DIY Dyno
Geoff's homebrew dyno page.
Update: 3/2/04
Added new page of dyno graphs and results.
Gave Greg Percival's FZR400 racebike a good thrashing, which proved inconclusive due to lack of time. While I had help, I loaded the BMW (K100RS-16v) onto the dyno. An interesting exercise -
what a pig - this thing weighs tons, and the new ones are even heavier. BMW need to put them on a diet.
Not much power, but a nice flat torque curve.
slight change in power as the runs continued. Engine temperature was approximately constant, so it might just have cleared it out after lots of slow running around town. Greg commented on a puff of smoke on the first run.
Dyno calibration needs recalculating and confirming when I get some time (Away in Christchurch at Uni AGAIN!). Reads consistent but low. BMW was reading around 64hp, and should be more around 75-80. This would also be in line with Greg's runs on the FZR vs. how it performed at the Nationals the following fortnight, and my run on the RD350LC when I first installed the software. As this was my test hack originally, I have a fair amount of data on it. Unfortunately due to the kludgy way that Andy's software worked with inertia, I can't transfer the figures across directly, and the added flywheel is another change. This won't take long to sort out if I can get some time together.
The high power Dyna ignition coils on the FZR caused lots of spikes in the ignition pickup readings. The BMW was a lot cleaner.
Update 3/10/03: I have now got Performance trends Datamite up and running on the dyno, and a damn fine thing it is too. More on this when I get some time to update the piccies.
Introduction
The following page will cover some of the things I have found out while designing and making my own rolling road dynamometer for motorcycles. In theory, much of it will apply to a car dyno as well, but due to the higher horsepower requirements, cost and size will increase substantially.
Excuse some of the fuzzy photos. I have just got an old digital camera, and they take a bit of getting used to.
Is it worth it?
Think carefully before you embark on this voyage, lest you to become bald like me. It has it's frustrations, and it is not cheap. A large drum or flywheel is expensive, and so are the rest of the parts. I did it because
I could
Dynojets are rare and expensive in New Zealand. They are relatively common and cheap in the US, so in many cases, it is better to buy something "ready to go" rather than DIY. This goes double if you don't have an engineering shop or the skills to do the fabrication.
There aren't many other dynos to choose from here. Most second hand dynos in NZ are old Sun and Vane brake dynos, which are large and require a mounting in pits in the floors. Some of the water brakes need a cooling tower as well.
My total (not including my fabrication labour) for this project so far is around $NZ5-6000 over 8 years, or $US3500 or so. Cheap if you are serious about development, but an expensive toy if you don't use it.
First step: Read Blair's SAE article and the TDK karts page before you start.
Introduction
I have had a hankering to make an inertia type dyno for a long time, and this is the cumulation of a lot of work and contributions from all over the world. i have had advice from the US, UK, Spain, Australia, New Zealand and a lot of other places I have no doubt forgotten. I would like to thank all those who have helped me with this project. In fact, it is a testament to the power of the Internet, as it couldn't have happened without it.
The original idea came from John Fitzgerald of Thunderbike Engineering in 1996 when he described to me via e-mail how his home brew dyno worked. In his case, he had modified a waterbrake to inertia type by adding a large flywheel and disconnecting the water pump. I went called in to see him later that year and he showed it to me and how it worked, and did a couple of runs (on a Triumph triple IIRC). The idea got canned for a couple of years after that, as I went back to university, and hence had no time or money. I still kept an eye open for ideas, and was trying to find a cheap and elegant way to collect the data.
Towards the end of 1997, I finally got things underway, by ordering the timer and a copy of the dyno software card from Andy Cree at Canterbury University, which he had originally written for John. While in theory, the computer program and data acquisition are easy, according to the experts, I have trouble programming my VCR. The computer side is the key to a successful inertia dyno and it is what you really pay for when you buy a Dynojet, etc. Fortunately, with the decreasing cost of computer hardware, the availability of general PICs like the BasicStamp, and more people who can program the black box, the prices are dropping.
Mechanically, fabricating the dyno has been relatively trivial. I subcontracted out the fabrication of the drum and dynamic balancing to Titan Marine Engineering, Railside Ave in Henderson, Auckland, New Zealand. If you use a chain or direct drive with a flywheel, then this job is a lot easier and cheaper. The problem I had was finding a workshop with sufficient capacity to turn the drum after they had welded it together - there was no way it would fit in my Colchester. With a flywheel, then it can be done in a smaller lathe with a gap bed. The welding, fabrication and painting of the frame took a couple of weekends.
MyFZR400 on the dyno (before flywheel was added to the drum. A ratty Formula 1 muffler gave a consistent 1.5hp gain over the shiny Pipemasters can.
