KMX Tornado -- Recumbent Cruiser

Joined
Sep 8, 2019
Messages
433
Location
USA, CA, Bay Area
2021-09-18 17.13.10.jpg

Here's the starting point: a KMX Tornado that's been fitted with an underwhelming 36v system. I'm planning to swap over to a Phaserunner + Cycle Analyst (I just love the combo) pushing a GMAC 8T (gotta have that regen) powered by a 72v 35Ah battery.

Also, fun little fact, if you take the front boom out, you can pretty easily stuff this in the rear of a Chevy Bolt. Neat.

2021-09-18 17.09.00.jpg
 
2021-09-19 20.15.36.jpg

I don't plan to keep the drive train at the momemnt, may change my mind on that, but for now with it removed I still need a place to put my feet. A second "left pedal" from another kit flip 180 and with a "temporary" strap hold down, a nice place to rest my feet.

2021-09-21 17.34.12.jpg

I have some spare parts around, including a 52v 50Ah case battery and a massively large generic 80A controller. Decided I didn't really care about this little motor and would full send it. Worked quite well, frankly, but did top out around 22mph which is just too slow :) Though, I can definitely see how a steering damper comes into play at higher speeds. On my list of todos.

2021-09-28 18.49.04.jpg

Picked up this enormous, super bright LED "whip" to attach to the rear. We'll see how it holds up, but I imagine it'll be fine. Cool sideproject idea would be to integrate it into the brake/regen to have the controller show red tail light while braking.

2021-09-23 20.14.41.jpg

Started reclaiming the 14s10p pack to turn it into a 20s7p (same cell count, nice!) only to find this horrifying thing -- a pack with P groups capable of pushing 100A has the pos/neg lines on a whole 2 strips of nickel. Just...yikes.

2021-09-29 22.37.57.jpg
The cells all pulled up and (mostly) cleaned up.

2021-09-29 22.38.20.jpg
Kept the interior cell washers (is that what you call them?)

2021-09-29 22.38.25.jpg
Mess of old damanged PVC wrappings. Why rewrap them all? The previous pack had all the cells in the P groups touching, which isn't a problem for the pack generally speaking, however it does mean that as it got used and heated up, all the cells PVC fused together. Thus, as I was pulling the pack apart, every single cell had the wrapping tear in some fashion.

2021-10-01 19.41.49.jpg
Here I employ some shop helpers to get all the batteries ready for new heat shrink wrap.

2021-10-01 21.18.33.jpg
Wraps completed. I made a very frustrating mistake here. I should have dremel cleaned the cells pos/neg tabs BEFORE doing this. I had to come back later and a lot of cell wrappings now have dremel damage. Ugh.

2021-10-01 21.27.19.jpg
The pack laid out in a lovely checkerboard pattern.

----

I plan to put the battery in a clear plastic case behind the seat, something like this:

2021-09-25 23_33_54-Window.png

Should be an interesting learning experience. Kinda leaning towards acrylic at the moment. Everything I've read says the Polycarbonate is significantly stronger but it's less UV stable (kinda matters here) and it's nowhere near as easy to make actual boxes from compared to the near effortlessness of seam welding acrylic. I figure for how well "protected" this area of the trike will be, I can forgo the PC for the ease of use (and price savings).

2021-10-02 23_22_13-Autodesk Fusion 360 (Personal - Not for Commercial Use).png

At one point, I toyed around a lot with the idea of building a battery tray under the seat as I've seen other (adam & outrider usa) do, but I had to reel that excitement back -- I'm just not that DIY ready yet.
 
Pretty sure I've got a lock on the battery box -- I plan to "offset" the battery levels and build the box along a slant.

2021-10-04 22_48_55-Autodesk Fusion 360 (Personal - Not for Commercial Use).png

Side panels will NOT be glued on, but have a few bolts holding them in place so it's still serviceable. I'll use some XT150 bullets to connect the series, bms, etc through some internal cut outs.

I'm also waiting on a PCB I designed for doing parallel group balance leads more nicely. This is something I've done before and it worked out so well I'm doing it again. It's also just rad.
 
I also modeled and 3d printed a bracket to mount the CA on the front boom shaft.

2021-10-07 21_47_56-Autodesk Fusion 360 (Personal - Not for Commercial Use).png
2021-10-07 21_47_04-Autodesk Fusion 360 (Personal - Not for Commercial Use).png

Splits into two parts and bolts right on.

