A beginner’s guide to dynamometers


The complexity of an assembled engine means that its horsepower or torque output can easily be lied about. You’ve probably walked up to someone at a car show and looked into the vehicle’s engine compartment as the owner tells you about how this engine makes 600 horsepower. In most cases, it’s impossible to call their bluff. After all, what parts are inside the engine? Are they box-stock components, or did someone have a field day with a die-grinder doing port work?

An extremely experienced ear and eye can spot some tell-tale signs in an engine bay and call BS on a power claim with some confidence, but for the hard truth, you need an objective testing system. If you are lucky, the owner has a dynamometer (“dyno”) sheet to confirm their claims about the engine’s output.

But even dyno sheets can lie—seriously.

Here are a couple of easy ways to sort the truth from the full-throttle lies.

This thorough breakdown comes from Banks Power and explains not only how a dynamometer works but also how the operator can leverage some of the dyno’s reporting to create deceptive results. Fair warning: The video is dense with technical information.

First, we must define what a dynamometer is. Essentially, a dyno is a machine that measures engine torque. Pretty simple, right? Well, it can be. Measuring torque output is done by placing a load on the engine or drivetrain of a vehicle and evaluating how the engine reacts. For engine dynos, the load is often produced by connecting the back of the crankshaft to an external water pump. Chassis dynos—probably what you think of when someone says “dyno”—generate engine load less directly: They place the vehicle’s driven wheels on weighted rollers that resist the tires’ motion via an electrical or a hydraulic system. Whatever the method the dyno setup uses to put stress on the driveline, the goal is to measure torque, and that is when the math comes in.

Any dyno sheet has a mountain of information besides the two tracer lines—one for horsepower, one for torque—and the peak power figures. The first thing that Banks points out that most people don’t think about is the sweep time.

Sweep time tracks how long it takes for the engine to pull through the load placed on it by the dyno from low rpm to high rpm. A long sweep time would mean the engine is at wide-open throttle for an extended period of time. Think of trying to drive up a pass in the Rocky Mountains in a three-cylinder Geo Metro: The car will probably make the grade, but the engine will be at full throttle for a long time.

The longer the sweep time, the more realistic the power number on the results sheet. As the video points out, if a dyno operator says the sweep cannot be longer than six seconds or so due to intake or exhaust temperature, for instance, your engine is not capable of safely generating peak power for longer than six seconds. Sweep time should reflect how you intend to use the vehicle.

light duty vehicle on dyno
courtesy California Air Resources Board

Then there are the other math items that we don’t talk about much. The main one is the correction factor. This is a method of standardizing results across varying local conditions. This is what allows you to use dyno sheets for comparison purposes—if you know what you are looking for. Not all dyno operators use the same correction factor, since this is based on air density and thus factors in that day’s temperature, humidity, and ambient pressure. Using the wrong correction factor—or doctoring the input for the day the testing is performed—can yield an optimistic graph that makes for great bragging … until someone smart enough takes a look at it and points out the lies.

Another interesting item brought up in the video is smoothing. This is a well-intentioned edit to a dyno graph, designed to make the lines less jagged and easier to read, but smoothing can make for misleading charts. Essentially, this process involves editing the horsepower and torque curves to remove mathematical noise.

In short, dynamometers are incredibly complicated. It is easy to take them for granted and put too much trust in a casual glance at the results sheet. Luckily you now know more than you did—unless you clicked on this article already an expert. In that case, tell us what we got right. There is also much more information in Banks Power‘s video than we examined in this article, so be sure to give it a watch to learn even more.


Check out the Hagerty Media homepage so you don’t miss a single story, or better yet, bookmark it. To get our best stories delivered right to your inbox, subscribe to our newsletters.

Click below for more about
Read next Up next: New “Right to Repair” legislation introduced in the House of Representatives


    I have always wanted to have – or build – a dyno largely for tuning purposes. I still like the ol-fashioned carburetor, and the traditional method of tuning – multiple passes down a straight country road – can get you in some trouble. they are expensive, and (literally) disposing of that much energy from a home-made device seems a little dangerous, so for now I keep on wanting

    A friend of mine built a dyno from an auto trans with a lever and a spring. Did not give him numbers but allowed him to tune his F4 (70’s).

    Very informative. Also, very clearly an ad for Banks, with their name plastered all over the video. Still worthwhile. One thing not mentioned is that horsepower that appears only at very high RPM is of very little use in the real world where engines spend very little time at those elevated revs.

    Well, Banks created and produced the video so it seems fair that it also mentions itself a few times over the length of the content.

    Some years ago, I read that one of the popular chassis dynos – may have been Dynojet ? – originated in the motorcycle world, and that when used for car/truck applications, their power was calculated from the motorcycle application. The article implied that many grains of salt be used for the resulting numbers. Racers know that some dynos – or their operators – are “happy” and result in exaggerated power #s. Best use is to evaluate tweaks applied to one engine – timing & jetting loops, different carbs/headers/camshafts/etc – rather than comparing on engine’s out put to another; you can learn a lot about even simple things, like the shape of the air bell atop a carb. Not inexpensive, but obviously valuable for racing applications.

