Using RC motors on E-bikes [Archive]

[bobmcree]

+1
Might I request that u also correct in the appropriate thread about mixing chemistries,
http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=1637&p=22955
safe's repeated assertion of how it's okay to run different chemistries tied in series as a normal practice.
I could be wrong, but my understanding from everything I've been taught or read, I think he's crossed the line from laughable to dangerous (to others & not just himself this time), as a little knowledge can be, & would like your view on this.

i think the question has been answered in that thread, but basically, if you have two batteries of different voltages and one is enough higher than the other to overcome the charge potential, that one battery will drain into the other one, possibly overcharging and damaging it. battery voltages vary with time and temperature, and different technologies have different charge potentials. for example, if your sla is at 13v your other battery will need to be at 15v or so before it discharges into the sla, but if you have a LiFePo4 battery the other pack only needs to be 50 mv or so higher than it to discharge into it.

the safest way to combine batteries of different voltage is to use a dual diode with a common cathode. you put one battery string on each anode and the cathode becomes the output. if you use a 40CPQ100 it will handle 40A and 100v and will cost you about half a volt. with the diode whichever battery has the higher voltage powers the load. what happens in actual use is that you draw power from the higher voltage battery until it sags down to the level of the lower one, then the lower one supports the load and the batteries will share with no possibility of one discharging into the other.

in the case of NiCd and SLA, if the voltages are very close you might get away with connecting in parallel, NiMH not so good, lithium definitely don't do it. If your SLA discharges into the lithium it could destroy it and kill you and some innocent bystanders.

I leave the dual diode on my bike all the time with the second tap available so i can carry any extra battery i want as an automatically shared reserve tank. usually i supplement my 48v 20 Ah NiMH with a 66v A123 pack, and the A123 pack runs down a bit before the NiMH starts to work at all, then they share for awhile until the A123 gets low and the NiMH takes over. If i did not have the dual diode the A123 pack would destroy the NiMH when i connected them.

 

[safe]

I've done series connection and parallel connection with SLA and NiCads.

Presently I've been using the series connection (so I have 24 volts of SLA and 24 volts of NiCads) and the results have exceeded expectations.

SLA alone would tend to become imbalanced as the "runt" cell tended to bring the whole system down.

SLA and NiCads in parallel tended to create a system that did not "sag" as much but it did not protect against the "runt" cell problem. I don't recommend this setup and no longer use it.

SLA and NiCads in series with the SLA's having two sets of 12 volts in parallel with each other tends to create a self balancing system. The NiCads are themselves paralleled so they too self balance.

The results in the "real world" (actual use) have been EXCELLENT for series connection.

So I'm confronted with this gap between actual results and some abstract theory about how things should work. The SLA's and NiCads in parallel worked fine, but they didn't do any kind of balancing so I would often end the ride with a very low "runt" SLA cell. Basically the parallel configuration didn't gain any advantages except to reduce the "sag". However, if you did parallel AND added a balancer it would have possibly been worthwhile. NiCads and SLA are so durable that they can be abused somewhat. (however the voltages usually started out close enough that they were okay) Typical SLA voltage would be 13 volts * 2 = 26 volts. The NiCads in sets of 20 would start at about 26-27 volts. When you charge an SLA battery you go all the way up to 14 volts... so SLA is easily tolerant up to 28 volts. I would usually plug and ride and not leave the batteries sitting for long, so the drain was almost immediate. If they sat around for an hour it might have been worse, I'll never know. All I'm saying is that in actual use the parallel wasn't that bad. (but series has turned out much better)

My advice:

Actually go and DO some of these things and get a thousand miles of riding with them before you criticize. The series SLA and NiCad battery setup has been fantastic and I would recommend doing it if you have similiar batteries laying around. The autobalancing feature alone is worth it... but an added feature is that the SLA is now running at about 1C discharge compared to the 2C that it was before when there were no NiCads. It actually has delivered better performance this way.

