Blog, DIY-Dyno

DIY Motorbike Dyno

Leave a comment

So as usual, the reason I’ve been quiet is not because I’ve been idle, but because I was working on something and didn’t want to show it till I was reasonably confident it was going to work.

Some recent developments and projects have highlighted how useful it would be to have a dyno to test and tune on. I’ve always wanted one but didn’t really have the space before and hence was going to do an engine dyno, and never quite got around to it. Here I’ve got a bit more room to work with and with the help of one of my sponsors/partners, was finally able to take the plunge and attempt building one.

It took about a month working intermittently in the rare gaps between rainstorms, and because I kept having to wait for stuff to arrive as I realised I needed this or that from ebay. But eventually it all kinda came together.

I didn’t have the spoons to make videos and as I said, wasn’t sure it was actually going to work but took plenty of pics along the way.

Getting an initial idea of how things are gonna fit together
Getting an initial idea of how things are gonna fit together
I'd intended to do a single central beam but realised this design would be stronger and make mounting the front wheel chock easier
I’d intended to do a single central beam but realised this design would be stronger and make mounting the front wheel chock easier
trying to make sure it's as level as possible on quite un-level ground
trying to make sure it’s as level as possible on quite un-level ground
front wheel chock can now sit on the rails and potentially be moved back and forth for different sized bikes
front wheel chock can now sit on the rails and potentially be moved back and forth for different sized bikes
welding up the main frame for the roller
welding up the main frame for the roller
front wheel chock sits on the to backbone rails. Currently one set of the bolts foul the rails below but I'm planning to make some modifications to fix that, and also make the chock itself a bit stronger.
front wheel chock sits on the to backbone rails. Currently one set of the bolts foul the rails below but I’m planning to make some modifications to fix that, and also make the chock itself a bit stronger.
main roller frame welded up and bolted to the backbone with super sturdy 16mm bolts like used in mounting towbars
main roller frame welded up and bolted to the backbone with super sturdy 16mm bolts like used in mounting towbars
main roller frame welded up and bolted to the backbone with super sturdy 16mm bolts like used in mounting towbars
main roller frame welded up and bolted to the backbone with super sturdy 16mm bolts like used in mounting towbars
main roller frame welded up and bolted to the backbone with super sturdy 16mm bolts like used in mounting towbars
main roller frame welded up and bolted to the backbone with super sturdy 16mm bolts like used in mounting towbars
main structure welded together. about to put a bike on it to get the right position for the front wheel chock
main structure welded together. about to put a bike on it to get the right position for the front wheel chock
tested a bike on it at this point to get the right measurements for the wheelbase, as we're mostly (only) likely to be doing R125s.
tested a bike on it at this point to get the right measurements for the wheelbase, as we’re mostly (only) likely to be doing R125s.
The roller needs a way of being locked for loading and unloading the bike as it's a liability without. We also need little ramps up and down from the roller to make loading/unloading easier
The roller needs a way of being locked for loading and unloading the bike as it’s a liability without. We also need little ramps up and down from the roller to make loading/unloading easier
welding up the little ramps
welding up the little loading ramps
welding up the little ramps
welding up the little loading ramps
checking the ramps aren't too close to the tyre (spoiler: one is, but I wouldn't find out till later)
checking the ramps aren’t too close to the tyre (spoiler: one is, but I wouldn’t find out till later)
painting the frame to protect it and make it look nicer
painting the frame to protect it and make it look nicer
painting the front wheel chock, which has since had all its seams welded, and the offending bolts moved so it sits on the backbone rails nicely
painting the front wheel chock, which has since had all its seams welded, and the offending bolts moved so it sits on the backbone rails nicely
painting up the little on and off-ramp frames
painting up the little on and off-ramp frames
painting up the little on and off-ramp frames
painting up the little on and off-ramp frames
added tie-down hoops and assembled the whole thing for the first time.
added tie-down hoops and assembled the whole thing for the first time.
made and mounted the sheet metal covers (made from an old metal shelf)
made and mounted the sheet metal covers (made from an old metal shelf)
trying to measure the MOI (Moment of inertia) of the roller with a fixed weight running through a pulley. IDK how good these readings will be but they're a start
trying to measure the MOI (Moment of inertia) of the roller with a fixed weight running through a pulley. IDK how good these readings will be but they’re a start
and made matching little covers for the on an and off-ramps
and made matching little covers for the on an and off-ramps
surprisingly it looks pretty tidy now actually!
surprisingly it looks pretty tidy now actually!
whole thing assembled including the roller and roughly in position in the workshop
whole thing assembled including the roller and roughly in position in the workshop
soldering up the two ardyno shields I bought from 2strokestuffing
soldering up the two ardyno shields I bought from 2strokestuffing
completed ardyno shield mounted on an arduino uno
completed ardyno shield mounted on an arduino uno
hall sensor mounted next to the roller with a magnet on the roller which should hopefully give us one pulse every revolution of the roller
hall sensor mounted next to the roller with a magnet on the roller which should hopefully give us one pulse every revolution of the roller
hall sensor mounted next to the roller with a magnet on the roller which should hopefully give us one pulse every revolution of the roller
hall sensor mounted next to the roller with a magnet on the roller which should hopefully give us one pulse every revolution of the roller
hall sensor mounted next to the roller with a magnet on the roller which should hopefully give us one pulse every revolution of the roller
hall sensor mounted next to the roller with a magnet on the roller which should hopefully give us one pulse every revolution of the roller
by some amazing fluke the dimensions of the frame exactly match the rails of my bike trailer. It's within a few mm. I couldn't have made it fit better if i'd measured it (which I didn't)!. I need to remove a couple of tie-down points from the central channel to let the backbone drop into the central rail properly but even the "outriggers" ought to just slot into place. The front hoop from the central channel will need removing but that's easily done.
by some amazing fluke the dimensions of the frame exactly match the rails of my bike trailer. It’s within a few mm. I couldn’t have made it fit better if i’d measured it (which I didn’t)!. I need to remove a couple of tie-down points from the central channel to let the backbone drop into the central rail properly but even the “outriggers” ought to just slot into place. The front hoop from the central channel will need removing but that’s easily done.
by some amazing fluke the dimensions of the frame exactly match the rails of my bike trailer. It's within a few mm. I couldn't have made it fit better if i'd measured it (which I didn't)!. I need to remove a couple of tie-down points from the central channel to let the backbone drop into the central rail properly but even the "outriggers" ought to just slot into place. The front hoop from the central channel will need removing but that's easily done.
by some amazing fluke the dimensions of the frame exactly match the rails of my bike trailer. It’s within a few mm. I couldn’t have made it fit better if i’d measured it (which I didn’t)!. I need to remove a couple of tie-down points from the central channel to let the backbone drop into the central rail properly but even the “outriggers” ought to just slot into place. The front hoop from the central channel will need removing but that’s easily done.
Bike on the dyno in the workshop!!
Bike on the dyno in the workshop!!

