Controller Trips under load

Get all your technical information about electric bikes here.
steverichab   10 mW

10 mW
Posts: 21
Joined: Oct 02 2011 8:15am

Controller Trips under load

Post by steverichab » Mar 18 2020 12:46pm

Hello:
I have two E-Bikes that need new controllers and I have purchased 2 controllers from
Amazon: https://www.amazon.com/dp/B074FYSJKW/ref=cfb_at_prodpg

Here is a video of the setup & problem: https://www.youtube.com/watch?v=WE4Ks6s1VWE

Here is how both controllers are wired to 48 V, 1000 W (Magic Pie 3) hub motors with internal controllers removed:
1) 48 V LiFePo battery The read 52.2 V and have very little "sag" when running
2) Throttle
3) Hall Sensors
4) Phase wires
The jumpers installed are: Lock=>"power lock" (orange-red), High Speed=>"Stalls"(blue-black), Direction of rotation=> "inverted gear(black-grey).
Please tell me why both controllers stall/trip at 4 to 5 amps but run fine at less load? The vendor suggested these jumpers but still no go.

Any help is greatly appreciated
Last edited by steverichab on Mar 25 2020 1:06pm, edited 1 time in total.

User avatar
e-beach   1 GW

1 GW
Posts: 3281
Joined: Jan 10 2012 9:48pm
Location: Any Los Angeles area beach I am at. Or Santa Monica or possibly the south bay beaches.

Re: Controller Trips under load

Post by e-beach » Mar 18 2020 1:18pm

I am going to say that I don't think it is your new controllers, but your battery or bms.

What is your BMS rated for? How many amps? Your new 48v 1800w controllers may be triggering a LVC due to battery sag.

:D :bolt:
Favorite Quotes:
"This is L.A., sugar. There is no 'over the top." Chris Erskine
"At a certain point the entropy wins." Maria Helena Braga

Current Build: ProFlex 757 Expert full suspension. Yescomusa 36v 800w Rear DD, upgraded 10AWG solid core through axle phase wires. 15ah Headway, 1000+ cycles, 80% DOD 30A Tronsung controller.

Past: Trek 4500 Yescomusa 36v 800w front DD.
Liahona, Yescomusa 36v 800w generic front DD.
1992 Trek 800, Yescomusa 800w 36v front DD.

steverichab   10 mW

10 mW
Posts: 21
Joined: Oct 02 2011 8:15am

Re: Controller Trips under load

Post by steverichab » Mar 18 2020 3:16pm

That is unfortunate. These are not the first controllers I have tried on these bikes. Had a similar thing with a couple of 1200 W "KT" type controllers and I thought it was something to do with the jumpers/settings. I returned them because I didn't want to buy the expensive LCD screens because I already have Cycle Analyst's on these bikes.
I can still return these controllers and go another route. Could you recommend an inexpensive (<$100) no frills controller that would work? Great bikes/motors and batteries and I would like to use them.

steverichab   10 mW

10 mW
Posts: 21
Joined: Oct 02 2011 8:15am

Re: Controller Trips under load

Post by steverichab » Mar 18 2020 3:19pm

To answer your question I am not sure what the BMS is rated for but the sag when I hit the throttle seems small ( maybe down to 50V from 52V) hard to say because it trips so fast under load.

User avatar
e-beach   1 GW

1 GW
Posts: 3281
Joined: Jan 10 2012 9:48pm
Location: Any Los Angeles area beach I am at. Or Santa Monica or possibly the south bay beaches.

Re: Controller Trips under load

Post by e-beach » Mar 18 2020 4:15pm

Before you do anything with the controllers, lets make sure your battery is working as it should. It could be that you have a bad cell in one group that is causing LVC. Or you are drawing too many amps and the BMS is cutting off that way.

Do your best to dig up the spec on your magic pie and post the specs and the link so we can see if the controllers are too high powered for what the battery can do.

Also, get the specs from your battery so we can try to find out about the BMS.

Questions:
Can you get access to the individual cell groups in your battery?
Can you access the BMS since wires inside your battery?

