High-speed, long-distance (heavy "30^3") Q128 commuter build

Get all your technical information about electric bikes here.
mystryda
100 mW
100 mW
Posts: 38
Joined: Jul 11, 2017 10:39 pm
Location: Northern VA (metro DC)

High-speed, long-distance (heavy "30^3") Q128 commuter build

Post by mystryda » Oct 05, 2017 10:29 pm

Alright, so the batteries finally arrived (thanks Tumich!), and I'm ready to start figuring out everything else. (150) of his LG MJ1 - 3500mAh/10A 18650 cells just arrived.

I want to use the Q128C-135mm 500W CST with my road bike to help me average 30 mph on my commute. I want to build something similar to Chas58's "25^3," but more powerful. (I'm an experienced cyclist and will be peddling full bore, on a road bike, in a tuck, etc. I'm already accustomed to high speed, 40 mph+ descents, and I average in the low- to mid-twenties on my current commute, so I know what I'm getting into as far as the speeds are concerned.)

Based on the models, I'm thinking to run the 328RPM / 36V shred of the Q128C @ 48V, but I do have a nagging concern that 48V might not be the right voltage point. Most of the time I plan to ask for 250W and 500W, with possibly 750W peak, out of the Q128. At 48V, that's about 5A, 10A, and 15A, respectively, for those power levels. Even 48V isn't the best voltage point, I do think that it's fairly close.

I'm really confused about controllers and would appreciate some help here. Would you all point me to what I want? Here's what BMSbattery.com has:
--S12SH 800W 40A Sine Wave Controller with Light Output
--S12SN 72V1KW 40A Torque Simulation Sine Wave Controller
--S12SN 1000W 40A Torque Simulation Sine Wave Controller
--S12SH 800W 35A Torque Simulation Sine Wave Controller
--S12S 500W Torque Simulation Sine Wave Controller
--S12P 750W Torque Simulation Square Wave Controller
--S09P 500W Torque Simulation Square Wave Controller

I "think" that the S09P meets my needs, or at least pretty darn close, since the ~14A throughput that it does for 500W @36V is really close to the 15A peak for 750W @48V that I'm interested in.

I'd like to preserve the possibility of adding Cycle Analyst v3 at some point.

As far as display/interface, I'm leaning towards the S-LCD3 or S-LCD6, mostly because of the remote buttons, but here's what BMSBattery offers. Should I get the 72V version of the S-LCD3 in case I end up at 14S/52V or 15S/56V?
--S-LCD6 LCD Meter for S-Series Controllers
--S-LCD5 LCD Meter for S-Series Controlers
--S-LCD3 LCD Meter for S-Series Controlers
--72V S-LCD3 LCD Meter for S-Series Controlers

So what else do I need? I'm thinking:
--A Pair of EBike Torque Arm
--HWBS - Hidden Wire Brake Sensor 1pcs Two of these, of course, right, one for front and one for back?
--Ten poles PAS--Pulse Padel Assistant Sensor -- I am very confused about which of these sensors to use. There's also the Dual Hall Sensor 12 Signals Easy Assembling PAS
--Copy Wire for S Series Display Meter -- to set up a hot spare controller
--Thumb Level Throttle without Handle

I'm a year-round, all-weather (hot/cold, rain, snow) commuter, and reliability is a huge issue. I do plan on buying two of everything, so that I have hot spares. Is there anything that I should buy three of?

And I'm not married to buying from BMSBattery.com. I know that they don't have the best reputation, but they seem to be the only source for the Q128-C, and I thought that it would be easier to get everything from one place. I'm happy to hear suggestions of products from other sellers if they're superior. I hope not to spend an arm and a leg on shipping, though.

Thanks in advance.

User avatar
motomech
100 MW
100 MW
Posts: 2888
Joined: Sep 11, 2010 12:21 am
Location: Tucson Az.

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by motomech » Oct 06, 2017 4:04 am

mystryda wrote:Alright, so the batteries finally arrived (thanks Tumich!), and I'm ready to start figuring out everything else. (150) of his LG MJ1 - 3500mAh/10A 18650 cells just arrived.

I want to use the Q128C-135mm 500W CST with my road bike to help me average 30 mph on my commute. I want to build something similar to Chas58's "25^3," but more powerful. (I'm an experienced cyclist and will be peddling full bore, on a road bike, in a tuck, etc. I'm already accustomed to high speed, 40 mph+ descents, and I average in the low- to mid-twenties on my current commute, so I know what I'm getting into as far as the speeds are concerned.)

Based on the models, I'm thinking to run the 328RPM / 36V shred of the Q128C @ 48V, but I do have a nagging concern that 48V might not be the right voltage point. Most of the time I plan to ask for 250W and 500W, with possibly 750W peak, out of the Q128. At 48V, that's about 5A, 10A, and 15A, respectively, for those power levels. Even 48V isn't the best voltage point, I do think that it's fairly close.

I'm really confused about controllers and would appreciate some help here. Would you all point me to what I want? Here's what BMSbattery.com has:
--S12SH 800W 40A Sine Wave Controller with Light Output
--S12SN 72V1KW 40A Torque Simulation Sine Wave Controller
--S12SN 1000W 40A Torque Simulation Sine Wave Controller
--S12SH 800W 35A Torque Simulation Sine Wave Controller
--S12S 500W Torque Simulation Sine Wave Controller
--S12P 750W Torque Simulation Square Wave Controller
--S09P 500W Torque Simulation Square Wave Controller

I "think" that the S09P meets my needs, or at least pretty darn close, since the ~14A throughput that it does for 500W @36V is really close to the 15A peak for 750W @48V that I'm interested in.

I'd like to preserve the possibility of adding Cycle Analyst v3 at some point.

As far as display/interface, I'm leaning towards the S-LCD3 or S-LCD6, mostly because of the remote buttons, but here's what BMSBattery offers. Should I get the 72V version of the S-LCD3 in case I end up at 14S/52V or 15S/56V?
--S-LCD6 LCD Meter for S-Series Controllers
--S-LCD5 LCD Meter for S-Series Controlers
--S-LCD3 LCD Meter for S-Series Controlers
--72V S-LCD3 LCD Meter for S-Series Controlers

So what else do I need? I'm thinking:
--A Pair of EBike Torque Arm
--HWBS - Hidden Wire Brake Sensor 1pcs Two of these, of course, right, one for front and one for back?
--Ten poles PAS--Pulse Padel Assistant Sensor -- I am very confused about which of these sensors to use. There's also the Dual Hall Sensor 12 Signals Easy Assembling PAS
--Copy Wire for S Series Display Meter -- to set up a hot spare controller
--Thumb Level Throttle without Handle

I'm a year-round, all-weather (hot/cold, rain, snow) commuter, and reliability is a huge issue. I do plan on buying two of everything, so that I have hot spares. Is there anything that I should buy three of?

And I'm not married to buying from BMSBattery.com. I know that they don't have the best reputation, but they seem to be the only source for the Q128-C, and I thought that it would be easier to get everything from one place. I'm happy to hear suggestions of products from other sellers if they're superior. I hope not to spend an arm and a leg on shipping, though.

