Swytch Battery Questions

[Bikes4two]

• A friend with a Swytch Bike made the fatal mistake of not charging the battery over the winter and now it won't charge up and pressing the battery level button doesn't illuminate any LEDs
• He has asked for my help - I'm no novice with ebike batteries having built three of my own, but this is my first encounter with the Swytch Battery
• In the pic below I've labelled the six connector pins 1 - 6: pins 3 and 4 are the charger +ve and -ve connection points respectively - what are the other 4 pins for please - PAS and Display I'd guess?
• The voltage on pins 3,4 is 17Vdc
• More importantly, what is the best way to get into these packs so I can investigate further?

Have others encountered the same issue and managed to resolve it? If I can get into the pack I'll check out the individual cell voltages and take the problem from there.

Many thanks

PS - I understand there are several generations of Swytch Bike - any idea what generation this battery is from?

 

[saneagle]

The middle two are for the charging the outer ones need to be bridged have to be bridged to complete power out the circuit. There's a thread from a few moths back that gives the exact details.

This might be it:

Swytch Pro/Max yellow battery connector secrets revealed?

Disclaimer - you mess around with these battery packs at your own risk - what follows are my own observations that may or may not be fact! On the yellow connector on the battery: - the two centre connections (O and D shaped) are charging input only. There is circuitry before the actual battery...

www.pedelecs.co.uk

 

[Cisco-man]

The middle two are for the charging the outer ones need to be bridged have to be bridged to complete power out the circuit. There's a thread from a few moths back that gives the exact details.

This might be it:

Swytch Pro/Max yellow battery connector secrets revealed?

Disclaimer - you mess around with these battery packs at your own risk - what follows are my own observations that may or may not be fact! On the yellow connector on the battery: - the two centre connections (O and D shaped) are charging input only. There is circuitry before the actual battery...

www.pedelecs.co.uk

Not quite. Ignoring the charge “O and D” connectors in the middle, the outer flat pins are the battery outputs, and the inner pair of flat pins are shorted together by the yellow connector plug to turn the pack on.

 

[Cisco-man]

It’s a Swytch Gen 3 pack. If you attempt to open it, I would love to see pictures when/if you do. Most of the body of the pack is rubber, and I would think the front face (with the swytch logo on it) prises out, but not sure.

 

[Cisco-man]

I’m asking the battery experts here as I’m not one of them.

Would it be stupid to force a charge (for a brief time only), by putting a charge on the two output flat pins whilst shorting out the inner two flat pins to turn the pack on. I’m not saying to charge it fully, but just a little charge to bring the voltages up so the BMS works - if that’s what would happen.

standing by to be shot down for suggesting something potentially risky…

 

[Cisco-man]

I’m asking the battery experts here as I’m not one of them.

Would it be stupid to force a charge (for a brief time only), by putting a charge on the two output flat pins whilst shorting out the inner two flat pins to turn the pack on. I’m not saying to charge it fully, but just a little charge to bring the voltages up so the BMS works - if that’s what would happen.

standing by to be shot down for suggesting something potentially risky…

Actually I don’t think this idea would fly, because there are mosfets(?) for turning off the output pins that would get in the way of a reverse charge.

Pin 1 in the photo is +ve output, and pin 6 -ve.

The plastic “anvil” on the bottom edge of the battery is (in normal use) to hold the bottom edge of the battery when you fit it in the handlebar holder. The holder pincers the battery between the yellow plug on the top edge, and this plastic piece on the bottom edge. I’m not saying that there aren’t screws behind it, but I’m thinking not.

The controller for the bike is inside the handlebar battery holder. That’s where the connectors for display, pas, throttle, brakes, etc. emanate from.

 

[Bikes4two]

It’s a Swytch Gen 3 pack. If you attempt to open it, I would love to see pictures when/if you do. Most of the body of the pack is rubber, and I would think the front face (with the swytch logo on it) prises out, but not sure.

Well, I've got the battery apart - the rubber cover will prise off which leaves you with a hard plastic two-part case (one part yellow, the other grey) which is held together by 10 small screws.

Getting the case apart requires some fairly determined prising again and working your way around the case edge once, maybe twice. There is a silicon gasket between the two case halves which makes it tougher to get apart.

