AWD Quad build in planning: "Satanic Panic"

[gman1971]

@toecutter, you have a rear suspension on your KMX, and the frame acts as a torsion bar when one wheel wants to go up/down, so yours is probably as close to a full suspension setup is going to get without having to dor a full CV axle setup.

 

[The Toecutter]

@toecutter, you have a rear suspension on your KMX, and the frame acts as a torsion bar when one wheel wants to go up/down, so yours is probably as close to a full suspension setup is going to get without having to dor a full CV axle setup.

I'm very pleased with it too. It has a better ride quality than many cars I've driven, at the same speeds on similar road conditions. I might just build another with a RWD middrive layout with the weight distribution set up to handle like a mid-engine rear-drive sports car. But before I do that, I'm going to finish this one.

 

[PaPaSteve]

How much $$$$ for that quad? It looks like a live axle rear suspension.... or an articulated rear suspension? (ie. no suspension)

Rear axle drive works from two BMX ratchets on the center carrier.
In the cycle world it is known as "dominate drive" because when turning the outside wheel dominates the traction and the inside wheel slips (ratchets).
During a turn the rider feels some push back in the pedals due to the wider circle the outside wheel travels.
The equivalent in the automotive world is called a "Detroit locker".
Linkage on the rear is called a 3 link made popular on the Lotus Super 7 designed in 1957, it does articulate uneven terrain well.

Yes . . . the quad was pricey . . .

 

[The Toecutter]

It is the merging of bicycle and sports car that can create the most efficient form of individual, practical, personal transportation possible. So of course design inspiration was derived from the Lotus Super 7. 2-wheeled streamliners could be even more efficient, but they wouldn't be practical for normies to live with.

Want a car that gets 4-digit fuel economy? Take a velomobile and turn it into a high-performance microcar. Give it microcar axles, microcar suspension bits, microcar wheels/tires, ect. all built to safely handle high speed. In mass production, we could have an individual transportation solution that goes for miles per penny of electricity, accelerates and corners like a race car, and costs about the same as a mid-range diamondframe unmotorized road bike. The key to the price point being mass production: if you hand-build it, it will cost at least 5-figures, but even then, the performance may still more than justify the price tag. No new infrastructure needed: use the roads.

No wonder a chain of industries want to shut down e-bikes and heavily restrict/regulate them.

 

[PaPaSteve]

It looks like a live axle rear suspension.... or an articulated rear suspension? (ie. no suspension)

Perhaps I've misunderstood your comment.
I think of "live axle" in terms of a gokart where both driving tires are fastened to a common axle with no speed deferential mechanism between the two tires.

 

[PaPaSteve]

Toecutter . . . want to say, I appreciate and value your project.
Always wonderful to see the progress.

 

[Papa]

In the cycle world it is known as "dominate drive" because when turning the outside wheel dominates the traction and the inside wheel slips (ratchets).

You got it backwards - Because the inside wheel is rotating slower than the outside wheel when turning, its ratchet paws lock, effectively slowing the input RPM. The outside wheel (rotating faster) just freewheels, resulting in single wheel drive.

Two independent freewheels is NOT a true 'differential'.

On dry asphalt, a true differential applies uninterrupted torque to BOTH wheels... even when both wheels are rotating different RPM. And yes, if one wheel loses traction, torque applied to the wheel with grip is reduced or unavailable. But reality is, contact patch grip loss is generally a non-issue on hard surfaces under most 'prudent' driving conditions.

 

[PaPaSteve]

You got it backwards - Because the inside wheel is rotating slower than the outside wheel when turning, its ratchet paws lock, effectively slowing the input RPM. The outside wheel (rotating faster) just freewheels, resulting in single wheel drive.

Two independent freewheels is NOT a true 'differential'.

You are correct, I got the inside and outside wheels description backwards but not the effect of back pressure to the pedals.
Wider track widths result in higher feedback pressure to the pedals.
And . . . I"m fully aware "dominant drive" does not function the same as a "differential".
Puzzled how the words in my posts were misconstrued in that way.
The double freewheel ratchet is just a simple, low cost, low weight method to have two wheels drive a low powered vehicle.
Will this method survive higher power and higher speed is yet to be seen

 

[gman1971]

Its like what they'd call a "one-way diff" in RC car terms. It acts as a live axle when both wheels are planted on the ground, and allows the outer wheel to freewheel when taking turns, as described. But the nice thing is that when pedaling, if one wheel goes up in the air, the wheel that remains in contact with the ground still gets the full force from the pedals, so you don't feel like pedaling in air. An open diff will basically be very inefficient for HPV pedal b/c if one wheel goes up in the air, you'll effectively be free pedaling the quad/trike.

Yeah, live axle like in a pickup truck, the rear axle has a diff but both wheels are locked together, but seems like what you'd see in an articulated dumper/excavator, great for uneven terrain.

Highly recommended if you can do this trike to quad conversion..

 

[The Toecutter]

Highly recommended if you can do this trike to quad conversion..

Seconded. It just sucks that my only option for decent gearing is to either engineer my own freewheel to work with their components, or rig up an IGH as a mid drive somewhere on the frame. 200% is not enough range for 7sp.

 

[The Toecutter]

Took it out again. I decided to limit the CA3 to 1000W to see what top speed it would allow. The CA3 read 1,065W and I was doing 38.2 mph. Flat ground, smooth road, no wind. I was NOT expecting to be this efficient with no body or even farings of any sort(only wheel disc covers). That number just seems wrong.