2021-10-07 22.46.57.jpg
2021-10-07 22.46.49.jpg

Worked out well, I expect it to hold up reasonable well -- if not, I can always thicken up the design and print another.
 
Interesting build as E-only power.

Keep the MAC under 50A peak and 30A sustained. Without human assist the brief stints of high torque human output the motor will have to work much harder; granting it more current to overcome is a dangerous temptation so keep an eye on the winding temps.

It's now 6 years on a pair of MAC 10T's with over 6000 miles each on Scorpion FS 26's, gears (& clutch) are still in excellent condition.

Phaserunner, 35A peak, run hard, put up wet.

MAC10T_Planetary.jpg

MAC_10T_Coils.jpg
 
Triketech said:
Keep the MAC under 50A peak and 30A sustained. Without human assist the brief stints of high torque human output the motor will have to work much harder; granting it more current to overcome is a dangerous temptation so keep an eye on the winding temps.

Good stuff to know. Motor temp has been a concern for me, for sure. Given the gmac has no way to get heat out of the core very effectively (compared to DD hubs which can be ferrofluid'ed and hubsink-fin'ed) it's definitely on my mind. On the plus side, the motor has the integrated heat monitor, so I should be able to have the controller keep things reasonable, even if I forget.

Finally, if this really doesn't pan out, I have a DD45 I can swap over too; I'm just trying to keep the weight of the trike down since I gotta lift-and-load it for transport.
 
3d printed a light mount attachment, added a rivnut to the front to attach it:
2021-10-10 22.35.14.jpg

Controller attached to the side. Easy access, good airflow, mostly out of the way. Again with the rivnuts.
2021-10-11 20.31.38.jpg

The more complicated bits machined out of 13mm ("1/2in")
2021-10-14 19.53.53.jpg

Bottom supports test fit, measures up very nicely
2021-10-14 20.36.14.jpg

Main assembly mocked up with good 'ol painters tape
2021-10-14 20.46.17.jpg

And it fits!
2021-10-14 20.49.12.jpg
 
Glued it all up and did a quick test fit. Works pretty well, sadly can't just slide in from the side, so I'll have to unhook the seat from the rear supports and lean the seat forward to drop it in, but that's not really the end of the world, eh?

2021-10-15 22.36.01.jpg

2021-10-15 22.31.49.jpg

2021-10-16 10.41.42.jpg
 
chuyskywalker said:
Good stuff to know. Motor temp has been a concern for me, for sure. Given the gmac has no way to get heat out of the core very effectively (compared to DD hubs which can be ferrofluid'ed and hubsink-fin'ed) it's definitely on my mind.

Convection cooling may be better than you think. The faster the rotor speed the thinner the boundary layer which is like a thermal insulation blanket. Higher internal air velocity of a geared hub penetrates the boundary layer and transfers more heat via forced convection. At about 220 RPM wheel speed approaching full boundary layer penetration (impingement); Convection efficiency increases 5-6X over light flow convection.

How much difference? Can't say, never measured it on the hub, thats the theory behind the science only. But I do have a pretty good background in thermodynamics in the aircraft electronics industry and have run similar tests.
 
chuyskywalker said:
Picked up this enormous, super bright LED "whip" to attach to the rear. We'll see how it holds up, but I imagine it'll be fine. Cool sideproject idea would be to integrate it into the brake/regen to have the controller show red tail light while braking.

2021-09-23 20.14.41.jpg

At one point, I toyed around a lot with the idea of building a battery tray under the seat as I've seen other (adam & outrider usa) do, but I had to reel that excitement back -- I'm just not that DIY ready yet.

I've got the same manufacturer (i believe) LED light whip - the 5 foot model. Extremely visible at dusk/night.. can be noticed from a distance during the day - i've had someone remark they could see me coming from the other side of a Park. It's a solid unit, that hasn't leaked any moisture inside the tube. I've got mine mounted to the wheel-stay, and collared above in place by a 1-way restraining clamp to limit it's movement and flex.