Right now I can see the 10 cells (it's a 10s 1p battery) - the cells are bigger than 18650 so I'll measure them to see what is what.

The 10 cell structure is hot glued or masticed into one half of the case and this is as far as I've got - time to peel away more layers to get to the BMS pins etc.


(I'll put up some more notes later this evening with pics and greater details

 

[saneagle]

Well, I've got the battery apart - the rubber cover will prise off which leaves you with a hard plastic two-part case (one part yellow, the other grey) which is held together by 10 small screws.

Getting the case apart requires some fairly determined prising again and working your way around the case edge once, maybe twice. There is a silicon gasket between the two case halves which makes it tougher to get apart.

Right now I can see the 10 cells (it's a 10s 1p battery) - the cells are bigger than 18650 so I'll measure them to see what is what.

The 10 cell structure is hot glued or masticed into one half of the case and this is as far as I've got - time to peel away more layers to get to the BMS pins etc.


(I'll put up some more notes later this evening with pics and greater details

Once you have access to the cells, you can measure their individual voltages, which will probably reveal the cause of the problem. I' m going to guess that one or more have dropped below 2.5v.

 

[Bikes4two]

Once you have access to the cells, you can measure their individual voltages, which will probably reveal the cause of the problem. I' m going to guess that one or more have dropped below 2.5v.

I'm leaving further work until tomorrow as the 'man shed' is too hot for concentration, or at least the level of concentration required for battery fettling.

I have the cell pack clear of the case and once I remove some of the white mastic, I should be able to measure individual cells. The total voltage as measure at the BMS is 19.45v and assuming all 10 cells are equal (will check tomorrow) that's 1.95v per cell - in your view do you think the cells are recoverable? (they are some type of Samsung 21700 - can't see which model but the whole pack is described as 180Wh so likely the Samsung 50E currently £2.99 each at Fogstar - a re-cell then?).

Dismantling notes to follow.

 

[Bikes4two]

Getting the battery pack apart:
Removing the rubber casing:

• The casing is a very tight fit - I used thin 1" palette knives to prise the casing off.
• I started by prising the rubber away from the centre of the long edge working intially from 'inside' the connector void and then the top edge, as illustrated by the purple arrows.
• You can work in the same wy from the other side of the casing, starting at the battery securing 'indent'
• 

Splitting the hard case

• Remove the 10 small screws around the periphery of the two part case
• Using a pair of palette knives (or similar), start at one of the long edges to get the first pry point going and ease in the 2nd knife and keep prying and working along the edges.
• There is a silicon gasket between the two parts making the task less than easy but straight forward enough with a bit of patience
• 

Lifting the cell pack from the case

• I only did this to get easier access to the BMS connections
• There are 4 further screws holding the cell assembly to the yellow half of the case (see attached pic - I just pushed the screwdriver through the mastic to unscrew).
• Then using a tool similar to a broad bladed screwdriver (or indeed a screwdriver as long as your are catious about not piercing and cell cases), work in from the short end of the cell stack and gently pry away from the mastic that is beneath and holding things in place.
• Access to cell ends and the cell pack +Ve and -Ve terminals ('X' marks the spot) is now easier for maybe a bit of 'recovery charging' if @saneagle things it worth while (I have a variable DC power pack for this).
• 

 

[Cisco-man]

Brilliant photo’s and method! Thanks for such a great “how to do”. It’ll help a lot of people in the future to fix these packs.

 

[saneagle]

I'm leaving further work until tomorrow as the 'man shed' is too hot for concentration, or at least the level of concentration required for battery fettling.

I have the cell pack clear of the case and once I remove some of the white mastic, I should be able to measure individual cells. The total voltage as measure at the BMS is 19.45v and assuming all 10 cells are equal (will check tomorrow) that's 1.95v per cell - in your view do you think the cells are recoverable? (they are some type of Samsung 21700 - can't see which model but the whole pack is described as 180Wh so likely the Samsung 50E currently £2.99 each at Fogstar - a re-cell then?).

Dismantling notes to follow.

You can't measure the cells like that because the BMS is switched off. What you're measuring is leakage charge, and the value you get depends on yhe impedence of your meter. It's nothing to do with the pack voltage.