Thanks in advance.
I can only give you some things to think about, whereas I'm sure that D8veh will stop by and give you deffinitive answers, but here goes anyhow.
Ebike CA sim shows that a 328 in a 27.5 whl. on 48V (road bike-tuck)will top out @ 29 mph(at least 700 Watts) and a rider adding 150 to 200 Watts will raise that by a couple mph. So your power est.s for that desired avg. speed are way under estimated. Even high 20's mph avg. speeds seems a little high for 48 V.s. The problem is, sine wave controllers can be "picky" about the pack Voltage they see the first time, when the values are set.
Even if you stay w/ 48V to be compatible w/ a sine wave and a more sensible high 20's top speed is the goal, not a lot of amps are needed. In light of that, I would suggest you stay more in the spirit of chas58's build and use one of the new 20 Amp, 6-FET KT controllers from PWS Power. They can be shunt-modd'ed to 25 A, if you feel the need to. The S12 series controllers from BMS B. are huge and even the S9 series are a little large for my tastes.
I'd like to preserve the possibility of adding Cycle Analyst v3 at some point.
Why? The SLCD displays offer all the info a rider would ever want. In fact, as time goes by, less and less info is desired and you will quickly get to the point where you would galdly trade h.bar space for less info.
So what else do I need? I'm thinking:
--A Pair of EBike Torque Arm
--HWBS - Hidden Wire Brake Sensor 1pcs Two of these, of course, right, one for front and one for back?
--Ten poles PAS--Pulse Padel Assistant Sensor -- I am very confused about which of these sensors to use. There's also the Dual Hall Sensor 12 Signals Easy Assembling PAS
--Copy Wire for S Series Display Meter -- to set up a hot spare controller
--Thumb Level Throttle without Handle
Yes, 2 pr.s of TA's. One pr. will almost certainly work on one side of the motor. The other side usually will require a fabb'ed link in conjuction W/ the BMS B. axle piece, or perhaps a different TA altogether. I took the easy route and used the pricey, but very useful V4 from Ebike CA. Also, for ease of installation, i use their 2-piece, snap together PAS disc. But if you don 't mind pulling the crank, you can use any of the BMS B. disc.s, although I read some reports about prob.s w/ the Dual Hall a while back(D8veh would know).
I have also read reports that HWBS units are junk, but at their price, you could try. Like most w/ disc brk.s, I have a fabb'ed reed sensor/magnet on my lever/lever bracket.
I don't believe the copy wire is for down-loading to the controller, but is for sharing info w/ a second display
You might order a left-hand, half-twist throttle to try as many perfer this throttle for less fatique, although w/ a nice PAS system, that is not much of a concern. At any rate, it's cheap and BMS B. is one of the few places who sell them.
Don't forget the spoke wrench.
Is there anything that I should buy three of?
No, 2 of everything should be plenty. But I would double ck the batt capacity you intend to carry for your proposed distances. Based on your power requirement estimates, you may have under-estimated that. I am of the mind that one can't carry too much batt. capacity up to a reasonable weight. But I use LiPoly, so I can carry lot's of capacity(25 Ah of 12S) for little weight penality.
100_0066.JPG
100_0066.JPG (66.38 KiB) Viewed 1291 times
Ebike CA torque arm on Ezee V1
100_0087.JPG
100_0087.JPG (93.72 KiB) Viewed 1291 times
Reed switch epoxied to brk. lever bracket
magnet is stuck on lever w/ sticky tape.
Motomech

'03 Rocky Mountain Edge 2WD 260 Q100H frt and Ezee V1 rear 2 Elifebike 20A & 25A 9-FET controllers 12S/20Ah Multistar Lipo rear 5Ah Turnigy frt Luna Cyclops Extra lite Alex 24DM rims, 2.4 Holly Rollers run ghetto tubeless. 25 mph. Mean Well HLG-320H-48A
https://endless-sphere.com/forums/viewt ... =3&t=83430
'07 GT Idive 4 4.0, Q100C 201 12S LiPoly elifebike 9-FET 17A controller. 20 MPH.
https://endless-sphere.com/forums/viewt ... 8#p1237928

mystryda
100 mW
100 mW
Posts: 38
Joined: Jul 11, 2017 10:39 pm
Location: Northern VA (metro DC)

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by mystryda » Oct 06, 2017 7:09 am

motomech wrote:Ebike CA sim shows that a 328 in a 27.5 whl. on 48V (road bike-tuck)will top out @ 29 mph and a rider adding 150 to 200 Watts will raise that by a couple mph. So your power est.s for that desired avg. speed are way under estimated. Even high 20's mph avg. speeds seems a little high for 48 V.s.
Oh, ebike.ca changed their simulator since I looked at it last. I definitely see where you're coming from. I do see that they changed the possible profiles to "Race Bike (tuck)." Leaving the 'e' out of "ebike," most "regular" cycling calculators are more optimistic for a road cyclist in an aerodynamic position than ebike.ca. But it is what it is. Estimating drag is hard, let alone modeling all the other variables, and I don't have a dyno, so I'm just going to have to wait and see what happens when I get this thing built. I'll be certain to post my results in the end.

What about running at 72V and trimming back the controller? Is there a downside to the higher voltage? If nothing else, the batteries would be easier to put together. (Tumich's packs came in 36V/10S5P.)
motomech wrote:The problem is, sine wave controllers can be "picky" about the pack Voltage they see the first time, when the values are set.
Do you mean the first time that it's turned on each day?
motomech wrote:Even if you stay w/ 48V to be compatible w/ a sine wave and a more sensible high 20's top speed is the goal, not a lot of amps are needed. In light of that, I would suggest you stay more in the spirit of chas58's build and use one of the new 20 Amp, 6-FET KT controllers from PWS Power. They can be shunt-modd'ed to 25 A, if you feel the need to.
I forgot that you mentioned this before; thanks. Is this the one you mean?
36V/48V 500W 6Mosfets 20A Brushless DC Sine Wave Controller
I see that they have a "torque simulation" controller that's 20A, but that's 9 FETs. I don't get the difference between sine wave vs torque simulation, other than torque simulation doesn't use the hall-effect sensors (yes?).
motomech wrote:
mystryda wrote: I'd like to preserve the possibility of adding Cycle Analyst v3 at some point.
Why? The SLCD displays offer all the info a rider would ever want. In fact, as time goes by, less and less info is desired and you will quickly get to the point where you would galdly trade h.bar space for less info.
With two sets of lights and the cycle computer, I'm already out of bar space on the drop handlebars on my road bike, so I understand your point. That's one of the reasons why the LCD6 display seemed more attractive: it looked slimmer than the LCD3.

I thought that Cycle Analyst 3 had a fairly mature cruise control capability, though, which seemed like it might be an attractive later upgrade.
motomech wrote: You might order a left-hand, half-twist throttle to try as many perfer this throttle for less fatique, although w/ a nice PAS system, that is not much of a concern. At any rate, it's cheap and BMS B. is one of the few places who sell them.
This is spring-loaded like a motorcycle throttle, right? I'm not sure how I'd get that to work on drop handlebars. Am I being unimaginative?

I know chas58 talked about a doorbell switch on his road bike. If I understood correctly (and I'm not sure that I did), it was being used to toggle PAS. Single button would be awesome.
motomech wrote:Don't forget the spoke wrench.
Just a regular spoke wrench, yes? I'm planning on buying just the hub and building up the rear wheel. (Not my first.)
motomech wrote:But I would double ck the batt capacity you intend to carry for your proposed distances. Based on your power requirement estimates, you may have under-estimated that. I am of the mind that one can't carry too much batt. capacity up to a reasonable weight. But I use LiPoly, so I can carry lot's of capacity(25 Ah of 12S) for little weight penality.
Well, what I got from tumich is close to 2kWh. The plan is to charge to 4.1V at both home and work, and I only want to exercise about 30% of the capacity, so I'm trying to limit my typical trip to a total of 750Wh. But it is what it is.

Frankly, I'm still trying to figure out how to mount what I've already bought in a way that I can easily remove them at work to take in to charge. I can't find a triangle bag large enough to hold (150) cells. The FalconEV bag seems to be the biggest, but by my measurement it maxes out around a hundred cells. Best I've come up with so far is getting a pair of midsize triangle bags and mounting them on both sides of my headset, sorta like panniers-meets-tank-bag, but clearance is going to be a real challenge. I might have to do as many as I can in a main triangle triangle bag and then a smaller bag for another. Or maybe I'll end up buying a couple yards of canvas and foolishly try to rig up something myself that's detachable.

Thanks for the pics of your build.

User avatar
dogman dan
100 GW
100 GW
Posts: 34330
Joined: May 17, 2008 12:53 pm
Location: Las Cruces New Mexico USA

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by dogman dan » Oct 06, 2017 7:14 am

48v for sure, or higher, (14s 52v) if you want 30 mph.

You won't have that same speed the whole ride, because at the end of the ride you will be below 48v actual voltage after sag under load and lower voltage no load when the pack is not full anymore.

Figure on more like a 25 mph without pedaling speed average on the ride. 2 mph more from pedaling may be close to the max you get. Tuck helps, but you got no peloton to hide in. On the other hand, hauling ass in a city, you better be out of the tuck, watching for that guy left crossing you. He did not see you. Figure on they try to kill you 2-6 times a day.

