Hey Guys,
I’ve posted some pictures of my Yanmar dyno on photobucket. Basically the engine is connected to a hydraulic brake from a Honda Civic. The disk and caliper ride on the same shaft, but the disk is attached to the shaft with a keyway, while the caliper mount has a bearing that allows it to stay stationary. The brake caliper’s mount also has a lever that extends 12 inches from the center of the shaft and is attached to a 50lbs fishing scale which stops the caliper from rotating. Whenever I apply the brake, force is exerted on the scale giving me ft/lbs. I apply braking force until the engine loads down to the RPM I want to measure HP at. Since I have a tachometer mounted above the scale I know the ft/lbs of torque @ RPM.
Ft/lbs X RPM X 5252 = Horsepower
I’ve made an initial test run which worked pretty well. I may replace the brake lever with a screw type mechanism from a large C-clamp. This would allow a more gradual and steady increase in load, making it easier to hold an RPM. Also, I run the shaft at ½ the engine RPM to decrease the likelihood of an exploding brake disk and so I can use a bicycle speedo for a tachometer. I have not been able to get a speedo/tach to read higher than about 2100 RPM. I'll be adding machine guards soon.
Thats pretty cool , what kind of numbers did your engine make ? I have a tiny tach for diesels and it works very well , also keeps hours . now get the supercharger working and a tach that will display up to 3600 and watch the torque climb . nice work .
I did a full test run today. I'm not sure if I have the right number programmed in the speedo/tach. I punched in 167mm for the tire diameter using k/m. Take into consideration that the shaft being loaded and RPM measured is rotating at half the engine's speed, so its RPM is half of and torque reading double that of the engine. I've converted it below. Regardless the calculation is the same. Anyway, the torque readings reflect the factory curve fairly well, but the horsepower does not. I think the RPM may be off. May have to pull the Sendec tachometer off the bike. I'm also welding up a smoother, more gradual way to apply pressure to the master cylinder. Finally, too much vibration reaching vaccum gauge. RPM x Torque / 5252
For some reason I have an inverted horsepower curve. Although its hard to keep the RPM steady under load, I think this may have some other primary cause. Overall very excited. Any input is welcome. Thanks.
Last edited by dieselbikin on Fri Jan 22, 2010 10:58 am, edited 1 time in total.
I wonder if the engine is not getting enough fuel for higher RPMs and therefore dropping off power. My book on turbocharging diesels says that naturally aspirated diesels should make a little black smoke at full throttle/under load. Mine really doesn’t. Next time I get a chance I'll check and share the results.
Thats absolutely superb. I love it! Great ingenuity at a fraction of the cost of a commercial dyno!
With a couple of scavenged transducers out of things like bathroom scales I could imagine how that could work interfaced to a computer to draw nice graphs. There are a few options for other tachometers if you don't mind a bit of soldering too, at higher rpms the sensor (reed switch) used in bicycle speedometers may start missing pulses?
Is it easy enough to hold the engine at the rpm you want? I'd imagine the brake would heat up pretty quick and need a change in holding pressure to keep it about right. I wonder if an ABS modulator could be adapted to electronically control the braking force...
Great stuff.
N
Smart engined 800cc turbo diesel triumph tiger. 100mpg (imp)
Belfast to Kathmandu overland, 2010/2011 - http://www.suckindiesel.com
Bangkok to Sydney ???
Wow. Nice job! I'd really like to see what kind of numbers the Yanclone and other engines produce vs their advertised HP and torque. I can see that your dyno's biggest merit will be for checking the results of engine modifications. Some people may scoff at the friction brake but that was the way they used to test huge steam and gas engines, simple and very effective. I don't think you'll overheat the rotor or caliper with 10 HP, but you could always devise some kind of cooling, maybe water mist using compressed air/water?
I was thinking of making an electric dyno a while back with a DC generator operating into a large resistance, that could run direct up to 5,500 RPM what the generator was rated for. It was a good sized 24 volt military generator (have a 60 amp output, a 100 amp and a 400 amp) and I was going to fit it with a variable field current and output load resistance to vary the "braking". It would be nice to have the whole thing computer controlled.
