Single Gear Stealth Build - Battery in Frame Tubes

[LawrenceEaden]

@i2000 Kinda. I'm going to put some 1mm plastic on each face of the bag to fix components to. Still yet to be fully resolved the easiest way to do this and keep the bag accessible and be able to take it off if I need to.

The controller has an ignition switch wire on, so the switch won't be turning the battery on and off.

Thanks.

 

[LewTwo]

The controller has an ignition switch wire on, so the switch won't be turning the battery on and off. Thanks.

Then something like this would be less challenging.
http://www.ebay.com/itm/271913588633

 

[i2000s]

@i2000 Kinda. I'm going to put some 1mm plastic on each face of the bag to fix components to. Still yet to be fully resolved the easiest way to do this and keep the bag accessible and be able to take it off if I need to.

Thanks.

Maybe a "fiber fix" tape with some cling film could make a hard case in a cheaper price. But it's not easy to be untangled if you want to remove them... Anyway, all looks promising for now.

 

[i2000s]

The controller has an ignition switch wire on, so the switch won't be turning the battery on and off. Thanks.

Then something like this would be less challenging.
http://www.ebay.com/itm/271913588633

Is the working current too low?

 

[LewTwo]

Is the working current too low?

No, if it is a "soft switch" (TTL level.). Worse case is it provides battery power to the board power supply which should still be minimal current at 42 volts DC (max).

 

[LawrenceEaden]

Yes that's the kind of key switch i'm going to get. I'm trying to find a nice metal one.

I was considering the controller heat sync and thought why not just modify the casing it came in.

One side already has the holes to screw into the metal block the FETs are mounted to, so it's ideal and a quick job to cut and chamfer the ends to look a bit better when mounted on the outside of the frame bag.

It's quite small, so I'm not sure how effective it will be, but the vanes will be positioned so lots of air should go over them as I ride.

I've been revising the packaging of the frame bag to try and get it as small as I can. This is a template on the frame to help me understand how it's going to look.

I'm happy with the silhouette on the bike. I think even with the bag it should still be in the 'stealth' category. I've bought a second hand leather jacket on ebay (£9) to make the bag from, I am still on a budget making this.

The colour looked very similar to some leather seats and bar tape I've seen before, so i'm hoping I can get some that are close to matching. I'm now trying to choose the rest of the colour scheme. Currently a gunmetal frame with a dark rust colour for the wheels is prevailing. I may do a few photoshop renders to help decide.

 

[LewTwo]

As long as you have the extra leather material have you considered recovering an old seat .. perfect match. Might also be able to form some hand grips as well by glueing the leather to some neoprene rubber or cork sheet (gasket material) and making a tube.

 

[jimbo7890]

Lawrence - awesome build, really inspiring. May I suggest that you finely sand your frame cut smooth. Although I see the FEA shows low stress, bicycles can experience a severe amount of stress cycles which could cause failure well below yield strength due to fatigue. Removing stress risers caused by discontinuities should greatly reduce this likelihood.

https://en.m.wikipedia.org/wiki/Stress_concentration


Sent from my iPhone using Tapatalk

 

[xadmx]

Yes that's the kind of key switch i'm going to get. I'm trying to find a nice metal one.

I was considering the controller heat sync and thought why not just modify the casing it came in.

good idea about cutting the case.
i am running a small 12 fet 800w controller in an enclosed scooter frame because of the space with a 10s battery, just the strip of alloy on the fets when you remove it out the case.
its been like this for 2 months now and runs fine.

 

[MadRhino]

The stealth goal is not that easy to achieve, and impossible with a hub motor on a fixie that has by definition: Minimalist wheels and components.

The only solution is not to fit the batteries in the tubing, but the motor instead. Gruber, Vivax... That kind of kit that is inserted in the seat tube, with the battery in a saddle tool bag or a bottle. That is neat enough to build a 250w fixie. I have seen some built on light weight road bikes, that are almost impossible to see. Of course, using the top and down tubes for batteries is a very neat complement to a seat tube motor but I would prefer it in a bottle for ease of maintenance.

 

[LawrenceEaden]

@LewTwo That's a great idea! I'll try and cut the patterns out to leave myself enough left to do that.

