High power RC motor and drive unit production

Between DIY electric car forums and RC forums, I've read quite a bit about different chemistries. I simply haven't used anything but LiPo using some charger other than the one that came with the battery. FWIW, my current charger can handle any of these chemistries, only it only goes to 12v.

No matter what chemistry I use, I will not settle for something which does not have a BMS. For LiPo, I was thinking to just get a balancer and leave it plugged in even when I drive. Not sure about the current handling of that though, not sure if the balancer can keep up when a cell starts to get weak. Anyone know? Also an LVC would have to be in there, if the balancer doesn't do that.

A123's, as far as I knew, didn't come as free cells. Which means you buy a bunch of drill packs, take them apart and then make your own battery. Just an additional step, aside of the lower energy density thing.
 
recumpence said:
Oh no, here we go on the battery arguement! :D

No arguement, I just think it is important that people reading this get all the facts.

recumpence said:
Gary, you have me there on the cost. However, I have never seen a LiFe pack that is as light as a LiPo pack or as small. From my research, they are typically twice the weight and size of similar LiPo pack. My recumbent pack is 5 pounds exactly with wiring and it is a 48 volt, 10ah pack. Also, A123 cells can put out huge current, but they drop voltage further than good quality LiPo to do it. If I pull 10C out of my LiPo pack at full charge, I only lose 3 volts under load.

First of all, your 12s setup has a nominal voltage of 44.4V, not 48V, and gets charged to 50.4V. A "48V" 16s LiFePO4 pack has a nominal voltage of 52.8V, and gets charged to 58.4V. The reason these 16s LiFePO4 setups are called "48V" is because a 48V SLA setup gets charged to about the same level (2.45V per cell x 24 = 58.8V...), but SLA cells have only a 2.0V nominal voltage, so 48V. Since many LiFePO4 packs end up as drop-in replacements for SLAs, they can use the same charger. There is an added performance boost in that the bike's electronics need to be able to handle the voltage fresh off the charger (i.e. -- 58V...) but SLAs will quickly come down to the nominal voltage, even lower under load. LiFePO4s will start out at the same 58V, and will come down to their nominal voltage pretty quick, but the nominal voltage is a full 4V higher.

As for voltage drops, I just don't see a big difference in 10Ah-size packs. Maybe you are thinking of single "p" configurations, like in RC models? There I would agree a new generation 2500 mAh LiPo pack is going to see less voltage drop under load than a single string a123 pack. That is just not been my experience, however, with the a123-based ebike packs that I've been using, most of which are in either 4p (9.2Ah) or 5p (11.5Ah) configurations. I don't see any more than 3V drop on any of them and that's pulling the same 80-100A loads. Unless the pack is empty, I just don't ever see less than 49-50V. That, by the way, is what your pack is, fresh off the charger.

recumpence said:
One thing I neglected to factor in (in my previous post) is charging. LiPo packs require an expensive charger to be charged safely. That dramatically increases innitial cost. The chargers are also large. So, they would typically not be carried with the bike. Whereas Ping packs typically use a small, lightweight charger that can be carried along easily. That, along with the safety factor, makes Ping packs and other LiFe packs more desireable than LiPo for the majority of riders.

Yes, this is true, but what you are also forgetting about is that the BMS on LiFePO4 packs also include the balancing function. More accurately, what the BMS circuits do is to shunt current off to the next cell so that if a cell is full the next ones will still get current supplied, so that they can "catch up". This way each cell gets a full charge and can do so at its own pace. It is more like having individual chargers for each cell. RC-type balancers typically just drain the high cells down to the level of the lowest. The BMS also contains an individual LVC circuit for each cell. This protects any cell from being over-discharged. You can use a pack-level LVC function only if cells are closely balanced.

recumpence said:
I will repeat my warning on safety here..........

LiPo packs can be hazardous! I have had one explode when overcharged (bad charger). They are NOT for the beginner or the faint of heart. I have been running them for years, though, and am thoroughly familiar with their behaviour. That is why I run them.

