Orbea SSS (Sunder Semi-Stealth) Q100@1kw

[Sunder]

After months of fartarsing around with gears, derailleurs and lacing the rear wheel, I've finally finished my semi-stealth project. To give you a bit of background, my previous bike was a steel framed no suspension mountain bike with a Magic Pie copy on the front wheel. While it did the job, my complaints were that the bike was heavy (A bit over 30kg), and didn't let me pedal at full speed. (~45km/h). It was also obviously an electric bike.

The new project was designed to rectify those issues by replacing the front motor with a rear Q100. The photo you see above is the result.

The bike now weighs just over 18kg with a 12S2P (10Ah) Lipo battery attached. While it is substantially lighter, the balance now feels a little "off" with almost all the weight behind the seat post.

First impressions was that it is very, very torquey - and very noisy at low speeds. Get above 15km/h though, and it's a near silent whine.

The bike feels a little heavy without power, but nothing like the old DD motor, but this is not a surprise, since my road bike weighs less than 7kg. I can conceivably pedal this bike meaningful distances, and keep up with normal cyclists on flats and very low hill terrain. They'd leave me for dead on any hill though. A tiny amount of power though, and I'm effortlessly doing 30km/h.

Full throttle on this thing is about 30-35km/h. I haven't fitted the bike computer yet, but will update this post when I do. Once on full throttle, I can easily pedal to high 40s, which is actually about what I can sustain on my road bike. I'm surprised with a 44/11t, I can only pedal up to the second last gear (44/13, I think), otherwise my legs start burning... I guess I need more training. I would have thought a 44/11t would top out at high 40s, but then again, my estimates could be wrong.

I intentionally did a very steep hill - maybe a 70-80m gain over about 700m. On full throttle with a bit of pedal assist, I was easily doing 20-25km/h, and the motor felt only slightly warmer than the bike at the end - caveat, it was raining and the motor was clearly wet, so I intend repeating this a dry warm day. It could also be that the outside is (most likely unintentionally) thermally isolated from the parts that get hot. Still need to be careful until I know for certain it can cope.

Overall, I would rate the satisfaction on this bike a 7.5/10. It could still be far better, but considering I rated my last bike only a 5/10, it's a step forward.

Pros:
Very hard to tell it's an electric bike
Much lighter than my last bike
Silent at high speed
Can be easily pedaled
Very torquey up hills and from start.

Cons:
You're not fooling anyone if you power from start. (grrrrrrrreeerrrrr!)
Would have liked a top speed maybe 5-10km/h faster

My wife would kill me if I told her I was upgrading again though, so I think I will stick with this bike for another year or so, then maybe look at another geared hubbie - slightly bigger, maybe hand re-wound with stronger magnets and filled with ATF. That might raise me to a 9/10 satisfaction.

 

[Drunkskunk]

Nice looking bike. Very stealth!

 

[MattyCiii]

Awesome bike & build!

 

[Sunder]

Thanks guys Hopefully I get some time to put the bike computer, and maybe some other meters on there to see how it is really performing. I suspect I'd be one of few running the Q100 over 750w.

 

[d8veh]

Are you running sensorless because that would account for the low-speed noise? They run much smoother and quieter in sensored mode.

You can improve the handling a lot by moving the battery forward. Your aim should be to get it right up against the seat-post and as low as possible. You can modify those struts at the front of the rack by: cutting them down, hammering the ends flat and drilling new holes. You might have to do something similar at the bottom to get it lower. It should be something like this one that has a 20aH battery, and it has no noticeable adverse handling effects.

You can make a better fitting rack out of aluminium tube (4 pieces cut to size) and thin plywood. Hammer the ends of the tubes flat (where necessary), drill holes and file the end to a nice shape. That's how the above one was done. Once you've made the rack, you unbolt the plywood and put it inside the bag and bolt right through so that the bag can't be removed. A zip-tie on the zip makes it tamper-proof. The bag is a football boot bag - much cheaper and better fitting than those from a bike shop. You cut the plywood to the exact shape of the inside of the bag.