The main sections are built out of 40mm steel box-section, with 30mm braces and the smaller parts made out of 15mm or thereabouts. It’s held together by some almighty 16mm bolts, as you’d find in a towbar mount.

The physical structure is now together, and whilst there’s a couple of little changes needed, as you can see I was able to put a bike on it, the roller goes round, the bike runs true and doesn’t try to veer off to one side or whatever.

You may notice the little ramp which goes before the back wheel is missing, which is because whilst it cleared when I measured it, once there’s load on the bike the tyre deforms a bit and under load and because the ramp was quite close it would scuff the tyre a little. So I got rid of that ramp for the moment as the bike is on there for the moment. That ramp needs a little bit of editing but other than that it all pretty much worked as intended.

Because the roller is higher than the front wheel chock the bike does sit at a slight forward angle which isn’t ideal but this dyno frame design was easier to make and stronger than what I had planned, and didn’t require me to dig up my workshop floor to sink it into the ground. The oil pickup on the bike is at the front anyway so I don’t *think* it should be a problem.

I got the electronics all wired up, figured out the capacitive lead for the motor rpm signal and even got the software (simpledyno) talking to it all ok.

However there are still some things to figure out. Whilst the capacitive lead (cable wrapped round the HT lead) reads the motor RPM signal just fine, the rpm sensor on the roller works, but randomly fluctuates wildly.

So it might be doing 200 rpm and then will momentarily freak out and claim 5000000 rpm. Obviously that plays havoc with any readings on the computer and needs figuring out.