@Jimbo7890 Thanks! I'm planning to do my own FEA on a model of the frame with the cuts, but stress concentration isn't something I'd considered, really great info to read, thanks. I'm actually considering brazing the inside of the joint between the tubes, just lightly to fill any gaps behind the weld, all yet to be decided.

@xadx Good to know! My battery can only output 10A so I'm hoping it won't be and issue.

@MadRhino impossible? Stealth does not equate to 'invisible'. I think saddle bags and bottles are unsightly. That's why i haven't gone an bought an eBike that uses these common solutions. I wanted to try something different. The seat tube motors do look interesting, but they are not worth the cost, for something that through London in the winter can't really be my only mode of transportation.

 

[LewTwo]

"It doesn't snow in London." or so some of my Scottish friends told me.
I froze my butt off in London snow :x :x :x :x
... be sure to get a GOOD lock !

 

[LawrenceEaden]

I always use a D Lock. Whenever I've used anything else and left a bike locked up on the street anywhere in London it's disappeared. (I learned after the second time). I'm hoping the key ignition will add some security as well over just a on/off switch.

Frame Bag Mock Up

I picked up the leather jacket I'm going to use to make the frame bag yesterday and even though it's small, there's ample material. I managed to make an almost full mock up of the bag using only one sleeve!

After making this mock up I think i'm going to have to make a few, most of the project is purely function with hidden form, but this bag is going to affect the aesthetic of the bike quite heavily. It's a small detail I really want to get right.

I've never worked with leather or many textiles before, but it's so thin it doesn't really affect the design of the bag at all. The mock up is only super glued together with a 3mm seam all round.

Getting to look at the heat sink on the bag is encouraging. I think I will paint it a rusty red colour, the same I have planned for the wheels. I've done a mockup of the colour scheme I'm after. Metallic Gun Metal frame and a deep rusty red for the wheels.

In other news, the motor is arriving Friday!!! I had better finish the battery for bench testing.

 

[MrDude_1]

neat.. you could make that bag even smaller... I like the style.

 

[d8veh]

If you keep the controller in its aluminium box, it'll be OK inside the bag. No need for any cooling holes. I installed a lot of controllers in bags and boxes. Never had a problem with heat. Nearly all the OEMs mount their controllers in tight boxes with no cooling holes as well.

Instead of the throttle, you should concentrate on the PAS, which gives a much better riding experience on a light bike. The controller you got has a bad PAS function though. You need one that gives different levels of power and current control rather than speed control.

 

[LawrenceEaden]

@mrdude_1 I like your thinking. I'm going to wait and see if i need the room for the connectors. if it's not full sardines I may make a smaller one.

@d8veh the ally case makes the footprint so much larger in all directions, i want to keep the bag as small as possible. The PAS will be the next step, this is a way to get the project going, i doubt this will be my first and last eBike build so I'm trying to do it on a budget, fail fast and learn a lot. I'm most interested in whether it will work how I planned in my first post, whether i can get the top speed from pedalling and gearing the drivetrain so, and using the motor to get me off the line at the 20 or so traffic lights on my commute. PAS doesn't work very well when you first start i've heard? If i was trying to take off up a hill i could go nowhere.

Motor Modifications

After making the mistake with the first motor I ordered of getting a 36 not 32 spoke hub I couldn't find a hub motor the right width, so I took a punt and got a 135mm wide motor hoping I could modify it to fit. It arrived today! I'm happy I made the swap, it's about half the weight of the 8fun one I started with. Apparently the same specs...

It came with a free hub to fit a cassette, I wasn't sure what fitting out be underneath, hoping it would be a male thread I could put the single gear freewheel that came with the bike on. Instead there was a female thread which I couldn't find any information on, i'm guessing these components are unique to hub motors? Has anyone had any experience with these?

inspecting further I saw that the freewheel had two bearings that positioned on the shaft and secured it to the motor, which had to stay. So i cut it down as much as i could keeping this in tact.

I then just had to file down the shaft a little to fit in the frame properly, i wasn't too worried about tolerances on this cut because the torque washers would still be seated on the original flat reducing any chance of movement under load.

This all worked out pretty well for centring the hub in the frame, i did a rough measure and found the centreline of the motor hub only to be 2mm offset from the frame centreline.

