Safe's Electric Bike Project #001

[dirty_d]

Uhmm....got car headlights?

but is that as fun as sticking coils of bare wire in water and running high power through it?

 

[safe]

Uhmm....got car headlights?

Hey that's an idea... use jumper cables to the car and then switch from high beams to low beams.

 

[safe]

but is that as fun as sticking coils of bare wire in water and running high power through it?

Maybe you could test this first? The developer of the theory should be the guinea pig... :wink:

I sleep on all of this and see if I can think of an easy way by tomorrow.

 

[fechter]

I have a cheap, but rather large 12vdc to 120vac inverter. I use this for putting a load on batteries for testing. I plug in as many 120v light bulbs as I need into the output. It also has a LVC so I can't overdischarge the batteries too much.

 

[safe]

Just Not Sure...

The morning ride was underpowered, but not that bad and now the afternoon ride was virtually back to normal. The power seems to fade in and out and I'm having a hard time believing that it's strictly temperature related. A battery test will make it easier to find the real problem. (by process of elimination)

I'm just not sure that something else isn't going on other than the battery. On one ride it will work fine, then on the next it seems gutless... hmmmm... wiring? controller?

 

[fechter]

It's hard to tell without making some kind of measurement.
Stick a voltmeter across the battery input to the controller and see what happens to the voltage when you put a load on the motor. You want to isolate whether the problem is with the batteries/wiring or inside the controller.

 

[safe]

I think I'll be needing some good test equipment in the future anyway as I will be trying to figure out the MCL circuit.

So I've got about 7 miles to go before 2,500 miles on the bike and then I'm officially "there" and "done".

After that I can start to fiddle around with it... :wink:

 

[fechter]

You can get a voltmeter for something like $15.

 

[safe]

Water Over The Dam

My intuitive feel of what is going on (lacking the load tester) is that my batteries are pretty large compared to the current that is being drawn. However, my batteries are also getting old. So what's happening is that like a large reservoir and a dam if the water level is high enough there is essentially unlimited current available to draw upon and most of the time you will leave the spill over mostly closed because you don't need the water. However, once the water level drops below the dams top then your large capacity underneath the dams top is inaccessible. This behavior seems to be typical of older batteries.

Inside my garage it was roughly 78 degrees this morning, but outside it was about 60 degrees, so as I rode the temperature of the batteries would likely have gone down a few degrees. Yesterday my morning ride was like this one and not lacking in capacity, but the peak current was low. Yesterday afternoon it got to about 86 degrees in the garage and on the third ride of the day the power was very good. (pretty much full power)

The "bottom line" is that it does seem that my batteries have a "threshold temperature" where as long as the temperature is above this level (it appears to be about 85 degrees) the bike can utilize it's full controller determined current limit. Below 85 degrees and bike cannot get the current flow it needs and the powerband starts to feel more like a 30 amp controller rather than the 40 amp one that it is.

One problem with this is that any attempts to implement an MCL controller on this battery that uses a higher current limit will fail because the battery will artifically cut the current limit based on it's inability to deliver on the demand.

Now that the temperature is going down the batteries will slip into a sort of "suspended animation" and not age very much until it warms up next year. So by next year in the middle of summer when it's 90 degrees all the time the bike will be fun to ride again.

The only other option would be to figure out some way to prop up the batteries with some sort of extra battery. If I added some extra voltage in the form of an extra battery then that will move all of my shifting points around, but it might deliver more power at a lower current than I'm doing now (at a higher rpm of course). Next summer I would get a double whammy of extra current (because of temperature) and the extra voltage.

At this point I could reduce the motor size and end up with longer range. It might be a good time to try going down to another 750 watt motor since it draws less peak current.

I passed 2,500 miles this morning...

 

[xyster]

Safe, if I was intent like you on squeezing every last ounce of life out of my SLA, I'd parallel a single string of emoli's or A123's to handle the extra current (probably A123's would be best for a 1p string). Then over time as I could buy more, I'd complete the lithium switch-over. You probably wouldn't even need a separate charger -- just let the parallel boost-pack voltage float with the main SLA voltage.

 

[safe]

Another Broken Electrical Link

Seems that after 2000 miles with the new harness the bouncing has broken more than one electrical connection. I was in the process of fiddling around with some 14.4 volt drill battery packs thinking if I added them they could give me the extra boost I needed and while holding a segment of the electrical harness I noticed that some black electrical tape had melted. "Ah ha" I exclaimed... well not really, but it sounds more dramatic... and so I dug down through the tape and found that the solder had become disconnected from the Deans connector.

