column B, put down the corresponding Perfometer readings. Column C is the Column B figures multiplied by 20. The manual you get with your instrument explains that some models give you the Column C figures direct.

Thty do, but the figures are toe hard to read in a running car. Take my advice and use only the outer dial figures and multiply by 20.

Column D gets right down to brass tacks�it evaluates your car's performance on the basis of its weight. Without that weight, we cannot tell how we are beating old Mr. G. You need the weight in tons, so divide the actual running weight by 2,000 (in America, we use the 2,000-lb. ton; in England, they use the long ton of 2,240 pounds). On the chart, the car tested was a Singer 1500 weighing 1.1 tons, so Column D figures are those found in Column C multiplied by 1.1 The result is pulling power in pounds per ton.

Do exactly the same with your figures taken on the deceleration runs when you drifted along with clutch out The next step requires the use of graph paper. You can buy a pad of it at any stationery store or you can make up your own by ruling off !4" squares checkerboard style. The "boughten" type is better as it carries faint rulings that will divide your readings into tenths.

In either case, arrange your sheet approximately like the illustrated graphs with the pounds-per-ton figures along the left margin and the miles per hour on the bottom. You will have to experiment a bit to see that your figures are placed so as to encompass the results you have to set down. Spot off on the vertical speed lines the figures taken from Column D for both the acceleration and deceleration test runs. Run a fair curve through both sets of spots. If you have a draftsman's French curve, this should be easy. If you do not have one, make a thin wooden batten and bend it so it touches the spots while someone else draws the line. Don't worry if some of the spots straddle the curve a slight bit. This is natural. Your original figures were taken from a jiggling needle. You increased any possible error when you multiplied by 20 and again when you multiplied by the car weight in tons. That is why you must rely on a fair curve that, when intelligently drawn, will average out the inevitable variations in readings. Label the power curve "Pulling" and the resistance curve "Drag."

Next draw a Total Force curve. It is the sum of the Pulling and Drag curves. You can plot it by either of two methods. One is arithmetical and is obtained by adding the Column D figures for acceleration and deceleration at any specific speed. Thus, in the example, the Total Force figure for 30 mph is 275 Pulling plus 55 Drag, or 330. The second way to do it is graphically. Take a strip of paper and lay it along the vertical speed line for, say, 30 mph. Tick off the distance from the bottom of the chart to the Drag curve and then transfer that distance above the Pulling curve. You will hit the same spot you found by arithmetic.

Either method will provide the spots for .the Total Force curve. Even if you were to go no further with the tests, you would have a picture of performance. Try curves plotted with lighter and heavier loads, different fuels, varying adjustments. Inflate or partially deflate your tires. Try a brake adjustment to find whether your shoes were dragging or not. Soup the car up with race cams 64 and twin pots and try her again. You may think she is doing better, but the cold, hard facts will show up on the curve sheets. You will have taken the guess out of your performance.

But we haven't stopped yet. It is easy to figure horsepower. If you are gullible and believe the manufacturer's rating, you are in for a shock. If your engine pulls close to the advertised power, you are indeed fortunate and should chip in for a monument to the maker of your bus. Don't kid yourself, the horsepower you are going to find is the real McCoy. It's also simple to figure.

Take the readings from your Total Force curve on each of the 10-mile-inter-val speed lines. Multiply the figure by the speed and divide by 375/For example, the Total Forte of the Singer at 30 mph is 330. Multiply 330 by 30 miles and divide by 375 and you get 26.4, which is the horsepower actually developed at 30 miles an hour. To plot the curve, you will have to lay out bhp figures along the upper right edge of your graph sheet. Again you may have to alter the exact layout to get spacings to fit the sheet.

At this time, an additional set of figures will help. If your car has a tachometer, it will be easy. If not, you can still dope it out. What you want are figures for the rpm at which the various speeds are obtained. You will note this row of figures directly under the bhp curve in the Singer graph. Once you have these figures, you can draw another valuable curve showing torque. The formula is again simple arithmetic. Multiply the bhp for each of your speeds by 5,252 and divide by the rpm.