What we want to do is check all the cell groups after a long charge, like a 12 hour charge And then let it sit overnight to see if one or more cell groups drop in voltage.

Also, meticulously carefully check all your wiring to make sure nothing is loose.

:D :bolt:
Favorite Quotes:
"This is L.A., sugar. There is no 'over the top." Chris Erskine
"At a certain point the entropy wins." Maria Helena Braga

Current Build: ProFlex 757 Expert full suspension. Yescomusa 36v 800w Rear DD, upgraded 10AWG solid core through axle phase wires. 15ah Headway, 1000+ cycles, 80% DOD 30A Tronsung controller.

Past: Trek 4500 Yescomusa 36v 800w front DD.
Liahona, Yescomusa 36v 800w generic front DD.
1992 Trek 800, Yescomusa 800w 36v front DD.

steverichab   10 mW

10 mW
Posts: 21
Joined: Oct 02 2011 8:15am

Re: Controller Trips under load

Post by steverichab » Mar 18 2020 9:57pm

Thanks E-beach,

Because these batteries and hub motors have worked together for the past 3 years I was focused on the new player but I now see that I need to shift my focus to the batteries. I assumed that because both batteries behaved the same regardless of which controller/motor/battery was in play. These batteries are in a metal case but I can get to the BMS if I cut away some of the heat shrink holding the pac together. I'll keep you posted on my findings
Thanks

User avatar
e-beach   1 GW

1 GW
Posts: 3281
Joined: Jan 10 2012 9:48pm
Location: Any Los Angeles area beach I am at. Or Santa Monica or possibly the south bay beaches.

Re: Controller Trips under load

Post by e-beach » Mar 18 2020 10:01pm

OK, what you are looking for is accuses to the since wires of the bms. By skillfully metering the wires you can meter the cell groups. This is the info we need to start the process of elimination.

:D :bolt:
Favorite Quotes:
"This is L.A., sugar. There is no 'over the top." Chris Erskine
"At a certain point the entropy wins." Maria Helena Braga

Current Build: ProFlex 757 Expert full suspension. Yescomusa 36v 800w Rear DD, upgraded 10AWG solid core through axle phase wires. 15ah Headway, 1000+ cycles, 80% DOD 30A Tronsung controller.

Past: Trek 4500 Yescomusa 36v 800w front DD.
Liahona, Yescomusa 36v 800w generic front DD.
1992 Trek 800, Yescomusa 800w 36v front DD.

User avatar
amberwolf   100 GW

100 GW
Posts: 29009
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Controller Trips under load

Post by amberwolf » Mar 18 2020 10:56pm

if the bms was shutting down, power to teh entire system would be shut off, and the cycle analyst would also shut off.

if that is not happening, the bms is not shutting down, so the battery is not the source of the problem.

steverichab wrote:
Mar 18 2020 12:46pm

Please tell me why both controllers stall/trip at 4 to 5 amps but run fine at less load?
when you say "trip", or "stall", *exactly* what do you mean?

do you mean that the motor is operating perfectly normally, but then suddenly and completely stops operating?


if so, does this also happen offground, or is it only when attempting to ride the bike? meaning, if you have the wheel offground but slam on the throttle (which should give a surge of high amps for just a moment), does the motor operate exactly as expected, or does it have some undesired operation? if so, exactly what does it do?





fwiw, with the insufficient information provided so far, it sounds like the phase / hall wiring is in the wrong order. matching colors (or any other "random" matching) doesn't work very often. there are a number of ways to find the right order, but the simplest is to leave all the hall sensor wires the same, and then start swapping the phase wires until you get a smooth rotation and very low (1-2a) current draw when throttling up (wheel offground), and then continue to throttle up until it gets to full throttle (wot); it shoudl stay at low current draw. throttle up only very slowly each time.

if the rotation is backwards, then if the controller has a reverse wire, use that to fix it. if not, swap any two hall wires, and then do the phase wire thing again until you find the same smooth movement and low current, but in the correct direction.

