Teardown of the GasGas ec250f

Now that the motor for the NoGas project has been liberated from the previous conversion the fun is about to begin. But before I can start constructing and building I’ll have to remove everything gas-related from the bike..

Since I’ve never owned a GasGas before it took a bit of fiddling to get everything off the bike. First to go was the seat and tank..

Then goes the plastic parts..

.. to get further the subframe is next on the removal list ..

To make the electric motor mount as similar to the original sprocket position I took a picture to try to reference the placement when that’s up. Since the shaft on the electric motor is in the center of the motor I won’t get the sprocket as close to the swing arm without using a transfer shaft, which I don’t want to do.. so..

There’s quite a lot of stuff on a bike like this..

.. and I’m just getting started!

To get to the engine I removed the rear swing and shock absorber..

.. and once that was done there were only two more bolts holding the engine in place..

I could’ve stopped here and started making a mount for the QS motor but since the paint on the frame was in quite a bad state and the steering bearings didn’t feel too good I decided to do a proper restoration of the bike while I’m converting it.. so, off with the rest of the stuff..

Now that the frame is naked I sand blasted all the rusty areas and removed all the loose paint. Here I’ve just primed it for paint and this is the state it’ll be in until I’ve got all the welding and modifying done. Then I’ll do a proper paintjob to get it all protected from rust and whatnot..

Since a lot of the stuff I removed was in a pretty bad condition I’m getting a lot of new stuff to go on the bike, like footpegs, levers, bearings and stuff like that. More on that when I get to it. The next step is making a mockup motor since the QS is just too heavy to play around with. Then the manufacturing of mounts will begin. I’m going to try to modify the frame as little as possible but I’m counting on having to make some welding and modifying of the frame to fit the QS.

There’s quite a lot of cleaning still to do, teadious boring work that I won’t write about so next post will be making electrical stuff fit.

To be continued..

Project start, electric conversion of the GasGas ec250f

So, since the petrol engine is running lean and I’d rather make the bike electric than spend time calibrating carburetor, I’ll start the project now – a few months earlier than planned.

This is the victim. A 2012 GasGas ec250f with a 290cc kit installed. I bought it just to do the conversion but had planned on running it as a gas-bike this summer as there are a lot of other projects I have to do… but, well.. I can’t say I’m sad to have to push this project up the priority list. 🙂

I’m going to run a QS180 90h motor paired with an APT96800 controller running at 72v. The reason for chosing 72v instead of 96v is that I’ve got quite a few 72v packs already and got the chargers and BMS:es I need to make this work. There’s nothing preventing me from upgrading to 96v in the future if it turns out it’s too dull at 72v.
I’ve got a pair of 30Ah batteries that are good for around 300A a piece that I’m going to run in this bike for now. Connected in series they should give quite a lot of *umph*, and last – well, hopefully for a bit of fun at least.

The first hurdle in this project is that the QS motor is here at the moment:

Buried under a lot of stuff in the Avant tractor I converted a few years back. The thing is, I need the Avant to keep running for other projects, so I’ll need to kick off replacing the QS with a LightningRods XXL motor I bought a while back for this purpose.

I got a “special edition” XXL motor with fan cooling that’ll be perfect for the application as it’ll sit stationary under the seat on the Avant. First sub-project will be making the belt wheel fit the axle and keyway of the XXL.

To do this I need the manual mill to be running, and I fried the motor on that one a while back milling unsuitable materials.. In the last post I repaired it with an 800w motor, all that was left to do now was to set it up properly. After a bit of fiddling I got it square enough and set the wheel up in the mill chuck I got just for this purpose, but when doing the first conversion.

Since I don’t have a DRO on the mill (I’ll have to get one later) I’m using a live center to center the part on the mill. I locked down the table to prevent it from moving and bolted down the chuck while keeping it centered with the quill.

Then all I had to do was drill the hole to just under size and ream the hole to size. 12mm hole for 12mm shaft. Perfect!

All that was left to do now was the keyway and I’ve got the tools for that.

Unfortunately I didn’t have the sleeve for 12mm holes, and I wanted to make the keyway now… so, sub-subproject: Making the sleeve.

Just had to find a 12mm axle and a tube with 12mm ID. Welded them together and set it up on the mill to make the proper slot.