[pulpit]
I wholeheartedly agree. LiPos are very powerful, and if proper precautions are taken, they can work quite well, and can be trouble-free. The problems arise when these precautions are not followed. In the RC world, we've all had this beat into us enough that people treat these packs with the required safe procedures. LiPo use over in this world, however, is fairly new. Worse is the fact most people's experience here with Lithium-based packs has been with the much safer LiFePO4 variety, so there's no built-in "worry factor". One early example I've seen is that some people here are simply using simple fixed voltage power supplies/chargers on these packs, instead of the inherently much safer RC chargers, most of which will at least monitor the voltage of each cell as the pack is charging, and will terminate the charge if any one cell gets too high (above 4.25V...). With LiFePO4-based packs, you can let cells go above the charge limit of about 3.7V and the worst that will happen is that you may eventually lose some capacity if you do this on a regular basis. You let a Lithium-Cobalt-based LiPo cell get above 4.3V and the fireworks will begin. :shock: The only safe way to bulk charge a LiPo pack is with an RC charger that monitors each cell.

I'm sorry if I sound like a broken record, but I feel strongly that we can't be too cautious in this area, and those of us with lots of LiPo experience should not miss a chance to preach the safety aspects of how to handle, use and charge these packs.
[/pulpit] :wink:

-- Gary
 
I agree with you on all points, except one;

My 12S pack has never been below 44 volts, ever. It comes off the charger at 50.3 volts and stays up at 48 volts nearly the entire ride. The last few miles of my pack may get down to 46 volts under normal riding and 44 volts if I accellerate with a relatively low pack. But, according to my Eagle Tree, it runs between 46 and 48 volts pretty much all the time.

If I pull 100 amps with a fresh pack, it will drop to 45 volts, but no further. These cells have performed wonderfully. :D

Matt
 
1clue, i don't know jack about chemistries of batteries but i would say if you are capable of using the type matt uses then they can't be beat on size and weight as far as i know, 5lbs for 48v10ah is very impressive.
However if your not comfortable with lipo "risk", like me, then i understand that the 123's are very well behaved should they zero out/etc so for me it was a no brainer, safety first!! ,not that lipo isn't safe if understood and respected but rather i'm a lazy sod and have to calculate this when thinking about battery maintenance :lol: .
initially i had a couple of prebuilt lipo's but they weren't great and allways used to give me the willys sitting in the spare room :shock:, at the end of me using them for a while one of one had gotten a bit fatter the other due to pouch puffage :shock: but see point above about the lazy calculation and figure this may have contributed.
so i bit the bullet and bought the drill packs and had them made to a pack, i haven't worried about the batteries on my bike since, and the performance is great, like Gary i have very little voltage sag right to the end.
i'm not posting for 123's here against lipo but just telling you why i chose 123's obviously it will depend on what you want.
i wouldn't be put off from individual cells though, the nice thing with a custom pack built from individual cells is it can be any shape :)

my two cents, which is good as thats the about right on the value of my opinion on bateries.
hope it helps :lol:

Cheers,

D
 
D's drive is done. :twisted:

It should ship tomorrow.

For those who want to know, the freewheel is inside the pulley. That is so a lower ratio can be used withou a two stage drive (smaller sprocket on the output shaft). This crazy pulley took me 4 hours to mod this way. So, unless you really need it done, don't ask. :mrgreen:

Man, I would chuck a part in the mill, machine it, take it out and check it, put it back in the mill, machine more, take it out, etc, etc, etc. What a pain. Every part but one had to be modded for this drive. But, it turned out beautiful!

Enjoy the pics!

Matt
 

Attachments

  • DD2.jpg
    170.7 KB · Views: 1,329
  • DD4.jpg
    137.1 KB · Views: 1,310
  • DD5.jpg
    148 KB · Views: 1,314
Beautiful! :)
 
Uhhh, D just moved. The new address is.... :wink:

But seriously that is a thing of beauty. It would be cool if somebody actually did a test to see how much power it is capable of putting out. I thought I would like the anodizing, but black kinda spoils the craftsmanship. That ENO looks right at home inside the pulley though.
 