Design objectives
One of the objectives of the design of this dyno is that it had to be reasonable portable. At the very least, it had to be able to disassemble so that I could pack it away in a corner when it wasn't in use, as I am rapidly running out of space in the workshop. I have done this by making the frame that supports the front of the bike unbolt from the drum assembly. The carriage that the front wheel straps onto to lock it into place also unbolts, as does the front support frame. The dyno drum unit is on wheels so I can roll it away.
Description of parts.
The dyno drum is fabricated from 500mm steel pipe with a 75mm shaft through the centre. I have found this is a little light for 70hp+, so I added a flywheel to one end to increase the inertia. In your case, I suggest making the end plates of the drum from thicker material (eg 100mm) to achieve the same end.
The frame of the dyno powerhead is 75*75 RHS mounted on wheels. The frame could actually be smaller and lighter, but I wanted to overbuild it, as having the frame collapse as the drum is doing 1700rpm would really ruin your day. The wheels aren't fitted to the front of the dyno frame at the time the picture was taken.
The front frame is 50*50 RHS with 100*100 angle to provide a guide for the front wheel to run in up to the carriage, which also slides back and forwards in it.
The carriage has a screw adjustment, but you could do it by undoing bolts and sliding it to suit. The screw adjuster (pics below) makes it a lot easier to set up by yourself, as you can strap the bike down and crank the bike along to suit on top of the drum. An electric motor would be nice here, as there is a a lot of cranking to get the carriage from one end to the other. I didn't put one on due to increasing weight (semi portable remember!) but if you had a permanent setup, it would be easy to do, by using a motor and gearbox instead of the crank handle.
As noted above, I suggest you make it heavier / more inertia by increasing the size of the drum and/or having a flywheel or thicker endplates. Dynamic balancing after machining is a MUST - it will take off on you otherwise. My drum has no vibration at all at full speed. It has 2 weights welded onto the endplates to balance it.
I have added a 560*75mm flywheel (top of page) onto the end, which has made a huge difference to the inertia (currently around 12kg.m^2). This was a press fit over the shaft and welded to the drum. I didn't have to rebalance the assembly after welding thanks to careful fabrication.
Other changes...
Aside from a motor to wind the crank (not actually much on an issue), the other changes I would make for a permanent installation include a starter motor (for race bikes with no starter) and a brake. The starter is almost a must have if you are doing a lot of race bikes, the brake is nice but not really an essential.
An important point with brakes is to make sure the pads don't drag on the disk or drum when it is running, or the extra drag will lead to strange results and incorrect data.
if I made another drum, I would make it solid steel. This would be easier to fabricate, and be a smaller diameter to make the dyno easier to load. The choice of the larger drum was to reduce tyre slippage. Changing from the grooved drum to a heavy knurl should reduce this in any case.
Tools
Besides the drum, this project is well within the reach of a good engineering shop. I used the following tools:
Arc welder
large drill press (capable of 20mm holes in steel) and drill bits
Tractor and front end loader to lift the drum into place (no chain block yet)
Lathe (optional - I used it to make the axles for the wheels and to turn the ends of the threaded rod of the carriage screw, but could have done without it)
1/2" hand drill
Various hand tools, clamps and a level
Skilsaw and jigsaw for making the boards on the top (not shown in the photos above)
Data Acquisition.
The data acquisition system is the key to the dyno, and is what you pay for with a commercial system. The drum speed is via a data acquisition card and a magnetic switch on the drum. The card also has a handheld trigger to start the run if you can't reach the keyboard
I originally bought a turnkey card and software from Andy Cree for $NZ800 (approx $US400). It was very basic but workable. Importantly for me, it was a ready to go system - I am no computer programmer. Main bugbear - it doesn't pick up engine RPM directly, rather it calculates from the ratio that you enter between drum speed and engine rpm. DOS based means the computer hardware is cheap.
Later, I bought the Performance Trends Datamite, (also used by TDK) with the inductive pickup and optical isolator. I recommend it as a good, well made package. It comes with an instruction manual which covers pretty much everything. A more dedicated review in due course.
Ontrack digital do a data acquisition that has been used on at least one dyno that I know of. it picks up the ignition pulses from the HT plug lead to measure the drum or engine rpm.
Factory Products sell retrofit kits for their eddy current dyno
Hyperpower also may be able to help.
Land and sea dynos
DynoSoft KD (Belgium)
Steve Muller's home made dyno in Australia - uses basicStamp
Odds and sods:
Read
Reading dyno charts for faults
If your software doesn't do temperature corrections, look here