 

[Sunder]

Thanks D8veh. With regards to the balance, it's not an issue when riding, only when lifting or pushing the bike, so I'm not going to do anything about it. Thanks for the idea though.

With regards to the sensor/sensorless mode, I definitely have the halls hooked up, but as the controller was scavenged from my old bike, I had to make up custom joins, rather than connectors. Is there any way without breaking the connections that I can find out if the halls are connected, ahd in the right order?

 

[motomech]

You are using the controller from the Magic Pie set-up?
Can you monitor Amps/Watts?
That could explain why your Q100 is not meeting you expectations.
You should have gotten the KU-63, it's cheap and matches the Q100 perfectly.

 

[Sunder]

You are using the controller from the Magic Pie set-up?
Can you monitor Amps/Watts?
That could explain why your Q100 is not meeting you expectations.
You should have gotten the KU-63, it's cheap and matches the Q100 perfectly.

Huh? A KU-63 is a 250w controller, that would be seriously underwhelming. I doubt I could get the kind of pull I am getting out of my current controller with that. It probably would be quieter though, after all, if I just turn the throttle a quarter of the way, it doesn't do the low speed growl.

 

[motomech]

You are using the controller from the Magic Pie set-up?
Can you monitor Amps/Watts?
That could explain why your Q100 is not meeting you expectations.
You should have gotten the KU-63, it's cheap and matches the Q100 perfectly.

Huh? A KU-63 is a 250w controller, that would be seriously underwhelming. I doubt I could get the kind of pull I am getting out of my current controller with that. It probably would be quieter though, after all, if I just turn the throttle a quarter of the way, it doesn't do the low speed growl.

There is a certain irony to your avatar. You realize that is what you are doing to us who know the Cute inside and out.

Huh? A KU-63 is a 250w controller

Yeah, if you use a 22V battery. If you were to read any of the dozens of posts the D8veh and I have made on setting up the Q100, you will see that the Stock Ku-63 puts out 14 Amps and we do a simple shunt mod to bring it up to 17 Amps, which is 750 peak Watts @ 12S.
This is the SWEET SPOT for the Q100, above which it gets growly and lets one know it is not going to last.

You talk about exporing the limits of the Q100, Well, I have already done that, going as high as 22 Amps and 15S, although not together.
My Lyen Mini-monster was way too much @ 12S and I suspect it would be @ 10S as well.
I found, for the 328, going up over 12S will add top speed, but not much to perceived torque.
You also talk about adding ATF, I have done that as well and it poured right out the unsealed axle bearing(although the Brg.s are sealed).
All this is linked in my build thread below.
You have the right to run your system any way you want to, but don't do what you are proporting to be a review of out favorite little motor and slam it if you are running a poorly matched controller.
99% the folks who would buy this motor, would go for the kit, which works beautifuly and to lead them to believe that the Q100 has bad habits is doing the motor(and kit) a disservice.

 

[Samd]

+1. Get a Ku63 and shuntmod it. The guys above have done Q100 to death and know their stuff.

In fact, if you want a Q100 setup to match a mountain bike with appropriate torque (and not a flat bar road-roller) then get two KU63's and another Q100 and dual drive it.

Torque is a function of radius size of the motor amongst other things, but the q100 does do modestly ok despite it's smaller diameter. About half the torque/acceleration of a large geared hub for same amps/volts.

I think Justin has added it to the ebike.ca simulator now so you can simulate torque of these little motors versus both midsized (SWXH/K etc) and big geared motors (MAC/Ezee/BPM 8funs). Roughly 40 newton meters per motor at 36v on a 26" on takeoff at 20 amps. Ok for the city commute.

Thinking of adding a rear one to my city single speed - just for the odd burst to get out of the way.

 

[Sunder]

You are using the controller from the Magic Pie set-up?
Can you monitor Amps/Watts?
That could explain why your Q100 is not meeting you expectations.
You should have gotten the KU-63, it's cheap and matches the Q100 perfectly.

