My second e-bike: Reality Check? (Pretty please?)

[Lebowski]

is the 30A battery current or motor phase current ? Cause 100A motor phase current instead of 30A would up the dissipation by 10...
Make sure not to confuse motor phase current and battery current.

But I agree with the conclusion, not much difference in the FET's. Even 100A phase current
gives an i^2 of 10000, an i_rms^2 of 5000, so 2500 per FET (low and high side so divide
by 2). With an R_on of 4.5 mOhm that's 11W per FET of dissipation. At full throttle
100A phase is around 70A battery current, so with an 80V battery 6 4110 FETs
dissipate 66W and deliver roughly 5500 Watts to the motor. Enough I would say and much
more than what a 6 FET controller is pushed to normally.

 

[Joe Perez]

is the 30A battery current or motor phase current ?

I can't see why RMS phase current should be higher than battery current. (perhaps you are looking at peak current?) Assuming that the controller does not contain a transformer, then the RMS current of the sum of all phases should be equal to the RMS current on the battery. (This is where I'll freely admit that my lack of practical experience with inductive circuits comes into play.)

It's also important to remember that any one particular FET in a six-FET array is active for only 1/3 of the total drive cycle (assuming a three-winding, six-phase system.) So the average current through that FET (and thus, the average power dissapation) will be less than what you would get if you assumed a 100% duty cycle. I'm obviously ignoring losses due to transition through the off-on and on-off states, but I think the fundamental point remains that if we are simply comparing loss through a 4110 as compared to a 3077, the overall magnitude of the power dissipation relative to the overall system power and the thermal ratings of the components is extremely trivial.

I really don't want this thread to derail too far into EE theory, as I greatly appreciate all of the practical help that people have been providing thus far. So if it helps matters, I'll gladly concede that for a 48v system, running 3077 FETs is "better." Personally, I wish to allow myself headroom for future expansion.

 

[dogman dan]

Back to a simpler discussion, the amped dd motor you have now is a 2807 winding. Climbs 7% ok on 48v, can climb 10% with brisk pedaling, But 15% will choke it. I has to go 15 mph to avoid overheating, and will still heat up plenty even at 15 mph. But at 15%, you'll slow to about 5 mph if you are only letting it have 1000w. A 2810 would still climb at 5-7 mph, but it's going to climb fine if you can pedal it back to 10 mph.

Even with the gearmotors, 1000w is only going to move but so fast up a 15% grade. So you may be wanting to push the wattage a bit higher even if you go with a bigger gearmotor. If you choose the 10T, you will have to volt up to reach 25 mph I think. But the 8 t on 1500w might do the climb good enough. Your hills are not miles long, so you can heat that sucker up for a short time with no problems. It will cool again on the way down the other side.

As an example of slow motor high wattage on a short hill, see some of my hill climb vids. This is a 2812 motor on 72v with a 40 amp controller. Near the top, I'm likely turning 2000w straight into heat, but only for a short time.
http://www.youtube.com/watch?v=6RpXuIyAnkY&feature=channel&list=UL

 

[wesnewell]

A little late to the party, but have you considered just doing a shunt mod on your current controller, and/or going to 48V? Either will increase the torque of your current setup from a little to almost double the power. Just going to 48v with your current controller will give you about 25% more torque for climbing those hills, along with about 6mph more top end speed. Seems ashamed to waste a perfectly good setup when it can be made to do what you want for nothing .

 

[dogman dan]

Going from 36v to 48v will help a ton. But a truly 15% capable bike takes a bit more. 5 mph more speed at the bottom of a short 15% hill might do the trick.

I'm still not clear myself, wheter the existing kit works or not, or just won't work on 15%. A new controller and 72v would go a long way to perk the existing motor up. No big deal if it eventually cooks it. I'm just suggesting 1500w, 72v 20 amps. With more speed avaliable, you could just charge shorter hills and take em on momentum.

 

[Joe Perez]

Even with the gearmotors, 1000w is only going to move but so fast up a 15% grade. So you may be wanting to push the wattage a bit higher even if you go with a bigger gearmotor.