When I was fooling around with hot rods and racing cars we used to go to a dyno shop to test engines. Theirs ran a big water pump (a "water brake dynamometer") With about 300 horsepower churning into it the water was steaming, they had a series of 55 gallon drums out back to flow the water through, sort of as a radiator I imagine. It was a "factory built" dyno with many modifications as I remember. The hot water would overflow all over the driveway, eventually flooding into the old run-down garage it was in. About 15 years ago there used to be a small shop in the town across the river that was part of American Bosch (only about 10 people left out of a peak of about 2000 employees that were in the Chicopee, Massachusetts plant) and they designed and tested fuel injection pumps there. I had the cook's tour and they had a nice computer-controlled dyno that I think could handle about 2000 HP! I didn't get to see much of it as I was more interested in looking at a couple of test engines they were selling. Unfortunately that place has packed up and left along with the rest of the industry around here.
I wanted to either build a hydraulic or water brake dyno, but the cost was too high for me. I may build one later. I would like to be able to load the engine for longer periods of time to measure EGT , supercharger bearing temps, ect. Any information on interfacing with a computer would be great. The generator dyno sounds interesting too. I appreciate the input.
Interfacing to a computer should be do-able. What do you want to link up so far - brake arm force and rpm's? The ABS ring off a car with sensor would give very accurate rpms...
Cooling could allow longer running at load as smokeyjoe says? I read on another website someone talking about the possibility of submerging the brake rotor in oil, not sure how that would affect things.
n
Smart engined 800cc turbo diesel triumph tiger. 100mpg (imp)
Belfast to Kathmandu overland, 2010/2011 - http://www.suckindiesel.com
Bangkok to Sydney ???
I scanned a good article to photobucket that tells how someone built a water brake dyno. Most of meat of article is in the last two pages and the pictures inside the article. Several points of interest to me were how little water it seemed to need, assuming the wash tub underneath was the only water involved and that you didn’t have to know the horsepower per say, you just had to able to consistently measure an increase/decrease in power. For example, the centrifugal pump pressed against a hydraulic piston. The harder the pump pressed against the piston the higher the pressure reading. Even if you couldn’t calculate “horsepower” with 300 psi @ 1500 RPM you would know you had increased the engine’s torque if it could only generate 275psi @ 1500 RPM before you made a modification. Dude,my engine's pullin 350psi @ 3600RPM!
All interesting stuff. To capture the data on a pc the only thing you'd really need is 2 output voltages from your dyno. These would be torque (or pressure for the pump based ones) and rpm's. You could perhaps capture rpm via a simple dc electric motor (though you'd have to calibrate it). I notice one of your pics has a load cell connected to the torque arm to measure the torque force which would be ideal.
These voltages would need to be scaled suitably using a simple circuit with a couple of resistors before being fed to the computer. The simplest way to get the data into the computer is via the sound card line-in jack. Depending on the soundcard a little more scaling might be necessary but this would all be done in the software and could be calibrated for horsepower/KW/whatever with enough faffing about. I could write the software for this. If you wanted to capture more data than just power it might be necessary to build an interface as the soundcard route only gives you 2 channels.
I like the idea for the water brake dyno. It would be handy to be able to hold the engine at a given load while you can play with various settings.
n
Smart engined 800cc turbo diesel triumph tiger. 100mpg (imp)
Belfast to Kathmandu overland, 2010/2011 - http://www.suckindiesel.com
Bangkok to Sydney ???
Just remembered soundcards have a filter on them that means they can measure AC but not DC. Can be overcome a few ways. An ABS ring and sensor would work a treat though for RPM's...
N
Smart engined 800cc turbo diesel triumph tiger. 100mpg (imp)
Belfast to Kathmandu overland, 2010/2011 - http://www.suckindiesel.com
Bangkok to Sydney ???
This is average of 3 dyno runs.
Header/K&N/Light black smoke under load
Dyno Sheet Average
RPM HP
3600 8.9
3000 8.4
2500 7.5
Replaced lever with screw from C-clamp for better pressure control After breaking the hook off of the scale twice I replaced it with a bolt.
I'm still not completely satisfied with the dyno's ability to hold a steady RPM. A friend suggested putting some transmission or brake fluid on turning rotor to crystallize the brake shoes. He said the shoes would lose some of their friction or grip, but would be more consistent or have less “pulse” as they got hot. If that doesn’t work I may use a power steering pump out an old Ford truck an engine brake. He suggested that vehicle because the Twin I Beam suspension requires more force to turn. I could run the pump through an adjustable pressure valve with a bypass back to tank. Hydraulic oil dynos are supposed to stabilize the fastest. However, keeping the expensive fluid cool would be an issue. A water brake takes longer to stabalize, but uses free water. We'll see.
Overall very excited. 
Any input is welcome. Thanks.