You must already know how much you will have to stay on top of wheel tuning riding that fast a lot. With the motor wheel, you will have to watch those spokes even more on that wheel. Particularly the first 100 miles, till they stretch in and settle down. It will be like race riding every day, so you will have to stay on top of everything, spokes, brake adjustments, etc.

If your gear is not above 52 t up front, try to find a larger one, if it will fit the bike without frame rub. You'll need a high gear to cruise 30, and still pedal with strength. You'll tend to end up spinning and only really putting in 50w if the gear is not high. 56t is great if you have only 14t in back.

mystryda
100 mW
100 mW
Posts: 38
Joined: Jul 11, 2017 10:39 pm
Location: Northern VA (metro DC)

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by mystryda » Oct 06, 2017 9:20 am

dogman dan wrote:48v for sure, or higher, (14s 52v) if you want 30 mph.
Well, in terms of waste and balance 15s would work out really well, since I could make a 15s10p with the cells that I have right now. I'm really happy to consider any operating voltage, since I'm building the battery from Tumich's cells. My only concern there is sourcing the charger.
dogman dan wrote:If your gear is not above 52 t up front, try to find a larger one, if it will fit the bike without frame rub. You'll need a high gear to cruise 30, and still pedal with strength. You'll tend to end up spinning and only really putting in 50w if the gear is not high. 56t is great if you have only 14t in back.
I currently top out with my 53/14 at 31-32mph in the flats on a good day, but it's a hard cadence to hold. I'll have to do some measuring to see what fits. I'm on a 52cm R500 (bikepedia post here) and I've put an outrageously long stem and seatpost on it, since I should really be on something like a 54cm. One of the things I'm considering is procuring something that fits me better, which might also make it easier to a) fit the batteries and b) ride a larger chainring.

User avatar
motomech
100 MW
100 MW
Posts: 2888
Joined: Sep 11, 2010 12:21 am
Location: Tucson Az.

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by motomech » Oct 06, 2017 12:36 pm

What about running at 72V and trimming back the controller? Is there a downside to the higher voltage? If nothing else, the batteries would be easier to put together.
Nooooooo :shock: The Q128 is basicly a bigger Q100 and we well know the Voltage limits of that motor as they have been tested to destruction. Phase wires will start to melt first and by the time the Voltage gets into the 60's Volt range, the winding are likely to start cooking.
(Tumich's packs came in 36V/10S5P.)
Can you reconfigure them to 48V? There is the Q128 that is rated 328 @ 48V, which is an even higher speed motor, but I fear too high for what you want to do. What you would have w/ this motor on 36V would be like the Q100 328(which is rated @ 36 V), that chas58 , crossbreak and a few others use. But these are super riders. probably in the top 10% on human powered bikes, so they can make high-speed systems work without controller over-heating while climbing. I'm from DC, so I know there are plenty of hills, so unless you want to reroute to avoid the steepest of them, you need to ask yourself;, "how strong of a rider am I really?"
motomech wrote:
"The problem is, sine wave controllers can be "picky" about the pack Voltage they see the first time, when the values are set."

Do you mean the first time that it's turned on each day?
The Sine waves w/ their SLCD displays are involved and should be your next field of study. But suffice to say, they are Voltage "auto ranging". The first time they "see" the pack Voltage, the values are fixed for the LVC and "fuel gauge" and non-traditional pack Voltages can skew their readings.
I forgot that you mentioned this before; thanks. Is this the one you mean?
36V/48V 500W 6Mosfets 20A Brushless DC Sine Wave Controller
Yes
I see that they have a "torque simulation" controller that's 20A, but that's 9 FETs. I don't get the difference between sine wave vs torque simulation, other than torque simulation doesn't use the hall-effect sensors?
They both use Hall sensors.
"Torque Imitation" refers to power limiting by limiting the Current, rather than the more crude approach of speed limiting
All Chinese sine wave controllers are Torque Imitation, while some square wave controller are T.I. The more powerful the system and resultant greater road speed ranges(You), the greater the benefits the T.I. system. The 5-speed limiting offers more flexibility and reduces any gaps between the ranges. The Current limiting reduces any "surges" when the PAS is engaged or the speed range is changed. When this is done on the speed range system, the motor wants to "rush" up to the next higher limit. The TI PAS system offers the best PAS experience short of the very expensive "strain-gauge" systems like the Ebike ca system or the Bionx.
With two sets of lights and the cycle computer, I'm already out of bar space on the drop handlebars on my road bike, so I understand your point. That's one of the reasons why the LCD6 display seemed more attractive: it looked slimmer than the LCD3.
I agree, the SLCD-3 mounts w/ 2 legs, so one needs two avail. flats 4 or 5 inches apart. Something hard to come by on busy drop bars.
I thought that Cycle Analyst 3 had a fairly mature cruise control capability, though, which seemed like it might be an attractive later upgrade.
W/ a good PAS system, cruise need not be used. Just because the pedals are turning, one could be "faux" pedaling, which is really almost the same as not pedaling.

This is spring-loaded like a motorcycle throttle, right? I'm not sure how I'd get that to work on drop handlebars. Am I being unimaginative?

I know chas58 talked about a doorbell switch on his road bike. If I understood correctly (and I'm not sure that I did), it was being used to toggle PAS. Single button would be awesome.
I.M.O., single button is a bad idea. W/ the better PAS systems avail. today(not around when chas58 did his build), just mount the thumb where you can and use the PAS almost all the time. W/ the 5-speed T.I. system, you can stay in PAS all the time and change your speed by rowing the limit button up and down. That's why you will want the display controls close to your free hand.

Just a regular spoke wrench, yes? I'm planning on buying just the hub and building up the rear wheel. (Not my first.)
You will likely use 13 ga. spokes, not a bicycle size. Unless you want a wider rim, it's hard to justify lacing your own. The problem for me has been, I usually end up having to buy a box of 100 spokes to get the size I want, which negates any saving on the shipping. At least the DT Swiss's don't get surface rust like the Chinese spokes do. But, I haven't had any problem w/ the BMS B. whl. ASM's.
Well, what I got from tumich is close to 2kWh. The plan is to charge to 4.1V at both home and work, and I only want to exercise about 30% of the capacity, so I'm trying to limit my typical trip to a total of 750Wh. But it is what it is.

Frankly, I'm still trying to figure out how to mount what I've already bought in a way that I can easily remove them at work to take in to charge. I can't find a triangle bag large enough to hold (150) cells. The FalconEV bag seems to be the biggest, but by my measurement it maxes out around a hundred cells. Best I've come up with so far is getting a pair of midsize triangle bags and mounting them on both sides of my headset, sorta like panniers-meets-tank-bag, but clearance is going to be a real challenge. I might have to do as many as I can in a main triangle triangle bag and then a smaller bag for another. Or maybe I'll end up buying a couple yards of canvas and foolishly try to rig up something myself that's detachable
Stay w/. the center bag, much cleaner and easier(you can stuff extra wire length in there and you will have some). What you may not have thought of this, but the frame bags expand outward in width limited only by your knees starting to hit. My XL EV Falcon bag is huge, to the point that it's weight, not space that dictates what I can carry. But I bought my XL framed Rocky Mountain with the idea of fitting the XL bag.
100_0064.JPG
100_0064.JPG (73.17 KiB) Viewed 1245 times
I also use a container on the frt. downtube to carry extra capacity;
100_0076.JPG
100_0076.JPG (115.41 KiB) Viewed 1245 times
On my other bike, I made my own "Water bottle" batt w/ LiPoly bricks. It would work even better w/ round cells;
100_0107.JPG
100_0107.JPG (89.08 KiB) Viewed 1245 times

I have been thinking, if you truely are stuck w/ a 36V system, you might think about using a BPM or MAC. You would have more motor speed options and more importantly, you could run much higher Amps, like the S12's from BMS B. I would not run more than 25 A or so on the Q128, even @ 36 Volts. You probably have to use a DNP free wheel, but they are not as bad as people say and you will only be using 2 or 3 grear anyway. BTW, most folks never use the frt. shifter and remove from the H.bars making the space for your display.
Motomech

