Custom KMX recumbent trike with QS205

[tonystark20]

I had done some adjustments to the rear triangle support (adding length so that the back had more space between the bar for the trailer and teh tire. (sat a little to close to my liking.) i thought I had it pretty tight when reassembled but inspection of the axle does show some marring that would indicate the ability for it to rock back and forth.

I've already taken the one side off the motor and the axle does have a snap ring right up against the core assembly, I'm going to remove the other side and see how it's assembled, but it is looking like the v3 axle assembly might be replaceable. if i find that it is I'll see if i can order one from QS or if I'll have to model it and take it to a machine shop.

The brake right now is only on/off, I had indented to work on an upgraded wiring system that would help increase simplicity and cut weight and allow for more waterproofing but I'm a ways away from that (the upgraded harness would allow for adjustable regen)

 

[amberwolf]

but it is looking like the v3 axle assembly might be replaceable. if i find that it is I'll see if i can order one from QS

If you do end up able to order one from QS, you might see what it would cost for multiple units, and also if the same axle is usable in multiple versions of the QS205, (and which ones it would fit).

I bring this up because there are at least a few of them around ES with sheared axles that could be fixed if it's as easy as undoing the snap rings and tapping out the old one, and tapping in the new one (if it doesnt' require a press and tooling to remove or install).

If it doesn't take a press, and the cost is low enough I'd be up for buying one from you, if you'd be willing to ship it to me down in Phoenix, AZ after you receive them.

If it's going to take a press, or it's expensive, I'd just replace mine with a completely hollow tube of the same diameter on it's entire length, and use spacer tubes over it between the covers' bearing ID rings and the stator, and probably woodruff-key it to the stator for rotation lock. Then round clamps (maybe with woodruff keys as well) on the "dropouts" to secure it to the frame. This eliminates all of the stress risers where axles always break, and would also allow all wiring to easily be run within the large-ID axle pipe, protected all the way out to the end. Then a right-angle housing (like electrical conduit stuff, etc) can cap the end and allow safe protected routing to the controller.

The brake right now is only on/off, I had indented to work on an upgraded wiring system that would help increase simplicity and cut weight and allow for more waterproofing but I'm a ways away from that (the upgraded harness would allow for adjustable regen)

Ah, well the on/off creates two problems:

The first is that it slams the axle backwards in the dropout if there's any movement allowed, and the next acceleration rocks it forward again. If it's a hard acceleration it will nearly slam it forward (but probably never as hard as the braking does). If there's no movement allowed, then the on/off braking just twists the axle itself from the dropout point inward to the stator mount.

The second is that it tends to break traction much easier than ramping the braking force up, even if it's done quickly.


So the analog regen will help with both problems you have had.


FWIW, if you need an analog braking lever system, there are at least three methods posted in this thread:
https://endless-sphere.com/forums/viewtopic.php?f=2&t=105460
the first of which I"m using on my SB Cruiser trike successfully. (a more refined version will be implemented "soon" once I get this other project completed:
https://endless-sphere.com/forums/viewtopic.php?f=30&t=105711
where I can use the Lebowski brain to run the Honda IMA Inverter's powerstage.)

 

[tonystark20]

currently the brake handles I have are modified with a hall effect sensor - they are hooked up to an arduino that controls all the lighting on the trike. (one of the functions I had planned to add was for an emergency stop the brake lights would flash - currently programmed but not turned on as the system has a whole host of other issues)

The ASI controller allows for many things including torque ramp up and down for the regen, engine torque % for regen, battery current max during regen as well as a cut off option (no regen) and and separate control pin option for regen. and regen can be controlled with on/off, analog, and via the network (can bus and TTL Rx/Tx) trike is also already equipped with a second throttle that I had intended for using with the brakes I just have to wire the hall effect inside of it.

I took the cover off the other side of the motor, it's really hard to tell if it does come out, even if it did there is lots of sealant to remove along with re-running wire.

