motor cuts off

[jeffaker]

 
I'm having the same motor cut off problem.  I set mine up with 3 x 12v 18ah PowerSonic PS12180 NB.  I would think that these 18ah bricks would have enough guts not to sag.  It drops out when it's under the highest load (going up hill), which is where I want the power.

Since these are brand spanking new batteries, will these SLA's improve in stability after a few cycles or will it get worse? 

I bought the PowerSonics after some testing with some free 7ah batteries gave me cut offs.  I got 36 12v7ah batteries FREE from a 1 year old UPS system that was never installed.  I thought they were toast from sitting for a year....I'm wondering now if the cut offs that I experienced the the 7ah's was the voltage dip.

Anyone know if I can hook up some of these 7ah's in paralell with my 18ah's?  Can I mix'n match different ah's?  I'd have to think adding more cells would help smooth that out.  I don't care much about the weight, as long as the motor's doing the work.

I had quite a challenge setting my bike up.  I was attempting to put it on a fully suspended mountain bike, but he suspension components proved to be too much of a barrier... too few mount points, forks too narrow.

I ended up with a new Walmart special 1 speed cruser for $95.00, aluminum frame with steel forks.  I've got my batteries mounted in 3 .50 cal ammo cans, which I plan to paint if I'm feel happy with the set up.  I think the olive drab looks a little ghetto.

Regards,
Jeff

[OneEye]

You can parallel different capacity SLA's.  This shouldn't cause problems.  How many amps are you pulling when your 18Ah's sag down to cutoff?

[jeffaker]

No clue on what I'm pulling.  I'm going to pick up a multi meter.  Probably a good investment at this point.

I'll post my draws in a day or two when I hook a meter up to it.

I had another idea on this today.  Using a momentary a/b switch on the left handle bar, to switch between my primary 18ah's and my 7.5ah's.  Then I could just pop over to the 2nd pack for a minute or so while the 18's bounce back.

[OneEye]

I think it would be more effective to parallel the batteries if you are going to carry them all anyway.  The packs bounce back almost instantaneously when the current draw is reduced.  The voltage sag phenomenon is an effective internal resistance.  You need to take a theoretically perfect battery and place a resistor in line with it to model its real-world performance.  As the current increases the voltage lost to the equivalent internal resistance increases.  As the current is reduced, the voltage comes back.  This is why you will often hear of an intermittent cutout or stuttering, as the controller hits the LVC it stops drawing amps, the voltage springs back and the LVC signal is no longer asserted, so the controller pulls the amps again.  A 'smarter' controller might reduce its amp limit each time it hits the LVC point in order to back off from this point, but I can see that causing as much trouble as it solves.

I'm a bit surprised you are hitting cutoff with an 18Ah setup.  It seems like other users have had success with as few as 12Ah of SLA.  MyElectricBike was running an 18Ah setup and the only trouble he had with an LVC was when one of the batteries he bought was a dud.  Once he figured out which one, he was able to replace it under warranty.  You might want to check the batteries individually using a 120W lightbulb as the load.  See if one of them has a more pronounced voltage sag at 8A than the other two.

[jeffaker]

I'm a little surprised that I'm getting with such a beefy pack too. 

I think you are onto something about having a bad battery.

I'm in the printing by trade so I'm not much of an EE, so I am learning a lot.

I think you are onto something about having a bad battery.

I got a cheapo multimeter today, but I think I need to take it back because it's max amps is like 250ma, I think I'd be pulling more than that.  I think on that can read 10amps should be in order.

So to be sure I understand the light bulb test.  I'd hook up the bulb to the battery, then hook my probes up on either side of the bulb?  or would I still measure the amps in the battery by measuring between the terminals under load?  Sorry, if that's a dumb question.  I'd then identify the dud by the battery which yields the lowest amp reading.

thanks for the help.
Jeff

[philf]

Therein lies the rub...

These puppies suck WAY more than 10A at peak - but worse...  They take a huge gulp (spike) of power when you first switch 'em on.  The idling controller (no throttle) only sucks about 20 or 30 milliamperes.  When you first switch on the power (or just connect the thing to the battery), there is a gulp that causes a great "ZOT" when the contact is closed.  This is directly attributable to the (two) capacitors they have installed at the input to the regulator for the 5V logic.  There are two 63V, 470uF caps there, and they "drink" when you first apply power.  No caps, no "ZOT".  But... the way the thing is built, they gotta be there.

Anyway, an inexpensive multimeter isn't up to snuff for measuring the kind of current draw you're likely to experience with these controllers/motors.  I know I've made the mistake of trying to improperly bridge the connection between my meter and the controller during this initial startup (popping the fastblow fuses in my multimeter in the process) enough times to have permanently marked this a "bad idea".

The defacto standard for measuring voltage/current (and therefore wattage) seems to be the "Watts-Up" meter that's been made popular in the R/C world.  It's resolution at the point in the scale we're talking about with an e-Bike is something to be questioned - my personal take is that it's more accurate at lower voltages - but it's still a useful tool, nonetheless.

You still sound to have an unusual issue - hopefully, as a collective, we'll figure it out..

[OneEye]

For the sag test I am suggesting, you will be using Voltage, and not directly measuring the current. 

Start with all the batteries fresh off the charger.  It is best if you use a 12V charger to ensure they are all at the same initial state of charge.  Hook one battery to the light bulb and measure the voltage across the terminals while connected.  Compare that with the other two under the same test conditions.  If one of them is significantly lower than the others, you have found your culprit.  The idea is to see if one of them is significantly sagging under a reasonable load rather than attempting to duplicate exactly what is going on under your peak current draws.

If you need to know the current in a circuit, you can indirectly measure it using Ohm's law.  If you have a known resistance and a known voltage drop, the resistance is the voltage divided by the resistance.