'03 Rocky Mountain Edge 2WD 260 Q100H frt and Ezee V1 rear 2 Elifebike 20A & 25A 9-FET controllers 12S/20Ah Multistar Lipo rear 5Ah Turnigy frt Luna Cyclops Extra lite Alex 24DM rims, 2.4 Holly Rollers run ghetto tubeless. 25 mph. Mean Well HLG-320H-48A
https://endless-sphere.com/forums/viewt ... =3&t=83430
'07 GT Idive 4 4.0, Q100C 201 12S LiPoly elifebike 9-FET 17A controller. 20 MPH.
https://endless-sphere.com/forums/viewt ... 8#p1237928

mystryda
100 mW
100 mW
Posts: 38
Joined: Jul 11, 2017 10:39 pm
Location: Northern VA (metro DC)

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by mystryda » Oct 06, 2017 4:32 pm

motomech wrote:
What about running at 72V and trimming back the controller? Is there a downside to the higher voltage? If nothing else, the batteries would be easier to put together.
Nooooooo :shock: The Q128 is basicly a bigger Q100 and we well know the Voltage limits of that motor as they have been tested to destruction. Phase wires will start to melt first and by the time the Voltage gets into the 60's Volt range, the winding are likely to start cooking.
I'm afraid that I've confused the operating voltage question. I'm fine with arranging the cells into whatever configuration works best, They're just a bunch of cells, so I can put them into whatever configuration I want. (I'm an elec. eng. with bench experience, so that won't be a problem. I know about tack-welding vs. soldering, etc.)

My original plan was 13S/48V, but I thought might end up at 14S/52V or 15S/56V. 15S @ 4.1V would be topped off at 61.5V--is that too hot for the phase wires? I know that it's right below the spec on a lot of 63V nominal caps.
motomech wrote:There is the Q128 that is rated 328 @ 48V, which is an even higher speed motor, but I fear too high for what you want to do. What you would have w/ this motor on 36V would be like the Q100 328(which is rated @ 36 V), that chas58 , crossbreak and a few others use. But these are super riders. probably in the top 10% on human powered bikes, so they can make high-speed systems work without controller over-heating while climbing. I'm from DC, so I know there are plenty of hills, so unless you want to reroute to avoid the steepest of them, you need to ask yourself;, "how strong of a rider am I really?"
"I'm Spartacus." Really, I'm a very strong rider in my own right. I'm also around DC (Northern Virgina: this build is for an assignment at Fort Belvoir that begins this fall), and everyday I maintain a pace in the mid-teens while pedaling up a grade that varies between 8% and 10%. I'll turn off the motor and climb on my own instead of burning it up, if it comes to that. But I understand your concern. I'm a guy like chas58.
motomech wrote:"The problem is, sine wave controllers can be "picky" about the pack Voltage they see the first time, when the values are set."
The Sine waves w/ their SLCD displays are involved and should be your next field of study. But suffice to say, they are Voltage "auto ranging". The first time they "see" the pack Voltage, the values are fixed for the LVC and "fuel gauge" and non-traditional pack Voltages can skew their readings.
Would this GT Power LCD RC 130A Watt Meter Power Analyzer Ver.2 give me the information I need if the SLCD display ends up confused about a non-traditional voltage?
motomech wrote:Stay w/. the center bag, much cleaner and easier(you can stuff extra wire length in there and you will have some). What you may not have thought of this, but the frame bags expand outward in width limited only by your knees starting to hit. My XL EV Falcon bag is huge, to the point that it's weight, not space that dictates what I can carry. But I bought my XL framed Rocky Mountain with the idea of fitting the XL bag.
I agree, it'd be a superior option. I just don't know if I can fit all the cells in.
motomech wrote:I also use a container on the frt. downtube to carry extra capacity;
With a regular road bike (racing, not touring geometry), there's no room below the downtube.
motomech wrote:I have been thinking, if you truly are stuck w/ a 36V system...
I'm not at all planning or stuck with 36V. I can configure the (150) 18650 cells into whatever voltage works best.

User avatar
motomech
100 MW
100 MW
Posts: 2888
Joined: Sep 11, 2010 12:21 am
Location: Tucson Az.

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by motomech » Oct 06, 2017 5:58 pm

OK, that all sounds good. This is going to be one fast, stealthy Ebike.
How about going back to the original plan, with a twist;
The 328C @ 36V motor.
The KT controller w/ the SLCD-6 display
A 14S pack, and when you connect it for the first time, do so with the cells near total discharge state so the Display will read it as a 48V battery because;
Would this GT Power LCD RC 130A Watt Meter Power Analyzer Ver.2 give me the information I need if the SLCD display ends up confused about a non-traditional voltage?
The controller and display must be used together and once energized, stay together.
Trust me on this, you will love the combo and want to make it work for you.
Start reading the owner's manual that can be downloaded on the product page @ either the BMS B. site or the PWS P. site. All the features, the kool blue back-lighting, the smooth and quiet performance, it's the best value in Ebiking.
For the spare controller/display, go w/ the KT/SLCD-3 for options. If you can mount the display somewhere, the remote control might be placed next to your left thumb. Shift w/ the right, select speed w/ the left. Hands never leave the bars, Very nice.
I agree, it'd be a superior option. I just don't know if I can fit all the cells in.
Now I'm thinking you are thinking you need more battery than you really do.
I'm a "brick" guy, not a round cell guy, so I tend to think of my LiPoly capacity in terms of Ah's. I get 1 1/2 miles per Ah cruising in the mid 20's mph w/. moderate pedaling. I would think, w/ your stronger legs, you would get the same w/. a high 20's mph cruise.
So, w/ a 14S/6P configuration(84 cells), that would a 21 Ah battery, right? Or 30 miles. Not a bad range if you couldn't fit 7P(98 cells).
Too me, I think the bigger(or related) issue is pulling the pack to charge, something I know I would hate. Could you do a "charging station" w/ the Mean Well HLG 320H-54A industrial power supply. That's the charger I use and it's awesome. For LED commerical applications, it's fully potted, waterproof, CC/CV, dead silent(no fan) and Voltage adjustable to 58 V For Li-Ion cells, it would be "plug it in and forget it. As it approaches the set top Voltage, the Current attenuates to mini-Amps, where it will safely stay,....forever. Only a very soft "switching" can be heard. It's faster than one would think by the spec.s. Unlike a balance charger, it does not go thru switching cycles to read the cells. Refilling 10 Ah's of my pack takes about 2 hr.s.

100_0016.JPG
100_0016.JPG (94.07 KiB) Viewed 1227 times
Here charging my "frt." 12S/2P 5Ah LiPoly pack.
Motomech

'03 Rocky Mountain Edge 2WD 260 Q100H frt and Ezee V1 rear 2 Elifebike 20A & 25A 9-FET controllers 12S/20Ah Multistar Lipo rear 5Ah Turnigy frt Luna Cyclops Extra lite Alex 24DM rims, 2.4 Holly Rollers run ghetto tubeless. 25 mph. Mean Well HLG-320H-48A
https://endless-sphere.com/forums/viewt ... =3&t=83430
'07 GT Idive 4 4.0, Q100C 201 12S LiPoly elifebike 9-FET 17A controller. 20 MPH.
https://endless-sphere.com/forums/viewt ... 8#p1237928

mystryda
100 mW
100 mW
Posts: 38
Joined: Jul 11, 2017 10:39 pm
Location: Northern VA (metro DC)

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by mystryda » Oct 06, 2017 9:50 pm

motomech wrote:A 14S pack, and when you connect it for the first time, do so with the cells near total discharge state so the Display will read it as a 48V battery
I'm still struggling with this one. It's got some sort of nonvolatile memory that retains the first voltage that it ever sees? But, ok, I can handle that.
motomech wrote:Start reading the owner's manual.
Yep, have done that. Doesn't all make sense, but I suspect plenty of things are illuminated by experience.
motomech wrote:For the spare controller/display, go w/ the KT/SLCD-3 for options.
Does the SLCD-3 do anything that the SLCD-6 doesn't?
motomech wrote:Now I'm thinking you are thinking you need more battery than you really do.
Well, I'm planning to aximize cell longevity by avoiding high-voltage fatigue (charging to 4.1V instead of 4.2V), avoiding high currents, and only exercising 30%. Assuming 50 weeks/year, 5 days/week, roundtrip, then I'm looking at 500 charge cycles per year, and I don't want to continuously dump money into batteries. My calculations only call for using about 500Wh/trip of the 1.9kWh, but that's without any wind, and I've built in a bit of reserve or engineering margin in case I've been optimistic about the drag coefficients or find that there's an unfavorable prevailing wind half the year. Time will tell.
motomech wrote:So, w/ a 14S/6P configuration(84 cells), that would a 21 Ah battery, right? Or 30 miles. Not a bad range if you couldn't fit 7P(98 cells).
In 14S, then the deepest battery I can build with the (150) cells is a 14S/10P. At 3400mAh/cell, that's 34Ah, but 1,714 Wh (3.6V nominal).

