mgeoffriau

iTrader: (7)

Joe, any particular reason you're looking at MTB's with long-travel forks like that? My impression was that you were riding mostly on roads, sidewalks, and paved paths.

Joe Perez

iTrader: (8)

1: Even here in Carlsbad, the roads are not paved with glass. Some kind of suspension is nice to have, especially given the fact that I'm removing a lot of lightness. (I'd opt for a full-suspension frame if it didn't chew up the space I need for the battery.)

2: The width between the dropouts is greater on MTB frames, so it's easier to accommodate a motor.

3: In general, parts availability is quite good for 26"-wheeled MTBs.

rleete

iTrader: (21)

So, assuming the bike is in the $500 range, what's the budget for this build?

Also, assuming all new battery, controller, etc. what is the expected range?

Joe Perez

iTrader: (8)

No budget. Just whatever it takes.

I hate it when people come up with ideas and then constrain them with a budget. If you're cash-limited, then you probably shouldn't be spending money on whatever it is you're doing in the first place.

I haven't run all the numbers yet. Maybe $2k total? Maybe more, maybe less.

rleete

iTrader: (21)

No, range, as in how far will you go under power? Obviously, this is extended the more you pedal.

Joe Perez

iTrader: (8)

Estimates vary, and depend on what depth-of-discharge you are willing to accept. For a 50v, 11Ah battery, most people claim something in the neighborhood of 20 miles, plus or minus maybe 25%.

That's with zero pedaling, and assuming a motor whose stator windings are optimized for low-end torque rather than top-end speed.

Joe Perez

iTrader: (8)

rleete

iTrader: (21)

I like that arrangement MUCH better than the cheesy milk crate on the back.

What's the shaded area on the back?

Joe Perez

iTrader: (8)

Actually, the milk crate's primary function had nothing to do with the batteries. It's there to carry groceries. This is, after all, supposed to be a practical commuter vehicle.

The BMS. I only had accurate dimensions for the primary body of the battery pack, so I just eyeballed the BMS and drew it in.

Figure an extra 1-2cm all around for the nylon bag to support the battery, and it'll just about fill out the triangle perfectly.

This is what the battery itself looks like:



I just placed the order, and the supplier is in China, so it'll be a couple of weeks before it shows up. Trying to get all the little mechanical stuff taken care of in the mean time.

fooger03

iTrader: (2)

What's the chemistry of that thing?

Joe Perez

iTrader: (8)

LiFePO4, made from genuine A123 26650 cells. Internal configuration is 16S5P (5 parallel strings each consisting of 16 cells in series) for a total of 52v at 11.5Ah. And more importantly, the BMS on this unit is rated for 30 amps continuous, 50 amps peak, with a big ole aluminum plate for a heatsink.


Datasheet for the cells:

fooger03

iTrader: (2)

wowzers

What is the charge method going to look like? Do you have to worry about balance as in a LiPo series? That's got to be a pretty amazing BMS if it's balancing 80 individual cells.

15 Amp charge at 57v for 45 minutes, or 50 Amp charge for 15 minutes - those would be pretty amazing charge times, but damn that seems like a lot of current. Are there any heat concerns with these cells at full charge/discharge being grouped together in a pack of 80?

Joe Perez

iTrader: (8)

The charger is 58.4v at 6A. It's nominally configured as a series (bulk) charger, and that's where the BMS comes into play. It monitors each parallel cell group, and shunts across that group when the voltage rises to the cutoff point. So it's a "top balancer" rather than a "bottom balancer" as the RC guys typically use. Conceptually, it's the same as my old one, but with a higher capacity and a properly-engineered BMS.

I seriously doubt that there will be a heat issue. This style of battery pack is quite common among the e-bike crowd, and nobody has really complained when operating them at reasonable current levels. Actually, they usually group the batteries together into a rectangle, which would have even worse thermal performance, as the surface area in such a pack would be less.

This is the BMS that will be used:



In this picture, it's installed onto a rectangular block of prismatic cells, rather than a triangular block of cylindrical cells, but the basic idea is the same.


There are a lot of folks using RC-style LiPo batteries, and while not all of them have burst into flames and burned their house down, they seem to be much more of a science-fair project than I want to deal with. This battery is larger, heavier, more expensive, and seems to be sufficiently well deigned that you can just install it and forget about it.

fooger03

iTrader: (2)

Fair enough - my hi-tech rechargeable battery experience comes from RC - my RC LiPo is a 6S that runs near 25v. I'm sure it uses a hell of a lot less current pushing around 4lbs of plastic than your bicycle will, as I can run it for an hour on a charge - and usually get bored before the batteries give up. Fukcer will do 60mph in a straight line too, and if you run it for a full hour, those batteries get to be nice and toasty.

Joe Perez

iTrader: (8)

Honestly, I'm not quite certain where I stand on the whole LiPo issue.

The high power density certainly appeals to me. And while clear information on proper balance-charging (and discharge protection) is harder to come by than MS3 schematics, it does seem that it's possible to cobble together a system which will perform both individual-cell balance charging and individual-cell undervoltage protection.

Of course, by the time you've bought all this fancy hardware, you've exceeded the cost of a LiFe battery that will deliver twice as many charge / discharge cycles for an equivalent DoD.

And such a system strikes me as an accident waiting to happen. LiPo batteries, by their nature, want to fail in the most dramatic way possible. Sure, you can put together a really complex system with lots of individual failure points to try and prevent them from going up in an expensive fireball, but is it really a good idea to do so, particularly in an application where failure will result in a large explosion directly between your legs while you are traveling at 25MPH across pavement on an unstable vehicle surrounded by moving cars?