20210522_220218[1].jpg

20210908_161348[1].jpg

I've done the Under seat battery tray thing also. I run dual 48v19.2ah packs, on a bespoke frame mount assembly. It took about a months worth of fairly intense fabrication, to pull off. Check my build out at https://endless-sphere.com/forums/viewtopic.php?f=6&t=110973

Left side battery services the 1000w Hub motor, while the Right side battery services the 12v electrics, run through a 48v->12v down converter located in a ventilated box behind the seat.

20210501_023014 (1).jpg

I was similarly inspired by the Outrider 422 Alpha setup, for battery location to keep weight low.

Your project caught my eye.. as i am in process of gathering ideas for an eTrike Mk2. I'm considering the frame build to be out of square tubing, vs round, as much easier to work with in terms of dimensional measurement and cutting in fabrication.
 
HackD said:
I've got the same manufacturer (i believe) LED light whip - the 5 foot model. Extremely visible at dusk/night.. can be noticed from a distance during the day - i've had someone remark they could see me coming from the other side of a Park. It's a solid unit, that hasn't leaked any moisture inside the tube. I've got mine mounted to the wheel-stay, and collared above in place by a 1-way restraining clamp to limit it's movement and flex.

That looks like a solid way to set that up. I may end up having to remove the rear rack, or find a way to not have it attached to the seat as I need to be able to lift the seat to get the battery on and off; a feature I really want to take advantage of for loading/unloading purposes.

HackD said:
I've done the Under seat battery tray thing also. I run dual 48v19.2ah packs, on a bespoke frame mount assembly. It took about a months worth of fairly intense fabrication, to pull off. Check my build out at https://endless-sphere.com/forums/viewtopic.php?f=6&t=110973

That looks like an INTENSE build process; wow.

HackD said:
Left side battery services the 1000w Hub motor, while the Right side battery services the 12v electrics, run through a 48v->12v down converter located in a ventilated box behind the seat.

Curious -- why not parallel them both? You'd get more mileage out of the combined packs due to lower draw on each.

HackD said:
Your project caught my eye.. as i am in process of gathering ideas for an eTrike Mk2. I'm considering the frame build to be out of square tubing, vs round, as much easier to work with in terms of dimensional measurement and cutting in fabrication.

The KMX frame has been great from that perspective -- in fact, it was one of the main selling points in picking it out is that it's extremely easy to fabricate around the box tube.
 
HackD said:
Left side battery services the 1000w Hub motor, while the Right side battery services the 12v electrics, run through a 48v->12v down converter located in a ventilated box behind the seat.
chuyskywalker said:
Curious -- why not parallel them both? You'd get more mileage out of the combined packs due to lower draw on each.

My only excuse is equal parts KISS, and lack of motivation to go that further step, to date. Original thought was just to physically switch out the packs from one mount shoe to the other, if one or the other exhausted. I have come close, but i've yet had to perform that step during the ride, ergo no motivation thus far to do so.

There's a few electrical remainders i have yet to do, but were left because it was 'good enough', and the wiring process to that point was already onerous enough. Electrical is a weak point, as in Sir Lucas Prince of Darkness, letting the smoke out of the wires, weak. Once working well enough, i'm reluctant to mess further, without chancing Fate.

HackD said:
Your project caught my eye.. as i am in process of gathering ideas for an eTrike Mk2. I'm considering the frame build to be out of square tubing, vs round, as much easier to work with in terms of dimensional measurement and cutting in fabrication.

chuyskywalker said:
The KMX frame has been great from that perspective -- in fact, it was one of the main selling points in picking it out is that it's extremely easy to fabricate around the box tube.

Good to hear confirmed. It makes logical sense, since i don't really have an accurate way to round tube-index for precision.
 
Created the battery pos/neg leads and connections from the first half of the pack to the second half. Soldered up the BMS leads onto the distribution boards, and got them all spot welded up with tabs, ready to go on the the pack. Got the pack all spot welded up -- goes really fast when you've got the ladder strip and it's always consistent. Finally, a bit of fishpaper out of an abundance of caution, kapton everywhere, and a small bit of padding and they are ready to go!

Some shots over the course of getting everything installed. I clearly made ALL the cables far too long, but it does make it easier to move things around. I had originaly designed the plug in the rear of the pack, but forgot to machine it in before I did the glue up. Plug is inconvenient on the left side, considering the controller is on the right, but I just couldn't fit it on the right side panel. Ended up improvising a 90 degree xt90 adapter. Not pictured here, but strips of neoprene foam were added to keep everything in place.