I don’t know if the reamers are ment for 12mm holes, but there’s not a lot of material left after milling the slot in the sleeve. It’s quite hard to hold on to while taking the last few passes, but it doesn’t have to be perfect.. Good enough is .. good enough.

Having made the keyway and made sure the wheel fit on the XXL motor it was time to disect the Avant.

I’ve got to service this machine, it’s been running for years without doing anything to it, and it’s leaking hydraulic fluid from the filter.. That’ll be a later post.

After disconnecting all the wires, removing the battery packs and the controller I could finally get to the motor. Seems I did quite a proper job converting this machine as it’s really servicable. I am however going to try to fit three battery packs instead of the current two when I put it all together to get it running a bit longer. I’ve got two spare packs sitting on a shelf doing nothing constructive at all, better to put one to use.

Having removed the QS motor from the Avant I just had to make a size comparison. 🙂

The smallest one here is the Lightning Rods XL motor. It’s a bit more powerful than the BigBlock I’m using on the runbikes, and those do over 110km/h and are insanely powerful.. super fun toys! The XL is supposed to push about 50% more power than the BigBlock..

The center motor is the XXL that’s going into the Avant. I don’t have the proper numbers for it but it’s a larger diameter motor and quite a lot larger overall. That should give quite a lot more torque and it can push quite a lot of power. The Avant averages about 5kW, so this should be plenty.

To the right is the QS180. It’s a proper beast! I’m super stoked to see what that motor can do on a bike!

After having measured the offset of the belt wheel on the axle I set it up on the mill again to bore the hole larger to fit the retaining clip on the motor axle at the right height.

Time for a quick test fit and then CAD to make a conversion plate between the welded QS-motor bolt pattern and the XXL.

The first mockup is on the 3D-printer right now. After test fitting to see how it fits I’ll mill it from aluminium..

To be continued..

Bike season closing in

Today we finally got some sunshine and took a ride on our RunBikes on the snowmobile tracks.
Since we’ve had super cold weather the last few weeks the snow hasn’t gotten hard, so the tracks are super soft and tricky to ride. For february the conditions are super though.
We did a few minor (and some a bit more than minor) crashes but no injuries and apart from a lost rear fender the bikes stay together and run great.
Unfortunately I got a crappy angle on my gopro camera so the video sucks, but I managed to get a few good clips from some of the crashes.. so when I get around to it I’ll post something here.

Super nice to start riding again, let’s hope for a great spring this year and loads of riding!

CNC plasma CAD files

On popular request I’ve exported the CAD to STEP format for sharing. I’ve done a quick cleanup of the file and all parts named _NOTUSE are not to be used. Most of the custom parts I’ve milled from aluminium but a good 3D print should work just as fine.

If you download and use the file please look at it as a template and feel free to improve wherever it needs improving. My plasma is running with a CUT60 chinese plasma cutter and working fine.

Also, if you build the thing I’d be happy to hear back from you. Please post a comment here or on my youtube channel, https://www.youtube.com/@marcusrunsten and post a link to your build.

All the files and the code for the project is on my github: https://github.com/marcusrunsten/cnc_plasma

The second bike, progress

So, the footpegs I put on the bike were cheap, but not awesome..

They were flimsy and didn’t give any confidence at all. They were also too close to the frame, and too far forward.. So, new solution:
I got some proper footpegs at a great price, so I bought two pairs. These are a lot sturdier and more like the ones I had on my rally bikes. I’ll be mounting them about at the same position as where the BB would sit if I hadn’t put a custom subframe and footpegs on the bike. Feels a lot better when trying them out at that position.
Today I welded some mounts for the new footpegs. Hopefully I’ll get time to finish them tomorrow so that we can try them out. Since the frame is all powder coated and nice these will be bolt on, after finishing, sanding and paint of course.
So, this is about how it sits at the moment. Not too much left to do. I’ll have to wait for the new 99T rear sprocket and will try out a custom milled 11T front sprocket. It’s milled from aluminium at the moment which probably won’t last. When it fails I’ll either put an 12T I bought on ebay on there or I’ll mill the 11T from steel. I tried to make an 10T but the 12mm shaft is just too big, there’s no material left around it with a 10T sprocket.

