Cromotor build - lurching motor

[ChubbyQuark]

Hi everyone,
First post here but long time learner from the great knowledge available. I've searched other threads but can't find any symptons that match what is described below.

This build started about 6 years ago, with the following components:

Cromotor (I believe it's a v3)
Battery - Wallen power 72v 20AH, 60A continuous discharge current
Greyborg frame (concentric bottom bracket)
Twist throttle
Controller 01 (Currently on the bike. 18 Fet, I think it's an Infineon, see pic. below)
Controller 02 (I think it's a Lyen 18 Fet 4115, this has connectors for a cycle analyst and programming and 3 speed setting)

So the bike has been tested with both controllers, controller 01 gives plenty of torque, but hesitates slightly at low throttle(around a third)/RPM. Controller 02 spins the wheel very slowly but also pulses like controller 01. Controller 02 is programable but I'm not sure if the software and cable required for this are available anymore?

Strangely when the bike gets up to speed the pulsing seems to fade away, do you think this is because the controller needs programming to match the battery and motor?

I've checked the hall wires all read voltage of 0.4-4.9 when tested to ground when the wheel is spun backwards, so they seem good.

Any advice on what these controller are would be greatly appreciated, and if it's possible to program controller 02? Perhaps I'd be better off buying a more update controller to math the system?

Thanks in advance!

Controller 01





Controller 02

 

[neptronix]

Here's the lyen/em3ev/infineon clone control software.
It requires a USB programming cable ( driver for this included ), if you have one.

I looked up your battery and would guess it is inadequately rated for a motor this size like many cheap batteries.
The cell spec is not mentioned in any of the ads so we should assume the usual extreme over-rating of the battery.
This in itself could be the reason for the pulsing etc.

You will want to measure maximum voltage drop from 0% to 100% load to rule the above problem ( massive voltage drop ) out.

 

[ChubbyQuark]

Great, thanks for the software. Do you know if the USB cables are available anywhere?

Assuming I can program the controller, could it be set to match the output of the battery, I dont need/want the full power of the motor (this project was started when i was after a different build) so would be happy running it at a reduced power, as long as that won't cause damage?

 

[neptronix]

Great, thanks for the software. Do you know if the USB cables are available anywhere?

Assuming I can program the controller, could it be set to match the output of the battery, I dont need/want the full power of the motor (this project was started when i was after a different build) so would be happy running it at a reduced power, as long as that won't cause damage?

I would suggest asking in the for sale / wanted, they're rare

Yes you could set it to match the output of the battery. The main worry is overloading and therefore overheating the battery, shortening it's lifespan..

A multimeter, RC hobby watt meter, or cycle analyst can help you determine the max load amps and the max voltage dip underload.. i would get that info before purchasing anything else.

 

[ChubbyQuark]

OK, thanks. I have a multimeter already but will research the watt meter and various tests I need to do. I was going to ask about the cycle analyst as this also seems to be a good, albeit pricey, way to set power profiles.

I'll try and post a short video of the issue as the accomponying noise may also give some clue.

 

[ChubbyQuark]

Video clip attached. Just noticed the flat spot on the tyre from where it's been standing so long!

When the motor gets to half speed the noise decreases and the pull is more consistent.

Motor Noise

 

[neptronix]

OK, thanks. I have a multimeter already but will research the watt meter and various tests I need to do. I was going to ask about the cycle analyst as this also seems to be a good, albeit pricey, way to set power profiles.

I'll try and post a short video of the issue as the accomponying noise may also give some clue.

The CAV3 is a piggy back controller with a noticeable degree of latency in 'making a power profile work' by altering throttle signals etc;
For a high torque motor like yours i would suggest not using this as the primary way to limit power.

Motors usually get less noisy at cruise.
Pull more consistent in the mid range is telling, here's some possibilities:

1) battery sagging a lot less due to higher efficiency in that part of the motor's efficiency curve.
2) not enough phase amps to battery amp ratio ( want about a 2.5:1 for this motor ), hence diminished low end torque.
3) some combination of both

Measurements needed!

 

[ChubbyQuark]

OK, will do some tests tomorrow. Looking at some other threads do I need to check for voltage drop when throttling along with current draw?

Edit: looks like a watt meter, as you suggested, is the way to go for this.

 

[neptronix]

CA v2 is a great option because it just does measuring ( I have a fleet of them <3 )

When i was so broke i couldn't even afford a CA v2, i used one of these:
Turnigy 180A Watt Meter and Power Analyzer

Interestingly the price on those hasn't changed in 12 years!

 

[ChubbyQuark]

Hmm, just looking at those but they are only rated up to 60V. Looks like one with an external shunt or the cycle analyst V3 is the way to go?

 

[neptronix]

oops i didn't consider that sorry;

Looks like the CA v3 has a standalone model ( low current ) and a high current model. These can be used without the piggyback functionality and are a good substitute for the v2

 

[amberwolf]

Remember that the CA can be used without any limiting features...just don't hook the throttle up to it, send that directly to the controller, and it can't affect the operation.

If you *do* want the CA to do limiting, then it does have to get the throttle first, and then send it's own to the controller instead.


There is no motor programming in any of those "generic" controllers, only really in FOC controllers (which are generally complex to setup and get running, about as much so, or more, than the full CA setup is to tune and get tweaked just right, in many cases.)

Some of the programmable generic controllers (in which I include the Lyen and similar ones based on my experiences with some of them) have user-alterable current limits; in those cases you can lower it to below your pack's actual capability.

Another thing to verify is that the LVC of the controller is set correctly to match the pack, because if it's too high then everytime the pack sags in voltage it will shutodwn the controller.