User avatar
e-beach   1 GW

1 GW
Posts: 3281
Joined: Jan 10 2012 9:48pm
Location: Any Los Angeles area beach I am at. Or Santa Monica or possibly the south bay beaches.

Re: Controller Trips under load

Post by e-beach » Mar 19 2020 11:58am

I agree, more information is required. As for the BMS shutting everything down the OP did mention a hard reset. However he didn't mention if the CA, or the whole system stops working. Back to more info.

To AW's post, does all power to the whole system get cut off or just the power to the wheel?

I still think LVC is the suspect, caused by either the BMS or Controller cutting off.

I still think the battery groups should be tested even though the two batteries reacted the same.

The controller the OP has an: "Under Voltage Protection: 42 ± 0.5V" according to this site post https://crankdriven.com/product/tdpro-4 ... cle-moped/

Did the other controllers have the same feature?

Does the magic pie have a proprietary controller system that won't work with other controllers?
Favorite Quotes:
"This is L.A., sugar. There is no 'over the top." Chris Erskine
"At a certain point the entropy wins." Maria Helena Braga

Current Build: ProFlex 757 Expert full suspension. Yescomusa 36v 800w Rear DD, upgraded 10AWG solid core through axle phase wires. 15ah Headway, 1000+ cycles, 80% DOD 30A Tronsung controller.

Past: Trek 4500 Yescomusa 36v 800w front DD.
Liahona, Yescomusa 36v 800w generic front DD.
1992 Trek 800, Yescomusa 800w 36v front DD.

User avatar
amberwolf   100 GW

100 GW
Posts: 29009
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Controller Trips under load

Post by amberwolf » Mar 19 2020 11:00pm

e-beach wrote:
Mar 19 2020 11:58am
Does the magic pie have a proprietary controller system that won't work with other controllers?
a number of people on es have removed the internal mp controllers (usually after those died of heatstroke ;) ) and replaced with external controllers, successfully, so i don't think the mp motor is incompatible with non-mp controllers.

steverichab   10 mW

10 mW
Posts: 21
Joined: Oct 02 2011 8:15am

Re: Controller Trips under load

Post by steverichab » Mar 25 2020 12:55pm

Hello and thank you all for your input!
Well, there are a lot of "need for more information" requests so here goes.
When I say "trips" I mean:

1) When the wheel is off the ground it spins perfectly and the Cycle Analyst indicates `1-2 Amps, even if I jam the throttle. This is true on both bikes!

2) If I try to ride the bike or even drop the back tire to the ground and the Cycle Analyst "sees" 4 to 5 Amps everything goes dead(Analyst & controller) until I turn the key on the battery off and on again. This too is true on both bikes!

As I said I have 2 controllers/Cycle Analyst/MP motors and 48V 10Ah batteries (all identical). All combinations of these devices have the same symptom.

I disassembled one of the battery cases and here are the readings:

Cell dimensions, 16 each=> 18mm X 65mm X 130mm, Prismatic LiFePO4 cells

BMS=> BWPB-013A-16S-30A-023D-10G4E (48Volt, 30A)

Total voltage read from pack 52.7 Volts fully charged.
Each cell reads 3.29 V but cells# 5 &#12 read 3.28 V and Cells# 15 &#16 read 3.33 V

I did not take the second battery case apart but fully charged it reads 52.9 Volts

The batteries are about 3-4 years old but only have 65 charge cycles per the Cycle Analyst.

Any feedback is appreciated but this project is truly on a tight budget and if you recommend replacing selected cells I would need a link to a source (I haven't found yet)
I have found replacement BMS's. Could that be it? I keep thinking that, how could 2 BMS's go bad? With the age of the batteries those were most likely but with these readings ...

Thanks again, Steve
Last edited by steverichab on Mar 25 2020 2:40pm, edited 2 times in total.

User avatar
e-beach   1 GW

1 GW
Posts: 3281
Joined: Jan 10 2012 9:48pm
Location: Any Los Angeles area beach I am at. Or Santa Monica or possibly the south bay beaches.