Also, I made stickers since I kind of like the black/grey color scheme.. What do you think?

When the bike is finished I’ll post a youtube video of the entire build. I’ve also got a lot of questions about what components I’m using and the total cost for the build. I guess I’ll put a BOM together and take on the dreaded task of figuring out how much $$$ these two bikes have cost me.. :/

To be continued..

Finishing steps and final assembly

So, after finishing the sand blasting of the frames we took a break and celebrated easter in the cabin with some nice snowmobiling and relaxing.

I spent more or less every night in a full week trying to get the paint off the frame. Considering the poor quality of the original paint it was surprisingly hard to remove.
When we got back from the cabin my dear friend Nils had put new powder coating on all the parts with a stunning result!
I must say I agree with my son, this frame looks much better in black! However having the nice looking frame I couldn’t mount the raw aluminium parts so I had to do something about that.
After properly degreasing and cleaning the parts I let them soak in lye for half an hour or so.
After the lye bath, inhaling loads of healthy hydrogen atoms in the process, the parts were properly rinsed and then the anodization process began. This is simply a bath of sulfuric acid and an adjustable power supply. The parts are submerged in the acid, the negative lead is connected to an aluminium sheet with a surface area equal to or larger than the part, and the positive lead is connected to the part to be anodized. This makes an oxide layer on the part that is a bit porous. The porosity and thickness of the oxide layer depends on the current flowing through the part and the time it’s left bubbling in the acid bath.

Since I wanted a thick uniform layer of color I let the parts anodize for about an hour at quite low amperage. The required current depends on the surface area of the part to be anodized but I usually kind of just guess and mostly it turns out OK. If not I’ll just redo the process from the lye bath and try again with different settings. =)

When the anodization is done I’ll just put the parts in a vat where I’ve mixed water and textile color, in this case black. The thickness and porosity of the part, the properties of the dye and how much color you want the part to get decides the temperature of the dye and the time it’s left soaking. Since I don’t know the properties of the dye and want the parts to be fully colored black I left the parts in the dye for as long as it took to clean and anodize the next part. This again -usually- turns out OK, but sometimes it doesn’t and then it’s back to the lye again..

To seal the porous oxide layer the parts are boiled for about 15 minutes – again depending on the properties of the oxide layer and so on. I’ve precviously had the water wash some of the dye out of the part so now I boil the parts while still in the dye bath.

It’s quite a time consuming process but the result makes it totally worth it! When I had the anodization going in the garage I anodized all the parts for the second bike as well. I’ll post more about that frame later.
Now that all the parts and the frame had the right color it was time for the final assembly. First the shock was installed to link the motor mount and rear swing to the frame.
After that I installed the updated motor and temporarily put the seat on…

This was to be able to test the fit of the 3d printed rear fender, which was too small and required a couple more prototypes before I had a good fit.

Once it’s there it looks pretty good.. (but later, after adding the aft part it looks a bit wierd, so a redesign of the rear fender is on the todo-list)
Front fork, wheel and handlebars installed..

Rear wheel, chain and so on..

Now we’re getting an idea of what the finished bike will look like..
Time to install the electronics. This one will run the ASI BAC2000 with the High Voltage map for the BigBlock motor.
The battery and all the charging- and powerleads are installed. To hide all the cables and connectors the cover that’s supposed to house the controller in installed. The BAC2000 is way to big to fit in the controller housing so it’ll reside inside the battery compartment.

A while back I got lazy and decided not to make my own footpegs, so I ordered footpegs from Amazon for cheap. To mount these I made steel brackets on the manual mill and painted them black.
The side covers are mounted. I’ll remove these later and make a proper seal between the cover and the frame to make it waterproof. When doing that I’ll also install a drain tube at the bottom of the battery compartment to get rid of all the moisture and water that might make its way into the box. No matter the steps you take to make the battery compartment waterproof water _will_ get inside and it’s important to let it escape somewhere..
I’ve made a few attempts to make a front fender but for now I’ve decided that this part isn’t necessary. 🙂

Instead, for the time being, I wrapped the cable cover box in protective vinyl wrap. This’ll make it less sensitive to flying pebbles and sand.

