Help with 'juddering' cyclone

[El Duderino]

Hello everyone!

I've finally found the time to start doing some long range testing on my 1200w cyclone with A123 battery and I'm getting a weird effect. Any chance that those of you with experience could help diagnose the problem?

Hot off the charger at around 53v, bike is nippy in any gear. Pulls well and goes as fast as I'd expect (tops out at 33). I'm
Running the bike for about 30 mins, covering about 6-7 miles. This is a nice loop and is about the right amount of time for me to take without being missed at work...

The next time i take the bike out (without a top up), i get what can only be described as a judder under acceleration. If this was a car, it would be a slipping clutch. The bike will still go to close to full speed, but i have to 'baby' it.

The bike is always cold on second outing.

I cant believe its a lack of juice. 20ah cells seriesed (is that a word) should provide a lot longer range.

Could it be controller or bms related? Throttle? Any ideas? Please?

Cheers for any help!

Chris

 

[darkone040]

is your bike old.. when i put the chain on the smallest gear the chain skips on it as the gear is a little worn out.. possibly..

 

[El Duderino]

Hi,

Thanks for the reply, but no, new chain, cassette etc. Probably got 50 miles on the bike. The judder is definitely coming from the motor.

Cheers,
Chris

 

[Gregory]

How many cells in series do you have to get 53v hot off the charger?

If 16s that is too low, and it may be the bms kicking in because the cells may not be balancing properly.

Also checking individual cell voltages at the end of riding would confirm this.

 

[El Duderino]

Hi Gregory,

Thanks for that. Ill investigate. I thought 53 would be about right for a 48v battery. Didnt think to check.

Cheers,
Chris

Edit: nope, 16 A123 cells in series looks pretty much bang on for 53v. They're 3.3v per cell.

 

[Gregory]

Not really, "hot off the charger" it should be around 58V or 3.6V per cell. That's where the balancing is done by the BMS. This 0.3V surface charge disappears quickly under load or may self discharge overnight.
So I guess you didn't actually mean hot off the charger. But that's just nitpicking.


I was thinking since you only got 7 miles from a 20Ah battery, it seems the battery or BMS are a likely culprit. One of the cells may be low and tripping the BMS LVC function prematurely. Hence the need to check individual cell voltages.
And the other common fault is loose connections/backed out plugs.

 

[dogman dan]

As always, there is the possiblilty that a hall sensor wire came loose, or something like that. Or is a cyclone a brushed motor?

Phase wires that are partially cut can also cause a juddering under load. So a bad plug on the phases could also be the problem.

But if you only charged to 3.3v, there is a definite possibility that your bike was only half charged. Yes indeed, 3.3v per cell will be the voltage you will see on a large portion of your ride. Question is, did your charger stop at the full end of where the battery voltage reads 3.3, or at the half full part where the battery reads 3.3v.

You gotta charge to at least 3.5v - 3.65v to be sure the cells are truly topped up.

Odd though, that it would make a juddering effect. Perhaps the lvc flashes on and off on the controler? So as soon as the load is off, or lowered as you reach speed, it stops poppling. If the problem goes away after you charge, then the battery is likely to be the problem.

 

[wineboyrider]

Proper gearing and chainline maintenance are a must for cyclone setups. Sounds too me like you have a huge load from the improper gearing setup and that is what trips something.
I used too have a cyclone setup and abandoned it due to maintenance issues that I no longer have to maintain. A good setup when working properly though.

 

[dumbass]

Many years ago when lifepo4 cells were just starting to be used Cyclone USA did a test using a 500w motor and a lifepo4 24v pack. The rider eventually started getting a jitter feel from the motor expecially under higher loads. What they eventually relized was the pack was low and the controller was turning off/on very rapidly in sink with the voltage loss and gain from the controller cutting the power. As the voltage went down the controller cut off. But the as soon as the voltage came up the controller powered the motor again. It was actually doing it so rapidly it caused the motor to jitter. This report may still be on Cyclone USA's web site. They have an area of motor tests.

Here's the link to Cyclone USA's test I am talking about. Remember this motor (500w) has abuilt in controller.

http://www.cyclone-usa.com/500wLi-Ion26.php

If this is the problem I agree with others you need to check each individual cell before, during and after a ride to fine the runt cell or cells.

Bob

 

[d8veh]

I'm with Dogman. Juddering on brushless motors is often a bad phase connection. It feels like a chain jumping. I've had it a couple of times on my hub-motors.

 

[El Duderino]

Thanks for all the replies fellas. I'm sure it was only reading 53v right off the charger, but that was read with a cycle analyst. I'll check with a decent multimeter on monday.

I'm well aware of the chain line issues... ask me how :lol:

But this is definitely at the motor, you dont feel it so much as hear it. So looks like I've got some digging to do.

My first thought was that the controller was somehow losing the timing as revs increased, but the fact that it only happens after the bike has had some use threw me. Thinking it through, the bad cell combined with controller or bms cutoff sounds more likely because i can turn higher revs with no judder on a fresh charge.

Thing is. The controller is designed for 48v no? So wouldnt even 50v still be ok as far as it was concerned? Doesnt this point to a duff bms or cell?

Cheers,
Chris

 

[dumbass]

Most or all 48v controllers should easily handle well over 50v. 16 Lifepo4 cells will settle out at around 53v. I'm still betting it's a LV problem and ether the BMS or controller is rapidly cutting in/out. Can you install 2 CellLog8s on the pack to monitor the individual cells?

Bob

 

[d8veh]

The controller is designed for 48v no? So wouldnt even 50v still be ok as far as it was concerned? Doesnt this point to a duff bms or cell?

No
Did you check you phase connectors? - especially if you have those cheap bullet/barrel oness. Pull them apart, squeeze the barrel so that it grips tighter, and re-connect.

 

[El Duderino]

Hi d8veh,

Not yet, bike is at work. That'll be the first thing i check on Monday though. Not using bullets. Have andersons, and they're safety wired to avoid pulling apart.

Cheers,
Chris