Triple Rewind of Unite 500W Motor

[safe]

Ebike As Air Conditioning

It's 94 degrees and high humidity that places the heat index into the 110 degree range. Oddly, if you ride an ebike in this the fact that air is flowing past the body (exposed skin) it evaporates and so it "feels like" only 94 degrees rather than 110. I reverted back to my bicycling outfit rather than the leathers at this temperature. I rode a few of the back roads with lots of trees and shade and it was actually pretty nice.

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

I got around to installing the newest NiCads I have into the place where I was trying to use up some of the older ones. I really abused that older set of NiCads so here are some "tips" that I think will make the NiCads last longer.

Tip One:

Let the NiCads cool a little after the ride. Obviously there are times when you need to ignore this rule if you want a rapid turn around time for charging, but cool batteries will charge better.

Tip Two:

You need to cycle NiCads several times before they come up to full strength. The best way to think of it is like exercising a muscle, start out easier and gradually build up. At some point the NiCads will peak in their performance. Try to get most of the charge used up in each cycle, but don't bother draining the cells to zero... there is no "memory effect" to worry about as some have thought.

Tip Three:

If you followed "Tip One" and let the batteries cool then when the batteries charge they will gradually rise in temperature close to the end. The charger needs to see this rise in order to cut off or else it delays cutting off until it reaches another higher charge level at which it cuts out based on voltage alone. You don't really want to get to the second cutoff because heat is what wears out the cells and lowers capacity. (these chargers actually have heat sensors, but I'm not using them, so you would be better off if you used them and then the charger would cut off sooner)

Tip Four:

Check the NiCads as they charge. I usually check the time I start and then come back about half an hour later when they are usually close to done. If one of the sets has completed, then check the temperature of the others, if any seem warm, then just cut them off manually. Check the voltage afterwards and the odds are that the charger missed the first cutoff signal.

Tip Five:

Check the NiCad sets after they are done. For my setup they all end right aroung 27.4 volts in the first cutoff. If it goes to the second cutoff the ending voltage will be 28 volts or above. Sometimes a set will cutoff too early (usually because they were too warm at the start and as they cool they give the charge a "false positive" signal) and if this set reads a voltage that is low you just have to restart it.

...in the end all the sets need to settle on some resting value. For me the resting value is about 27 volts. Over time you just sort of get a feel for what is going on. The main thing is to try not to cook the NiCads any more than needed. Excessive heat reduces their capacity and can create a "runt cell" that drops to zero and needs replacing.

Anyway, that's my two cents worth on the topic...

 

[safe]

The Good and Bad of Parallelism

In my most recent wiring I have five sets of 24V connected in parallel which then is connected to a second parallel set.

The Good

On the positive side the 24V tubes will self balance across the parallel connection making better usage of the capacities available.

The Bad

On the negative side if one of the 24V tubes develops a "runt cell" which drops to zero volts then the overall voltage enters into the parallel set with a significantly lessor value. This makes the other tubes work harder as they try to backfill the bad tube. Today I noticed that one of the tubes was not able to hold a full charge above 27 volts and when I rode it I found that the battery worked better overall if I disconnected the offending tube completely. When I got home I opened up the tube and found a "runt cell" in it. (these are a new set of cells, so you expect a couple to start off bad)

...parallelism can be the "low budget" BMS that does automatic balancing for you, but if a single "runt cell" has failed in any of the tubes the entire parallel set is dragged down with it.

Checking the starting voltage can identify the "runt cell".

(and when you take the cells out of the tubes they are always "all or nothing" so there's little doubt which went bad)

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

I simply can't imagine soldering a battery together and expecting it to last for very long. The odds of at least one cell going bad is high enough that the expected lifespan of the battery overall is not going to be very good. In my opinion your options are to either do something like I've done and allowed for easy replacement of bad cells or to use a full blown BMS that monitors everything. Soldering a large many celled pack together permanently is doomed to fail over the long run, but this is a common practice in the ebike world because it's cheap to do and sells the ebike. The seller does not care about long term reliability, but just to sell the ebike. Once the owner begins using a soldered together battery they are responsible for it and when it goes bad they just assume they need to buy something else, many times thinking that the problem was the chemistry of the battery or some other thing which is not really accurate. All chemistries have predictable wear rates if you take good care of them.

It's important to keep your batteries "safe".

 

[safe]

Chain Harmonics

The fact I'm able to worry about this says that the rest of the bike is doing well. The NiCads are all functioning nicely and starting and ended each ride in a state of being balanced, so for the moment that issue is resolved.

The Problem

At about 50% of peak motor rpm the chain gets into a harmonic rhythm which upsets the power delivery. Below 50% and things are fine and above 50% and things are fine. The problem is that typically when you upshift you drop the motor rpm and that exposes the problem. So it "runs rough" at certain rpms.

As best I can tell the chain vibrates from the releasing side of the motor sprocket until it gets to the rear derailler. Since higher gears mean that the derailler gives less spring tension the problem worses as I get into higher gears. I have modified the spring inside the derailler to give more tension, but I might need even more of that or even better would be to find some stronger spring.

What creates the oscillation (I think) is the motor sprocket because I made it myself and it's only 12 teeth. The sprocket didn't turn out perfectly round.