15S, then, is essentially a "free" upgrade, since then I wouldn't have any cells leftover in 15S/10P: still at 34Ah, but 1,836 Wh (3.6V nominal). 15S looks really appetizing, then.

And all this makes me think about one of the classic xkcd cartoons, called "Bag Check:"
Image

It had the alt text "A laptop battery contains roughly the stored energy of a hand grenade, and if shorted ... hey! You can't arrest me if I prove your rules inconsistent!"
motomech wrote:To me, I think the bigger(or related) issue is pulling the pack to charge, something I know I would hate.
Yeah, this weighs heavily on my mind, in terms of logistics. I know very little about the building that I'll be commuting to, other than there's a bicycle rack. I can't assume that there will be power there, so that's why I've been thinking in terms of how to disconnect the battery to bring into the office to charge. At home it's not an issue, because the bike lives in the living room and I'll charge it right there.
motomech wrote:Could you do a "charging station" w/ the Mean Well HLG 320H-54A industrial power supply. That's the charger I use and it's awesome. For LED commerical applications, it's fully potted, waterproof, CC/CV, dead silent(no fan) and Voltage adjustable to 58 V.
This looks like a really good option. Thanks.

What are the little blue widgets with the silver buttons? Cell monitors for 7S?

mystryda
100 mW
100 mW
Posts: 38
Joined: Jul 11, 2017 10:39 pm
Location: Northern VA (metro DC)

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by mystryda » Oct 06, 2017 10:53 pm

motomech wrote:Could you do a "charging station" w/ the Mean Well HLG 320H-54A industrial power supply. That's the charger I use and it's awesome. For LED commerical applications, it's fully potted, waterproof, CC/CV, dead silent(no fan) and Voltage adjustable to 58 V.
Mean Well has a couple similar commercial LED power supplies with adjustable output voltage that can get above 58V. In terms of charging, is that all I need, just a supply that's settable to correct multiple of 4.1V? I guess it might be worth mentioning that I don't plan on using a BMS.

User avatar
wturber
1 kW
1 kW
Posts: 324
Joined: Aug 23, 2017 8:52 pm
Location: Fountain Hills,AZ
Contact:

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by wturber » Oct 07, 2017 2:05 am

mystryda wrote:
Well, I'm planning to maximize cell longevity by avoiding high-voltage fatigue (charging to 4.1V instead of 4.2V), avoiding high currents, and only exercising 30%. Assuming 50 weeks/year, 5 days/week, roundtrip, then I'm looking at 500 charge cycles per year, and I don't want to continuously dump money into batteries. My calculations only call for using about 500Wh/trip of the 1.9kWh, but that's without any wind, and I've built in a bit of reserve or engineering margin in case I've been optimistic about the drag coefficients or find that there's an unfavorable prevailing wind half the year. Time will tell.
I could be wrong here, but if you use only 30% capacity to work and don't charge at work and then use another 30% charge going home you will be putting the batteries through half the number of charge cycles. This doesn't (according to BatteryUniversity.com) quite make up for the stress of the deeper (but not super-deep) discharge, but it nearly does. The benefit is that you don't have to worry about the logistics of removing batteries and charging them at work. Your batteries should still last quite a long time. Maybe some more experienced folks here can confirm or set me straight.

Also, I don't think either of the two LCD displays gives Amp Hours or Watt Hours used. So that little $16 might be a good idea to have temporarily installed with the batteries. Use it to get a good idea of how many Watt Hours you are are using and how that relates to the battery voltage. You don't need to mount it on the handlebars, just leave it with the batteries and check it after a ride (maybe make notes) until you are comfortable that you know how much power you are using etc.
"Commuter - DC Booster"
Iron Horse 3.0 hardtail - 48V / 1000W / 470rpm generic Chinese DD Hub motor (ebay)
7 x 36v 4.3ah 10s 2P battery packs - Wangdd22 1500W 30A DC Boost Converter delivers 54v and about 1175 watts peak
53T/42T Sakae Road cranks - 30mph+ on flats
https://endless-sphere.com/forums/viewt ... =2&t=90369

User avatar
dogman dan
100 GW
100 GW
Posts: 34330
Joined: May 17, 2008 12:53 pm
Location: Las Cruces New Mexico USA

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by dogman dan » Oct 07, 2017 6:38 am

If you do want to charge to 4.1v max, then it does get easier to run 15s. The limiting factor is typical 48v controllers have 63v rated capacitors in them, so 15s x 4.2 is 63v. Bottom line, run more than 54.6v full charged voltage if you can. 8)

4.1v leaves a lot of capacity on the table, but if you really only need 60% for the commute, then that is not a problem. Except for those days, when you have a cold headwind going to work, followed by a weather change that gives you an even worse headwind going home. On that day, you will need 125%. :roll: That's the reality I learned in 5 years of commuting to work for a 30 mile round trip.

My best advice for this, is have a second charger at work, just for those days you realize you do need to charge to get home. Bear in mind, your battery will lose about 20% of its capacity if the weather is near, or below freezing. You can charge full 4.2v with much less effect on lifespan, if you ride it right away. Its the time sitting there full, plus the temp, that matters. Cold ass winter day, less effect on capacity to charge it full.

My personal approach has been charge to 4v overnight, then just before the ride, bump it to 4.15v. Closer to full out the door, but not a long wait to start. With no ride planned, I let them sit around at whatever it was a the end of the ride. this is generally 3.7-3.8v, for lipo which is empty at 3.5v. If I empty it, I'll just bring it back to 3.8v to store it.

You are good to go with your gearing, it will be plenty for 25 mph average cruise. But as you know, a sprint cadence and a cruise cadence are two different things. You will want a bigger front ring, eventually. But for now, you are good to go. When you get the big ring, mount your 53 as the middle ring.

docw009
10 kW
10 kW
Posts: 552
Joined: Aug 02, 2015 7:43 am
Location: Chicago area suburbs.

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by docw009 » Oct 07, 2017 8:24 am

In your first post, you mentioned the S09 controller. It is a sensorless controller. I have one. I also have the S12S controller, which I haven't installed yet except for brief tests, but that one requires a motor with sensors.

Both are rated for 36V and 48V, and I often switch batteries on the S09P and other KT controllers that I own. I'm not sure what was being discussed earlier about voltage compatibility. I can mount a 36V or 48V in the morning, and the controller recognizes the voltage. Now if I take off a 48V battery and immediately connect a full 36V battery (which is 42V), there's enough charge retained on the controller capacitors that it sometimes thinks it's still at 48V. Then it powers up, with a low battery icon. They also run on 52V, but I've not run the systems long enough to see if the right LVC gets called up. With the LCD3 dislay, all can display voltage.

I agree about how a display becomes redundant after watching it for a week. I mostly have LCD3 displays, but I have one simple LCD1. I don't miss not seeing instant power, or the temperature inside the display. I do care about battery status, but experience tells me what's happening. Not interested in a cycle analyst.

Edit: I do use a $15 wattmeter to see how the various batteries I own will work. The AH reading seems reproducible, but the WH conversions were too variable, so I just compare AH usage.

User avatar
motomech
100 MW
100 MW
Posts: 2888
Joined: Sep 11, 2010 12:21 am
Location: Tucson Az.

