Customisation of the Yosepower kit or other ebike kits

[PC2017]

Just thought I would start a new thread were the Yosepower kits are concerned and the generalised field of customising any ebike ekit is concerned...

My current project has hit a wall - Customising a Yose 350w rear kit (36v 26" 15amp) for the sake of future proofing and to allow a 48v battery upgrade when needed.

I had a few issues with the kit from the start, as I was slightly experienced with the general feel of conversions and having used a Yose 250w front for a good year with my homemade dodgy battery wire connector alongside a Dillenger 350w front the year before that, I knew what to expect from the YOSE kit...

The batteries I am using to this day are the Dillenger supplied batteries, one 13ah Samsung XX and the other 14.5ah Samsung (only stated in the sales pitch at the time & not confirmed) both now at least 3 year old with 152+ cycles each, the 13ah holding at 42.0V and the 14.5ah holding at 41.2V.

After installing the Yose kit as is, after several test rides the motor/controller started briefly cutting out, this never happened with the Dillenger (dolphin mount controller) nor with 250w(15amp) Yose. I asked about and moaned to YOSE, they kindly sent me 2 motors(without wheel) and neither one really solved my problem, so I went and did the next best thing --- Aliexpress --- Determined that either the YOSE motor, the YOSE controller or my dodgy battery wire connector was at fault, I purchased a dolphin mount (15amp) controller(which eliminated my dodgy battery wire connector) , fitted it and have been running smoothly for the last so many months....

With an influx of cash and saving I started to buy new stuff, tyres, wheel truing, chain, freewheel, hydraulic v-brake, suspension seat post, 22 amp controller and a new 1to4 cable all in anticipation for spring....

(apologies for those who know this bit) So I installed the 22 amp controller, using my dodgy battery wire connector and the C5 setting to limited the amp draw, armed with a fully charged battery(14.5ah), C5 set to 4, I went for a ride and it was perfect, just as expected and for January surprisingly warm!

Next day, using the same battery, I nipped to the shops, got to a mini circle and hit the throttle and (not allowed to swear) it CUT out, a quick switch off and on again (best practice protocol) it lived for a few meters and cut out again, undeterred I went home to really think about this....

One 15 amp controller connected directly to the battery - no problems
One 15 amp controller with my dodgy battery wire connector on the YOSE 250w - no problems
New 15 amp Yose controller 350w rear INTERMITTENT cutting out...
New 22 amp Aliexpress controller 350w rear WITH HALF CHARGED BATTERY cutting out...

So I tested a theory, what if it's LVC - armed with a fully charge battery I headed out full of optimism and not one cut out, therefore I have concluded that my problem is RELATED to lvc, voltage sag or battery wear.

Am I on the right track here, or have I missed a key factor and without electrically testing every part of the setup which I am limited on ability and knowledge, is there any way of diagnosing my actual problem before I purchase a 48v battery?? Or would a high spec expensive 48v battery actually solve my problem??

 

[Sturmey]

With the LCD3, by hitting the center button twice, the battery voltage in real time is displayed instead of km/miles distance. It will stay on throughout your journey You can clearly see any battery sag as battery voltage falls. Also, when LVC cutout happens, the LCD3 remains on but battery bars outline flashes in a certain way. (This has happened to me on occasions towards the end of very long journeys.) So I think it should be obvious if you are getting lvc cutouts. If the LCD3 cuts out with motor, it could be a BMS cutout. The default value for lvc is 30v on my two kits but can be changed.
Is it possible that the LCD3 is at fault? Its possible to run bike without LCD3 by strapping the plug. I have tried this and it worked . Alternatively, try swapping the LCD3 with another.
As a general note, I have used both Yose 350 and later the 250 front version. The 250W version although slower, seems very strong compared to the 350w version and is much better on hills. I was disappointed with the 350W version for this reason. The KT 7/15 amp controller gets very hot once the motor comes under any pressure e.g strong winds/hills. The motor itself doesn't seem to get hot. I would say the 9 mosfet larger controller is probably a good idea.
Anyhow best of luck.

 

[Nealh]

My inPut is the 14.5 battery is at fault and knackered or at least a Parallel cell group is very tired or has failed cells. The fact that it is down to 41.2v is an indication that it is loosing capacity and cells have high IR, voltage collapse is a sign of IR which means they also can't deliver the load any more (except maybe in PAS1 - 3).
The fact that it cuts with throttle suggests the battery simply can't supply the current asked of it.
How does it fair using PAS1 - 5 ?