    Anyone who watched that video and didn’t learn at least a few things was already a well-learned person on the subject of dynos. I knew quite a bit and had even knew that corrections were applied, and why. But how they were applied and the ways that dyno results can be fudged was enlightening.

    From the old school days hoursepower was just one factor. Vehicle weight, sprung and unsprung, tires, ambiant temp and surface conditions were all a factor when titles or money were on the line. A line at the dragstrip was “figures lie and liers figure” lets see the tail lights.

    Torque is what makes the car go. Horsepower is a calculated number that is pretty much meaningless. More torque at a lower RPM is what it is all about.

    Great article and video. In a short time, did a great job of explaining process and pitfalls. I enjoy Motortrends Engjne Masters and can now appreciate their results more.

    I have had a little experience with dynes, tuning my Mustang 5.0L engine. I added alloy heads, E303 emissions cam, Cobra intake, larger throttle body and calibrated MAF, shorty headers and a custom built cat X pipe. We were testing parts, exchanging exhaust systems, intakes, and measuring emissions before changing the cylinder heads to find out how much each part was worth, compared to the advertised gains.
    Two dynes were used, of two different designs. One was a dual roller, the other was a single roller. What we found was that there was a discrepancy of around 50 HP at the 300 HP level between the dynos, the highest HP reading (375HP) being the single roller model very popular at that time. The double roller gave a reading of around 325 HP, which we thought was low, but…
    When we went to the drag strip (another method of measuring HP by clocking the trap speed), we found that the 3400 lb launch weight Mustang went precisely as fast as 325 HP would indicate. We checked the information on a couple of different “drag strip dyno” apps on my computer, using the launch weight and trap speed. BTW, we found that a max effort drag strip launch was not necessary, as the trap speeds were consistent over a .5~.75 difference in ET.
    This pretty much left me to think of using the dyno only for tuning chores (AFR, mainly, and “relative” power increase or loss during a specific tuning session), and not really for accurate HP readings.
    When I want an accurate HP reading, I go to the drag strip, make a couple of passes after weighing the car, and plug the numbers into one of the drag strip dyno programs off of the Internet. While that number may not be quite the number from an SAE certified dyno run by an SAE certified technician, it’s close enough to get an idea as to why and how much, the car is faster.
    If anyone is old enough to remember Cook Neilson and Phill Schilling’s Ducati will note that they used extensive drag strip testing to determine HP, and the success of their modifications. Anyone noticing this would know that they didn’t have extensive factory help, and that Old Blue was really a “California Hot Rod”.
    I have a nice hill where I live, and often check out my latest tune with a stop watch and a couple of landmarks. This saves time, travel and dyno shop charges. I call this sophisticated testing (using a very complex mechanical timing device, along with my well tested butt in the seat) the “ass dyno”. Mostly accurate, no temperature factors to worry about, and very portable and relatively cheap to operate.

    Nice video – however their definition of “Torque” is incorrect.
    In the opening, the presenter states that “Torque is the amount of turning force applied to an object or surface over given distance.” Wrong. Applying a force over a given distance is the definition of Work and the unit of measure is Foot Pounds (ft-lb). For example lifting a one pound object one foot = 1 ft-lb of work.
    The unit of measurement of Torque is Pound Feet (lb-ft) which he correctly demonstrates with the one-pound weight acting on a one-foot lever. The result is 1 lb-ft of force.
    That nit has been picked. 🙂

    Having a job tuning on a DynoCom chassis dyno for several years I learned a lot by experimenting. Many customers did not like the results at our shop because an hour away was a Mustang dyno that gave higher #’s than our stingy dyno. I showed a few customers how just tightening the straps down more holding the car to the rollers can change the #’s. The correction factor became known as “The Fudge Factor”. LOL.
    All Big Block cars are pushing 500hp and small block cars must be pushing 400hp. Just ask at any cruise night. The dyno has busted many egos over the years. I enjoyed all the local people that would not believe the dyno sheet for my wife’s car. The car ran 11.0@120mph thru the mufflers with a 400 small block, fast burn heads and .485″lift hydraulic cam. Car made 375hp on the chassis dyno. I was I was told I was a liar, it must have at least 500. Being 3000lbs with driver I explained 400 flywheel hp will put a 3000lb car in the tens with proper traction. That being said the car did 60′ 1.52 on 235×60 drag radials.
    More proof is in the sealed crate motor classes and the and the times and mph they are running.

    I work in the racing performance industry.

    #1 everyone will tell you they have 500 hp.

    #2 the best use of a Chassis Dyno is to get a base line test and then make changes and measure the change difference.

    Too often the numbers between types can be different and true calibration is not a hall mark of most Chassis Dynos.

    Measuring the gains between the first and second test is the best reading you will get unless they really keep the Dyno in spec.

    Someone I know explained to me how some people mistakenly got high dyno numbers. They ran the car in third gear on a four speed transmission to measure torque and used engine rpm’s. This gave a much higher torque value and hence a much higher horsepower number.

Leave a Reply

Your email address will not be published. Required fields are marked *