In every measureable way the series connection combined with the SLA's in parallel (lowering the C rate and adding self balancing) has improved the overall perfornmance of the bike.

------------------------------------

And let me add that my first attempt at series was incorrect.

At first I connected the 12 volt SLA's so that they made up units of 24 volts. That connection did not protect against the "runt" cell problem. So I scratched my head and realized that the 12 volts sets need to be in parallel units of 12 volts THEN combined together to make the 24 volts. "Obvious?" I thought to myself. (that was actually how I intended to do it in the first place, but I had some wiring that was already made so I tried it... sometimes laziness and opportunity can teach you something too)

Once corrected the pack now perfectly self balances without needing to resort to a balancing circuit.

(I never have "runts" anymore) 8)

 

[bobmcree]

My advice:

Actually go and DO some of these things and get a thousand miles of riding with them before you criticize.

my dear friend safe,

i was not criticizing you. i was merely trying to pass on some of what i have learned with my thousands of miles and thousands of hours in school and studying on my own.

if you have batteries in series you can combine anything you want so long as the lowest capacity battery can handle the full current and you never drain the lowest capacity cell down to its cutoff. parallel is different. when you transfer power from one battery to another you lose some energy in the process. sla is worst for this with an efficiency around only 2/3. This means you will lose 1/3 of the energy you feed from the nicd into the sla. lithium is much better and you lose very little.

if you spent a small fraction of the time you spend on graphic arts actually learning some basic electronics theory you would not look quite so ridiculous.

 

[safe]

All I know is that it worked.

-----------------------------

This discussion should NOT be here however because it's about battery topology, so I'm going to start a thread over in the battery area and if you want to contribute to it then do so.

This thread is supposed to be about RC motors.

Postscript:

When I first started here at Endless-Sphere (two years or so by now) I was constantly having mud thrown at me because I challenged the "Hub Motor Establishment" that dictated that anything but a hub motor was stupid. Now we have a section devoted to non-hub motored bikes and it's not considered such a stupid idea anymore.

Now I've come to realize that "Self Balancing" battery topologies are definitely possible and this seems to be the "New Taboo" around this place. You don't need to balance a pack if you have sufficient levels of parallelism in the battery topology. (at runtime at least)

There are always forbidden topics I guess...

 

[erth64net]

When I'm ready to convert a motorcycle, it looks like there's already 30KW brushless outrunners on the market

From:
http://www.rs-e-motoren.de/joomla/index.php?option=com_content&task=view&id=86&Itemid=79

 

[Miles]

When I'm ready to convert a motorcycle, it looks like there's already 30KW brushless outrunners on the market

If you have $5,400 .............

 

[safe]

If you start with the same number of cells for a battery the need to get all the way up to 72 volts means that you need to favor the series interconnections within a battery pack over the parallel. Doing this mandates the use of balancing circuits.

As the voltage requirement lowers (which is the case with RC motors) there becomes a unique opportunity to increase the amount of parallelism within the battery pack design.

This "could" mean that the RC motor might allow pack designs that differ from those of other ebikes because the self balancing that comes with parallel cells means that maybe you can skip the balancing circuits.

RC motors appear to be a "win win" because the lower voltage allows for more battery design options.

Anyway... by now people probably got the "heads up" on this one, but that's the concept.

Yet another reason to go with the RC motor... (this really does apply here)

 

[erth64net]

Anyone seen the Scorpion 55mm series motors yet?

With Kv's between 150-225 and rated for between 2500-4000Watts, these could hit a sweet-spot around here pretty quickly!

 

[recumpence]

The problem with those motors is the shaft. The shaft is coming out the top of the end bell for airplane prop drive. We need a base exiting shaft like a heli would use.

Do, they make a heli version?

Matt

 

[safe]

They look great that's for sure.

 

[erth64net]

The problem with those motors is the shaft. The shaft is coming out the top of the end bell for airplane prop drive. We need a base exiting shaft like a heli would use.

Can't you just poke the shaft out the other end? Scorpion shafts, at least on their existing models, are easily removed, and flipped-around.