Well, 52v at 30A will yield a maximum power of 1,560 watts, which is just a hair over the claimed power in the video that chroot posted earlier depicting a 10T scaling a much steeper incline than I have to deal with. And again, I don't need to be able to make tremendous speed up the hills, I just need to be able to make it over them without blowing anything up, which is the one thing that my present system is not capable of doing reliably.

If you choose the 10T, you will have to volt up to reach 25 mph I think.

Volt up relative to what? Higher than 52v?

Like I said at the outset: speed is not the #1 concern in this build. If I wanted to be going fast, I'd be riding my SV650 or driving the turbo Miata. Speed is nice, but just being able to get over the hills at all is the real goal.

A little late to the party, but have you considered just doing a shunt mod on your current controller, and/or going to 48V?
(...)
Seems ashamed to waste a perfectly good setup when it can be made to do what you want for nothing .

I did briefly experiment with 48v by installing an SLA pack in series with the primary battery, and while there was a notable difference in top speed, hill-climbing ability was not dramatically improved.

As to current, that is the real problem that started this whole thing. With the AmpedBikes lithium bottle battery I have now, I have blown the BMS in it twice just by operating it at the standard 36v and 22 amps. The most recent feedback which I've gotten from Danny at Amped is that their battery just isn't suitable for operating in that range (mind you, the battery, motor and controller were sold to me as a complete package), and that after having the battery repaired, they want me to downgrade the controller to 15A.

So that's why I've basically conceded that the system I have now is not even worth upgrading, and that I'm better served by just tossing it out and starting fresh. I consider it to have been an expensive learning experience.

5 mph more speed at the bottom of a short 15% hill might do the trick.

I'll be sure to convey that to whoever programs the traffic lights here in Carlsbad. 8)

I'm still not clear myself, wheter the existing kit works or not, or just won't work on 15%.

At present, it does not work at all, as I blew the BMS for the second time this past Tuesday afternoon. This is what it looks like at the moment:

Prior to this, it worked well except on the 10-15% hills, where it was only just barely adequate. And since their proposed solution involves downgrading the current through the system to a level that will make it not work at all for the hills which I really need it for, I'm not really considering any sort of upgrades that involve retaining the hardware which I have in any way. Sorry if I wasn't as clear up front as I could have been on that. I didn't want this to be a thread where I bash on a particular vendor.

 

[dogman dan]

Ok, so you still have a motor. It's just a battery problem. That means running that motor at 72v could be a possibility, with a 72v controller of course. Sadly, it may not help as much as I thought, if you have a light on the hill. But even from a dead stop 72v 20 amps is still a solid 1500w of go power. More watts helps for sure.

 

[o00scorpion00o]

How about selling the existing kit ? Or do what dogman said, run higher voltage higher current through your existing motor. But that would mean a 72volt batter, but you wouldn't need the speed that would be capable of, but faster wind motors need to spin faster meaning more voltage and when they start to bog down they heat up too much. That's why the slower wind motors are better. And the geared hubs allow the motor to spin faster anyway making them more efficient for climbing.

If you had to get a new battery anyway then get a new motor, the 48 volt may be just fast enough for you in the 10T, and cell-mans battery will be more than capable!

The mac kit would make the bicycle feel a lot more like a bicycle, if you like to pedal as I do, then it's the best choice.

The mac kit should also eat hills much better. Again , I strongly recommend a temperature sensor in any motor, especially for hilly areas, then you can adjust your controller for more power if it's not heating up much.

That battery has to go. Get a new bms, and sell it!

The problem will be deciding the winding of the motor, the 10 T may well be good enough, even the 8T, but the 12 T would probably run the coolest, but would also be the slowest!

As I said, the 8T is the perfect balance for me because it's fast and has huge amounts of torque and has got me up almost every hill I could imaging. I just didn't take it up the 16% mountain slope where I killed the halls in the magic pie coming back down. But the pie survived and it was the wiring after that fried, which in turn fried the halls. I wrongly used cat5 network cable after I re-wired the motor for an external controller. If I kept off the regen it would have had a good chance to cool on the way back down.

The 8T will get its day at the mountain when cell-man sends the new stator with the temperature sensor, then I will know how far I can take it.