I checked this on the webpage i mentioned earlier for calculating spoke lengths on hub motors.

http://www.ebikes.ca/tools/spoke-calc.html

Ta Dah! There is a bit of tolerance in my measurements, but it's come out at 50:50 tension. I don't know if there are any issues caused by putting the spokes right on the right hand side of the hub flanges.

I cut off the thread on the freewheel which usually holds the cassette on, my plan is to file a groove on the freewheel which i can put a circlip on. I can then position a single sprocket with spacers either side to get the chain line right, like a lot of single gear conversions I've seen.

I think I've now got everything I need to give the whole system a bench test, so that will be my next step early next week.

 

[d8veh]

The best solution is PAS and throttle.Throttle is good for quick starts and hill starts, but PAS is much better for everything else. After you try a good PAS system, you wouldn't want to have a bike without it.

I like what you're doing. I think you will get what you want. I will do something similar for my next project. I will start with a reasonable quality disc brake road bike and change the handlebars to carbon fibre straight ones do that I can fit hydraulic brakes. I will also use a Q100C motor. It will be a 36v 201 rpm version, which I will run at 48v or 52v to get a cruising speed of about 20 mph.

Don't forget to take the label off the motor if you live where there's a 250w limit.

 

[motomech]

I cut off the thread on the freewheel which usually holds the cassette on, my plan is to file a groove on the freewheel which i can put a circlip on. I can then position a single sprocket with spacers either side to get the chain line right, like a lot of single gear conversions I've seen.

You're welcome :lol:

motomech wrote;

When I had problems fitting one of my Q100C's, I toyed w/ the idea of converting it to a single gear and it looked quite easy w/ no or little machine work.
If the cassette is not used and the inner threads for the cassette locnut are not needed, about 12M/M of the spline can be trimmed off.
Then, using a cheap cassette(the kind that falls apart when removed), select a gear(12T & up) to use.
Locate on the spline w/ some fab'd round pipe(2 pieces? inner & outer) & secure the ASM w/ a C-clip(cut a grove on end of spline).
(Sheldon Brown actually sells gear spacers, but the amount needed would be pricey.)

 

[Santa'sLittleHelper]

I like it! Subscribed and watching the talented work...

 

[LawrenceEaden]

@d8veh you're right. I've put wiring in along my battery pack in the seat tube to fit a PAS sensor in the future. It would be good to calculate the mechanical resistance of the motor and map automated pedal assist to make it ride like a regular bike when off the throttle.

@motomech. I forgot about that! Great minds, did you eventually do the modification?

@santa'slittlehelper Thanks!

Controller Confusion

I've been trying to get my head around the controller. The wire colours appear to be completely different to what stated on amazon where I bought it. Some of the connections seem crazy to me! There's a two wire connector that is is 36v with no protection connecting to a rail which then makes a 36v live wire on another 3 wire connector.

I've taken it back to bare bones and I'm trying to identify the connections by the control boards annotations. So far I have the following

• Hall Connections
  
  • U - Yellow Phase Wire
  • V - Green Phase Wire
  • W - Blue Phase Wire
  • GND - Ground
  • 4.3v - 4.3v supply

• Unknown 3 Pin connector
  
  • HBD - Orange wire
  • LBD - Grey Wire
  • GND - Ground

• Unknown 3 Pin connector
  
  • VSP - Pink wire
  • SP - Green Wire
  • GND - Ground

• Unknown 2 Pin connector
  
  • DSP - Green wire
  • GND - Ground

• Unknown 2 Pin connector
  
  • DSP - Green wire
  • GND - Ground

• Unknown 2 Pin connector
  
  • D+ - Power + (Battery, 36v)
  • L - Orange Wire

• Unknown 3 Pin connector
  
  • US - Green Wire (connected to the blue motor phase wire)
  • L - Orange Wire
  • YK - Brown Wire

• Unknown 1 Pin connector
  
  • US - Blue Wire

• Motor Phase Teach
  
  • 12/6 - White Wire
  • GND - White Wire

• Unknown 1 Pin connector
  
  • HH - Purple Wire

• Battery connector
  
  • D+ - Battery Positive
  • D- - Battery Negative

• Charge connector..?
  