 

[Nealh]

Irrelevant of the hub used I have used a 25a KT 9 fet with both the Yose (Bafang hub) and an AKM128 both at 36 & 38v with no issues with Panny cells.
The 6 fet KT in the main I stopped using as they got very hot (too hot to handle) esp using max amps and climbing steep ascents for some distance. Though once cooled they work again but continual use will end up with toasted fets.

 

[Nealh]

Ideally with the 9 fet controller good 10a cells are needed and the more in P the bettery. If thinking of a new battery build DIY consider going down the 21700 route you could use Sam 40T or if 18650 my current pref would be LG29 or MJ1.

 

[PC2017]

by hitting the center button twice

Thanks for the input, yes I did this, I always watch voltage rather then watts and the battery indicator. When the battery is half charged the voltage was standing at 37V then after the motor, either on PAS or throttle would drop very fast cut out the controller/motor but the LCD would stay on and the battery indicator does not alter a part from loosing a bar, which is why LVC never occurred to me as a possible issue because I have only seen the battery indicator flash as you rightly mention, the only other possibility is the LVC is coming from the BMS not the controller... However I will put your idea of swapping out the LCD on the LIST, but like Nealh as mentioned I am eliminating the more obvious possible outcomes first as this does seem to happen at half charge only and Nealh has confirmed my next item on the LIST as part of my process of elimination - The 14ah battery....

How does it fair using PAS1 - 5 ?

Thanks for the input.

The new controller is 9 Mosfets 36/48V 22 amps KT (julet)

When half charged cut out is instant regardless of PAS1 or up to 5 or with throttle... I had a similar conclusion to you about the 14ah, it never occurred to me in the past that swapping batteries could be part of the issue as I have always done this when ones empty I used the other simple, but now I think if I only had one battery I may have come to point where the battery would of had to checked as part of the issue... So my next theory is to run the balanced battery (13ah) until or if I experience any problems and eliminate or conclude that the 14ah battery is at fault...which I hope it is then I can draw a line under this issue.

Thanks for the battery ideas, I was so close to just buying the YOSE 48v but now it's back to saving and looking for build(er) with the cells you mentioned, I may look at enerPower again.

One quick question regarding battery to controller wire- The wire coming from my pass through battery mount connector is 16 AWG(18 amps) but the wire going in to the controller is 14 AWG(13 amps) why is this?

 

[Nealh]

The controller 14awg can handle 32a continuous so plenty for the 22a rating.
The battery 16awg is again suitable to handle the current it can supply which is approx. 22a continuous, I doubt the battery is rated for mor ethen 15 -20a discharge.
All AWG for short bursts can handle about double there rating.

 

[Sturmey]

Thanks for the input, yes I did this, I always watch voltage rather then watts and the battery indicator. When the battery is half charged the voltage was standing at 37V then after the motor, either on PAS or throttle would drop very fast cut out the controller/motor but the LCD would stay on and the battery indicator does not alter a part from loosing a bar, which is why LVC never occurred to me as a possible issue because I have only seen the battery indicator flash as you rightly mention, the only other possibility is the LVC is coming from the BMS not the controller............

Just to clarify.(what I think).....
BMS cutout This happens in the battery. It happened to me and was caused in my case by a broken balance wire. In this case, the power is interrupted to both the display unit and the controller and to any external or display lamps. The bike may only go dead momentarily.

LVC cutout. This happens in the controller when the battery voltage is low. The controller switches off the motor power (mosfets) only and gives an indication on display. The display unit and any lamps (including the display unit lamp) will remain powered. On my LCD3, this can be reset immediately by pressing the center button provided the battery voltage come back up to about 32-33v in my case.

 

[PC2017]

All AWG for short bursts can handle about double there rating.

As you guessed I am still concerned about my battery wire connector, so thanks for the info!

BMS cutout

Yep true I never thought about it that way... I went out this morning with 13ah balanced reading 42v from charged, c5 @ 5, no cut outs on a quick journey, during this I watched the voltage on PAS5 on throttle full speed, slow moving (12mph) to full pelt(18mph+)the voltage moved from 40.7 to 37.7 then back to 40.7 when off throttle to 41v at standstill. Going out thisaf for a 6 mile run maybe squeeze a fe more miles in weather permitting - I will report my findings

 

[Nealh]

As you guessed I am still concerned about my battery wire connector, so thanks for the info!