 

[recumpence]

That depends on the motor. Some motors are easy to alter and some are not. I am not sure about that particular motor.

Matt

 

[safe]

Are prices dropping?

https://www.hobbycity.com/hobbycity/store/uh_viewItem.asp?idProduct=4316&Product_Name=hexTronik_PRO_120A_BESC_w/_PC_Programmability

$60

https://www.hobbycity.com/hobbycity/store/uh_viewItem.asp?idProduct=3890&Product_Name=HXT_63-74_200kv_Brushless_Outrunner_(Eq:5240)

$66

Can you seriously put an RC motor together for $126. (plus extras... but still)

What's wrong with this choice?

 

[MitchJi]

What's wrong with this choice?

http://endless-sphere.com/forums/viewtopic.php?f=28&t=8160&p=125821#p125821

Like most things, you get what you pay for with RC motors. The cheap imports use cheap steel, thicker stator plates and lower quality magnets. The net result is they can run hot, mainly due to eddy currents which generate power-robbing heat. Higher quality motors, such as the Hackers, NeuMotors and AstroFlights are much more efficient, run fairly cool, even at high power levels but do cost a bit more.

 

[safe]

Perfection is not my goal. I know that there is always going to be the "top of the line" product, but if the resistance of the motor is far less than the brushed types I'm presently using it's still a big improvement.

Typical brushed (Unite) motors are around 0.2 - 0.3 ohms

Typical RC motors are around 0.02 - 0.08 ohms

...also, I plan to be "legal" and only want to use 1000 watts of power. I'll be watt restricting these things.

Given that information... Why Not?

Is now the time to switch?

-----------------------------

Does anyone know the resistance value on this one? (HXT 63-74 200kv)

 

[recumpence]

The issue I have is the controller more than the motor. Everyone I have seen or heard of trying a low end ESC fires it pretty much instantly.

Matt

 

[safe]

The issue I have is the controller more than the motor. Everyone I have seen or heard of trying a low end ESC fires it pretty much instantly.

So you approve of the motor? I might seriously get one so I want to be sure this isn't some kind of error. It's a little bit shocking... this motor is so much better than the Unite and yet at this price point it's actually the same price. I had come to believe that the RC motors were going to cost in the $150+ range.

As for the ESC... I'm going to do the "Wattage Restriction Circuit" on the battery side if I do this and that would limit the total power to 1000 watts. (keeping it legal) Given that these motors pull upwards of 2500 watts if allowed to I would guess that the ESC would be seriously under used. With that expectation I was even looking at this one:

https://www.hobbycity.com/hobbycity/store/uh_viewItem.asp?idProduct=7341&Product_Name=SuperSimple_100A_ESC

$30

-----------------------

With the wattage restricted to 1000 watts this should be good enough right?

So it would be:

Motor - $66
ESC - $30

...for a grand total (not counting the other circuits I have to build) of only $96.

That's my kind of price range. :lol:

 

[fechter]

Keep in mind that the amp ratings on many of those speed controllers are figured out by the same guys that rate car stereo amplifiers. Peak is not anything like continuous. With proper limiting, however, nothing should be blowing up or melting.

 

[safe]

A 1000 watt restriction at 24 volts translates to 50 amps. So it seems to me that when the spec says:

Constant Current: 100A
Max Current: 120A

...that I would be "safe".

--------------------------

This is looking very good... I've been so enthusiastic about the RC motor watching it evolve here that it would be great to hop on board the trend.

So any last comments about why it could be a bad idea?

Anything?

My 24 volt battery pack would work perfectly and would be able to deliver a theoretical peak of:

12 (sets) * 2.0 Ah * 10C = 240 amps

...so in that area I'm way under my limits too. (a very conservative design)

---------------------------

I've been thinking about converting my old mountain bike (full suspension, lot's of good parts to start with) over to one of my road racers and this RC motor would be ideal for it. Maybe get that running by 2010.