One of many concept prototypes for the front fender. The CAD model of this bike has been improved all through the project and now it’s a real good reference to get measurements and model parts into.
The rear of the bike gets an inner fender to prevent all the water and dirt from sticking to the riders behind. I’ll make a fender that’ll sit closer to the wheel, protecting the motor from the muck getting tossed around by the rear wheel later. For right now the new seal on the motor will have to suffice to get some test riding done.
Finally with the rear fender added the bike is pretty complete. There are parts to add and I’ll have to get rid of the Amazon footpegs and make my own since these were ridiculously flimsy – but they’ll work for now. I’ll also have to add a chain tensioner and I might have to make a smaller front sprocket to make the bike slower.. 🙂

Re-assembly of motor and frame preparation

So, since the magnets in the rotor was super strong I’ve been a bit nervous about re-assembling the motor, but I got an idea..

With three M5 all-thread rods and a few nuts..
Threaded the rods into the top part of the motor and added nuts at a similar height on all rods..
These would then support the stator allowing me to gently let it all the way down over the rotor without slamming together.
To further protect the rotor I inserted a sheet of OH plastic in the air gap between rotor and stator. This is totally unnecessary but still..

So, now it was just a matter of easing the stator down all the way. Most of the way I could just turn the nuts by hand but it’s the last centimeters that are hardest to control but the three-nut-solution made it super easy.

I let the motor sit for the liquid seal to cure for a day or two before adding some conformal coating to the hall sensor circuit board and to the thermistor wires to keep them out of the way of the rotor. Then I added the bottom cover and it was all done. Now I just need to do the same to my second motor.

Frame preparation

Before rebuilding the bike I wanted to get it painted and since a friend of mine has a complete powder coating setup we’re going to powder coat the entire thing. To do that I need to do some preparation to the frame and since it’s a bit cumbersome I decided to powder coat both my frames at the same time. The paint that’s on these frames are sketchy at best..
I’m going to make a video of the entire build of this second bike. The first step in preparing the frame is plugging all the holes in the battery box. I simply welded all the smaller holes and made inserts for the larger ones.
I drilled and tapped the mounting holes for the motor mount.
After doing a bit of cutting and welding the atmosphere in my workshop is toxic, to say the least. I need to build a filtering solution to be able to breathe while working.
After welding the subframe the second frame is ready for sand blasting and paint. I decided not to weld the mount for the rear shock onto the frame. Instead I’m going to make a bolt-on solution where I make a mount that bolts onto the frame at the right height. This will make the build and installation easier and I can drill the holes later when assembling the bike.
So, now all the parts are ready for sand blasting..
I’ve never used a proper sand blaster before but it turns out some of the paint that’s on the frame is real tough to get rid of. I spent a good 4 hours last night trying to strip the paint from all the parts and didn’t even finish two parts. Today I’m going to rough the paint up with a grinding wheel before sand blasting, hopefully this’ll let the blaster work better on the parts.

To be continued..

Bike repair and improvements

So, after testing the bike for a few trial runs it stopped working. Connecting to the controller I got an error saying “Post static gate test” which usually means there’s a problem with the controller having a mosfet fused to ground or the postitive terminal.

So, a bit displeased I started taking the bike apart but before putting in an RMA I decided to test another controller I’ve got sitting around, and got the same error with that one.. so, the problem seemed to be with the motor, which was a bit wierd. Connecting my second motor to the controller confirmed that the motor was the issue since it ran without any problems.

After doing some measuring I was a bit confounded as all the resistances seemed to check out and sure enough when I connected the motor windings to the controller it all worked. However, when I connected the hall sensor array it stopped working again.. Hmm..

I just had to take the motor apart to find out what’s wrong with it, and on opening it up I found a bit of water inside that’s not supposed to be there. After drying the motor the error on the controller disappeared again and everything seems fine.. although, where the motor sits on the bike, just in front of the rear wheel, I’ll need to make sure water cannot get into the motor. Even though I’m adding all kinds of fenders the motor is going to be suscepted to moisture and water..

Taking the motor apart the seal that sits between the stator casing and the end bells on the motor crumbled to dust which is probably why the water got in there in the first place.

I decided to take the motor apart properly to check for damages and it turns out there was none. Getting the rotor out of the stator was insanely hard though as the magnets in the rotor is immensely powerful on this motor!