I love being able to do 40 mph, and climb hills, and pedal too. And it even has decent hill climbing at about 500 watts. Sometimes on a hill I set cruise and pedal along with the motor. It's amazing what you will get up if you help the motor along at 500 watts.

Above all the 8T allows me to carry a much smaller battery, 16S Lipo, than if I was to volt up a slow wind motor! But you may find for you the 10T is better, which would be the case if you don't want the speed!

 

[Joe Perez]

Ok, so you still have a motor. It's just a battery problem. That means running that motor at 72v could be a possibility, with a 72v controller of course. Sadly, it may not help as much as I thought, if you have a light on the hill. But even from a dead stop 72v 20 amps is still a solid 1500w of go power. More watts helps for sure.

I appreciate that you're trying to help me save some money here, however that's not really a concern. My primary goal here is to do a completely new build, optimizing every part of the system (frame, brakes, motor, battery, etc) to a level which I deem more optimal. I'm not going to go out and buy a $3,000 carbon-fiber frame, however I have no problem at all spending an extra $400 to really optimize the motor.

Switching from my existing motor to a geared MAC 10T motor will have the advantages of lighter weight and complete freewheel ability, and from everything I've read and seen thus far, it would seem that the MAC is also a much better choice for hill-climbing. If there's a specific reason why some other motor (even the one I have now) might be superior to a MAC 10T in terms of performance then I'm certainly open to this, however the savings of a few hundred dollars does not weigh into my thinking on this. I have no interest in cheaping out if it means compromising the primary goals of the build

How about selling the existing kit ?

At the moment, I am attempting to pursue a refund from Amped on the kit as a whole, with the desire to part ways from them on good terms. Up to now, our dealings have been quite pleasant, and I'm hoping that we can resolve this issue amicably. If that succeeds, then great. If not, then I will probably do exactly as you suggest- fix the BMS, downgrade the controller, and then sell the whole thing. Either way, I don't plan on any part of it winding up on my new bike.

And the geared hubs allow the motor to spin faster anyway making them more efficient for climbing.

Hence my strong desire to eliminate the motor I have now from the build.

If you had to get a new battery anyway then get a new motor, the 48 volt may be just fast enough for you in the 10T, and cell-mans battery will be more than capable!

The mac kit would make the bicycle feel a lot more like a bicycle, if you like to pedal as I do, then it's the best choice.

The mac kit should also eat hills much better.

This is what I'm hoping. At this point, I've really just come down to whether the 10T is the best choice for a motor, and if so, what voltage I should run the system at.

Based solely upon the video that was posted earlier, my impression is that the 10T is adequate and that I don't require the 12T. So for the battery, it comes down to whether I use Cell_Man's 52v "triangle" battery, or trade Ah for voltage, perhaps with a 20s 4p arrangement. That would give me the sane number of cells, and thus the same weight, with 65v @ 9.2 Ah. Given that my 36v 10Ah battery has never been problematic in terms of range, this might be an appropriate solution.

Again , I strongly recommend a temperature sensor in any motor, especially for hilly areas, then you can adjust your controller for more power if it's not heating up much.

I'll do this. I assume that some sort of PTC thermocouple is used- any suggestions on an appropriate display gauge?

 

[o00scorpion00o]

I'll do this. I assume that some sort of PTC thermocouple is used- any suggestions on an appropriate display gauge?

The new cycle analyst is due soon, I don't know when ? it will have the ability to display temps and limit current if you exceed a certain temperature. You will be able to plug it into cell-mans controller. Just make sure the controller has the connector for the cycle analyst.

The other device is this

http://endless-sphere.com/forums/viewtopic.php?f=3&t=22919

 

[wesnewell]

Here's my suggestion. You can have everything you need to get going in a few days.
http://www.ebay.com/itm/48V-1000W-26-Rear-Electric-Bicycle-Engine-Kit-Bike-Hub-Conversion-Scooter-Motor-/290694031625
Two of these for 12s.
http://www.hobbyking.com/hobbyking/store/__20647__ZIPPY_Flightmax_8000mAh_6S1P_30C_USA_Warehouse_.html

This will get you about the same range as your 10ah 36V battery since they have about the Wh.
This will provide up to 1500W to the motor. If that's too weak, a simple shunt mod can get that up to over 2000W.
Top speed will be about 28mph on 12s lipo. If you want more than that, you can run up to 15s on the controller for about 33mph. More than that will require a new controller that cost $35 shipped.
If you want more details, ask. I've been running this motor for about a year now, mainly on 18s lipo, but up to 24s without a problem. 18s carries my 275lbs up some pretty steep hills.

