Your curve fittings don't fit the measurements. You need at least a 3rd order poly to fit the measurement with its onset and steeper rise. The power numbers will be higher (and more precise) as a result I guess. But a 3rd order poly is "just barely enough". The best fit will most likely be given by a special exponential function but I don't know how advanced the curve fitting tool in the dyno software is. Hit me up if you want me to make a very precise power curve using a zero-phase lowpass filter on the RPM data instead of fitted curves. Therefore you'd know if it's the fit or something else that's keeping the power numbers down. (Friction and deformation in the gears might be another big factor that's adding to this) I'd LOVE to produce a nice power plot for you in Matlab, but in case you want to do this yourself, you can get the power of your dyno by this formula: T = 1/2 * m * r² * dw/dt P = T * w T: torque, P: power, m: mass of flywheel, r: radius of flywheel, w: angular velocity, dw/dw: angular acceleration You need at least a first order lowpass to smooth w (angular velocity) because the derivative of w amplifies high frequency noise linearly. Without filtering, the T and P (torque and power) curve would get noisy and the max. power reading would get unreliable. If you have RPM data from the wheel freewheeling I can also calculate the power loss from friction and air resistance of the wheel so that you can incorporate that into your model to compensate measurement errors. But I suspect the gears have a bigger impact.

@@TimeFadesMemoryLasts Dang, I can’t read half the words your saying

Almost like the decimal needs to be moved 1 spot to the left

You have to calibrate the app before your runs This numbers are absolutely wrong For calibration just click on the Dyno button on screen After that you can fully calibrate your dyno

@@fat_cat8457 Haha, don't worry. TLDR: I can make a very detailed, exact and professional torque and power curve from these dyno runs.

Yup. It should take a few seconds, at least, to get through the pull.

Smaller flywheel. 12k rpm during dynk pull is only half of that motors rpm range.

@@Poizdrc Smaller for sure. The engine is trying to rotate too much mass. Gotta make sure you can use the full rpm range.

@@Poizdrc but this was the roller rpm? It's seems he used a low gear so it wasn't the engine rpm. In a car you often choose a gear which is near 1:1 with tire rpm. But don't ask me why 🤣

Bigger wheel or remove the gearbox. Likely the engine's own ramp up time is affecting the numbers. Also rpms look unstable.

Yeah, and/or heavier, the engine seems like it barely has to work to get that flywheel spinning, that's why it only produced that little horse

Yes the longer it takes.to sweep from idle to.max rpm the more accurate it will be you want to load the engine down as much as possible. Also I think you can get very small eddy current brakes they would be far more accurate

I was also going to say this I don’t know if you are taking into account the reduction from the flywheel to the dyno wheel if you multiply the hp numbers by the same as your reduction does that make the hp numbers seem more accurate?

Did you consider the 2:1 gear ratio in the software? It's best to have a 1:1 in these types of dynos (open loop) - usually the larger the mass the better - or you should switch to a closed loop/brake hp dyno design with feedback (eddie currents) and field drivers (but maybe overkill for such low power).

You need to hold the engine back…. Doesnt seem like the engine is working at all but still cool

It's part of the instructions for Simple Dyno

2 comment and Lot of like

yeah it’s common, it is cool to see though

I thought the same thing. I was like, "Huh? No way! I never would've thought of that." Good stuff always from JQ90!

It's a simple magnetic pick up, most if not all modern engines use them for crank and cam position sensors. On those the magnet is stationary and a toothed wheel is what spins to induse an a/c signal and considering he used headphones he may want to invest in in an A/C to D/C converter considering computers only understand D/C digital signals and not A/C which is what he was producing. Also as a note add more magnets for a better much cleaner A/C signal and it will happen more often giving better data.

👍

Yup especially with a slower four stroke, clutch slip would be my primary concern.

brake dinosaur

Everything about your idea is good I reckon, except measuring current to figure out torque. There'll be unknown losses in there for sure. It's much more accurate to mount the motor on some kind of hinged rig that lets you measure directly the torque on the motor can. Jeremy Fielding did it with a cheap kitchen scale, but you could use a load cell for the professional look!

Motors aren't 100% efficient. Also their efficiency varies with rpm *and* load making them virtually impossible to be used as the dyno without complex measurement post processing

@@AntonBabiy isn't that what I said? 😆 Anyway yes you're right. You can use a motor as a load for sure, but you need to measure torque directly rather than inferring it through current.

motors vary in efficiency at both speed and torque so it will never be a 100% accurate calculation im afraid

@@oscarzt1652 If you're measuring torque separately then motor efficiency doesn't come into it. You're just using the motor as a variable brake, and not relying on its linearity or efficiency or anything else. Power = torque x rotational velocity so if you can measure both torque and speed accurately (and neither is hard to measure) then you're sorted.

you should probably install one at each wheel.. or maybe two.

Make a tank with one engine and reverse transmission for each track. ❤️

Or a enlarged powerwheels

There would be. Ideally it would be a direct drive with the flywheel able to freewheel after the pull. Maybe an automotive style clutch could help? Also, air/fuel ratio could be throwing off the numbers, plus the engine hasn't been broken in. So, a list of things to do would be: -Lap the valves -Tune carb -automotive style clutch rather than centrifugal -play with fuel mixtures

not really, how harder you spin it the better it grips, it slips on lower revs yeah but it will catch up and you would hear that in the engine if it hold mid rpm for some time before reaching max rpm. (my old moped had a 2 speed automatic with centrifugal clutches that even sat in a compartment filled with oil , and like i said they would let you know if they are slipping based on rpm behaviour. and you could see if the actual clutches are glazed by slipping and burning it) his graph shows a smooth line in the dyno, if it slipped it would have shown a flat bit in the graph

@@Awkward_Fox Imo, these isn't a need for a clutch, having a direct drive engine to flywheel is my idea of a solution.