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by motomech » Oct 07, 2017 10:23 pm

What are the little blue widgets with the silver buttons? Cell monitors for 7S?
Those are Battery Medics, one half of my BMS.
8 Years of using LiPoly has taught me be a few things. I have gotten away from smaller bricks strung together to make a Ebike Pack and use the largest that will fit my space. My current main pack is 4) bricks of 6S/10000 mAh MultiStar configured 2S/2P for 12S/20000 mAh. The great advantage is there are only 4) balance taps, so no mess of wires that would be needed if I was to be Balance charging. Every time the pack is charged, at mt storage Voltage value of 3.90 or 3.95V/cell, I'll ck. the cell values and 95% of the time that's it. In the very rare event, the cells have strayed, I'll plug in 4) B.M. and dis-charge down to the lowest cell in the pack. The BM's are really slow and I have tried to speed them up a bit, but the housings start to melt. NBD, because balancing my main pack is like a once or twice a year event. This is mostly for 3 reasons, the LiPoly these days is so good, I never halve to "break" my pack to charge and, most importantly, I stay within the 80% discharge rule. Top charge of usually 4.10V and almost never below 3.80V. My controller's LVC is 42V which leaves me @ 3.65V/cell, which is perfect for LiPoly, the cells are just starting to stray. This is why I need to stay w/ 12S, but that's fine 25 mph is fast enough and having a foolproof LVC has saved me on a couple of ocassions. The LiPoly for this pack cost me $270 shipped(on sale @ HK) and weighs 13.6 lb.s. I luv LiPoly :lol:
Motomech

'03 Rocky Mountain Edge 2WD 260 Q100H frt and Ezee V1 rear 2 Elifebike 20A & 25A 9-FET controllers 12S/20Ah Multistar Lipo rear 5Ah Turnigy frt Luna Cyclops Extra lite Alex 24DM rims, 2.4 Holly Rollers run ghetto tubeless. 25 mph. Mean Well HLG-320H-48A
https://endless-sphere.com/forums/viewt ... =3&t=83430
'07 GT Idive 4 4.0, Q100C 201 12S LiPoly elifebike 9-FET 17A controller. 20 MPH.
https://endless-sphere.com/forums/viewt ... 8#p1237928

User avatar
motomech
100 MW
100 MW
Posts: 2888
Joined: Sep 11, 2010 12:21 am
Location: Tucson Az.

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by motomech » Oct 07, 2017 10:41 pm

Well, I'm planning to aximize cell longevity by avoiding high-voltage fatigue (charging to 4.1V instead of 4.2V), avoiding high currents, and only exercising 30%. Assuming 50 weeks/year, 5 days/week, roundtrip, then I'm looking at 500 charge cycles per year, and I don't want to continuously dump money into batteries. My calculations only call for using about 500Wh/trip of the 1.9kWh, but that's without any wind, and I've built in a bit of reserve or engineering margin in case I've been optimistic about the drag coefficients or find that there's an unfavorable prevailing wind half the year. Time will tell.


In 14S, then the deepest battery I can build with the (150) cells is a 14S/10P. At 3400mAh/cell, that's 34Ah, but 1,714 Wh (3.6V nominal).

15S, then, is essentially a "free" upgrade, since then I wouldn't have any cells leftover in 15S/10P: still at 34Ah, but 1,836 Wh (3.6V nominal). 15S looks really appetizing, then.
Well, I thought about this overnite and I thought I might have been assistance because of the motor and controller selection and my experience w/ the Q motors and sm. sine wave controllers. But because you are not willing to compromise your goals(Not saying you should)of high Volts and high speed, the best advice I can give is, you need to step up the level of your componates and move up to the MAC, Infineon/CA and THUN PAS. The Q128(or tbe BPM, for that matter) just won't be reliable @ 30 mph over any sort of time.
I get the feeling you really want an engineering project and those pieces would make a better basis for that.
Good luck
Motomech

'03 Rocky Mountain Edge 2WD 260 Q100H frt and Ezee V1 rear 2 Elifebike 20A & 25A 9-FET controllers 12S/20Ah Multistar Lipo rear 5Ah Turnigy frt Luna Cyclops Extra lite Alex 24DM rims, 2.4 Holly Rollers run ghetto tubeless. 25 mph. Mean Well HLG-320H-48A
https://endless-sphere.com/forums/viewt ... =3&t=83430
'07 GT Idive 4 4.0, Q100C 201 12S LiPoly elifebike 9-FET 17A controller. 20 MPH.
https://endless-sphere.com/forums/viewt ... 8#p1237928

mystryda
100 mW
100 mW
Posts: 38
Joined: Jul 11, 2017 10:39 pm
Location: Northern VA (metro DC)

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by mystryda » Oct 08, 2017 7:07 pm

motomech wrote:Well, I thought about this overnite and I thought I might have been assistance because of the motor and controller selection and my experience w/ the Q motors and sm. sine wave controllers. But because you are not willing to compromise your goals(Not saying you should)of high Volts and high speed, the best advice I can give is, you need to step up the level of your componates and move up to the MAC, Infineon/CA and THUN PAS. The Q128(or tbe BPM, for that matter) just won't be reliable @ 30 mph over any sort of time.
I get the feeling you really want an engineering project and those pieces would make a better basis for that.
Good luck
Cross my heart, hope to die: I'm really not trying to make this harder than it has to be. As a self-admitted engineer, it's hard for things not to turn into "engineering projects," I suppose. But I am trying to KISS--keep it simple--and hopefully avoid turning it into a money pit.

Here's how I tried to come at this: I used accepted models from road cycling to determine that somewhere around 600 Watts are required to go 30 mph while in the drops. I then reasoned that if I could contribute a steady 200 Watts that I could then get the contribution from the motor down to 400 Watts applied, or about 500 Watts out of the battery when considering the whole system to be roughly 80% efficient. Folks here suggested that the Q128C is hardier than the Q100.

My concern with a larger geared hub motor like the MAC is that it'll be too much motor, and that I'll pay a penalty not only in weight but with only feeding it 350-500W then also by operating it in an an area of very low (<75%) efficiency. Am I laboring under mistaken assumptions?

I think I've been neglecting the voltage/rpm relationship, so here's a go at taking that into account:
I'm assuming that the Q128C 328 is geared similarly to the Q100 328, which you mentioned a couple years ago has a no-load speed of 280-290 rpm.. Since I can just about always peddle to 18 mph and would love if together the motor and I could cruise at 30 mph, it seems like I should target a no-load speed of 430 rpm/35 mph. Then, 36V * (430rpm / 280rpm) = 55V, which is pretty close to 56V/15S.

And I'm back at the same place.

How about a different approach: what limitations of the Q128C do I need to simply accept?

User avatar
motomech
100 MW
100 MW
Posts: 2888
Joined: Sep 11, 2010 12:21 am
Location: Tucson Az.

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by motomech » Oct 09, 2017 2:26 am

Cross my heart, hope to die: I'm really not trying to make this harder than it has to be. As a self-admitted engineer, it's hard for things not to turn into "engineering projects,
:lol: Yes, sometimes engineering types can be a bit perplexing to us rooted in the world of a lesser skill-sets.