 

[Joe Perez]

The new cycle analyst is due soon, I don't know when ?

I just found the thread in which it's being discussed. Even though I have no desire for throttle re-mapping or pedal torque sensing, that's some very interesting stuff.

I contacted Jason, and he informed me that they'd just done another mini-build, and pointed me to the ordering site for them, and I picked up one of the beta units as well as the appropriate shunt and cable. So that's one thing taken care of.

You will be able to plug it into cell-mans controller. Just make sure the controller has the connector for the cycle analyst.

Really? I figured that was an ebikes.ca exclusive feature. Ah, well. I don't at all mind doing a bit of extra wiring- it's my job, after all.

The other device is this

http://endless-sphere.com/forums/viewtopic.php?f=3&t=22919

Clever, but I've worked with this kind of thing before, mounting Windows-based tablet PCs in cars for the purpose of ECU tuning. Even with the top up, daylight readability is poor, and with the top down, forget about it. I do own an Android phone, however when I'm outside on a sunny day it is completely unreadable unless I hold it right up to my face and squint at it, and this assumes that I've removed my polarized sunglasses first. I think I'll stick with a traditional monochrome reflective display.

Here's my suggestion. You can have everything you need to get going in a few days.
http://www.ebay.com/itm/48V-1000W-26-Rear-Electric-Bicycle-Engine-Kit-Bike-Hub-Conversion-Scooter-Motor-/290694031625
Two of these for 12s.
http://www.hobbyking.com/hobbyking/store/__20647__ZIPPY_Flightmax_8000mAh_6S1P_30C_USA_Warehouse_.html

I think that I'm going to pass on the idea of e-bay sourced drive components and Hobbyking LiPo batteries for now.

 

[wesnewell]

Here's my suggestion. You can have everything you need to get going in a few days.
http://www.ebay.com/itm/48V-1000W-26-Rear-Electric-Bicycle-Engine-Kit-Bike-Hub-Conversion-Scooter-Motor-/290694031625
Two of these for 12s.
http://www.hobbyking.com/hobbyking/store/__20647__ZIPPY_Flightmax_8000mAh_6S1P_30C_USA_Warehouse_.html

I think that I'm going to pass on the idea of e-bay sourced drive components and Hobbyking LiPo batteries for now.

Ebay doesn't source anything. They're just an advertising outlet. Nothing you buy on ebay comes from ebay. The kit is sold by yescomusa.com. If you want to pay more for it, you can buy it from their website. They are located not far from you in CA. You'll just pay a little more.
http://www.yescomusa.com/product.php?productid=729&cat=48&page=2
I also venture to say that they sell more hub motor kits than most if not all the other dealers in the US combined. They certainly have a large selection and I've bought 3 from them with complete satisfaction of price and product. All 4 are still working fine after at least 8K miles in the last year or so. But, since I don't like to pay more than I have to, I bought all of them though ebay ads. Happy motoring.

 

[Joe Perez]

wesnewell, yes, I understand how eBay works. I've been using the service since '97, back before it turned into a flea market. All else being equal, I prefer to deal with vendors who actually know something about the product they're selling and can provide datasheets and specifications. From looking at Yescom's website, it's pretty obvious that they don't fit into that category. (But I'll keep them in mind if I ever need a cheaply made tattoo machine or chicken coop.)


At any rate, moving forward. I picked up a new bike on Friday. Nothing fancy- a Giant Revel 1, but a huge upgrade from where I'm at now. Aluminum frame, disc brakes... It's amazing how light this thing it, and a shame that I'm about to add so much weight to it.