Huh? A KU-63 is a 250w controller, that would be seriously underwhelming. I doubt I could get the kind of pull I am getting out of my current controller with that. It probably would be quieter though, after all, if I just turn the throttle a quarter of the way, it doesn't do the low speed growl.

There is a certain irony to your avatar. You realize that is what you are doing to us who know the Cute inside and out.

Huh? A KU-63 is a 250w controller

Yeah, if you use a 22V battery. If you were to read any of the dozens of posts the D8veh and I have made on setting up the Q100, you will see that the Stock Ku-63 puts out 14 Amps and we do a simple shunt mod to bring it up to 17 Amps, which is 750 peak Watts @ 12S.
This is the SWEET SPOT for the Q100, above which it gets growly and lets one know it is not going to last.

You talk about exporing the limits of the Q100, Well, I have already done that, going as high as 22 Amps and 15S, although not together.
My Lyen Mini-monster was way too much @ 12S and I suspect it would be @ 10S as well.
I found, for the 328, going up over 12S will add top speed, but not much to perceived torque.
You also talk about adding ATF, I have done that as well and it poured right out the unsealed axle bearing(although the Brg.s are sealed).
All this is linked in my build thread below.
You have the right to run your system any way you want to, but don't do what you are proporting to be a review of out favorite little motor and slam it if you are running a poorly matched controller.
99% the folks who would buy this motor, would go for the kit, which works beautifuly and to lead them to believe that the Q100 has bad habits is doing the motor(and kit) a disservice.

Okay, now I'm getting some helpful information out of you. It would have helped had you mentioned that the KU63 was intended to be over-driven, otherwise its stock spec is 24V, 250W.

I read plenty of threads on the Q100 before buying it including plenty from you and D8veh- though now that's nearly 6 months ago, I forgot a fair bit. The things that did stick in my mind was that most people found the motor safe at 750W, but at least one had been running it for a few months at 1000W - hence, why some of my comments above, about being one of the few running above 750W.

Everyone has different requirements - my commute is quite short, and the longest, steepest hill is only 10% grade, and 700m long, which I pedal a fair bit up. I have no doubts that with my current setup, I could ride to the Blue Mountains and do 22% grades for 3km, and kill it in no time flat. However, given that I don't use it that way, I know I'm running the risk, but it's an acceptable risk. If the motor blew, I wouldn't blame the motor at all. I pushed it way out of it's design spec.

On that note, I'm not slamming the motor at all - I gave it a 7.5 out of 10 for my application, which by most grading systems would be a distinction - quite a flattering review, I would have thought.

But back to the issue at hand. If it's "growling" at low speeds, even on flats, but not at high speeds, even on hills, this doesn't sound like the noise is over the 750W "sweet spot". The noise isn't a lot louder than my old DD, so I'm not too concerned that it's telling me it's not going to last - and the fact that it doesn't make that noise when I'm at high speed is even more reassuring. But if it is a preventable noise from a bad hall set up or easily solved with ATF, I'd like to know.

Seriously, no sarcasm here - I appreciate all the things you and d8veh and a few others have written about the Q100 in the past. I probably read for 10-12 hours on this board before deciding to go the Q100 over some other mini-motors. However, also be aware that my needs are not your needs, and I'm using your wisdom as guidance, not as law. I know that I'm not likely to ever climb long or steep hills on my bike, and I'm a pretty fit, low weight cyclist to give aid to the motor. Just because I'm doing things that destroyed your motor, and may well destroy mine... Well, I can't choose whether you bang your head against the table or not. I overclock my computers. I mod my cars. And yes, I overdrive my electric bikes. I haven't killed a computer, car, or electric bike yet. But I'm willing to. A new Q100 would be less than 2 hours wage. I'm happy to experiment, accumulating on the wisdom you've collected.

 

[motomech]

+1. Get a Ku63 and shuntmod it. The guys above have done Q100 to death and know their stuff.

In fact, if you want a Q100 setup to match a mountain bike with appropriate torque (and not a flat bar road-roller) then get two KU63's and another Q100 and dual drive it.