D,

Congratulations on your expected visit from the stork. It's a beautiful baby ... e-drive. :D

You have me almost exactly backwards. The only cells I have any experience with at all are more dangerous than what Matt uses, and smaller. I had planned on using those since I DO know the precautions to take, and I would monitor them like a mother hen watching her chicks. I've never used any which were nearly that big, though, so it will be an adjustment no matter what chemistry I wind up using.

However, since people on this site have an opinion and some knowledge of the other chemistries, I'm listening with an open mind. I don't exactly want to ride around with a bomb stuck to my bike, but I think I know how to take care of them, and will no doubt spend hours on research before I finally start buying.

Real experience with batteries, used in the same application, is what is really valuable. Anyone can go read some info off a web site, but if it was put up by the guy doing the selling, what can you believe?

I'm really busy lately, so I haven't been able to do any research.
 
Wow, that looks so good I think you may be able to use it to travel back in time 8)
I may have to steal your idea of slots adjusting the motor for belt tightness. It just looks better than what I did. good job Matt, you rock! :mrgreen:
 
Looks good on the scooter.

You will notice the narrower belt...... That was to increase the efficiency of the drive. It will still handle more than enough power. With the motor you are running, that 9mm wide belt is perfect. :D

Looking forward to your impressions of it!

Matt
 
Hi,

The mount is not centered in the drive. For mounting between the pedals does the Goodrum Clamp Mount â„¢ need an offset so that the wide side of the drive (pulley side) doesn't interfere with the pedals?

I think Matt said the belt stage could be configured up to 5:1 which would be 16t-80t. I think Miles recommended that D use a 19t, then OK'd 17t. What sizes of small pulleys are ok for what power levels?

What is the difficulty of using pulleys larger than 80t?
 
Both mine are setup for 17T/70T. I believe D's is 19T/66T. Looking at his, I'd say it will be fine. Mine is right at the hairy edge, I think, as Miles predicted. I may have to go to a 19T if I get any slippage at high power levels. I think also that 66T is as small as you can go and still fit a an ENO FW inside, although Matt is not engcouraging any one else to go this route as it took him 4 hours to get D's to fit. Don't expect to not pay hansomely extra for this. :)

I should get my second drive today, I think. This one will be for the Hacker and it will go on my 16" folding bike. The Hacker has a slightly higher kV (149 vs. 135...) than the 3210, but this gets offset by the smaller diameter wheels. I've decided to use the 3210 setup on my 20" Mariner folding bike, replacing the Cyclone 1000W setup that is currently on there. First, however, now that my 16s5p a123 is finally finished, I'm going to run the Cyclone setup with my Eagletree eLogger, and get a baseline of the Cyclone performance. Then I will swap out the Cyclone for the 3210/eDrive assembly, and run the same tests. For now, I will use a Kontronix PowerJazz controller, in place of an HV110 controller, until I have the time to verify that the latter will work with a 16s a123 pack, fresh off the charger. 15s works fine, but I've not yet tried 16s. I have an HV85 that will be the "sacrificial lamb" this weekend for the 16s test. The PowerJazz is good for up to 15s LiPo and 18s a123 setups.

-- Gary
 
I was able to squeeze the freewheel into a 60 tooth pulley for D's drive. He wanted it as compact as possible. The final drive between the e-drive and his pedal sprocket will hvae to be calaulated properly to get the performance he is looking for.

The problem is, a minimum number of teeth need to be engaged on the motor pulley for each 1000 watts of power. When you get up around 3 to 4KW, the tooth count needs to be up at 19 tooth or so. My recumbent uses a 22 tooth motor pulley with a relatively tight belt to handle 5kw+.

Anyway, why does D vanish right when things get good? :wink: :mrgreen:

Matt
 
hahahaha, im here, im here. work has been a biattttchhhh. im not sure what my company think they are doing not leaving me enough time to browse the forum during office hours, maybe i'll drop the union a mail about this infringement on my rights.