Re: Controller Trips under load

Post by e-beach » Mar 25 2020 1:48pm

steverichab wrote:
Mar 25 2020 12:55pm
Hello and thank you all for you input!
........Cell dimensions, 16 each=> 18mm X 65mm X 130mm, Prismatic LiFePO4 cells...........
Are you saying you have 16s batteries?

16s LiFePO4 fully charged is 58.4v. (16 x 3.65v = 58.4v)

If your LVC is say, 3v per cell, that would be a 48v trigger from your bms. LiFePO4 is known to sag under load, it might be causing the bms to shut down.

So are you running 16S batteries?

:D :bolt:
Favorite Quotes:
"This is L.A., sugar. There is no 'over the top." Chris Erskine
"At a certain point the entropy wins." Maria Helena Braga

Current Build: ProFlex 757 Expert full suspension. Yescomusa 36v 800w Rear DD, upgraded 10AWG solid core through axle phase wires. 15ah Headway, 1000+ cycles, 80% DOD 30A Tronsung controller.

Past: Trek 4500 Yescomusa 36v 800w front DD.
Liahona, Yescomusa 36v 800w generic front DD.
1992 Trek 800, Yescomusa 800w 36v front DD.

steverichab   10 mW

10 mW
Posts: 21
Joined: Oct 02 2011 8:15am

Re: Controller Trips under load

Post by steverichab » Mar 25 2020 2:23pm

Yes e-beach,
These are 16S LiFePo4 batteries. From somewhere in this post I thought I saw that LVC for this controller is 42V and I have never seen these packs much above 53V even when new.
It would be very sad if all 32 of these cells were bad after only 60 or so charge cycles.

Any thoughts on how I could bench test these battery packs for voltage sag?

john61ct   1.21 GW

1.21 GW
Posts: 3637
Joined: Dec 18 2018 2:06pm

Re: Controller Trips under load

Post by john61ct » Mar 25 2020 2:27pm

For longevity, LFP should not be charged beyond resting 4.35Vpc

A charging voltage of 3.45-3.50Vpc is plenty, very little capacity is added by going any higher. The exact setpoint can vary a bit, depending on the current rate. Just CC (Bulk) is fine no CV / Absorb stage needed.

Around 3.2V is the 50% midpoint, at rest.

LVC should really be well above 3.0Vpc, but again depends on discharge C-rate.

The curve is so flat, a coulomb counting wattmeter can be useful for helping guesstimate SoC% if loads vary a lot.

john61ct   1.21 GW

1.21 GW
Posts: 3637
Joined: Dec 18 2018 2:06pm

Re: Controller Trips under load

Post by john61ct » Mar 25 2020 2:31pm

Test for capacity will give State of Health.

I use 80% as EoL, properly treated should be thousands of cycles.

If quality cells to start with, lots of crap out there.

What is your max C-rate discharge?
steverichab wrote:I thought I saw that LVC for this controller is 42V
That is murderously low almost 2.6Vpc, should never let them get anywhere near that!

john61ct   1.21 GW

1.21 GW
Posts: 3637
Joined: Dec 18 2018 2:06pm

Re: Controller Trips under load

Post by john61ct » Mar 25 2020 2:33pm

Link to the LFP cells used

steverichab   10 mW

10 mW
Posts: 21
Joined: Oct 02 2011 8:15am

Re: Controller Trips under load

Post by steverichab » Mar 25 2020 2:45pm

The batteries I have are both Golden Motor

https://goldenmotor.bike/product/48v10a ... d-battery/
The golden motor site say max continuous discharge is 30A (3C) and peak discharge 50A(5C).

The 2 new controllers have these spec's listed:
Specifications:

Rated voltage: 48V DC

Current limit : 32A

Match motor: 48VDC 1800 W

Conversion efficency: 95%

Under Voltage Protection: 42±0.5V

Thanks

User avatar
e-beach   1 GW

1 GW
Posts: 3281
Joined: Jan 10 2012 9:48pm
Location: Any Los Angeles area beach I am at. Or Santa Monica or possibly the south bay beaches.