For example, again at 30 mph, the bhp is 26.4. Multiplied by 5,252 you get 138,652.8. The. chart shows that 30 mph is obtained at 1,900 mm. Divide by those rpm and you get a figure so close to 73 there, is no sense in breaking it down any closer. The fair curve will again iron out the minor variations. Readings for torque in foot-pounds are noted at the lower right edge of the chart. They can, however, be placed anywhere room is found for them and for the curve.

About now you may be wondering what sort of a cripple we tested as we carry the curves but little beyond a mild 60 mph. There is no need for figures much higher unless you have an out-and-out road roarer that does not even preen its feathers until you get above 60. Most cars, even pretty snappy sports cars, will show curves that begin to run rather vertically instead of horizontally after you get above a reasonable touring speed. Unless you have made original readings with micrometer accuracy at the higher speeds, you will gain little aside from arithmetical migraine. If you carry your figures throughout the general driving range and find them happy, you can rest assured that the Occasional spurt into the speed stratosphere will reflect the results obtained at the normal velocities. That is, they will be good or bad just as performance is hot or cold up to 60 or 70 per.

What you have, if you have followed us so far, is a picture of the performance of your car that really means something if you care enough to give it study. No other test�aside from one made on a chassis dynamometer�will give you anything like comparable results. What you now have can also be a measure of your own driving ability. If you wiggle all over the highway or if your foot tap dances on the go-pedal or you ride your clutch, run some tests with a better man at the wheel. The results may surprise you.

If the above seems a bit complicated, remember that you are trying to grasp it all at once. Like eating a whale, it is better to bite off a small chunk and get it down before shoving in the next mouthful. An entire set of tests for any car, including the simple mathematics and the laying out of a chart, should not take over an hour. If you can use a slide rule for the calculations, you can cut the time in half.

Allow me to stress one important point. The accompanying charts are not published with any idea that they represent the virtues or faults of any particular make or model of car. You may own a gar of the same male and year and yet get quite different results. Please remember that the cars for which test sheets are shown were not tuned up in any way. They were run-of-the-mill buses driven by their owners under weather conditions that were as similar as possible. All cars were run over the same course and the readings were set down in the same tabular fashion.

In each case, the Perfometer was "leveled in" on the same pre-tested garage floor. These precautions were taken so that the same cars, over the same road, with the same drivers and observers can be similarly tested in the future to see whether they vary dug to ajustments or other changes as time goes on. For the writer to intimate that the charts are true evaluations of other cars of similar make and model would be a gross injustice, not only to the cars of others but also to the instrument used in evaluation. Thus this article and the accompanying charts must not be construed as being representative of anything but the results obtained by individual cars ruh under conditions that were sufficiently similar to provide our particular group of car owners with figures that will be ^ invaluable to us. For others, they should be considered solely as illustrative of what can be done by any owner with a Perfometer.

There you have it. If the end does not seem to justify the means, you have wasted your time reading this and there is some question whether or not you care as much about your car as you make out. It is none of our business if you waste gas and have dragging brakes, no pick-up, and front wheels that are greaseless. Let her fall apart; the dealer needs your order for a new car anyway. But if you are really fond of your rubber-footed companion and would like some accurate data to boost your boasting, here is your answer. If you have the ability to back a car into an empty ten-acre parking lot, you can prepare an invaluable performance portrait of your own and your friends' gasoline go-buggies. You can also have a lot of fun by flashing your charts in the snoot of the know-it-all bird who got all his car information straight from the dealer. He won't know what he's looking at�a result well worth the slight trouble and expense of getting up the sheets.

Also see:


Brooklands Raceway
Land Speed Records
Honour Roll - Founding Fathers of the Automotive Industry
Automotive Industry - The Pre-War Era
Automotive Industry - Between The Wars
Automotive Industry - The Post-War Era
A Brief History Of The Four Wheel Drive
Latest Classic Car Classifieds