I'm going to physically measure my tire circumference to make sure I have it set correctly. I set it based on a chart for my tire size. That could change the results if I'm getting an overestimate of the speed readout.

 

[99t4]

I'm going to physically measure my tire circumference to make sure I have it set correctly.

Best method I've found (you may already know this) is to put a dot of paint on the tire tread and roll the bike (loaded for best accuracy) over smooth pavement, then measure the distance between the paint dots left on the pavement.

 

[The Toecutter]

Best method I've found (you may already know this) is to put a dot of paint on the tire tread and roll the bike (loaded for best accuracy) over smooth pavement, then measure the distance between the paint dots left on the pavement.

I just did this with some dijon mustard. My wheel was actually undersized by 0.5" in the computer. I had it set at 18.9", when it was actually 19.4". I was doing over 39 mph on just over 1 kW, and more like 56.0 mph 2 weeks ago, not 54.6 mph. I might try and see if I can hit 90 mph on the bike trail today.

 

[The Toecutter]

There were too many people out on the trail for me to do a top speed run. I usually slow down to 2x walking speed when I see them.

I did find that travelling in one direction has a slight downslope and the other direction an upslope. I thought it was flat, but it's not. The readings make more sense in this context. Maybe a 0.2-0.5% gradient, it's not much

Going the other way, a similar 1,054W got me to 34.1 mph. So somewhere around 36-37 mph @ 1,000W on true flat ground, throttle-only no pedaling.

That's still really good for not having a body on it. When it was a 3-wheeler with the coroplast shell on it, throttle-only would have used ~500W at the same speed, and I'd normally pedal and get by using ~9 Wh/mile in those conditions.

My next body shell is going to fair all 4 wheels so it will be a lot more efficient than the last one. Not as efficient as the Milan SL I own, but much closer to the Milan aerodynamically than it would be to my last body shell design with outboard wheels was. I want to do 60 mph on ~1500W and ~100 mph on ~4500W.

 

[The Toecutter]

I found out why the CA3's power reading was so low. I didn't double the RShunt value to accommodate dual controllers.

The CA3 was only reading half the power consumed. So 36-37 mph actually takes more like 2 kW, not 1 kW. This is also in line with expectations: I thought this vehicle would have more aero drag and more mass than the mountain bike, and therefore consume more energy. Lo and behold, it does. I was getting incorrect data.

With that said, I can't explain the almost 9 kW reading two weeks ago under hard acceleration, because holding the above logic consistently, that would imply 18 kW draw, and the two Phaserunners combined are incapable of that peak
@ 72V. At least I know my battery can handle it and that even if it somehow did manage to draw that, it would have been well below the battery's continuous power rating. Perhaps over a certain amount of power demand it starts to read accurately and it actually was 9 kW. Similar issue when I hit an indicated 54.6 mph at slightly below 6 kW indicated two weeks ago. It was still accelerating quite hard and I programmed the limit to 12 kW with each controller programmed to 6 kW peak in the Phaserunner Suite, so there's no way it was actually double the 6 kW @ 12 kW at a steady speed, since when I had the throttle in all the way at that speed it was still accelerating hard prior to that steady cruising and having a lot more speed to offer. I'm reasonably certain it was very close to 6 kW @ 56 mph(54.6 mph indicated). But at lower speeds, I was definitely using double the power the CA3 indicated.

Lesson learned, and not the hard way: I need to put a high current shunt in this bitch before I blow something up and/or fry the CA3 and/or fry my torque sensing bottom bracket.

 

[gman1971]

Ic.. 40mph on A-1, with the fairing it had before the quad conversion was ~2kW... so 36 @2kW seems more plausible. Glad you figured it out... a 50% error could blow stuff up! also give range expectations that are just wrong. Usually measuring avg power at the charger could help track these sort of discrepancies. I use GPS speedometer on all my vehicles as well as the CA one too.

G.

 

[99t4]

dijon mustard

Haha, more convenient that cracking open the paint can of course. Now why didn't I think of that? Well next time I'm going to use mayonnaise!

 

[The Toecutter]

An example of a Nexus 8 on a Quad
Built a few of these with a Rolhoff.
Note : when using a IGH powered by a motor there needs to be a momentary power cutoff to allow a shift.
View attachment 386935

Would you mind getting some closeup photos of how the Rolhoff IGH is set up? I am trying to see what parts I need to order and/or custom fabricate in order to set this up on my KMX frame. I haven't ordered a Rolhoff yet, but if there is a variant and/or existing parts that allow for me to have two chain systems on it instead of built as a wheel hub, then I'd like to avoid the headache of making custom parts in favor of a solution that pieces together as much as possible.

 

[PaPaSteve]

Rolhoff was used on the first few builds, the later builds were Nexus 8's.
This photo shows a dust covered Nexus with sprocket bolted to the spoke flange with a spacer used for a best chain line.
Sprocket could be bolted to either flange.
Dropouts are vertical slots and chain tension maintained with spring and pulley.
Trying to imagineer your application, a possible solution is a roller clutch on the motor shaft.
I have a bunch of sprocket drawings for many different application that are easy to mix or match in CAD.
Used to make them in either Grade 5 Titanium or 7075 aluminum at work on a waterjet.
These days I use SendCutSend . . . often their delivered price is competitive to what I can purchase the raw materials for.
Think it, Make it, Ride it.

 

[PaPaSteve]

Instead of a roller clutch a freewheel adapter on the motor shaft might work.

Freewheel to Axle Adapters - ElectricScooterParts.com

Freewheel to axle adapters.

electricscooterparts.com

 

[A-DamW]

Think it, Make it, Ride it.

This is the way...