If the current is too high on the controller for the pack, and it causes too much voltage sag, that's a common problem. That can cause a slow form of "stuttering" of the motor, where it runs, loses power, runs, loses power, etc, all while holding throttle steady with a steady load, as the votlage drops, rises, drops, rises...

If it's so high that it trips the BMS, then the whole system will lose power until the BMS is reset. That is usually a slower process than the controller resetting from LVC, and some BMS don't turn back on at all until they are disconnected from all loads or a reset button is pressed, or the power button on them is cycled, or a charger is connected (and plugged to the wall as well so the charger is actually running).

 

[ChubbyQuark]

So while trying to figure out the best way to get a watt meter on the bike I did a quick test:

Took the voltage across the ignition switch at rest which measures 80.9 (72V battery, freshly charged), pulsing the throttle (half throttle) with the wheel off the ground the voltage was dropping to 78.2V

 

[ChubbyQuark]

Great info. thanks amberwolf. The power doesn't completely go, more like a dead spot every revolution or two.

Would an updated controller be beneficial? Conscious the ones I'm using are quite old, I inherited them when I started the build a few years ago, I'm sure things have moved on quite a bit, the VOTOL controllers seem to get good reviews and have a programming interface.

I can't currently check any of the controller settings as the software only works on Windows 7 and I don't have the cable to connect to the PC.

Edit: I meant to add neither the battery or motor are warm after 5mins use, a mix of flat and 5% hills, but the controller is warm to the touch.

 

[amberwolf]

So while trying to figure out the best way to get a watt meter on the bike I did a quick test:

Took the voltage across the ignition switch at rest which measures 80.9 (72V battery, freshly charged), pulsing the throttle (half throttle) with the wheel off the ground the voltage was dropping to 78.2V

No load current ought to only be a couple of amps, maybe a few.

If you're getting 2.7v of sag at only that current, there is probably a huge drop in voltage at actual load.

So what is probably happening is your battery can't handle the load and drops in voltage so much that the system cuts out (controller LVC, if it's not actually powering off). Load gone, voltage rises, controller turns back on, restarts motor, load comes back, votlage drops....etc.

As the load decreases the voltage stops dropping quite so low, and the controller stops cutting out. Load can decrease with speed up to a point, where the air resistance begins forcing more power to be used to overcome it.

If the battery is as old as the frame, the cells may have simply aged so much they cant' do what they did when they were new.

It might not be a battery problem, but could be a connection problem--too high a resistance between battery and controller.

 

[ChubbyQuark]

Hmm, that battery is new, bought from Wallen power, specs below. I thought this would give enough power to the system but I was a bit too keen to get the system running and probably didn't do enough battery research.

Can the cycle analyst v3 be programmed to override the default controller settings? I'd quite like to get one anyway as they seem like a good investment for future builds.

If it stays dry today I'll take it on a longer test run,

Thanks again.

Fit for Motor Power: 72V 750W~3300W（Up to 3300W）
• Nominal Voltage: 72V Rated Capacity: 20Ah
• Combinations:20Series 4Paralles（20S 4P)
• Cells:Panasonic 3.7V 21700 cells
• Component: Li(NiCoMn)O2
• Watt Hours : 1440Wh
• Way of Charge : CC/CV(Constant Current；Constant Voltage)
• Standard Charge Current : 3A
• Maximum Charge Current : 5A
• Charge Cut-off Voltage : 84V

• Continuous Discharge Current (BMS) : 60A
• Peak Discharge Current : 180A
• Discharge Cut-off Voltage : 56V
• Cycle Life : ≥1000 Cycles with DOD 80%

 

[amberwolf]

Hmm, that battery is new, bought from Wallen power, specs below. I thought this would give enough power to the system but I was a bit too keen to get the system running and probably didn't do enough battery research.

If it's new, then as long as the company builds them from new cells (not recycled junk or stuff laying around in a warehouse for years like some do), aging shouldn't be the issue.

Without knowing the current limits set in the controllers I couldn't tell you if the battery is capable enough for those, but no matter what the limits are it still shouldn't drop that much voltage under no load current.

What *is* the no load current, if you have any way to test this? (should be a couple of amps, typically)

Can the cycle analyst v3 be programmed to override the default controller settings? I'd quite like to get one anyway as they seem like a good investment for future builds.

If you mean the controller current limit? Sort of. You can use the current-limiting throttle (there are modes for Amps/current, Watts/power, Speed, etc:

Cycle Analyst V3 - Grin Products - Product Info

The Cycle Analyst display is our first flagship product released in 2006 and now in its 3rd generation. While the hardware design may be showing its age, it still has capabilities and functionalities that are unmatched by modern ebike systems with their pretty colour screens. This device also...

ebikes.ca

in the Throttle Input menu it has explanations of each mode.

If you use the current limit mode, then set whatever limit you want in the Power and Current Limit Setup Menu, along with any specifics of how you want it to behave in the event of limiting.

It doesn't actually change anything in the controller, but it modulates the throttle signal the controller gets so that it effectively still limits battery current in the system.

 

[ChubbyQuark]

I managed a longer test ride today and re-checked the volt drop. The drop is around 1-1.5V without load, this is tested with the wheel at rest and then applying 1/3 throttle. Strangely the pulsing noise that can be heard here doesn't correlate with the colt drop.

Motor Noise

I've not tested for current before, from what I understand I'll need a watt meter or multimeter in series with the power amd load?

On the test ride the pulsing/lurching only seems to happen with less than 1/2 throttle, at full throttle acceleration is consistent. Seems similar to this issue here,
E-S Greyborg owners' pics - Page 13 - ES 1.0 Graveyard but I can't see if this was ever resolved in the thread, but my guess would be it's a controller issue as it's not been set or designed for this system.