My soldering gun is pretty weak and I've had troubles getting enough heat from it to do the welds. What I really need to do is buy a soldering gun with twice the power so that I can be sure to get better connections. (that don't break free after time)

The Result

Yeah, you guessed it... "my mojo is back!" and now where I was only able to pull 30 mph on a straight away I'm back to my usual 38, 39, 40 mph.

I do think that the temperature is a factor and in the hottest summer months you do seem to get that little "extra" of about 10% like the charts predict and when you get down to below 70 degrees you lose about 10% and more as the temperature drops.

My first suspicion was that it was either the controller or the wiring harness because it came on a few days before the temperature really dropped.

So the batteries are not that old yet... they still have some life, so I could go on and test the MCL and see if that works, but I'll have to get a load tester to make sure that the batteries can deliver when asked to deliver. (the battery aging issue is now in focus, so if nothing else I'm prepared for the day when it really does arrive)

At this point the health of the batteries seems okay.

Onward to 5,000 miles!!!

 

[CGameProgrammer]

You're never supposed to use "pure solder" joints if it can be avoided; you should always ensure a strong mechanical connection independent of the solder. That's hard to do with Deans connectors which is one reason not to use them. Andersons can be crimped on wire easily, and the large size (50+ amp) holds together very strongly so they won't decouple either.

 

[safe]

You're never supposed to use "pure solder" joints if it can be avoided; you should always ensure a strong mechanical connection independent of the solder. That's hard to do with Deans connectors which is one reason not to use them. Andersons can be crimped on wire easily, and the large size (50+ amp) holds together very strongly so they won't decouple either.

On the battery side of the connections I use a crimped-to-circular connection that bolts to the battery. Those are as solid as a rock. The Dean Connectors would be much better if they had something other than the flat plate to solder to, but I like the fact that they don't wear out easily. My understanding (I haven't used the Andersons) is that they don't have the same wear charactoristics and also lack the positive/negative configuration that prevents getting things connected wrong. That was one of my main worries, that I'd accidentally connect the wires wrong and create a short. Anderson Powerpoles are single connectors and not double.

Or am I wrong on that?

After 2,000 miles of hard riding to have 2 out of 29 soldered connections fail isn't too bad of a record. Chances are that these were the "weakest links" and that things should be more solid from now on... either that or more connections are soon to let go from fatigue...

 

[xyster]

My understanding (I haven't used the Andersons) is that they don't have the same wear charactoristics and also lack the positive/negative configuration that prevents getting things connected wrong. That was one of my main worries, that I'd accidentally connect the wires wrong and create a short. Anderson Powerpoles are single connectors and not double.

Or am I wrong on that?

You're wrong on that.
http://www.flyrc.com/articles/using_powerpole_1.shtml
"Powerpoles have curved contact pins that snap into place and hold the connectors together. When two pairs are snapped together, it takes approximately three pounds of tension to pull them apart. Anderson rates these at up to 10,000 no-load connect/disconnect cycles!"

 

[safe]

Powerpoles have curved contact pins that snap into place and hold the connectors together. When two pairs are snapped together, it takes approximately three pounds of tension to pull them apart.

So what does three pounds of tension feel like?

Does that mean you have to apply a lot of pressure to get them to come apart or that they come apart pretty easily? (wait a second... don't they mean the ends snap together and not the "units" of positive and negative? This still isn't clear to me yet. What is snapping with what?)

Also, it does look like there's a color coding scheme, so you could safeguard yourself from error with color coodination.

My Deans Connectors cost about $3 each side to the best of my recollection... so $6 per positive and negative wire pair. Compare the price... what would that translate to in Andersons?

Remember... $6 per pair of positive and negative wires... (looks like 4 pieces equivalent)

Okay this might solve the "join together" question:

 

[dirty_d]

are you sure you're soldering this stuff together right? ive never had any solder connections break, the wire or connector always breaks first.

 

[xyster]

Powerpoles have curved contact pins that snap into place and hold the connectors together. When two pairs are snapped together, it takes approximately three pounds of tension to pull them apart.

So what does three pounds of tension feel like?

Like lifting 3lbs off the ground by tugging on an attached string.

Does that mean you have to apply a lot of pressure to get them to come apart or that they come apart pretty easily? (wait a second... don't they mean the ends snap together and not the "units" of positive and negative? This still isn't clear to me yet. What is snapping with what?)

They snap on; they snap off. Single pairs snap apart easily; triple pairs snap apart with a bit of wiggling; sextuplet pairs don't snap apart without pliers.

Also, it does look like there's a color coding scheme, so you could safeguard yourself from error with color coodination.

Sheez man, look at the picture again. Besides the color coding, individual connectors snap side-by-side together in either of two directions such that it is not possible to connect pairs incorrectly, regardless of the color scheme.