Torque is a function of radius size of the motor amongst other things, but the q100 does do modestly ok despite it's smaller diameter. About half the torque/acceleration of a large geared hub for same amps/volts.

I think Justin has added it to the ebike.ca simulator now so you can simulate torque of these little motors versus both midsized (SWXH/K etc) and big geared motors (MAC/Ezee/BPM 8funs). Roughly 40 newton meters per motor at 36v on a 26" on takeoff at 20 amps. Ok for the city commute.

Thinking of adding a rear one to my city single speed - just for the odd burst to get out of the way.

Exactly.

One could do two "slow winds"(201's) together like D8veh's beautiful trail bike. It does about 22 mph on 12S and I'm sure it climbs very well.



or

Go with two "fast winds"(328's) like I did. More of an urban "assault:" ride, it can reach 27 mph on 12S, but still climb most public road hills with ease. The motors are mounted in 24" wheels.



I don't think combining a slow wind and a fast wind would be great.

 

[Samd]

I found mine a bit noisier on takeoff from the factory and it bedded in a little after 50 km or so. I suspect the gears are cut cheap rather than real smooth but for the price it is fine. I wouldn't expect helical cut german gears for the price.

The KU controller's logic when syncing on startup is coarse from a frequency perspective. Higher frequency controllers growl a bit less. And a pure sine controller would be fun to test. All that would be left would be mechanical noise, not the controller growling as it tried to make a 3 phase sinusoid from chunky linear approximations...

 

[Sunder]

+1. Get a Ku63 and shuntmod it. The guys above have done Q100 to death and know their stuff.

I think Justin has added it to the ebike.ca simulator now so you can simulate torque of these little motors versus both midsized (SWXH/K etc) and big geared motors (MAC/Ezee/BPM 8funs). Roughly 40 newton meters per motor at 36v on a 26" on takeoff at 20 amps. Ok for the city commute.

Thanks Samd, I'm not doubting their wisdom at all, but I'm also not running my bike in the same way they may be. Not going to rehash my last post to Motomech, but trust me, I'm not dissing their experience at all - I relied on a lot of it to choose the Q100 and feel safe reusing my 12S battery and 20A controller.

Re: The sim, I plugged in the closest I could get, this was what I got out. If I understand that correctly, I should be able to get 42.4km/h under ideal conditions? If you consider that the road is never ideal, It's pretty damn close to my "butt dyno" of ~35km/h no pedal numbers. (Still yet to confirm those numbers using a bike computer, could be slightly underestimating)

 

[Sunder]

I found mine a bit noisier on takeoff from the factory and it bedded in a little after 50 km or so. I suspect the gears are cut cheap rather than real smooth but for the price it is fine. I wouldn't expect helical cut german gears for the price.

The KU controller's logic when syncing on startup is coarse from a frequency perspective. Higher frequency controllers growl a bit less. And a pure sine controller would be fun to test. All that would be left would be mechanical noise, not the controller growling as it tried to make a 3 phase sinusoid from chunky linear approximations...

Thanks Samd. Any ideas where we could get better gears for this motor?

 

[Samd]

It's a fair indication on the simulator - although tyres are a significant driver from there. But when comparing two different powerplants you'll usually have the same tyres for each?

You'll see the takeoff torque for each on the left of the chart in blue.

I told a lie above - I usually order the Ku60 controller, which is the 350W one, stock at about 17amps.

I haven't bothered with pursuing other gears, I thought mine were ok once run in. A couple of the guys here are playign with full sine controllers, maybe they'll be standard in a few years to come. But I kinda like the takeoff growl.

Your peak is a fair call, those tyres of yours look damn efficient. Bear in mind that your vehicle is probably getting 1500 to 2000 mpg equivalent in dollars per km terms, while cars are struggling to be fractional of that. So efficiency is good, but some people here get a bit obsessive about an already brutally efficient machine!

 

[Sunder]

Fair enough. After a week, I should have done 50km, and will see whether they bed in.