But the why else would the snap ring be there? Plus I can see a spot behind the snap ring that looks to be a keyway.
The strongest hammer I currently own however is a standard ball peen. and it didn't look like anything was budging. might try some penetrating fluid and placing the motor in a way that the stator won't want to move around so much within the magnets. But I have a feeling it may require a press to get out, the machining looks to be of tight tolerance.

 

[amberwolf]

currently the brake handles I have are modified with a hall effect sensor - they are hooked up to an arduino that controls all the lighting on the trike. (one of the functions I had planned to add was for an emergency stop the brake lights would flash - currently programmed but not turned on as the system has a whole host of other issues)

What kind of issues (for those not already listed in the thread)?

The ASI controller allows for many things including torque ramp up and down for the regen, engine torque % for regen, battery current max during regen as well as a cut off option (no regen) and and separate control pin option for regen. and regen can be controlled with on/off, analog, and via the network (can bus and TTL Rx/Tx) trike is also already equipped with a second throttle that I had intended for using with the brakes I just have to wire the hall effect inside of it.

FWIW, if you're already used to using regular brakes to control the amount of braking you use, it's going to be more natural and more reflexive in a panic stop, etc., to use only the brake lever itself to control the braking.

If you are setting it up to use the brake lever to engage the ebrake, but then use a separate device (throttle, etc) to then control the *amount* of braking, you're increasing your reaction time and complicating the reflex process, potentially enough in the right situation to prevent you from braking sufficiently in time.

If the lever does all the work just like it normally would, then it should allow reaction times and reflexes you already have to work.


The way I did mine in the version presently operating is to use a cable-operated throttle that's pulled by the cable-type brake lever, as the analog voltage source. The details are in that linked thread for why it is the way it is, but it basically uses the ebrake switch built into the lever to turn a set of relays on. One of the relays switches out the actual throttle, and switches in the one in the COT, so that as I pull the lever, the COT outputs it's voltage instead of the regular one, and the mechanical stuff is setup so the COT range is the same as the lever range, giving me matched mechanical vs electrical range on the lever.

The way it *will* work when done, because the Lebowski brain has a second analog input that can be used for analog ebraking, is to directly run the COT to the second input so the brain just reads it directly, and no switching/etc is needed, *and* a different response curve can be used fro the brake vs the throttle.

I took the cover off the other side of the motor, it's really hard to tell if it does come out, even if it did there is lots of sealant to remove along with re-running wire.

I fully expect ot have to redo wiring; on mine it's already damaged at the axle exit from the axle breakage. Sealant is relatively easy to deal with, if tedious.

But the why else would the snap ring be there? Plus I can see a spot behind the snap ring that looks to be a keyway.
The strongest hammer I currently own however is a standard ball peen. and it didn't look like anything was budging. might try some penetrating fluid and placing the motor in a way that the stator won't want to move around so much within the magnets. But I have a feeling it may require a press to get out, the machining looks to be of tight tolerance.

Well, in the coming week I'm hoping to get to the QS205 project (along with many others), and if so I'll see if I can get the snap rings off, and use the rubber mallett to try to at least see if the axle will easily move.

First I will have to build a support for the stator itself, securing the stator support core directly surrounding the axle, so that:
--the windings cannot be damaged
--the alloy stator support cannot be damaged

It may be necessary to heat the stator support to cause it to expand, after first freezing the axle to cause it to contract.

but I'll post what I find.

 

[tonystark20]

but I'll post what I find.

I look forward to seeing if you come up with anything, good point about heating and cooling to try and getting it to move - might try that myself as well.

What kind of issues (for those not already listed in the thread)?

Chances are I've already explained the issues, but I've had issues with driving in the rain - can't figure out what's being damaged in the rain to cause things to stop working, and then there's the fact that I had to switch to using arduino nano's to control lights vs my original plan to use i2c compatible pwm drivers. at this point I have so much wiring pushed into the electrical box combined with the multiple microcontrollers makes it difficult to program new functions and diagnose why some work but some won't.

had intended to implement the second throttle for ebrake control in such a manor that i'd mainly be using it on downhill control, now the initial plan was with the sabvoton controller I had where on flat ground it was great but on longer downhills it was too powerful. (i tend to use the brake downhill to control the speed as the area I'm in we get some crazy corners on the downhill. still fun to drive and interesting to see drivers reactions considering I can out maneuver at a high speed on those roads than they expect.