After getting it on to the trike, every time I plugged in the XT90, it would spark -- which makes no sense as I have AS connectors on the plug. There's gotta be a way to use a multimeter to figure out if the ones I have are legit. For now, though, I tossed an 80A resetable fuse in between the battery and controller which does the trick; it's just not pretty.

I also discovered that the xt60 jack on the phaserunner is VERY loose, like, plug falls out upsidedown loose. I added a bit of elctrical tape to the side of the plug to make it more snug, we'll see how that holds up...

Things left to do:

* Double throw brake lever on the right side
* Maybe a fender over the rear tire
* 12v battery for the rear LED whip
* Adjustments, tweaks, test ride, etc.
* Figure out the foot holding situation, these free rolling pedals simply won't do

PHOTO TIME

----

2021-10-19 20.52.15.jpg
2021-10-19 20.55.13.jpg
2021-10-22 08.58.55.jpg
2021-10-22 09.08.08.jpg
2021-10-22 10.49.46.jpg
2021-10-22 11.41.32.jpg
2021-10-22 12.04.38.jpg


----

2021-10-22 13.58.40.jpg
2021-10-22 21.24.18.jpg
2021-10-22 15.31.28.jpg
2021-10-22 16.08.42.jpg
2021-10-22 16.08.58.jpg

----

2021-10-24 19.31.44.jpg
 
Very impressive craft work! :thumb:

(I have not yet built a battery but) I had thought safety common practice is to have the female connector on the battery side? Harder for a stray coin or metal tool to create a battery short.
 
99t4 said:
I had thought safety common practice is to have the female connector on the battery side?

That's absolutely correct...

...and not a single place makes a female flush/panel mount! Anywhere. I spent _days_ looking to find something. If you've got a good lead, I'm all ears.

Also, all the XT90 antiparks are all female, which makes a certain amount of sense since it'd be really hard to make XT90 male pins have an inline resistor.
 
every time I plugged in the XT90, it would spark -- which makes no sense as I have AS connectors on the plug.

I had a bad connector, the resistor was blown. Go figure.

Other items checked off:

- Got a double throw brake lever, remade the brake cables, works great. Brakes need some small adjustments, but overall good.
- Picked up one of these 6x 18650 holders which has 5v and 12v output -- so far really happy with it for powering the LED whip.
- Printed a taller light bracket so I can angle the front light downwards a bit more, less blinding people.
- Had an issue I thought was over current on hard acceleration -- nope! The thumb throttle was hitting 4.3v while pegged, and the CA was detecting this as a fault and cutting out power. The CA just waited, like, 5 seconds or something so it wasn't obvious that it was connected to the throttle simply being full pressed. Anywho, fixed the range on the CA and that's all well now.

On to the bad news....

The GMAC isn't working out so great.

The entire setup would go unresponsive (faulted) with :bolt: Instantaneous Phase Overcurrent error when hard regen braking. An error for which there is absoltutely no extra information from Grin nor any of the 12 search results google pulls up. (And now, this post too shall be among the unhelpful...) The error requires you reset the power to clear it -- not a great situation to be in while trying to stop! Unable to work this out, I had to find something else...

So, I swapped out for a DD45_FST I have from another build and...yeah, should have started here. Going to handle the power much better and it's significantly quieter -- I'll just have to get over lifting the bit of extra weight when loading/unloading.

With the DD45 onboarded and acceleration cutout resolved, I can let it rip -- well, up to about 30mph at which point the steering gets REAL squirrly. Next up: front suspension and steering damper -- should really make it a speedy cruiser.
 
chuyskywalker said:
The entire setup would go unresponsive (faulted) with :bolt: Instantaneous Phase Overcurrent error when hard regen braking.

Who knew that regen braking would be a troublesome pain in the ass?