:D :D :D :D :D :D :D :D :D :D :D

look at my baby 8)
Mr S, thank you for such a wonderful piece of engineering,not to mention the extra work you put in for me to appease my ocd and have everything just so, you sir have what is known as ...mad skillz :mrgreen:
im so impressed with my drive, compact,elegant ass kicker - where have you been for the last two years?
i can't convey how keen i am to see this baby run,i have everything i need now to ship the bike off to steves for final assembly bar my ruddy controller :evil: oooh there should be some hot videos this season :D
i love the black anodise Matt, not everyones taste i know but its simple, my seatpost is black and stealth is key, it looks a lot darker black than i thought which is really pleasing, i was expecting it to appear silvery under scrutiny but that looks great.
from the output shaft we're going 219 chain,11t to 80t, i have decided that i will use the 42t chainring because the extron needs to be on the outside to clear the frame, so the 52t chainring wont fit on the inside but the 42 will, this isn't a problem as 42 was used in the calculations for speed.
im stoked!! i'm really close to having the bike built and the using of the front shifter on the rear deraileur is also looking to be very good idea but i'll stick those details in my build thread with some pics of the new setup.

Again thx Matt, i can't believe how good this is :mrgreen:
k let me go take some pics and i'll try to post them later.


Cheers,

D
 
I went a bit farther than I planned on with your drive to take it as far as I possibly could.

The motor plate was machined with a curved relief allowing the motor to be moved very close to the jackshaft for the most compact arrangement possible. The freewheel completely disappears into the pulley. There is no protrusion past the edge of the pulley at all. The FW adaptor fits opposite direct into the FW. That allows a MUCH narrower profile. I had to do a lot of massaging to do that, though. It was a pain making the FW set screws accessable through the front side of the FW. But, in the end, it worked out.

I did not machine flats in the motor shaft. You guys can do that. I did not want to mess up your motor shaft and I cannot chuck it into my mill. I would recommend a Dremel tool or a die grinder to put set screw flats in.

Did you want the drive shipped to the address on your PP payment?

Matt
 
Hi Matt,

yes i really did notice how slim it was, top job 8)
im hoping to pop it in behind the seat post and it be solid, it'll be tight but i just might do it, if not i have backup plans to mount elsewhere but thats a last resort for me.
if the pp address is the 38 lower road address that would be the right one, i pm'ed you the full addy but i seem to have 3 messages stuck at the mo, 2 to john so i guess he doesn't know my hv isn't here yet :cry: ahh well can't have it all, can have the bees kness drive tho :D

Cheers,

D
 
I am looking forward to seeing the build come together.

Looking at at your bike, it seems the wheel will come closer to the seat tube under suspension compression. Also, you have tons of room under the downtube. I think it would look cool, and be much easier to mount, under the downtube in front of the cranks. :D

Matt
 
MitchJi said:
Hi,
The mount is not centered in the drive. For mounting between the pedals does the Goodrum Clamp Mount â„¢ need an offset so that the wide side of the drive (pulley side) doesn't interfere with the pedals?

We could certainly do that, if it was needed - either by laterally offsetting the "u-shape" or by having it at an angle.
 
recumpence said:
Gary,

Did you get drive #2 yet? How does that bearing tube/offset look?

Matt

Yes, it came yesterday, thanks. :) The tube/longer shaft is just right. I checked this on a couple of bikes, but it was just holding it in place. I'm working on mounts for both my 20" Mariner and 16" Dahon folding bikes, for the 3210 and Hacker setups respectively. In a bit I will be able to check the fit more exactly.

The Hacker mount has all four motor screw holes lined up perfectly, but I may need to lengthen the slots a bit, only about 1-2mm, in order to get enough tension on the belt. I could drop down a tooth or two in belt length, but I hesitate to do this as I'm already worried there's not enough contact on the 17T motor pulley. I may end up having to go to a 19T for both setups.

I wsa going to put the 3210 setup on my Townie but I'm going to wait for the clamps for this setup. I like the idea of doing the Mariner first, as it currently has the Cyclone 1000W setup so it will give me a reference for comparisons.

-- Gary
 
Back
Top