Safe's Electric Bike Project #001

[safe]

If you think this is frustration now, wait untill you have a battery pack with no solid connections between the cells.

Well, they would be pressed together tightly with a spring. The problem I just had with the wiring was with a rigid soldered connection, so my "real world" experience tends to favor a more elastic connection over a rigid one. Rigid things tend to break or fatigue... flexible things handle vibrations better..

you don't consider for a moment that your SLA's may be EOL?

The voltage is a little lower. At the beginning of the summer the starting voltage held at about 13.1 volts and it looks like I'm down to about 13.0 volts now. The year before it was 13.2 volts. I figure I've got another year and then it's probably time to spend another $140 for another round. Since I've already gone 2,500 miles for $140 it's still a good deal. For a car that would cost:

2,500 miles / 30 mpg = 83 gallons * $3 = $250

...any time you beat gas prices you are doing okay. :wink:

 

[xyster]

If you think this is frustration now, wait untill you have a battery pack with no solid connections between the cells.

Well, they would be pressed together tightly with a spring. The problem I just had with the wiring was with a rigid soldered connection, so my "real world" experience tends to favor a more elastic connection over a rigid one. Rigid things tend to break or fatigue... flexible things handle vibrations better..

Could be. My 300-cell pack hasn't had any apparent connection problems after I rebuilt it by crimping wires across the flexible tabs, soldering, and securing with elastic electrical tape drawn tight. My previous pack build suffered a couple failed connections where I had soldered the tabs together across their tops, into a rigid array.

Current, flexible method:

 

[safe]

I rebuilt it by crimping wires across the flexible tabs, soldering, and securing with elastic electrical tape drawn tight. My previous pack build suffered a couple failed connections where I had soldered the tabs together across their tops, into a rigid array.

To my way of thinking your new technique is superior to a rigid connection.

The "real world" seems to support your new design so far... (vibration happens)

 

[safe]

Increased Range, Lowered Heat... but less Power and Speed

Something is not right on my #001 bike. Even though there's a 30 degree drop in air temperature my bikes range has increased and my motor is not heating as much, but now I have less power and less top end speed. (which for a bike with gears means I can't achieve the gearing because there isn't enough power to pull it) On a downhill the motor still revs to the same rpm, so the voltage is there, but the current isn't. Either my wiring is going bad or my controller is dropping the current limit. I can still get around, but it feels like a different bike. (like it has a 30 amp current limit rather than a 40 amp limit)

I've got less than 50 miles until 2,500 total miles after that the I might as well do the MCL thing. Since it's likely I'll be tearing the controller apart and rebuilding it I might as well incorporate MCL at the same time. No sense in buying a third controller. These controllers seem to have a limited lifespan given the components that they are built with.

The battery is still testing fine and the voltage is still starting at 13.0 volts and ending at about 12.2 volts after 10-12 miles. That's as good as it's ever been +/- 10%. (normal)

 

[xyster]

Increased Range, Lowered Heat... but less Power and Speed

The battery is still testing fine and the voltage is still starting at 13.0 volts and ending at about 12.2 volts after 10-12 miles. That's as good as it's ever been +/- 10%. (normal)

Have you tested the pack voltage under load -- like while you are riding? I think it much more likely that, after 2500 miles, the SLAs are circling the drain, than your controller somehow gave itself a new, lower current limit.

 

[safe]

Have you tested the pack voltage under load -- like while you are riding? I think it much more likely that, after 2500 miles, the SLAs are circling the drain, than your controller somehow gave itself a new, lower current limit.

I really need to get an ammeter don't I?

Yes, that would be the best thing to do, find out what's really going on by having the ability to accurately measure it. I can measure the battery voltage through the charger and I suspect that's pretty good because I think my charger is pretty well thought out. Strictly going by "feel" it "feels" a lot like when I was having wiring troubles or a controller that failed. It happened SUDDENLY after the end of one ride so, that tends to make me suspect it wasn't a gradual battery decline thing.

It began before the temperature drop.

...so it's not likely to be the batteries fault. The battery is still showing very good performance as shown by the starting and stopping voltage and the range is actually going up, not down.

Since SLA's are supposed to be able to have a peak draw of about 5C then my 38 Ah battery SHOULD be able to draw a peak of 190 amps. 40 amps is only about 1C. I would be very surprised of the battery is unable to draw the small current that I need. The size differential is just too huge to suspect the battery limiting the current.