Here's how I tried to come at this: I used accepted models from road cycling to determine that somewhere around 600 Watts are required to go 30 mph while in the drops.
I revert back to the Ebike CA sim., which has proven to be remarkably accurate over the years, but It now has the "tuck" position which produces faster results that I have even seen.
Now let's plug in the parameters for the Q100(like chas58's bike);
Q100 "328" motor
Strong Ezee Li-Ion 48 V batt.
MTB "tuck"
700cc whl.s
I've left off the controller, because it's the key here and the sim's results can be misleading. If we use the default 20 Amp controller, which seems doable, we get nice curves;
The No-load speed sits right on the apex of the power line telling us the motor has become RPM limited @ the right time. Very good, but efficiency is low (76%), with much of the "meat" of it's curve past past the no-load line(29.6 mph). The problem with this scenario is that 20 Amp controller does not represent the sm. controllers we are likely to use and certainly does not represent what the Q100 can take. It we get more real and plug in a custom 15 Amp controller the picture changes dramaticaly.
The power line has now been "quashed" flat and the top speed is now 27 mph w/ the power line's apex extending way past the no-load line, the motor has become Power limited rather than rpm limited.
This is the conundrum of the minimotor in a big wheel. At no-load, the motor is screaming for more Amps, but at the same time, it really can't safely take much more. I have melted the cheap stock phase wire connectors when I ran a 328 in a 26" wheel and a "hot" 22 Amp Lyen Mini-Monster. Yes, I replaced them, but how much further behind were the phase wires themselves? The windings?
So why can chas58 make the mini/big whl. combo work and I can't? His legs. Again, back to the graph and see the power line extending, more or less flat all the way to 33.5 mph before dropping off. What this means when the power curve is "just below the surface", so to speak and the rider's leg power can be meaningful. Adding 200 Watts of leg power raises the raises the top speed an amazing 4 mph, from 27 to 31 mph!

My Q100 2WD experiments;
MTB(no tuck)
26" whl.s
44 Volts
Single Q100 (260)-22 to 23 mph
Dual Q100 (260)- 24 to 25 mph
Single Q100 (328)-22 to 23 mph
Dual Q100 (328)-27 to 28 mph
Here we see the second 260 motor can only add 2 mph to the rpm limited single 260 motored system, while the second 328 can add an astounding 4 to 5 mph to the power limited single 328 system. The second motor helps "pull" both motor higher up into the rpm range. A case where the sum is greater than the parts.
I ran the bike like this for a while and while the controllers would get hot, at least the load was being split between two sets of phase wires. But while the top speed was exhilarating, the mid-range was very leisurely and efficiency suffered. Since I only cared to go 25 mph, I went back to my current setup of two 260 motors. Acceleration is brisk to top speed and efficiency improved, but most importantly, single motor usage became practical again.

The Mini-Motor's Universal Law of Hill Climbing; Never allow the climbing speed to fall below 1/2 of the top speed.
At this point, winding saturation has the motor dumping more energy into heat than locomotion. This is also what is happening w/ a high-speed mini is being "lugged" at top speed and is a matter of total Wattage. This is why adding more Volts to a mini to achieve a higher top speed quickly becomes a "Fools errand". Adding more Volts necessitates an inverse reduction of Amps, further starving the motor for current. I have tried adding volts up to 15S. 13S(48 Volts) added a mph or so. 14S a little less and 15S only added heat.

So what does all this have to do w/ the Q128?
Well, I tend to think of the Q128 as a Q100 w/ 30% more mass. So it should be able to handle more Current and dissipate more heat, but 30% more? I'm not sure. But remember, the Q100 and the Q128 share the same phase wires.
So, a Q128 in a big whl. could be a satisifing bike for the rider who understands it's limits and w/ the legs to help the system when it needs it most, maybe not so much in the flats, but climbing. And to my way of thinking those limits are around 50 Volts and 20 to 25 Amps. I don't think the idea of pouring on more Volts and/or Amps to achieve a certain level of performance is sound. That's why I recommended moving up to the MAC, where one can run a high-speed motor, pour on the Volts and Amps and "let 'er eat", so to speak.
For the proposed build here, I think one thing leads to another.
They are;
W/out a practical throttle, the PAS action must be the best possible, w/ no delay.
The "best" affordable PAS system is based on the new KT controllers and their related displays.
The KT controllers are limited to 20 Amps.
So that, by default is the controller to be used. But,
can this controller handle a 52 V batt.?
To date, I have not seen a deffinitive answer to that question.
So, if we go back to the sim. and plug in a custom controller of 17 Amps(what I feel, thru experience and interpolation, is a more accurate "real World" number)things look pretty good. A top speed of 28.5 mph w/ 200 Watt added to bring it up to 31.5 mph, landing right on the apex of the power line. Back to the 52 V batt., it only adds 1/2 mph to the top speed, further illustrating the diminishing returns of adding Volts to a high-speed mini-motor and to my way of thinking, it's not worth potential trouble w/ the sine wave controller..
As always, the sim results represent numbers in the "perfect World" and the avg. results on the road should be reduced by 5 to 10 %.
Which brings us back full circle to where we began :lol:
Motomech

'03 Rocky Mountain Edge 2WD 260 Q100H frt and Ezee V1 rear 2 Elifebike 20A & 25A 9-FET controllers 12S/20Ah Multistar Lipo rear 5Ah Turnigy frt Luna Cyclops Extra lite Alex 24DM rims, 2.4 Holly Rollers run ghetto tubeless. 25 mph. Mean Well HLG-320H-48A
https://endless-sphere.com/forums/viewt ... =3&t=83430
'07 GT Idive 4 4.0, Q100C 201 12S LiPoly elifebike 9-FET 17A controller. 20 MPH.
https://endless-sphere.com/forums/viewt ... 8#p1237928

mystryda
100 mW
100 mW
Posts: 38
Joined: Jul 11, 2017 10:39 pm
Location: Northern VA (metro DC)

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by mystryda » Oct 12, 2017 7:03 pm

motomech wrote:
mystryda wrote: Cross my heart, hope to die: I'm really not trying to make this harder than it has to be. As a self-admitted engineer, it's hard for things not to turn into "engineering projects,
:lol: Yes, sometimes engineering types can be a bit perplexing to us rooted in the world of a lesser skill-sets.


Engineers just see the world differently. Would it be too bold if I inquired what your profession is? Just a matter of curiosity.

So I've spent quite a bit of time going back to the drawing board, trying to start from scratch.

Both the MAC 6T and 8T looked promising in simulation, but is it true that it only comes in a 7-speed freewheel? I'm reluctant to go from a cassette to a freewheel, and even more reluctant to transition away from 9-speed shifters. (Some might find this ridiculous, but I really do actually use the full range of the 9-speed cassette, and I don't want to replace my current ultegra 9-speed "brifters.)

The Ezee also looks good in simulation, and it can be had with a 9-speed freewheel, but that would mean going to a freewheel and, frankly, it's beyond my price point.

So what about the Bafang CST? Do you think that that would be a good match? The price point is higher than the Q128C, but not outrageously so, and it takes a cassette and fits in 135mm dropouts. It also seems to have better efficiency numbers, similar to the MAC and Ezee. Here's how's the ebikes.ca simulator compares them in my operating range:
Image

Thank you, btw, for showing how you can dial in the simulator by entering a custom controller. (I am mystified as to have the simulated motor amps for the Q100 ("System A") can be higher than the battery amps.)

If the Bafang CST is a good match would all the other components stay the same?

User avatar
wturber
1 kW
1 kW
Posts: 324
Joined: Aug 23, 2017 8:52 pm
Location: Fountain Hills,AZ
Contact:

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by wturber » Oct 12, 2017 8:08 pm

Quick question - is that 26" in wheel in your calculation correct?
"Commuter - DC Booster"
Iron Horse 3.0 hardtail - 48V / 1000W / 470rpm generic Chinese DD Hub motor (ebay)
7 x 36v 4.3ah 10s 2P battery packs - Wangdd22 1500W 30A DC Boost Converter delivers 54v and about 1175 watts peak
53T/42T Sakae Road cranks - 30mph+ on flats
https://endless-sphere.com/forums/viewt ... =2&t=90369

mystryda
100 mW
100 mW
Posts: 38
Joined: Jul 11, 2017 10:39 pm
Location: Northern VA (metro DC)

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by mystryda » Oct 12, 2017 8:35 pm

wturber wrote:Quick question - is that 26" in wheel in your calculation correct?
"No," because I'm running 700C x 23.

"Yes," because chas58 says 26" on the simulator most accurately reflects a 700C on a road bike (and that the 700C is more like a 29'er).
chas58 wrote:For the simulator, I use a ' 26" wheel ' for my road bike.

A 700c road tire on a road bike has approximately the same diameter as a 26” mountain bike tire.

I just ran some calculations, and it looks like the simulator uses these tires:
700c = 56mm tire
26” = 2.0” tire (or ~23mm 700c tire)

Unless your bike uses the so called 29” tire, you are probably better off modeling with the 26” tire diameter.
But thanks for checking.