Yep true I never thought about it that way... I went out this morning with 13ah balanced reading 42v from charged, c5 @ 5, no cut outs on a quick journey, during this I watched the voltage on PAS5 on throttle full speed, slow moving (12mph) to full pelt(18mph+)the voltage moved from 40.7 to 37.7 then back to 40.7 when off throttle to 41v at standstill. Going out thisaf for a 6 mile run maybe squeeze a fe more miles in weather permitting - I will report my findings

The battery to controller wiring if short will pose little issue.

The sag of 3v on throttle sounds typical for a 4 or 5 P battery, it is when as you mentioned battery is at 37v that cut out occurs. LVC will come in to play for sure on full throttle or PAS 4 or 5, it is prudent at this stage to conserve power and use low PAS2 for less stress on cells.

 

[PC2017]

it is prudent at this stage to conserve power and use low PAS2 for less stress on cells.

Words of wisdom - so far so good with the 13ah 7 miles covered and pushing it a bit more than a normal ride, standing voltage at the end was 38.9v, as you mentioned 3v -3.5v sag (noted while watching the voltage on the LCD) when setting off from traffic lights at PAS 5 (not normal practice) and no cut outs to report, testing continues tomorrow - If this 3v sag is expected with a 4p/5p arrangement would this sag be any less with a similar 48v with newer better cells like the ones you noted above? I am just thinking out loud here - less cells cheaper option or is it actually worth it to spend more and get that extra row of cells?

 

[PC2017]

Sorry people its this old chestnut again...

HaiLong PLUS Rahmenakku Li-Ion 48V 13Ah VTC5

ENERpower Rahmen Akku Li-Ion Hailong PLUS DIY 48V aus eigener Fertigung 13Ah (624Wh) mit Murata US18650VTC5 Für Motoren bis 950 Watt…

enerprof.de

us18650vtc5-vtc5 (murata or sony?? or same company??)

 

[Nealh]

Murata produced the cells for Sony hence the dual name, there was a licence agreement allowing Murata to do so. A few years ago Murata actually bought the Sony battery business lock stock and barrel.

 

[Nealh]

All cells will sag, some more then others. Sony/Murata are rated as a very good cell and will suffer about half the sag. The more cells in P then less sag but that means more weight and expense.

 

[PC2017]

Murata actually bought the Sony battery business lock stock and barrel.

Good to know thanks, I thought the VTC part sounded sony. These German guys tend to go for cells with high mAh but stay with 4p, the battery linked is the only one I found so far that is 5p - £344 ish with a 5 amp charger and that's with free shipping, £333 3a charger then 12 euro for shipping looks like the 5a charger is the better option, I best get saving again.

 

[PC2017]

Got to 37v on the 13ah today 10 mile covered c5=5 (16.54a) then it cut out, reduced c5=4 (14.66a) still cut out over PAS 2, reduced c5=3(11a) no problems...

A troubling thought is still plaguing me, why would I not see any cut outs on the dolphin mount 15a controller even at 37v??

 

[Sturmey]

You have never really given a good description of what exactly happens or what exactly you mean by 'cutout'. The reason I say this is because there is a sort of 'pause' condition on some bikes that is just caused by a delay built into the controller. For example, on my Bafang hub, there is a consistent and annoying pause for about 2 seconds when the bike returns back to 25km/hr after being at a higher speed (often after going down hill). This has also been reported by other people and I suspect that its a deliberate time delay to prevent wear on the clutch. In contrast, I have noticed a much shorter (about half second) time delay with my yose when the throttle is pressed after pedalling and been on PAS ( and a noticeable 'clunk' from the hub which is probably the clutch slapping in).

 

[PC2017]

I have noticed a much shorter (about half second) time delay with my yose when the throttle is pressed after pedalling and been on PAS ( and a noticeable 'clunk' from the hub which is probably the clutch slapping in).

This was the problem when I first got it, however IMO this is not how they should run, my 350w Dillenger front never missed one beat until the hall sensors gave out after 2000 miles then after replacing them it did another 2000 miles with never a cut out or a pause, then it died sadly, I know some have had 10000+ miles out of one motor but 4000 miles imo that was not bad for everyday use.