Re: Controller Trips under load

Post by e-beach » Mar 25 2020 2:59pm

The link you showed says 15S. Not 16S. Their is a difference.
Golden Pic.JPG
Golden Pic.JPG (96.14 KiB) Viewed 107 times
What do you have, 15S or 16S?

:D :bolt:
Favorite Quotes:
"This is L.A., sugar. There is no 'over the top." Chris Erskine
"At a certain point the entropy wins." Maria Helena Braga

Current Build: ProFlex 757 Expert full suspension. Yescomusa 36v 800w Rear DD, upgraded 10AWG solid core through axle phase wires. 15ah Headway, 1000+ cycles, 80% DOD 30A Tronsung controller.

Past: Trek 4500 Yescomusa 36v 800w front DD.
Liahona, Yescomusa 36v 800w generic front DD.
1992 Trek 800, Yescomusa 800w 36v front DD.

steverichab   10 mW

10 mW
Posts: 21
Joined: Oct 02 2011 8:15am

Re: Controller Trips under load

Post by steverichab » Mar 25 2020 3:22pm

e-beach,
These are 16S LiFeSo4 batteries as confirmed by the label on the BMS and the number of cells in the pack. I have include pic's this time of the BMS & cells (attached).

Good eye but that must be one of those "Chinesium Typos".

Thanks Again
Attachments
Golden Motor Battery.docx
(588.13 KiB) Downloaded 8 times

steverichab   10 mW

10 mW
Posts: 21
Joined: Oct 02 2011 8:15am

Re: Controller Trips under load

Post by steverichab » Mar 25 2020 3:39pm

e-beach,
These are 16S LiFeSo4 batteries as confirmed by the label on the BMS and the number of cells in the pack. I have include pic's this time of the BMS & cells (attached).

Good eye but that must be one of those "Chinesium Typos".

Thanks Again

PS I called the vendor I purchased the batteries from in 2015(Golden Motor Canada) and they confirmed they are 16S LiFePo4 batteries and there is a typo on the website.
Attachments
Golden Motor Battery (1).pdf
(445.88 KiB) Downloaded 7 times

john61ct   1.21 GW

1.21 GW
Posts: 3637
Joined: Dec 18 2018 2:06pm

Re: Controller Trips under load

Post by john61ct » Mar 25 2020 4:18pm

Well the longevity-promoting specs I gave (better than the vendors') was mostly on a per-cell basis in any case.

If you find the cells have died an early death, assuming can-style, then replace with A123, can't do better than that, until you get up to the larger prismatics from CALB, Winston etc

steverichab   10 mW

10 mW
Posts: 21
Joined: Oct 02 2011 8:15am

Re: Controller Trips under load

Post by steverichab » Mar 25 2020 4:47pm

Thanks John61ct,
The A123 are interesting would it be possible to fit 10Ah of these in the aluminum case that I show in the attachment on the previous post?
I found the 10Ah Prismatic LiFePo4 that I have now for $5.80 each on Alibaba but not sure on shipping. I really don't know yet if these are bad but I would like to know of a test that would let me know for sure.

Steve

User avatar
amberwolf   100 GW

100 GW
Posts: 29009
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Controller Trips under load

Post by amberwolf » Mar 25 2020 6:38pm

steverichab wrote:
Mar 25 2020 12:55pm

1) When the wheel is off the ground it spins perfectly and the Cycle Analyst indicates `1-2 Amps, even if I jam the throttle. This is true on both bikes!
then it is likely the whole system is wired correctly, as phase/hall combo being wrong between motor/controller usually makes that no-load current higher (sometimes a lot), and/or causes motor rotation to be jerky or noisy, etc.


2) If I try to ride the bike or even drop the back tire to the ground and the Cycle Analyst "sees" 4 to 5 Amps everything goes dead(Analyst & controller) until I turn the key on the battery off and on again. This too is true on both bikes!
that means the load on the battery is high enough to cause it to shutoff to protect itself.

the most likely thing is that one or more of the cells is defective, and under load it's voltage drops so far the bms shuts off the output to protect the cells from overdischarge.