My Deans Connectors cost about $3 each side to the best of my recollection... so $6 per positive and negative wire pair. Compare the price... what would that translate to in Andersons?

Remember... $6 per pair of positive and negative wires... (looks like 4 pieces equivalent)

Best prices I found for complete sets:
http://www.powerwerx.com/category.asp?CtgID=3578
and where I bought mine.

 

[safe]

Best prices I found for complete sets:
http://www.powerwerx.com/category.asp?CtgID=3578
and where I bought mine.

"Yellow/Black Powerpole Sets with 45 amp contacts (Sets: 25, Price per Set: $1.20)"

Yikes! That's fantastic!

I think you've changed my buying plans for my next project!

All the time I had assumed that the Andersons were really expensive... as I recall when I first looked them up they were about $10 per set, but obviously I just had not found the right links.

$1.20 per set is amazing...

(this would translate to $2.40 per positive and negative four piece set)

 

[safe]

are you sure you're soldering this stuff together right?

I have an inadequate soldering gun. (one of those little pen style ones that produce very little heat) What I need to do is buy a "real" full sized and strength soldering gun. Most of the connections were soldered very well, but every once and a while the solder could melt on and into the wire, but I wasn't able to really "cook" the joint. It's that last high threshhold "cooking" that makes a secure joint.

But switching to Anderson Powerpoles actually looks to be a cheaper solution and they've always had a strong reputation in the electric vehicle world. (so I don't think I can really go wrong either way)

It's also nice to have experience with as many different technologies as possible...

 

[CGameProgrammer]

Powerpoles come in both single-pole and double-pole variants. Even with the single-poles, you can hook an infinite number side-by-side, and since they don't connect upside-down, that would ensure you don't misconnect them.

The only disadvantage I can think of is that the smaller ones decouple pretty easily, and so they could loosen in a high-vibration environment. But tape fixes that, and the larger ones (50+ amp) hold together very strongly and won't come apart accidentally. Oh, and I think the resistance is a bit higher than with Deans.

Regarding price, some places charge much for them -- a local hobby shop sells them for $3 per set. But yeah, Powerwerx is $1 (more for the larger versions).

Buy a few and fool around with them.

 

[safe]

The only disadvantage I can think of is that the smaller ones decouple pretty easily, and so they could loosen in a high-vibration environment.

Other than my bad soldering job the Deans Connectors have been really great connectors.

But it will be fun to try out some Anderson Connectors next time...

 

[dirty_d]

you're using some sort of paste or gel flux other than just rosin cored solder right? as long as the metal is fairly clean and you have a good amount of flux on it you shouldn't have to heat it up that much just enough so it melts and the solder should flow everywhere and into the cracks and look nice.

 

[safe]

...you shouldn't have to heat it up that much just enough so it melts and the solder should flow everywhere and into the cracks and look nice.

The solder has some flux built into it, so it's not the solder. A very small soldering gun that doesn't have much power tends to dissipate what little it has into the wire when you apply the gun to the wire. You're supposed to heat up the wire enough so that the solder "seeps" inside it. If you just melt the solder on top of the wire it doesn't always penetrate deeply enough. Trust me, I know a good weld and a good soldering job when I see one and my soldering with this gun is often poor. I could do much better. The soldering gun also seems to sometimes get hot and sometimes not, like it's peak temperature is not very constant. (so sometimes it works fine and others are a struggle)

 

[safe]

Mojo Confirmed!

Given the fact that I noticed and fixed the broken electrical connection in the afternoon yesterday there was still the chance that on a cold morning the batteries would not deliver as promised. So I got up early and rode at sunrise with an outside temperature of 60 degrees. The bike worked as it is supposed to work and had power where it was supposed to have it. Due to the colder temperatures the peak was slightly off of the summer peak, but in the 10% range as it's supposed to be and not the extreme 30% drop I was having. (plus the electrical problem caused a altered powerband shape with a flattened power curve... not what you would expect of a linear drop in battery voltage strictly due to temperature)

So I can say with reasonable certainty that the batteries are fine... (whew!)

The lessons learned were:

1. If the powerband loses it's peak current carrying capacity you need to check for things like wiring or a bad controller before you assume the batteries are worn out. (or test the batteries with a load tester of some sort)

2. Batteries tend to lose their voltage peak as they get colder, but unless they are already worn out they don't seem to flatten the powerband out that much. In other words, as the batteries cool the powerband should come down in a proportional manner... if the powerband shape alters (making you lose a disproportionate amount of middle rpm peak power) then you need to suspect other things.

 

[dirty_d]

thats what i said, you need more than just the flux that is in the core of the solder, thats only good for parts that are already perfectly clean or tinned, for soldering large wire to some type of connector your going to need to put flux on it first.