Voltron
100 kW
100 kW
Posts: 1289
Joined: May 02, 2013 4:53 pm
Location: Santa Barbara CA

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by Voltron » Oct 12, 2017 11:08 pm

Were there some reasons you're not looking at a crank drive setup? It solves the problems of keeping your 9 speed, less weight on a skinny road rim, good climber for the weight, no torque arms needed, etc. Plus hub motors can have trouble swinging the extra diameter of a 700 on climbs.. I laced the same geared motor into a 24, a 26, and a 700c, and bigger wasn't better...

mystryda
100 mW
100 mW
Posts: 38
Joined: Jul 11, 2017 10:39 pm
Location: Northern VA (metro DC)

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by mystryda » Oct 12, 2017 11:17 pm

Voltron wrote:Were there some reasons you're not looking at a crank drive setup? It solves the problems of keeping your 9 speed, less weight on a skinny road rim, good climber for the weight, no torque arms needed, etc. Plus hub motors can have trouble swinging the extra diameter of a 700 on climbs.. I laced the same geared motor into a 24, a 26, and a 700c, and bigger wasn't better...
Yes, I ruled it out because of drive train wear. When I originally broached that idea on the boards a couple months ago, I was advised that putting 700W+ through the drive train for 250 miles a week, 50 weeks a year would lead to wear and reliability issues.

I'm not worried about the climbs. I can climb by myself in the mid-teens, and I'll take care not to burn it up. The terrain is somewhat rolling, but not for very long and not very bad.

User avatar
Alan B
100 GW
100 GW
Posts: 7028
Joined: Sep 11, 2010 7:43 am
Location: San Francisco Bay Area, USA
Contact:

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by Alan B » Oct 13, 2017 2:09 am

The motor controller is essentially a buck (voltage reducing) power converter. It down converts voltage and when doing so produces more current on the output. So the motor sees more current most of the time, decreasing as motor voltage (and speed) increases.

User avatar
chas58
10 kW
10 kW
Posts: 816
Joined: Mar 19, 2013 4:05 pm
Location: Detroit, Michigan, USA

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by chas58 » Oct 13, 2017 12:56 pm

mystryda wrote:I know chas58 talked about a doorbell switch on his road bike. If I understood correctly (and I'm not sure that I did), it was being used to toggle PAS. Single button would be awesome.
I use that as a power switch. As I have a fairly low power system, I can just use it as full on or full off. Not like a thumb throttle does a whole lot on a low current Q100 anyway. I generally prefer to pedal myself below 15 mph. Still, it would be easy to put in a rheostat to modulate the speed if I needed.

Motomec is right, PAS systems have gotten a lot better over the years.
mystryda wrote:
wturber wrote:Quick question - is that 26" in wheel in your calculation correct?
"No," because I'm running 700C x 23.

"Yes," because chas58 says 26" on the simulator most accurately reflects a 700C on a road bike (and that the 700C is more like a 29'er).
Good point. ;-) Your going to be well served by using 25 or 28mm tires though. They will have less rolling resistance, more traction, and more impact resistance.

But they recently did a revamp of the site, so I'm wondering if a 700c is now more of a road bike tire size now that they explicitly list 29" wheel (which of course is a 700c wheel too, just with fatter tires). Either way, a standard 700c road tire has similar circumference as a standard 26" tire (2097mm (23-622) vs 2070mm (54-559)).

The Q100 328rpm motors do spin at 328rpm. That original quote saying they are 280-290rpm is about 5 years old and was based off of some interpolation and false assumptions, not measurements. I bought and measured the RPM and the specifications are correct. Dev8 quotes the unassisted speeds at 15, 20 and 24mph (200, 260, 328rpm), and my measurements agree with that.
25^3 bike: 25 lbs, 25 mph, 25 mile range.
Road and Mountain Bike Cute Q100 builds:
https://endless-sphere.com/forums/viewto ... =3&t=49691

User avatar
motomech
100 MW
100 MW
Posts: 2888
Joined: Sep 11, 2010 12:21 am
Location: Tucson Az.

Re: High-speed, long-distance (heavy "30^3") Q128 commuter b

Post by motomech » Oct 14, 2017 12:02 am

Would it be too bold if I inquired what your profession is?
I'm from Md. too and started out as a Motorcycle tech in the summer and dispatch rider in DC in the winter. After 20 Years, tired of riding in the snow, I switched to golf course mechanic. in 2001, I got a call to go the Costa Rica to set up the shop and train the techs at a new Arnold Palmer/Four Seasons course and ended up staying 5 years. Then I went to work for a consulting Co. in Cancun and they sent me all over L. Amer. to help clubs improve their "tracks". I wore more "hats", not only working on the equipment end, but the turf management side as well. I became an "expert" on a new salt water tolerant turf grass called Paspalum. The challenge was always trying to "raise the bar" in countries where the standards are not very high. Retired now in Tucson and judging by the amount if time I spend here, need to find something to do.

First off, I'm not trying to dissuade you from going w/ the high-speed Q128, big whl. combo, but to point out there probably will be some concessions, in terms of power that can be applied to the motor, that will have to be made. But heck, I tried to discourage chas58 when he outlined what he wanted to do and look how that turned out. So take my advice w/ "a grain of salt".

Both the MAC 6T and 8T looked promising in simulation, but is it true that it only comes in a 7-speed freewheel? I'm reluctant to go from a cassette to a freewheel, and even more reluctant to transition away from 9-speed shifters. (Some might find this ridiculous, but I really do actually use the full range of the 9-speed cassette, and I don't want to replace my current ultegra 9-speed "brifters.)
I'm pretty sure an 8-speed free wheel fitted to the MAC will fit in a 135mm drop-out. There is only one free wheel w/ an 11T sm. gear and that is the Enoch DNP and it comes in 7, 8, 9 and 10 speeds. They are not too bad, just heavy. I'm running an Enoch 8-gear w/ a 9-speed shifter on my Ezee V1. It shifts fine although I seldom use more than 3 gears. The fact is, once the motor is installed, it's easy to get lazy and not bother since it's not needed. The more power, the less need to shift. My other "assist bike" runs a Q100C and I kept the power low, which forces me to pedal, and shift.
The Ezee, made in Taiwan I think, is beautifully made, real eye candy, but not neccesarily "better" than the more crudely made Bafang.
If the Bafang CST is a good match would all the other components stay the same?
The BPM CST, which is listed on the product page as a 270 rpm @ 36V motor, actualy is shown to be 290 rpm motor on the new sim. It should hit 30 mph on 52 plus Volts. IMO, the biggest between the BPM and the Q128, aside from power handling, is the form factor. The Q128 trades diameter for greater width, which allows it to hide behind the brk. rotor, a little more stealthy. On the street, I really don't notice the Ezee's 5 Kilo weight much more than I noticed the little 2 Kg. Q100.
It appears the BPM CST's are on sale;

https://bmsbattery.com/ebike-kit/589-ba ... e-kit.html

Paired w/ the recommended SO9P, S12P or S12S, it's probably the approach I would take if I was after a 30 mph Ebike. But then again, I wouldn't be pedaling at that speed with the MTB' donor bikes that I use. I couldn't get that kind of gearing. Not even close.

Bottom line;
I have read here something along the lines of, The BPM (Bafang) is good up to 30 mph and above that, it's best to go w/ the MAC and believe this is about right.
26 to 28 mph avg.-Q128
27 to 30 mph avg.-BPM
30 mph and over- High-speed MAC
Motomech

'03 Rocky Mountain Edge 2WD 260 Q100H frt and Ezee V1 rear 2 Elifebike 20A & 25A 9-FET controllers 12S/20Ah Multistar Lipo rear 5Ah Turnigy frt Luna Cyclops Extra lite Alex 24DM rims, 2.4 Holly Rollers run ghetto tubeless. 25 mph. Mean Well HLG-320H-48A
https://endless-sphere.com/forums/viewt ... =3&t=83430
'07 GT Idive 4 4.0, Q100C 201 12S LiPoly elifebike 9-FET 17A controller. 20 MPH.
https://endless-sphere.com/forums/viewt ... 8#p1237928

Post Reply