You have never really given a good description of what exactly happens

I don't really know what constitutes a "good description" without "you" actually experiencing it... The first time it happened was when I installed the YOSE 350w as is, motor, 15 amp controller all the YOSE cables, then I would of said you are quite right it seemed like a clutch slip, a slight in and out when throttle was applied at a stand still, at the time YOSE sent me 2 new motors without wheel after the last motor install, I decided to go back to my old 15 amp dolphin mount controller(KT) from my first kit that I had salvaged as the pins were off in the 1t4 cable and everything was perfect, hence the new aliexpress sourced dolphin mount when my old one died and up until I thought of a dual voltage controller for a future battery upgrade every was peachy - unless I had been swapping out a half charged battery for a full without thinking about and not actually noticing, I will reinstall the dolphin mount controller and test this hypothesis, it's on the list....

Now after running my new controller (22a) set at 16.54a I was happy however the battery was fully charged, when I tried a half charged battery, it cut off like you mention a clunk or click, LCD3 still on, no watts when pedalling, "assist" displayed, throttle on, symbol rotating, no watts, no motor, volts displaying 37v, battery indicator at half - then a quick switch off and on(LCD control) back up again for a brief period then the clunk and off again - however today I was far from home so I changed c5 to 03(11a) and was off again with power and 350w max at PAS5

I will endeavour to film it if possible...

Next on the list, install the 250w front and run it at 16.54a with the same setup as is now and see if the 350w motor is the problem as you mentioned the rear 350w is not as good as it should be!

 

[Sturmey]

it cut off like you mention a clunk or click, LCD3 still on, no watts when pedalling, "assist" displayed, throttle on, symbol rotating, no watts, no motor, volts displaying 37v, battery indicator at half - then a quick switch off and on(LCD control) back up again for a brief period then the clunk and off again -

Ok, thats the kind of detail I was looking for. Now, from what you describe, its very like an lvc (low voltage) cutout from the controller although the battery indicator outline normally flashes to indicate this and it can be reset by pressing the center button momentarily without actually turning off the LCD. So I am not sure. There may be differences in controllers.
But I would not rule anything out. Its a process of elimination.
You could also check the C12 setting which on mine is set to 4 (default). You could try changing it to 0, which will drop the voltage limit from 30v to 28 v just to see does it make a difference. (Is there a difference with the dual 36/48V I wonder?) You could also set C12=7, which would increase cut off voltage and make it worst if this is the problem.
A substitute good battery or possibly paralleling your 2 batteries would be a great way to prove as to whether the battery is the problem. I have a home made dummy load made from heating elements which is great for testing the battery discharge rate and also its capacity, and the battery can be tested on its own off the bike.
One thing I not sure of is what happens the controller settings when you swap controllers. I assume you have to check/change settings as they are all held in the controller and not in the LCD but I am not too sure about this. Perhaps some are held in the LCD? You can enlighten me on this.

 

[PC2017]

Its a process of elimination.

Yep hence the list.

C12 setting which on mine is set to 4 (default)

Yep mine was at 4 and I did change it 0 - with 37v & c5=5 c12=0 there was no click or clunk it just cut out as before lcd still on no watts etc

I assume you have to check/change settings as they are all held in the controller and not in the LCD but I am not too sure about this. Perhaps some are held in the LCD? You can enlighten me on this.

I always assumed there was no settings held in the controller a part from the basics it need to fuction ie LVC and Voltage detection, I have been just swapping out the controllers all KT and using the same LCD3 the only setting I changed initially when moving from the 15 amp dolphin mount controller to the 22 amp 9 fets was the amp draw c5, from c5=10 to c5=05.

I will run the front hub YOSE 250w at 16.54a tomorrow from a fully charged battery and log the results - I had not one problem with this motor or it's original controller, however when I tried this controller on the 350w the problems persisted ergo I automatically ruled out a controller fault, possibly in haste.

A substitute good battery or possibly paralleling your 2 batteries would be a great way to prove as to whether the battery is the problem. I have a home made dummy load made from heating elements which is great for testing the battery discharge rate and also its capacity, and the battery can be tested on its own off the bike.

far too technical for me.