If you get your fingers caught between the stator casing and the end bell on the motor I guess you’d have to leave your fingertips inside the motor. I’m not looking forward to putting it back together again.. I’ll have to come up with some contraption to make it not slam back cause I reckon that’d make the motor self destruct and take half of the garage, both my hands and my left liver with it as it goes…

As I had the motor apart I thought I’d add a thermistor to be able to have the controller reduce power if the motor gets too warm. The Lightningrods motors come with a thermistor if you ask Mike to install one, which I happeded to forget..

The normal way of adding a thermistor is to make a hole in the case and route a cable from the thermistor to the controller through there but I didn’t like the idea of adding another cable to the motor. Two is enough I think. So I instead opted to replace the 5 lead cable that goes to the motor originally with a 6 lead cable instead. To be able to fit the larger cable I had to make a new grommet for the motor casing though.

Fist a bit of CAD, and then I printed it from solid TPU.
Turned out OK and pushing the cable into the grommet was SUPER hard, so it’s a tight fit.
The grommet is a great fit in the motor too so it should make a good seal. I cut the original cable and soldered and insulated the new cable to the old, matching the colors as good as I could.
I also soldered the thermistor between the negative lead and the new brown lead (colors for future reference by me) and went on with tidying everything up.
I used epoxy to secure the thermistor to the motor as close to the rotor as I could and tucked all the wires away neatly. Taking some measurements everything seems to be in working order so the last step was to add molex crimp terminals to the other side of the added cable to fit the 6 pole connector for the controller.
All done for now. Tomorrow I’ll try to get the rotor back inside the motor. I’ll add liquid seal to the end bells before putting it all together again hopefully keeping the water outside the motor for the future.

Taking the motor apart I saw that the conformal coating on the hall sensor PCB was missing in part, probably allowing the problem to occur in the first place. I used epoxy to cover the PCB to hopefully make the motor more resilient to failure in the future even though some moisture might get inside.

Well, to be continued..

New battery for the Radon Swoop

So, after having used the current 20s5p molicel pack for a couple of years it decided to give up during a ride in the snow a few weeks ago. It suddenly lost voltage and would not recharge and when I took it apart two cell groups were sitting at 0V. As I had taken a shortcut bypassing the 60A chinesium BMS to get the power output I wanted from the battery I guess this was bound to happen. The BMS at least prevented me from charging the battery when any cell group were at a bad voltage, so that’s good.

I took the pack apart and put all the 18650s through a thorough testing program and capacity measurement. All cells in the 18 cell groups that were at an OK voltage were good but the cells in the bad groups were at very low capacities if they took charge at all.

Before this ride I’ve had the bike sitting untouched for several months so I guess it’s gotten slowly discharged during this time and when I took it out for a ride pulling 100-ish amps from cells already low at charge it eventually killed them. (This is just me guessing at what happened).

As the battery box I was using was a bit cumbersome to build and mount due to me trying to make a quick-change battery at the time I decided to design a new pack from scratch.

Fitting 100 cells in the swoop frame is tight fit, so I had to make some prototypes before finding the best shape to get an as-decent-as-possible battery layout that’d actually fit (and be able to mount) in the frame. To make it as sturdy as possible I could not fit the side covers before putting the pack in the frame so it’s tight.. 🙂

Since I’ve upgraded to the BAC2000, pulling A LOT more current from the battery (this might also be part of the reason for the failure of the old pack) I built the new pack as a copper/nickel sandwich that should be capable of providing all the amps I need. To save some money I only replaced the bad cells with cells I already had at home. Doing this it’s important to make sure the “old” cells aren’t mixed with new cells as the drain profiles might differ causing individual 18650s to see more load than the average cell which could lead to failure.
To give the pack the best chance to survive I installed an ANT BMS capable of 120A cont and 300A peak. This allowed me to route the power through the BMS even for the consumer and letting the BMS kill the power to the bike if any cell group goes low. The BMS can be monitored via bluetooth so I can keep track of the health of the battery too which is a nice feature to have.
The pack looks nice and being a fair bit wider than the old pack doesn’t matter. As it sits it’s not in the way of anything when riding. The pack gives a good punch when slamming the throttle and the bike wants to lift the front wheel all the time which is awesome.

Hopefully this pack’ll last a couple of more years before needing a rebuild again.