On a larger scale, that's the sort of mindset that results in lots of people being killed, manslaughter charges, and your very own segment on the History Channel's "Modern Marvels: Engineering Disasters."

Engineers hate risk. They try to eliminate it whenever they can. This is understandable, given that when an engineer makes one little mistake, the media will treat it like it's a big deal or something.

Examples of bad press for Engineers:

• Hindenberg.
• Apollo 1, Space Shuttles Challenger and Columbia.
• St. Francis Dam.
• Skynet
• Hubble space telescope.
• Hyatt Regency Hotel, Kansas City.
• Chernobyl, Three Mile Island and Fukushima.
• Titanic.
• Ford Pinto.

The risk/reward calculation for engineers looks something like this:
So, yeah. LiFe batteries look pretty attractive.

fooger03

iTrader: (2)

I haven't researched enough to see individual cell undervoltage protection (I would expect that bumping the ESC undervoltage protection up a couple tenths would suffice, as long as you have a balanced charge - it seems it would be difficult to discharge one cell in the pack fast enough relative to the other cells that you would damage the cell, as long as you had a balanced starting point.) I expect you're already well aware of the charging prinipals, but for others: In my 6S, it's simply a charger designed to balance charge the cells; there's the primary connector which carries full current, and then each individual cell has two leads which attach to a second connector. Both connectors attach to the same charger. The charger is then able to monitor the individual cell voltages while charging, and apply a slightly different voltage to speed/slow the charge process of the cells such that they all arrive at target voltage simultaneously. It has error protection in case cells are overcharged or over-discharged. If they are slightly over-discharged, it will only trickle charge the battery until all cells are within tolerable range, at which point it will begin charging at full current.

And no, I wouldn't want a LiPo exploding anywhere near me - especially between my legs while I'm doing 25mph on an inherently unstable vehicle surrounded by cars.

Bryce

iTrader: (24)

I Lol'd when I thought about some audiophiles. Budget? What budget?


My younger brother has a LiPo-powered rapid-fire airsoft gun. The thought of that battery exploding while in my hands hasn't really occurred to me while using it.

Joe Perez

iTrader: (8)

Shows what little I know about these packs.

I've never used one, but I had thought that the fancy chargers used the "main" connectors only for series-charging "old style" packs (NiMH, NiCad, etc) and the LiPo packs were charged using only the balance leads.

So, if one were to build a battery out of three 6S Zippy LiPo packs, and buy a charger such as this one, you'd need to connect the three balance cables to each individual pack's balance leads, and also disconnect all of the main leads and plug them into the series ports on the charger as well?

(does quick math)

So, if I wanted a single-plug solution to charge such a LiPo pack, I'd need a connector with a total of 27 conductors, six of them heavy-gauge?

Joe Perez

iTrader: (8)

Well, I like to think that I'm spending money in places that will actually make a functional difference as opposed to just gold-plating everything in sight and paying 42,000% markup for wire. (As a broadcast engineer, so-called "audiophiles" **** me off. They are to audio as pedophiles are to children.)

fooger03

iTrader: (2)

3x6S would give you about 66 volts.

Each 6S pack would have the + and - high current wires (which in reality could be shared across all 3 packs if you had enough current)

And 7 additional low current leads.

That's 21 balance leads, + 6 high current leads for a total of 27 connections.

P.S. - that looks like a pretty bada$s charger.

If you run a programmable ESC, you get to program the cutoff voltage in it, if you run a preprogrammed ESC, it will cut voltage at a relatively safe realm. If all three of your packs (and all 18 of your cells) came from the same manufacturer lot, you probably stand a higher chance to get struck by lightning than seeing a pack explode. I would wager a guess that damn near all pack explosions result from:
o Overvolting on the charge (because of a shitty / incorrect charger, no balance capability which can overvolt a single cell, or dumb person)
o Too high of a current draw on the battery (thus overheating the battery)

To completely annihilate the chance of overvolting a cell, you buy a good balance charger.

To wave goodbye to the possibility of exceeding current specifications, you buy a battery big enough to handle the amperage. (This part is slightly more complex in that it involves math - but the *math* part really isn't complicated.)

First, you want to find the "C" rating of the cells - High end LiPo cells can run around 150C (surge) nowadays. (5 years ago, top-of-the-line was 90C)

Then find the capacity of the cells - For simplicity's sake, we're going to say that our theoretical cells have a 10,000mah capacity (That stands for "milliamp hours" for the electronically challenged)

We now divide our 10,000mah capacity by 1000 to get amp-hours, this cell gives us 10 amp-hours. Multiply the amp-hours by the C rating (150) to get allowable current draw, so:
10 * 150 = 1500.

That, my friends, is a battery with a maximum discharge rate of 1500 amps.
Keep in mind, that 1500 amps is surge capacity, and in reality, you wouldn't want to pull more than about half that (75C or 750 amps continuous) on it.

Realistically, for your application, you're probably going to be pricing batteries in the 75-100C surge (or 40-50C continuous) realm (much cheaper than 150C batteries), especially if you won't be approaching the max current draw of the battery. A 10,000mah battery rated for 40C continuous still has a max continuous discharge capacity of 400 amps. A 5,000mah 40c battery would have a max continuous discharge of 200 amps.

When possible, you want to oversize your battery by about 20% - that is: if you expect to be able to pull 300 amps, then get a battery that is capable of a continuous 360 amps - this will greatly improve the cycle life of the battery.