It took about a week of emails and waiting, but I finally managed to place the actual order with Cellman as well. 10T motor, 9 FET controller, and the Triangle battery. With all the communications delays I'd been strongly considering just buying from ebikessf, but the battery was the real selling point. Should fit like a charm:

 

[dogman dan]

The other motor you might look into for climbing hills is the 9 continent in a 2810 winding. Run it at 72v, and it will hit 30 mph, and cruise at that speed without melting on flat ground. If your 72v is 40 amps worth, it can be melted on steep hills that use all 40 amps for long durations.

Get one from Methods, here. http://www.methtek.com/category/motors/9c/

I was really happy with this motor, and have several I run at 48v. I went to the 2812 rear on 72v for my off road hill climber, but it's no longer avaliable.

With the dd motor, you can push 3000w for short periods. There are a few guys out there running the 2810 on 72v that have been pretty happy, and haven't melted it down.

 

[o00scorpion00o]

Mac will be more efficient for the hills, no sense in having more power to do the same work

Sorry dogman,

 

[Joe Perez]

Yeah, I'm kind of soured on the whole idea of DD hubs right now. If I lived in an area with flatter terrain, then I'm sure even my 2807 would be just fine. Of course, if I lived in an area with flat terrain then I wouldn't need an electric motor in the first place.

We'll see how this MAC motor fares. Despite all of the positive things I've read, I still feel like I'm taking a bit of a chance pushing a tiny little motor with plastic gears up a 15% hill, but we'll see. I've seen some of the work that's been done recently with regard to "oil cooling" (drilling a hole in the thing and filling it with ATF / light oil) and I may give that a shot to buy some extra safety margin. I'll be curious to see how well the motor is actually sealed relative to the outside world, both the shaft bearings and the hollow shaft itself.

 

[o00scorpion00o]

Afaik it's not a good idea to oil cool a geared hub because of the grease ?

The new type gears in cell-mans mac will not break. I certainly couldn't kill them torture testing it. But I did kill the clutch. But I'm getting the new one with a new stator and temp sensor, new axle and keyway, so should take the power this time.

on 52 volts LifeP04 I don't think you will have a problem as cell-man sets the controllers to about 2kw max and my friend pat has the 8T mac and is running it at stock settings and it's still running perfectly well.

I would definitely get the upgraded clutch, and temp sensor and hook it up to some kind of display, that way you will quickly learn the limits of your motor and be able to back off the power or stop, This simple cheap device should be in every motor and we shouldn't except anything less at this stage in the game. A few cent to protect a 300 motor!

 

[Joe Perez]

Afaik it's not a good idea to oil cool a geared hub because of the grease ?

I would have thought that immersing the entire assembly in a liquid which is designed specifically to lubricate planetary gears and dissipate heat (which is what ATF is) would actually be preferable to simply throwing some grease into the assembly, sealing it up, and hoping that it stays in place.

I would definitely get the upgraded clutch, and temp sensor and hook it up to some kind of display, that way you will quickly learn the limits of your motor and be able to back off the power or stop,

What is this "upgraded clutch" of which you speak?

What I have ordered, specifically, is the 10T motor with upgraded dual phase-wires and LM35DZ temp sensor. I don't recall any mention of there being various clutch options in the MAC design. Are you referring to the newer BMC clutch as discussed here? http://www.ebikessf.com/V2-vs-V3-clutch

(For that matter, I've read conflicting reports as to whether the new-style BMC gear/clutch package is drop-in compatible with the MAC motors. Any thoughts? This would be convenient from the standpoint of service parts in the future.)


I've picked up one of the new r3 CycleAnalysts, so that part of the equation should be pretty well taken care of. I'm just not entirely certain what the actual "limit" of the motor is, temp-wise.

 

[o00scorpion00o]

I think it's to do with the fact hub motors are not so easy to seal, and oil could leak through the axle.

The new clutch is supposed to be a lot tougher and it can't fit into the mac without new gears, as the clutch is thicker, the gears are not expensive. I think it's the same as the new bmc one, but I haven't seen it yet. But it will be easy enough to fit new gears and clutch if the time comes.

You will greatly appreciate the temperature sensor. I don't know what to hook mine up to yet.

 

[Joe Perez]

I think it's to do with the fact hub motors are not so easy to seal, and oil could leak through the axle.