 

[xyster]

The battery is still showing very good performance as shown by the starting and stopping voltage and the range is actually going up, not down.
[/b][/i][/color]

I don't know about SLA, but I do know the first sign of age-related deterioration of lithium batteries is increased impedance as evidenced by decreased voltage under load, not decreased capacity -- which isn't measurable until later and is easily masked by differences in ambient temperature.

 

[safe]

I don't know about SLA, but I do know the first sign of age-related deterioration of lithium batteries is increased impedance as evidenced by decreased voltage under load, not decreased capacity -- which isn't measurable until later and is easily masked by differences in ambient temperature.

Yeah, I thought of that... but these guys suggest that the base level of resistance is low for SLA. The battery would have to be really dead, which I know it's not because it still holds the charge well.

From "Battery University"

"The internal resistance of Lead-acid batteries is very low. The battery responds well to short current bursts but has difficulty providing a high, sustained load. Over time, the internal resistance increases through sulfation and grid corrosion."

The easy way is to spend $40 and get a replacement controller and simply try it. The last time (last year at the end of the summer) I burned out a controller and when I replaced it it was like a whole new bike. The symptoms are identical... the range is good, but the peak power is bad.

It's the perfect time to redesign the controller for MCL... (what the heck I've got all winter to figure it out :wink: )

But the ammeter addition sounds like a good idea. If there really is a battery current flow problem then the ammeter would make that very clear. Plus, I really want to have an adjustable "Boost Control" on my MCL since it would be fun to on occasion crank the amps way up.

Ideally I'd like a 100 amp 36 volt controller and then trim it way down to 40 - 50 amps for general riding with occasional tests up to 60 amps. (heat is going to become too much of a problem above about 50 amps) If the AllTrax controller measures the motor side current I might just buy one of them.

Could it be possible that a battery that could once pull 190 amps can no longer sustain 40 amps?

And still show little sign of aging as far as voltage holding capability?

...I just don't think so. Not yet anyway.

 

[safe]

Here's a possible scenario...

The battery gains "liquidity" during the hot summer months and the current carrying capacity during that time is maintained at say 60 - 70 amps. During those summer months the hot weather and 2000 miles of riding deteriorate the battery. As the season draws to a close the amp limit slowly drops to 50 amps. Then the temperature drops and the "liquidity" that the heat provides goes away and now the amp limit drops below 40 amps. For the first time I begin to notice the deterioration. Up to this point there was always "excess" current to draw from, but now (with the temperatures dropping into the high 50 degree mark) the current limit is below 40 amps.

So let's assume this is what happened (very possible) how could I PROVE this is the case and not some other thing like a controller that went bad or wiring that has come loose?

It would be a dumb thing to replace batteries if that's really not the problem.

Is there some kind of battery analyzer that could evaluate (accurately) what my current carrying capacity should be?

 

[xyster]

Up to this point there was always "excess" current to draw from, but now (with the temperatures dropping into the high 50 degree mark) the current limit is below 40 amps.

Arrow So let's assume this is what happened (very possible) how could I PROVE this is the case and not some other thing like a controller that went bad or wiring that has come loose?

Hook up an ammeter and a voltmeter and watch by how much the voltage sags as you ride. An ammeter is very useful for controlling the current via throttle modulation. The practice of setting my own current limit to 5 amps, 10 amps, whatever amps, has become second nature as I've ridden my bike with an ammeter for almost a year. My hand modulates the throttle for a particular limit without my brain consciously thinking about it, and with my eyes only occasionally peeking at the ammeter. This is something you might consider before building your own fancy automatic current limiter.

 

[Lowell]

Safe, you need a CycleAnalyst

 

[xyster]

Safe, you need a CycleAnalyst

He's too cheap for that (and so am I for the time being).

 

[safe]

Hook up an ammeter and a voltmeter and watch by how much the voltage sags as you ride.

How's that going to make me know whether the current is being limited by the "system" (wires, controller, motor, throttle) verses the battery?

All that can do is tell me that it's not working to it's full potential, it's not telling me why.

Are any of these battery meters that claim to show the health of the battery any good? I'm not going to want to spend more than 25% of the price of new batteries ($150 * 25% is about $40) or I would be wasting money.

Can you imagine that?

"I just bought a $250 product that accurately tells me that my $150 batteries are old." :lol:

Dumb de dumb dum...