The reason I haven't used the ebrake analog signal directly is that during my normal driving I can almost always get away with using the regen to bring me to a complete stop, I've set the cable tension and code on the brakes to reflect this - the regen will kick in first and then there is a period of dead space so that the mechanical brakes don't work right way. the bb7 brakes I have are absurdly powerful (I've flung myself out of the seat once when I tested it. (pretty much stopped on a dime)

I've been considering a new trike design for quite some time and now with a broken motor I have some decisions to make.

I would like to build a trike with complete canopy from weather. to do this I'm thinking fiberglass / carbon fiber or I was thinking a sheet metal (with dimple dies and riveting) to build a space frame type design. I also wanted to experiment with motors to find one more efficient that I could combine with a solar trailer. (I'm getting 30-35wh/km right now) If I get a new qs205 I'd likely go with a slower wind (mine tops out at 60km/h which I almost never reach) 40 top speed would be nicer for my average use. and I'm sure the extra torque would be awesome for acceleration and hills. I also was think of trying mid drives and custom transmissions, among other hub motors. plus i wanted to swap out the control panels I have for ones that would use touch sensors (allowing me to make it completely waterproof) with the use of touch sensors i also realized that I could build a speech recognition module into my helmet to control lights (i actually find it surprising the about of times i want to change a setting but its an awkward time or position for me to take the hands off the controls. also had plans to build a cruise control module ath twould interface with the Rx/Tx of the cycle analyst to maintain a given speed but you could also tell it the distance you want to reach and it could offer a visual indicator if your driving lead footed or have an override mode where it would throttle you back so that you make a distance you've asked to to achieve. so many ideas but nowhere near enough of a budget. I have some things in my life that I'm hoping to achieve in the next few years that will give me long term stability.

 

[adam333]

Odd, your dropout look a lot like what I use on my trike.

I have the max power limited to 7.5kW and also use on/off regen and I never had this issue.

Only difference is that I am using a Cromotor V2.

 

[tonystark20]

Odd, your dropout look a lot like what I use on my trike.

that's actually what I based mine off of when machining, on the next trike I'm going to change the design a bit as it's quite the hefty chunk of aluminum.

I did however mess up when machining these so they are closer to 11-12 mm vs the 10 mm that they should be, I had flat stock stainless acting as a shim, but when i did the reassembly of the rear the axle very clearly wasn't in place correctly (i can see marring on both sides of the axle)

I (like many others) have been trying to order a new motor but haven't had much luck. So i wanted to take another crack at fixing the axle, I don't own a press but do have a 6" bearing puller on hand (was $40 on amazon canada) made a mark with my drill on the end of the axle to help align the tool, had the legs of the puller facing out (grabbing the outer section of the motor was easier than inside due to the shape of the tools arms), once snug as much as i could with the tool (used a second wrench to hold against) I smacked the motor assembly with the tool facing down, this lossended teh white glue and after that I was able to wrench the puller to completely remove the axle.

for reassembly, I'm not in college anymore and I don't own my own lathe yet so I'm going to take the axle and a drawing i found to a local shop to see if they can machine one for me. Anyone know what the white / silicone glue is, and what i can use to reseal the motor for reassembly?

I also think I may buy the hubsink and statorade form ebikes.ca might as well have the statorade at the very least while everything is disassembled.

I'd also like to add the I am working on designing a new version of my trike that will be fully enclosed for all weather driving and have suspension all the way around. I've been having issues designing proportions so i'm thinking I'll build a mockup with 1x2 lumber and foam panels (both cheap items to buy near me) I've also designed a mechanism that will give me full two wheel brake controls from either front handle bar (the mechanism will ensure that all cables are kept tight and that the breaks can activate at an even rate. and I've settled on a possible chain ring / gear set up that should allow for proper cadence and torque to allow for pedaling on long distance.