You can eliminate the *ZAP* of the initial connection by hooking up first with a large resistor in between the controller and the battery.  This limits the current while the capacitors charge up.  Then you hook up without it so you aren't wasting power when riding.  Some of the folks on the Endless Sphere forums have a setup where they use a switch between a resistor circuit and one without the resistor for this precharging.  They switch one way when they are connecting or disconnecting the battery, the other way when they want to ride.

[jeffaker]

I put my batteries on my RediPulse Pro-10 charger for a little more than 48 hours.

The test I cobbled together isn't exactly how you described, but I think it was effective.  Rather than using a light bulb, put my daughter's powerwheels jeep up on blocks, and dropped each battery in there.

Battery#1 12.57v off the charger/ 12.35v under powerwheels load
Battery#2 12.50v off the charger/ 12.25v under powerwheels load
Battery#3 12.54v off the charger/ 12.30v under powerwheels load

My meter accuracy is +/-0.7% DC so these readings are at my ability to measure differences.  I'm not seeing what I would call a meaningful difference there.

[OneEye]

Hmm.  Nothing significant noted there.

So we are left with a few possibilities
  1.  The no-load amp draw from a power-wheels motor is extremely small, so the differences are below measurement tolerances
  2.  The controller is cutting off at the wrong LVC
  3.  *something* else is going on.

I would look at #2 first.  Get a voltage measurement under high acceleration and see if it is happening at the right voltage cutoff or not.  Now that a multimeter is involved you can troubleshoot better.

[ahend]

You may want to revisit all your power connections. Make sure they are not loose or any bad crimps and or solder joints. Bad contacts will resist current flow, heat up and cause a voltage drop across it. After a ride with it cutting in and out, feel all your power connections and see if any are heating up. Perhaps there are poor power connections inside the controller, this has been documented to be quite possible.

Don’t give up on the batteries yet. See if you can assess a more taxing load test. If you can rule out the batteries as culprits, it may be time to roll up your sleeves and trace the power route.

[jeffaker]

I taped my multimeter to the frame of the bike and started testing each one under load.  But my testing was curtailed be getting called into work and then having to fly out of town for a couple days.

I'm sure my neighbors think I'm a total nut now, riding up and down the hill infront of the house.

I am very grateful for everyone's comments in helping me work through this.

I'll post the results of my testing as soon as I get some time back on the bike in a day or two. 

When I do my testing, I'm considering doing some testing supplementing my 18ah's with the 7.5 ah's I have.  Currently my array is straight forward, three batteries in series.  I'm a little confued as to how I should wire up something that is paralell and in series with six batteries.

Can anyone point me to a diagram on the web?  Is it just as simple as keeping my current wiring intact, and just connecting an additonal battery to each of my "main' three + to +, - to -?

Thanks!

[OneEye]

Is it just as simple as keeping my current wiring intact, and just connecting an additonal battery to each of my "main' three + to +, - to -?

Bingo.

You might want to start out with a measurement across the entire series of batteries under load when the stuttering happens.  That'll tell you what voltage is triggering the cutoff.

[jeffaker]

OK, so riding around with my multimeter taped to the frame cross bar, as far as I could tell, the voltage seemed to drop uniformly on all batteries.  With the controller kicking out when the volts dropped down to a bit more than ten.

I have a big pile of 7.5ah's, that show 13+ volts coming off the charger.  I took 1 7.5ah, and put it in series with each of my 18ah's.   I then took her for a spin.

It was friggin' awesome.  I rode it hard for about 30 minutes in the neighborhood, flat out, up hills.  I had no cut offs at all, even going up a hill steep enough where it would stall out on speed (twice up it).

I now need to make up a new wiring harness that is a little more permanent that what I cobbled together this morning.   I'll need to beef up the kickstand, it's a little tippy with all those batteries on there.  And I think I'm going to have to address braking.  I opted not to do regen braking, it seemed like more trouble than it's worth.  However, my $85 Walmart Cruiser single speed just has a coaster break.  I can really feel the mass on that little break.  I gotta put something up front.

OneEye, philf, ahend, Thanks for all your help!  I'm off an cruising.

Jeff Aker
Maple Grove, MN

[Dave]

OK, so riding around with my multimeter taped to the frame cross bar, as far as I could tell, the voltage seemed to drop uniformly on all batteries.  With the controller kicking out when the volts dropped down to a bit more than ten.

I have a big pile of 7.5ah's, that show 13+ volts coming off the charger.  I took 1 7.5ah, and put it in series with each of my 18ah's.   I then took her for a spin.

It was friggin' awesome.  I rode it hard for about 30 minutes in the neighborhood, flat out, up hills.  I had no cut offs at all, even going up a hill steep enough where it would stall out on speed (twice up it).

I now need to make up a new wiring harness that is a little more permanent that what I cobbled together this morning.   I'll need to beef up the kickstand, it's a little tippy with all those batteries on there.  And I think I'm going to have to address braking.  I opted not to do regen braking, it seemed like more trouble than it's worth.  However, my $85 Walmart Cruiser single speed just has a coaster break.  I can really feel the mass on that little break.  I gotta put something up front.

Congrats on getting things running. I just now read through this thread, and it sounded a lot like the issue I had with some weak wiring... bad soldered connection and loose spade connector. I'm still running a set of 7aH batteries that will get me to work with a little to spare, so big hills still require a bit of effort on my part.

Any photos of your rig? I for one would love to see how you hang that much battery on the bike.

You mentioned you work for a printer... Same here. I work for a label printing company in SW South Dakota and make it to the Twin Cities on occasion to visit vendors & whatnot. Good to know there's another geeky e-bike riding printer-type in the area (kind of in the area anyway!)