Total voltage read from pack 52.7 Volts fully charged.
16s should read around 58v fully charged, but if they've aged enough or weren't great to start with they may be unable to hold that, and end up just under 53v (around 3.2-3.3v per cell).

i would recommend first leaving the packs on their chargers at least overnight. if they do not increase then, leave them on for as many days as you can, while you are present to monitor them (in case something goes wrong). remeasure the total pack voltage at the end of each period you can charge them for. if it doesn't go up even a little by the end of a week or more, then it's possible the cells simply are unable to go any higher anymore.


if after that time the behavior is the same, it's likely you'll need new batteries.


you can attach extension leads to the cells themselves inside the battery, so that you can connect your multimeter to them such that you can see the meter while riding, and move the meter from cell to cell and retest under load for each cell, noting down the lowest voltage you see on each cell as the system shuts off. (this may require using a camera to record the meter, as it may not be possible to watch it during the ride).

if all the cells behave the same, then they'll all sag below lvc, which is usually around 2.8v or so for lifepo4. this means the whole pack is bad enough to not work correctly.

if only one or two cells sag below lvc, while all the others stay at about the same voltage they start at, then just the ones that drop in voltage are bad.

it is probably better to replace the whole pack if possible, but if it's not, you could at least replace the worst ones that cause the problem, and then use it until others also fail, replacing each as it happens.


if you find the voltages on each cell don't actually drop much under load, but that the total pack voltage does (just prior to the bms input), it could be the *connections between cells* that are causing the problem, where the hardware has loosened, or oxidized, and removing and retightening these connections may fix that.

User avatar
amberwolf   100 GW

100 GW
Posts: 29009
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Controller Trips under load

Post by amberwolf » Mar 25 2020 6:41pm

e-beach wrote:
Mar 25 2020 2:59pm
The link you showed says 15S. Not 16S. Their is a difference.

Golden Pic.JPG
i can't find anything on the site, or in that pic, that says how many cells in series there are, at all.

there is a "15s" in that pic, but it is in a current limit section, and refers to the time in seconds (s) that much current is allowed. ;)

User avatar
e-beach   1 GW

1 GW
Posts: 3281
Joined: Jan 10 2012 9:48pm
Location: Any Los Angeles area beach I am at. Or Santa Monica or possibly the south bay beaches.

Re: Controller Trips under load

Post by e-beach » Mar 25 2020 7:05pm

A few things,

Your battery seems to be charging and balancing to the 3.3 nominal voltage (+ or -) of LiFePO4. Full charge is 58.4v.

Your BMS is a Best Way Protection Board number 013A (BWPB-013A). I can not find exact data on that board, although it seems to be set at the factory to certain parameters depending on the battery chemistry. Ask your battery vendor if they have any specifications on the board you own.
BWPB-013.JPG
BWPB-013.JPG (73.1 KiB) Viewed 87 times
https://translate.googleusercontent.com ... ONcqn2jiwQ

As for a "bench test" of your system you would probably need a dynamometer for that. However you can easily test with a watt meter that has it's own power supply, like a small battery pack.

Although, your Cycle Analyst might have the information you need stored even if it goes dead. Go through your CA manual and see what functions it has.

What you would like to know, as AW said, is the actual voltages of each cell. However if you can only check the total voltage of the pack when everything shuts down, then at least you might garner a clue as to the problem.

The quick test is this: Do your stall test as you have done before, but instead of looking at the amp draw, look at the volt meter and see how low it goes and then report back as to the lowest voltage you have before it shuts itself off.

:D :bolt:
Favorite Quotes:
"This is L.A., sugar. There is no 'over the top." Chris Erskine
"At a certain point the entropy wins." Maria Helena Braga

Current Build: ProFlex 757 Expert full suspension. Yescomusa 36v 800w Rear DD, upgraded 10AWG solid core through axle phase wires. 15ah Headway, 1000+ cycles, 80% DOD 30A Tronsung controller.

Past: Trek 4500 Yescomusa 36v 800w front DD.
Liahona, Yescomusa 36v 800w generic front DD.
1992 Trek 800, Yescomusa 800w 36v front DD.

Post Reply