It could just be a noisy cable, or an issue with the sensor or something else. I built two of the ardyno shields so that I had a backup so might test and see if it happens with the other one. They both seemed to work the same though when I tested them after building them, so hopefully shouldn’t be any difference. I have no idea really so will just have to test my way through it to figure out what’s up with it.

Also whilst the capacitive lead from the ignition reads ok, for some reason the multiplier in the software isn’t working properly. So given it’s a four-stroke, single cylinder, it sparks one every other rotation, which should mean a setting of 0.5 in the ardyno software. I set that and it made no difference to the actual RPM reading though, so maybe I’m missing something or maybe something in the software is broken. I *think* there’s source files so if it’s broken hopefully I can fix it, as once again, we need those motor RPMs to be accurate, and I confirmed with the ECU directly that the dyno is reading exactly half the actual RPM.

Whilst it’s somewhat frustrating that it didn’t all just magically work out of the box, I was expecting the computer/electronic side to require some tinkering. The main milestone I was hoping to reach was having a physical device with a roller I could put a bike on and that the roller goes round as intended, bike doesn’t try to veer off, and nobody (especially me) got hurt in the process….which is does rather well!

There is also the question of whether the roller is heavy enough. It’s about 25kg and is a roller from a brake testing machine from an MOT station. I have four of them so if I need a heavier roller I can try filling one of them with either molten metal of some sort, or sand/refractory material, idk.

The roller weight has been my main worry all along as if this type of roller doesn’t work then the whole design needs rethinking, which would be expensive and a lot of work, but seeing the bike on the dyno I think we can probably make it work. I have three other rollers spare so if I decide to fill one up with molten metal of some kind, or refractory sand + plaster of paris, concrete or something else and it doesn’t work out, I’m not entirely scuppered.

The bearings on the roller are rated up to about 5000rpm, and at the top of third gear I was seeing values of about 2500rpm but that’s as high as I dared push it at the time.

If we check on gearingcommander.com it shows that the top of 3rd gear (with the corrected max rpm set to 10.1K rather than the incorrect 10.5K) is about 51mph, and top speed is early 80s, so there is potentially still enough leeway left in the bearings but I’m testing my way up gently.

If something in the setup lets go while it’s at full pelt that could be very dangerous indeed. A bike at full speed, or a roller at 5k rpm breaking loose in such a tiny space that also contains me would likely be game over so I’m proceeding with caution, and I would prefer the roller RPM readings be stable before I start pushing the envelope there too.

It may also be that we end up testing in 3rd gear or something rather than necessarily going all the way up to the top of 6th. Seeing the bike on the roller, the whole setup seemed to require more effort from the bike to keep it going round. than I expected. I’m not sure what the convention is for dyno testing, whether you normally go up through all the gears or just do pulls in a specific gear, but I suspect different gear ratios (and possibly rollers of different mass) would allow varying the amount of load on the system at given RPM.

The roller is also currently only partially “gritted” so the grip isn’t ideal but one can get “re-gritting kits” and I figured I’d wait till I knew it all worked and I had some idea whether the roller weight was correct before messing with that.

Overall though, so far so good, and the computer bits will just take some time to figure out, and even then it’ll take some additional time to figure out how to actually use the setup to best effect, but that was to be expected and I’m quite looking forward to working all that stuff out.

It’s also possible that my MOI (moment of inertia) measurements are somewhat out. The SimpleDyno software lets you specific the sizes of the parts of the roller but as mine came as one piece, and the values I entered didn’t seem to produce a credible MOI, so I had to fudge the values somewhat to get to the MOI i’d measured and calculated. It’s entirely possible those figures are way off, but with the roller RPM glitching and currently only reading half the motor RPMs, it’s still too early to say for sure.

We know the approximate power of the stock bike that’s on there, so hopefully once the sensors are reading steady I can tweak the settings till it all lines up a bit better.

The most challenging bit was (hopefully) having an actual physical structure with a roller to put a bike on, and that seems to be done \o/

Hopefully there will be plenty of updates to follow, as well as some videos of doing actual power runs on youtube once I’ve got all the computer stuff figured out.

I want to extend a big “Thank You!” to Tuneboss Malaysia (FSR technology) for contributing much of the cost towards materials for this project. You’ll see why this dyno is relevant over the coming months once the other secret project I’ve been working on is unveiled 😉

Leave a Reply