Yeah, I assumed that would be a problem, though in all the threads I've seen here, folks don't seem to have been bothered with it. Typically, the hub is being filled with oil only to a point which is below the axle-line, so it's not as though the axle is fully submerged.

I've yet to actually disassemble one of these motors, however I would assume that it wouldn't be all that difficult to glob some oil-resistant silicone sealant around the wiring at the point where it passes through the hole into the center of the axle, thus affording some measure of resistance to migration through that path.

The new clutch is supposed to be a lot tougher and it can't fit into the mac without new gears, as the clutch is thicker, the gears are not expensive. I think it's the same as the new bmc one, but I haven't seen it yet. But it will be easy enough to fit new gears and clutch if the time comes.

So, the complete package of clutch and gears will directly interchange between the two motors?

You will greatly appreciate the temperature sensor. I don't know what to hook mine up to yet.

If it's of a style which outputs a DC voltage (as opposed to being an unpowered resistive element) then you could attach it to any reasonably accurate voltmeter. The LM35DZ, in particular, is ideal for such a configuration as it is directly calibrated in degrees centigrade, at 10mv / °C. So 80°C will read as 800mv on your voltmeter. In a most simplistic configuration, all you'd need then is something like This or This to observe it.

 

[o00scorpion00o]

Yeah, I assumed that would be a problem, though in all the threads I've seen here, folks don't seem to have been bothered with it. Typically, the hub is being filled with oil only to a point which is below the axle-line, so it's not as though the axle is fully submerged.

I've yet to actually disassemble one of these motors, however I would assume that it wouldn't be all that difficult to glob some oil-resistant silicone sealant around the wiring at the point where it passes through the hole into the center of the axle, thus affording some measure of resistance to migration through that path.

So, the complete package of clutch and gears will directly interchange between the two motors?

If it's of a style which outputs a DC voltage (as opposed to being an unpowered resistive element) then you could attach it to any reasonably accurate voltmeter. The LM35DZ, in particular, is ideal for such a configuration as it is directly calibrated in degrees centigrade, at 10mv / °C. So 80°C will read as 800mv on your voltmeter. In a most simplistic configuration, all you'd need then is something like This or This to observe it.

Yeah I think I might hook it up to something like that until the cycle analyst III comes out, it should be comparable with that I hope ?

Afaik the new clutch with modified gears will fit, that's what I'm getting.

I wouldn't be too concerned about heat to be honest. You will be glad you went with the geared hum I can't wait to see what you think of it!

 

[Joe Perez]

Yeah I think I might hook it up to something like that until the cycle analyst III comes out, it should be comparable with that I hope ?

Well, I certainly hope so, given that I just bought one.

From what I can tell from the info that Justin has posted in this thread, an internal modification is probably going to be required in order to use the CA3 with an active device like an LM35- specifically the removal of an internal pullup resistor. Shouldn't be too difficult, assuming the software can be appropriately re-scaled.

I wouldn't be too concerned about heat to be honest. You will be glad you went with the geared hum I can't wait to see what you think of it!

The wait is killing me Having the bike down is a real bummer right at the moment, especially since I was in the process of re-painting my car when it died. Nothing like driving around town in a half-primered Miata with no rear bumper, trunklid or top. (although the trunk will be going back together this evening, the top is still a few days off. And it's raining right now, so the car is outside filling up with water as we speak.)

 

[o00scorpion00o]

Typical, there is rarely anything plug and play in this game!

You will like the mac. I doubt you will have any issues at the stock power levels.

Did you get the triangle battery ? can't remember if you said.

 

[Joe Perez]

Typical, there is rarely anything plug and play in this game!

Not a problem, fortunately. Wiring and circuit-board work is easy for me, compared to metal fab and machining.

You will like the mac. I doubt you will have any issues at the stock power levels.

"Stock" meaning 52v at 30A? Cuz that's the planned starting point.

Did you get the triangle battery ? can't remember if you said.

Well, I ordered it. An existentialist would point out that I have not yet "gotten" it, and then ask whether a person can truly ever "get" anything, wherein a nihilist would say that yes, they can, however it doesn't really matter in the end.


Also, you can purchase the Beta version of the CA3 here: http://www.ebikes.ca/store/store_CAV3.php