 

[recumpence]

The only other issues would be startup and sync. Many cheap ESCs have terrible startup software and do n ot stay in sync well. But, hey, give it a shot and let us know how it goes!

Again, I would think the motor would be fine at 1kw.

Matt

 

[fechter]

Well, for $30 it is certainly not a great risk.
Here's one of the reviews from the web site:

Customer Rated 11/9/2008
hcuellart
I ordered this ESC to be used on a relatively high voltage setup (8S), first I used it with a 6S setup and everything went perfect. I exceeded 80A with no heating and the motor running great!! I was very satisfied till I tried to use it on the 8S setup. The controller could not take such high voltage, it fried... It is being sent for replacement, so to anyone using it, please don't go over 7S as it will fry... Still its a great value controller!!

I guess that's somewhere around 24 volts. Extra capacitors will help.

 

[safe]

For Lipo it's:

1 cell = 3.7
2 cells = 7.4
3 cells = 11.1
4 cells = 14.8
5 cells = 18.5
6 cells = 22.2
7 cells = 25.9
8 cells = 29.6

...my NiCads come in at about 26 volts to 28 volts.

So it does look that it's near the maximum allowable, but slightly below it. With the current being limited the heat should be less than if you ran without restriction.

Heat is always the enemy and it's usually heat buildup over time that actually destroys things.

My goal is to keep the "average" power to 1000 watts.

This ESC is "rated" as 2400 watts continuous.

It is "rated" as nearly 3000 watts maximum.

---------------------------

I'm more worried about sloppy control logic in the ESC that produces imprecise motor control. But then again... the way I ride... it's basically WFO or nothing all the time. Very rarely do I use partial throttle.

--------------------------

I'm going to need to build one of those control circuits to overcome the servo control these things rely on... so that's more work.

--------------------------

With the economy as it is I might also just wait and watch for bargains. These are great deals (already) but sometimes they can get even better. New products come out and they have to sell off the old stuff. This product looks to be overstocked and probably of lower quality and interest than the newer ones so that's why it's discounted.

Hard to imagine the price going too much lower though... :?

 

[gwhy!]

I was only looking at this very motor a few hours ago thinking uhmm thats a good price, and maybe I can just about afford to get 1 with a cheap esc Thinking along the same lines as safe on this one :wink: . One question I would like to ask is how are you guys attaching the prime pulley to the output shaft on these motors considering the amount of power they can put out. Do you use a 2 setscrew setup screwed down on to flats on the shaft only or do you also bond the pulley on to the shaft ..

 

[safe]

You (and I) both got excited about the motor price and then realized quickly afterwards:

"Oh no... now how do you use it?"

...everything that recumpence has been doing for the last six months or more applies to solving the gearing issue. There are all kinds of additional work needed to make it functional on an ebike.

---------------------------

The MY1016Z3 with the 10:1 geardown is a simple bolt on to my bike and the sprocket they give with the motor fits perfectly into my gearing plans. So for the bike I'm doing now (Project #003 as seen in my present avatar) I'm not going to change anything.

But I'm starting to follow the RC threads closer and closer because the performance is just sooooooooo good with these that they are the eventual destination for all of us.

 

[erth64net]

The issue I have is the controller more than the motor. Everyone I have seen or heard of trying a low end ESC fires it pretty much instantly.

So you approve of the motor? I might seriously get one so I want to be sure this isn't some kind of error. It's a little bit shocking... this motor is so much better than the Unite and yet at this price point it's actually the same price. I had come to believe that the RC motors were going to cost in the $150+ range.

If you're fixing power down to around 1KW, then I'd be personally inclined towards something such as the Scorpion 4025-16...and you're only throwing another ~$30 at this project...for a major part. The Scorpion motors have N50 magnets that'll survive up to 200 degree (celsius!), and enamel thick enough to handle up to 180C...overheating at that point becomes less and less of a concern. Further, even with the "good" ESC's, the AXI motors tend to resonate (see about 2/3's down the page at http://www.recumbents.com/WISIL/shumaker/e-cumbent2.htm ).