  • D+ - Battery Positive
  • D- - Battery Negative

• Unknown 1 Pin connector
  
  • HL - White Wire

I don't really want to start connecting things up wildly. Any knowledge on this would be appreciated. I haven't been able to find any searching so far. The controller is a YK85s. I'm hoping that the board annotations are universal and someone can help.

Also I got to put my two motor options side by side and there's a huge difference between the Q100 and the 8fun for the same power rating.

Thanks!

 

[LewTwo]

You may want to start here ... https://endless-sphere.com/forums/viewtopic.php?f=2&t=48589

Also check this URL: http://bicycles.forumsee.com/a/m/s/p12-59751-052416--yk85-controller.html

 

[LawrenceEaden]

I just did lots of testing and I think I've got it sorted. Further updates tomorrow.

@LeoTwo I've read through that, my wires are different colours and it actually is wired differently.

 

[LawrenceEaden]

It lives! I got the motor to turn. I went all round the board with the multimeter and made a guess that there was a 36v ignition switch, which was wired differently to other YK85 controllers I had seen online.

I'll post my full diagnostic results later about what I believe all the pins do for future reference. But I managed to identify the throttle and that's the only functionality i need right now so I'm going to save these results and cut the redundant wires down. I got a little excited that i had the motor going.

https://vimeo.com/187892159

This is a pretty good milestone though the project. Bench test done, next I've got the spokes coming tomorrow to build the wheel.

 

[LawrenceEaden]

Controller Simplification

I managed to find another supplier of this controller, the YK85S (different to the YK85), to request an electrical diagram and got the following back.

This let me cut back and insulate all the wires I don't need, everything but the key switch, battery, motor, motor hall and throttle wires. Leaving an annotation on the cut down wires incase I need them in the future. I'm leaving out the brake sensors for cutting the motor because I'm not using pedal assist and my throttle is a thumb throttle which I think is safer than a twist throttle, so my cut offs are no throttle, then the ignition key in the frame.

I got the motor spinning up and wanted to check a few voltages. The throttle output ranges between 0.8v and 3.5v output with a 5v input. This sounds like I can't get full throttle if the controller sees full throttle as a 5v input. One of the motor wires measured 18v at full throttle. I'm assuming this is meant to be the same voltage as the battery pack? (37v). I have yet to test putting 5v to the throttle and measure the motor voltage. If this is the case I'll make a simple summing op amp circuit with a potentiometer on one input and the throttle output on the other so I can easily tune it.

Wheel Building

Lacing up the wheel is the last thing i have to do before i can do a test ride, so it's top priority. I'm planning on changing the colour of the rims, which I know would be easier with the wheel in pieces, but i want to get a proof of concept first before refining and polishing the details.

It's pretty simple and easy to lace up the wheel, the only tip i worked out was the only way to get the nipple through the double skin rim is to twist it on the end of a spare spoke and thread through the hole.

Motor Optimisation

When trying to choose my motor I was susprised how little technical information was available for each brand. Luckily, the motor that I had to end up getting due to the 32 spoke pattern did have a torque curve document.

I'm interested in the playoff of motor efficiency and torque. I'm hoping to be able to do some data logging at some point which I can map bag to this document to eventually create an hidden indicator of when I'm travelling at peak efficiency and peak torque or somewhere in the middle. None of this is really essential for the project, but it will be useful to understand for getting the most out of the bike.

I just found a very elegant build trying to achieve the similar kinds of stealth I'm after here https://endless-sphere.com/forums/viewtopic.php?f=6&t=64720
It's made me revisit my frame bag design, maybe even that may be too unsightly, I think i'll stick with it for now, then when I look at designing my own controller circuitry later down the line I want to explore the space under the seat, there could be enough of a void to keep it invisible from the side profile.

 

[flangefrog]

There are several tools to help screw the nipple on with deep rims:

The IceToolz 12S4 can be had very cheap on eBay:
http://www.icetoolz.com/ec99/ushop10074/GoodsDescr.asp?category_id=68&parent_id=50&prod_id=12S4

Some more that are adjustable such as the Problem Solvers driver:
http://www.wheelfanatyk.com/blog/wheelbuilding-tip-15-spin-those-nipples/