So the repair options are:

Start over and create a perfect 12 tooth sprocket. (too much work)

Increase the spring tension in the derailler to hide the problem.

Live with the problem because it's not all that bad.

...tomorrow the hellish heat returns. (argh)

 

[safe]

New Average Speed

The old battery setup gave me average speeds like 27.9 mph. (page 72)

The new battery setup gives me:

Distance - 10.26 miles

Time - 21 minutes and 35 seconds

So:

21 * 60 + 35 = 1295 seconds

10.26 / (1295 / (60 * 60)) = 28.5 mph average

The difference:

28.5 / 27.9 = 2%

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

Obviously it depends a lot on the terrain, but a 40 lb weight difference only achieved about a 2% improvement in a 10 mile loop I ride. Maybe I can improve on that by riding the loop faster next time, but I don't think there's much more I can get. Unless you are doing a lot of starting and stopping or hillclimbing there is little difference by dropping the battery weight. This is probably because I weigh about 185 lbs and the battery is a small percentage of that, so differences in battery weight don't effect the overall weight that much. Much of the loop is flat and I'm able to maintain speed most of the time.

It's going to be hot today, so I did this ride at sunrise...

 

[safe]

Faster

Distance - 10.25 miles

Time - 20 minutes and 50 seconds

So:

20 * 60 + 50 = 1250 seconds

10.25 / (1250 / (60 * 60)) = 29.5 mph average

(fast speed zones - 47 mph, 46 mph, 49 mph -5% slope)

(slow speed zones - 19 mph, 18 mph +5% slope)

This was a really fast ride this morning, but I'm still not able to crack the 30 mph barrier for this route. The best I ever did was with a stronger motor (about double the power) and the heavier batteries when I managed a 32 mph average. Power matters the most at higher speed because at the higher speeds wind resistance is the most dominant factor. Weight seems less of an issue than you would think for this fairly flat course. Since most Go Kart tracks are pretty flat the weight should only effect times when you are trying to change your speed. If it's possible to carry your speed through the turns then the weight matters less. This seems to suggest the importance of excellent traction because what you want to do is keep the throttle nailed wide open most of the time and just ride the turns flat out.

Skinny tired bikes that are light weight might actually lose to a heavier bike with super wide tires that give good traction as long as the heavy bike is able to get through all the turns without losing momentum.

Think of it like a spring... if you have skinny tires you have to slow for turns (compress) then speed up again to get out of them, but if you have fat tires then you never need to compress the spring because the weight is being carried at a constant higher rate. Ideally you would have a really light weight bike AND have wide tires... the best of both.

 

[Stan4d]

Hmmmmm... expected some replies...

Let me repeat what I've discovered here.

Basically the limiting factor for most electric motors is their heating. Heat is produced based on the formula:

Heat = Current * Current * Resistance

...so as you try to increase the current the heat rises dramatically. But resistance can also be a factor. If you could drop the resistance from the typical Unite motors level of about 200 mOhms - 500 mOhms down to the level that most RC motors run at (30 mOhms - 80 mOhms) then you could reduce the heat by that same amount.

The "trick" that I'm using is exploiting the difference beteween a series connected circuit verses a parallel connected one. When you divide the motor coils up into separate parallel winds this divides the current flow and that dramatically reduces the resistance. (for a Triple wind this means that only one third as much current flows through each of the coils in order to achieve the same magnetic field )

So a 500 watt Unite motor can all of a sudden pull 1500 watts and not overheat. (no need for goofy air blowers to solve the problem you fix it with a change in the motors design) However, you do not ever get something for nothing and the penalty is that the motor now DEMANDS a higher current just to operate. The no load current increases dramatically. All of a sudden you get the added power, but at the expense of "fuel economy".

People who want sheer power should study this.

If it's of any help, on the last Triple rewind I managed to get this bike up to 58 mph on flat land.
http://www.ebikehub.com/forum/download/file.php?id=265.jpg

Just a few questions:
1. Are you claiming that this is your discovery and no one else? Do you hold the patent?
2. Are you claiming to have built the ultimate PINO while deriding others for building bikes capable of the same with IC engines?
3. Do you realize that this is a public forum and all material submitted is public domain?

 

[safe]

It's a long thread...

At the beginning I was just excited about being able to rewind for different behaviors.

58 mph on flat land? To my knowledge it was 52 mph and that was with a little bit of a tailwind. Without the wind it was 50 mph and the motor was producing about 2 hp before falling apart because of the cheap glue the Unite motors are made of. Downhill was 58 mph and eventually 60 mph on that motor. I've switched to these really nice Currie motors on ebay that are made with Neodymium magnets and are not for sale to the public. I'm using the old bike which I limited to 1000 watts of power to test principles for EBRR. So it's a legitimate testbed.

I have two bikes, one is a moped (Triple Rewind) and the other is an ebike. This thread is about the old bike that is legal here in Missouri as a moped.

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

This morning I managed to run out of battery power after about 8 miles and had to push the bike home on the last hill. I didn't top off the charge before I started, so I think I was down a little. All ended well though because the voltages ended well balanced.

 

[spad4me]

The tube type battery holder is an RC component.
That did not work because of vibration.
The company went belly up years ago.

Rewinding a motor is easy see rc groups find any post by Ron Summerin.