Oh, wait... I knew that.
 
if your xt60 falls out of the phaserunner
you will melt the pins and get that error,
simple fix is to flex the pins open a wee bit for a proper fit.also causes that error code, let me go find the quote from Grin

not exactly what i was looking for but heres the paragraph from this troubleshooting page
https://ebikes.ca/product-info/grin-products/phaserunner.html?___store=canadian&___from_store=international


Damaged XT60 Connector: We've seen several instances with high power rigs where the XT60 plug for the battery cable has become badly damaged from arcing, sometimes even melting the contacts. The battery current draw on the Phaserunner is much lower than what the XT60 contacts are rated for, so we believe that this failure is mostly the result of mechanical stresses on the battery leads pulling the connector at an angle and causing interruptions and arcs in the power flow. It's possible to spread apart the 4 tongs of the male pin for a tighter mechanical contact, and we recommend ensuring that the battery cable is not being tugged (eg by cable ties) in such a way that it pulls the connector at an angle. Instead the battery connector/cable should float freely so that it stays aligned and in good contact.
 
goatman said:
simple fix is to flex the pins open a wee bit for a proper fit.also causes that error code, let me go find the quote from Grin

Interesting, I'll check that out the battery pin spreading option. Mildly leery of doing anything that might compromise the controller since they won't have any more for another year though! However, between that and the small piece of tape, should keep it well secured. As for the tugging nature, the cord is pretty well bundled into a relatlively stable "tail" of all the other cords coming straight out the rear for a good 3 inches, should satisfy the general advice.
 
chuyskywalker said:
I'll check that out the battery pin spreading option.

Well well well, that worked exactly as suggested in terms of creating a better grip on the connector. Maybe I'll swap the GMAC back in at some point and see if that resolved the regen overcurrent error, but for now I'll keep the DD45 in for a bit.
 
Cool project. How much would you charge to print me a copy of your Cycle analyst mount?

Also, you're very likely going to want to move the batteries towards the front or down low when you get the design down for either. I imagine it's quite tippy right now.

what kind of range and performance is it getting at the moment?
 
The Toecutter said:
Cool project. How much would you charge to print me a copy of your Cycle analyst mount?

Thanks -- I'll put it up on thiniverse if you'd like, but I'm not sure how long it's going to hold up, to be honest. It feels a bit thin, but I won't know till I've been using it a bit.

The Toecutter said:
Also, you're very likely going to want to move the batteries towards the front or down low when you get the design down for either. I imagine it's quite tippy right now.

It actually works out pretty well so far; the weight of the battery is sitting on the frame not but 2inches from where it also holds my weight. Took it out for a 25mi run and I didn't have any tipping concerns. Much more concerned about these 1.25" front wheels -- really wish I could have something a bit meatier there.

The Toecutter said:
what kind of range and performance is it getting at the moment?

So far, about 25-30 wh/mi -- so a good (72v * 35Ah = 2520wh / 30-to-25) will be about 84-100mi, but I'd really only count on 60 before wanting to charge up. As for performance, it has no issues hitting 20 quick enough and I haven't actually full throttled it yet -- but I still managed to hit 40mph when I knew I had a very straight, very smooth road.
 
Wish me luck -- I've just ordered a single lever, dual caliper hydraulic brake set from Alibaba.

Comes with a parking brake and electronic switch as well, which is pretty slick.
Code:
TEKTRO HD-E745
they claim to be. We'll see!
 
A few updates to the build:

1) I moved the LED whip location to on the left side of the rack

2021-11-02 19.15.32.jpg
2021-11-02 19.15.46.jpg

With it bolted in the previous location I couldn't fit the trike in my very small car without it scraping the roof. This is printed VERY thick in PETG and should hold up just fine. If it goes bad, I'll bust out the NylonG and make it again. (Getting the shape, measurements right for the support bars was a real pain; they're not symmetrical!!)

2) Been working on burning down the battery to get a more real feel for the range

2021-11-14 13.28.55.jpg
2021-11-14 13.28.48.jpg

Turns out, something that's got 50+ miles of range takes a little while to get through the battery. Around this point, though, I'm now having moments where if I ask for too much power, I'm getting enough voltage sag to trigger a power cut from the controller. Not really sure why the Phaserunner is doing that instead of just rolling back the juice, but it's not a huge deal.

3) Fixed the pedals

2021-11-14 16.53.42.jpg
2021-11-14 16.53.36.jpg

Some u-bolts, ally bar, a drill, vice, and hammer and these are rock solid now. Haven't gotten out for the longer ride in this state yet, but on the roll around the neighborhood it was really nice to have a solid, non-swinging place for feet. I used rivnuts to attach these to the boom; I may remove those and use a through-bolt instead; I'm not entirely sure I trust the screws to hold these on.
 
Back
Top