Home Made Dyno

[andyinchville1]

Hi All,

I race RC cars (electric) and I know they make a little dyno to test the electric motors....Does anybody here know of a good way to make a dyno to test performance increases as we modify our little bike engines?...Right now I am doing my informal test by riding the bike and having a helper time me BUT in order to make good charts to see performance increases/decreases throughout the RPM range I need a way to measure time to a particular RPM given a constant load.

All the factory made dynos are serious $$....I think the cheapest I was able to find was a setup used to test go kart engines and such....It was about $10K....a little....err lot outta my budget....

I was thinking of getting a bike trainer (you know the ones that allow you to
ride stationary indoors) and somehow rig up a timer and rpm meter ....the timer and RPM meter would be the hard part....the instrument or device(s) would need to log data so that one can make a power curve for the engine....I wouldn't even need it to tell me the exact HP or torque that is actually being developed.....I want to just know the relative power increases
each time I tinker with the engine (i.e. under a constant load if my newly modified engine reaches say 1500 RPM faster than it did before I did the mods I would be happy)...

As far as a constant load, if the bike trainer doesn;t offer enough resistance I could maybe hookup an old engine (small) and restrict the exhaust to make the constant load....or maybe a BIG electric motor could be used to provide rotational resistance?

Any ideas would be appreciated.

THanks!

Andrew

[ibdennyak]

Hey....good thinking.....most exercycles have a variable resistance, and if that doesn't work, mount it on a treadmill. See them laying by the curb and in the dump all the time.

[andyinchville1]

HI,
Yes that part was or would be the easy part (the resistance part)...just have to devise a method of gathering the time to speed info and be able to plot it on a graph....Hmmmm brain is burnin on that one!
Andrew

[ibdennyak]

LOL, an engineering mind at work. Considering that I studied to be an electronic engineer in my misspent youth, I should be able to help you. Unfortunately, in that point in time we had those....what do you call em....the big glass things you plugged into the sockets. I was thinking more on the lines of more or less, rather than how much more or less. Lots of those things have digital readouts, so I'm sure some one more knowledgeable can show you how to connect it to a computer. Another short coming I have. Guess I should have worn a helmet sooner.

Denny

[jpilot]

I've given this idea some thought too.

Chassis dynos use a heavy roller that is accelerated by the driven wheel(s). This is the easiest method since the weight of the roller, therefore the resistance, doesn't change over time. The dyno simply measures the time it takes to accelerate the roller. Since the weight and diameter of the roller is a known constant, the applied torque can be easily calculated. It seems to me that a large pipe filled with concrete and mounted to a a couple of pillow block bearings would fit the bill. This would be much better than the bike trainer or treadmill since the resistance would be constant and not subject to change from one run to the next.

The hard part is the electronics to log the data. Graphing the torque vs. RPM curve would require some custom electronics that would interface to a PC. It's not really that difficult to do but it would require some time and money to develop and build.

Then there is the problem of density altitude affecting engine performance from one run to the next. The performance of the engine under test can vary significantly from a warm, humid day to a cold, dry day. The high dollar dynos measure air temperature, barometric pressure, humidity, and pressure altitude to calculate a compensation factor so that the affects of density altitude can be taken into account.

Taken a step further, a wideband O2 sensor could be added to measure fuel/air ratio to aid in tuning. That would be way cool.

Any measurement is only as good as the test equipment used to make the measurement. Something that's just thrown together that doesn't give consistent results would be pretty useless.

[ibdennyak]

jpilot
I have to bow to your experience and knowledge on dynos. The one I had was just a watercooled hydraulic pump with a valve that applied a measureable load, and a couple guages. (the most important one was temp....I had the water boiling occasionally). The procedure was to increase the load until the engine dragged down to a rated rpm.

I have to question whether what Andy wants to accomplish is worth the time and expense of the added features though, the only constant in my world is that if it isn't raining now, it will be shortly.

Denny

[uncle_punk13]

Well in the old days they didn't have PC interfaces... So going back for a moment to the idea of innovative, low-buck, in-da-shop engineering... In the machinist industry they had r.p.m. indicator guages, and these are still available I'm sure, so with that hooked up to the homebuilt roller road, and a bit of math, this could be done relatively accurate; enough to see improvements or changes anyway. Yeah with out registering/ calculating all the variables it won't be like taking your bike to a shop and having it full on dyno tuned, but you will at least be able to see what's happening with (either positive, or negative effect) tuning...
Just a thought.

[andyinchville1]

Thanks for the input so far guys....seems like maybe the best I can think of so far is to measure the amount of time it takes the engine to spin up a to a certain speed given a fairly difficult load....Of course this method would probably involve me watching a speedo or tach and using a stopwatch...not the best way because of reaction times can vary....was hoping for a totally electronic method for info gathering....Yeah the weather thing is another variable too but if I can get the machine set up I think everything else would be easy.....

[hotsauce492003]

Maybe if you use a big dial for the rpm's and next to it a big wall clock and use a digital camara to record a video of the clock and rpm dial on the same take(video shot); you could have an accurate enough record of each run of rpms vs. time and review it for comparison with other runs.

[hotsauce492003]

For the turning weight set up, what about a heavy but well balanced truck tire mounted on a shaft with a sprocket conected to your bike engine on one side and a disc brake on the other to stop it(optional).
Now if that disc brake is mounted on a swingarm that pivots on bearings on the same shaft the wheel is monted on; and you hold that arm against a batroom scale at the right leverage ratio, you could have a decent dyno of two kinds. The inertial one and the brake one.
I think that would be enough turning weight for this little engines.

Parts needed for the inertial dyno: From a Junk yard: An old heavy 16" or bigger steel truck rim with a heavy (heavier) truck tire.
Also: A shaft and bearings at each end. A bicycle sprocket. A welded frame to support the shaft/bearings.
An extension on that frame to bolt the engine on. And the a digital bike speedometer to record the tire rpms; and a wall clock and the digital camara capable of video mounted on a tripod.
Maybe would be a good idea to build a cage around the tire for safety.