 

[dirty_d]

hook the battery up to a load and measure the current, and the voltage at the terminals, if the voltage drops a lot then they suck and have high internal resistance. compare the results with what the voltage drop should be for a good battery if you have one or have someone else do the same thing.

Edit: if you want to find the internal resistance and compare it to the manufacturers standard to see how worn out it is then you should be able to take a load that is around 3 ohms or so. hook the battery up to it with an ammeter measure the current and measure the voltage, voltage / current will give you the exact load resistance. now measure the open circuit voltage, now (OCV / current) - load_resistance should give you the battery's internal resistance. right..... fechter.......

 

[xyster]

Hook up an ammeter and a voltmeter and watch by how much the voltage sags as you ride.

How's that going to make me know whether the current is being limited by the "system" (wires, controller, motor, throttle) verses the battery?

All that can do is tell me that it's not working to it's full potential, it's not telling me why.

What dirty_d said. You can rule out a voltage-drop problem with the meters while riding, and then if there is a voltage-drop problem, rule out a battery-related problem with a separate load, like some lights. Then go from there.

Are any of these battery meters that claim to show the health of the battery any good? I'm not going to want to spend more than 25% of the price of new batteries ($150 * 25% is about $40) or I would be wasting money.

I just use two of these $12 analog panel meters with a $12 shunt (and you could make your own shunt if you wanted, and no doubt find even cheaper meters).
<a href="http://www.allelectronics.com/cgi-bin/category/385/Meters_(Panel).html"> all electronics meters</a>

 

[roguesoul]

Your battery can show a full charge and also have greatly diminished capacity.

http://www.windsun.com/Batteries/Battery_FAQ.htm

Why not remove your controller then connect motor and batteries directly.

love your designs.

 

[safe]

Your battery can show a full charge and also have greatly diminished capacity.

Yeah, I know that, but I was trying to discover the level of internal resistance that has been built up in the battery because of the aging process. I think the answer has been found using nothing more than an ammeter and a voltmeter. (which are nice to have anyway... things I'll be needing in the MCL debugging)

 

[safe]

The only trick is going to have to be getting good readings of BOTH current and voltage at the same time while you ride. I suppose if you stare at the gauges you could get a good approximation. It's not like getting a really reliable measurement, but "good enough" might be enough.

Either that or I could spend $10 (plus shipping charges) and get an accurate load value direct from the batteries themselves. The ammeter and voltage will give the resistance value and then you can estimate what the current "should" be, but the direct measurement of peak battery current reduces any guesswork.

With a 50 amp current tester it's sort of a "pass / fail" test... :wink:

 

[dirty_d]

well you can just take the readings with the batteries not even in teh bike, the load doesn't need to be the motor, just something that will draw a good amount of amps like 10 - 20 for a few seconds to take a good reading. you don't need to know the resistance of the load just an approximation you could make the load of some iron wire even, you find the resistance with the voltage and current through it.

 

[safe]

well you can just take the readings with the batteries not even in the bike, the load doesn't need to be the motor, just something that will draw a good amount of amps like 10 - 20 for a few seconds to take a good reading.

Like what?

How am I going to draw 36 volts at say 15 amps and 30 amps?

Lightbulbs?

I could test each battery one at a time... so 12 volts at a time...

12 volts * 40 amps = 480 watts

So a 250 watt bulb and a 500 watt bulb?

If the bulbs are of a known wattage then all I need to do is measure the voltage drop and I can calculate the current. Right? Or no?

 

[dirty_d]

i was thinking some thin iron wire wrapped around a rod, if it gets too hot which i think it will, submerge it in some water, that will keep the temperature pretty constant i think so the resistance should stay pretty constant. if you do it like i said instead of like in that picture you only need one load not two.

 

[dirty_d]

i think the resistance of the bulbs would be too high unless its actually a 12V light bulb. a 120V 100W bulb consumes 100W at only 120V, 100 / 120 = .83A, at 120V only .83 amps flows through the filliment, the resistance of it is 120 / .83 = 144.5 ohms, at 12V through the 100W bulb only .083A would flow, no good.

 

[safe]

i think the resistance of the bulbs would be too high unless its actually a 12V light bulb

They make DC bulbs don't they? Of known wattage?

 

[dirty_d]

yea they do but it would be easier and possibly free to just use wire.

 

[xyster]

i think the resistance of the bulbs would be too high unless its actually a 12V light bulb

They make DC bulbs don't they? Of known wattage?

Uhmm....got car headlights?