Maybe for the next evolution, I'll look at pure sine controllers, but as I said in post one, I'm sticking with this bike for a while to keep the wife happy. After all, even though the bike is mostly for fun, it was also to save money. It does however, cost less to mod a bike than it does a car

The tyres are not road tyres, but they're definitely not mountain bike tyres either. It's a hybrid bike I guess, and it's got hybrid tyres. (700x35).

 

[Sunder]

Okay,

I installed a bike computer last night, and gave it a run this morning.

At 47V (~3.9v/cell), the bike has a top speed of 38km/h, no pedal. With a sustainable pedaling pace, I can get that up to around 45km/h. At sprint pace, I can't quite break 50.

There are obviously some inefficiencies here stopping me from getting my estimated top speed. One of those may be wire thickness, and the gears may not be broken in yet - Samd, you're right, after only 25km, the gears are already sounding noticeably quieter. I hope it keeps getting quieter for the next 25km or so, that would make it near ideal.

I have a 14S 5Ah pack lying around somewhere, so I might try that and see if it improves the performance. Since most of my road riding is 50km/h, I would like to be able to pedal to over 50km/h. It really frustrated me this morning that I was doing 45km/h in a 50 zone, and a B double articulated, multi-tonner was tailgating me, then did an illegal/unsafe overtake. In a friggen 50 zone, only a few hundred metres from a school. That truck driver deserves to lose his license.

 

[Samd]

I'm showing my tyre bias, just fitted 2.4inch wide knobbies, so 35's on a 700c for me are right up the rolling thin end!!!

 

[d8veh]

Sunder, I'd be interested to know how many amps you're running at, and keep us informed about how you get on with you're motor because it's adding to our knowledge base. Your growling noise may be just due to the high current. I thought Magic Pies were up around 25 amps. Personally, I can't see it lasting long at that current, but I might be wrong. Also, I seem to remember that someone tested a Q85 and found that current saturated at something very low (14 amps?), so you might be wasting power by going too high. It might be worth trying a controller with less current to see if it makes a difference. You might find that you get the same torque and speed and less growl with a lot less current. For information a guy hooked up a 22 amp controller to his Q100 and it lasted about 5 miles.

 

[Sunder]

Sunder, I'd be interested to know how many amps you're running at, and keep us informed about how you get on with you're motor because it's adding to our knowledge base. Your growling noise may be just due to the high current. I thought Magic Pies were up around 25 amps. Personally, I can't see it lasting long at that current, but I might be wrong. Also, I seem to remember that someone tested a Q85 and found that current saturated at something very low (14 amps?), so you might be wasting power by going too high. It might be worth trying a controller with less current to see if it makes a difference. You might find that you get the same torque and speed and less growl with a lot less current. For information a guy hooked up a 22 amp controller to his Q100 and it lasted about 5 miles.

I've ordered a generic Ammeter from Aliexpress, which should be here in a few weeks, and I'll rig it up along with a similar voltmeter, to test for sag, etc.

So far, I've done about 30km on it, without a problem, a fair bit at WOT, but not much of it on hills. I generally also pedal, so the motor isn't working its hardest without help.

I have a feeling you could be right, that I'm wasting a lot of current as heat, so I may get the same performance with less power. Does the mini-monster have a programmable way of restricting current? Might that be an easy way out to test?

 

[d8veh]

Minimonster is programmable, once you get your head round how to do it, but be warned, it's possible to make mistakes, and you have to find the right wiring combo because I believe phase/hall wire colours don't match, which gives new possibilities to burn something.
I always use the KU63 or KU65 with these motors and solder the shunt. Standard current is 15 amps and you can get them up to about 22 amps by soldering the shunt. With a watt-meter it's very easy to do because the shunt is directly behind the end-plate of the controller. I connect up with my watt-meter; lift the wheel; apply full throttle; apply the brake to slow the motor right down and check the current; add some solder - say 25% of the shunt; check current again; repeat until it's right. You can always come back another time to add more. The whole procedure takes a couple of minutes - less time than it takes to program a controller. The Lyen controllers have better FETs, so probably run colder at high current, which would be my main reason for getting one.