@@Awkward_Fox tuning the carb adjusts fuel mixtures...also valve lapping isn't nessesary if the valve seat and valve face are properly machined

@@SlowSTEN I mean to say fuel mixture. Like nitro and whatever other fuel/oil is used

Decimal place value? 3rd grade. Out sick that day?

Old Xerox machines (the behemoths) had realistic solenoid clutches. I used to have one inside an RC car.

@@aserta Oh cool, thanks for that. I have a custom nitro rc trophy truck in the works also. I'll have to check that out.

Valve lapping isn't nessesary if the valve seats and valve faces are machined properly.

@@SlowSTEN he lapped the valves in the four cylinder and it gained quite a bit higher revs and sounded much crisper in the throttle

@@defaultdriftco00 Yeah I remember that one. Clearly they weren't manufactured to particularly good tolerances for THAT much improvement. Worth a shot here.

Please show us the valves and cams. I think this is a water cooled 2 stroke.

@@fromthebackofmymind its a dohc 4stroke. Much like the inline 4 he has 😅

I might be wrong, but I dont think hall effect would give any significant improvement here. He is just measuring pulses, he doesnt need an analog reading on the magnetic strength.

@@chemieju6305 hall effect sensor can give a digital signal so its easier to understrand and process to the computer

Dynos only measure torque. The software does a simple math equation to figure horsepower based off the torque numbers vs rpm.

Yes, the earbud has a certain cutoff frequency possibly making the signal lower amplitude at higher rpms. To really see if this is an issue he'd need to first see the signal on a scope but then keep in mind the audio input also has a finite bandwidth

i thought is was pretty dumb to use an earbud tobegin with. I 100% agree he should switch to a hall effect sensor. they can be configured to put out pulses when the magnet is on top of it or configured to output a digital readout.

Inertia will change with the square of the gear ratio. If you have 1:2 ratio and go to 1:1, the effective inertia will change by a factor of 4. Use a shield to catch anything that let's go. I can testify that it's no fun having a disc spinning 9,000 rpm zipping around your work area. I was testing if we could go to 10,000 rpm when a heavy coupling bent a 1.25" diameter dyno shaft 45 degrees, then come flying off. Once that starts, it happens fast.

A gear reduction is perfectly acceptable, as long as the math is done at the end to account for it. MANY full size automotive chassis dynos use gear reductions in them.

Real dynos use a water brake and strain guages to measure torque, and then calculate horsepower from the measurement of that, and the length of time of the sweep.

@kayo... 🌹 worst freaking attempt at marketing ever

And compare it to what? Nobody knows the "safe" EGT for one of these engines. Reading a spark plug would be a much safer way to judge the state of the tune.

Either double up the magnets or do a direct drive. The reduction gearing is causing a mis-read and the numbers to be low. And either add mass to the flywheel or ease into the throttle more slowly. Going ham on it is causing a short recording time for the data log.

I'm pretty sure the magnet and earbud was supposed to be a hall effect sensor.

@@bisurdaddy effectively, yes. However, using a discrete hall effect would work much, much better. Signal would be far superior.

Yep. A dyno that can vary the load to the engine and control the engines RPMs at WOT would be the best solution, but an inertia dyno can yield results that are close. The issue with an inertia dyno is getting the weight (flywheel) to be enough so the power output of the engine takes a reasonable amount of time to spin it up to the engines maximum RPMs. A rising RPM curve that happens too quickly doesn't give enough data points or "resolution" for an accurate calculation. If the engine takes longer to spin the weight up to max speed, there are many more data points available and the end result calculation will be much more accurate

I agree. I wouldn't use a clutch at all. A slipping clutch WILL skew the calculation in the end.

they use a laser to measure the vibration

@@Wulthrin that's a lot simpler than I was thinking.

Balancing the flywheel would be a simple process compared to the rest

A chassis dyno ISN'T a true 1:1 ratio though..... Varying tire sizes blow that "1:1" theory out of the water, and there is a different ratio for every car with a different tire size that goes on it. Yep, the tire size has to be entered into the dyno, but it's no different than having a gear reduction in the mix somewhere, and the dyno computer MUST use the ratio of tire size to the roller size in it's equation.

@@davelowets - agreed. this is why chassis dyno's are not the best way of doing business. IIRC, a full sized automotive chassis dyno needs 1:1 in the gearbox, so the tire height can be included into the computer to work that out... in this scale, that is not doable, so the next best solution is to save a set of tires that are the exact same size so as to make equal comparisons. Fair enough?

@@realbogus People run what ever tires are on their car. I've never heard of anyone running "dyno tires" before. For a small inertia dyno like the one here, I would run it 1:1.

@@davelowets in the scale model world, tire size is stupid important. In the real world, a tire might lose a .25", over a 26" tall tire, who cares?

@@realbogus A 29" tall slick may grow by at least 2" at top speed. It's a big deal. Obviously you've never drag raced with slicks before.