 

[Nealh]

1p of 50e cells isn't cutting the mustard, they are only rated for 10a and one ideally doesn't want to draw more then 6.5/7a from them for longevivty.
One will need at least a 20a if not a 25a /30a cell for a 1p config.

One will have to use Fogstar retail if only buying ten cells as the wholesale site has a min spend of £75.

 

[Bikes4two]

Thanks for the input @saneagle - I incorrectly said 'measured at the BMS' - I actually measured the voltage at the ends of the cell pack at the points marked 'X' in the last picture above.

Anyway, since then I've managed to probe through the mastic blobs (heck, that stuff really bonds and is tiresome to remove!) and measuring from the +Ve end of the pack, recorded the following cell voltages:
1. 4.08v
2. 4.08v
3. 4.08v
4. 4.08v
5. 3.08v
6. to 10. a few millivolts, so in practice, zero volts

In terms of researching (internet trawling) the resuscitation of those last 5 cells, opinion varies as to whether it can be done and even if it can, how good will the recovered cells be with respect to the first five?

Of all that I read last night I felt the most informed and practical advice came from a YT poster called Bigclive whose posts I used to read extensively, and in his discussion of recovering deep discharged cells (LINK) he thinks it is possible in the 'use case' he had before him, and in the YT video notes gives a link to an in depth paper (LINK2) that helped him to come to this conclusion.

However, the stresses and strains on an ebike battery are somewhat different to what he had on his bench and given that 5 of the cells (in this Swytch battery)are zero volts and five are not, makes me wonder as to 'repairability' of it all i.e. is there a BMS problem?

I am going to advise my friend to recycle (as in bin it) the battery - if it where my pack and the BMS was ok and I replaced/recovered the 5 zero volt cells, I'd never trust the battery to give me a decent power output and would always be wondering about future issues.

What do you guys think?

 

[saneagle]

I can guarantee that I've attempted to recover a lot more cells than Bigclive. I would never used cells recovered from zero volts in an ebike battery. Some will not recover enough to be charged to 4.2v. Those that do are normally very saggy (high internal resistance). Some appear to go back to being normal. Maybe it depends on how long they sit at 0v.

When I attempt to charge them, I use very low current - normally starting at about 100mA until over 3v, then I increase to 200mA for the rest of the charge.

The only ways I can think of that of that would drain cells to what they were in your battery is BMS failure unless the battery got water in it. It does look more like water than BMS because you have one half-way, which is otherwise difficult to explain.

I'd chuck that battery. It was pretty useless and over-stretched when it was new. It would be much easier to get any other 36v battery and wire it to the controller with a normal connector. That should give a battery that would be useful and would last.


What's that white thing glued to the middle of the yellow plastic side piece?

 

[Cisco-man]

A Saneagle suggestion that I found really useful for your remaining good batteries, is a power bank case (such as the link) that can be used for iPads/iPhones. Better than throwing them away.

4 Slots 21700 Battery Power Shell Charger Box LCD Display Power Bank DIY Case v | eBay

Can be installed 4x flat 21700 lithium battery(battery not included), Can be disassembled repeatedly. 1x 21700 Power Bank Case(Batteries not included). The power bank case needs to be welded battery, batteries in parallel.

www.ebay.co.uk

 

[Nealh]

IMHO any cell sub 2v is scrap , high IR and teriible sag can be expected.
Above 2v and one may be ok mothering them for other uses such as torches or in a small power bank.

What ever switch were thinking of using 1p of 50e, god only knows.
That config is sure to fail fairly soon.
For that config a 50S or P42A/P45B cell is needed.

 

[Bikes4two]

I've declared the battery pack scrap. For my own amusement I've cut away the BMS from the cell pack and in due course I'll find a use for the 5 good cells and out of curiosity I'll play around with one of the others to see what it's like after a 'recovery' but it'll not go anywhere near an ebike.

Apart from the choice of cells, the construction of the battery pack is very robust and weather-proof. I don't know if Swytch give the pack an IP rating but the outside rubber cover is an extremely close fit and the silicon gasket between the 2-part case is very good. I'd guess the only weak point, if indeed it is weak, might be the connector itself, but then again with the battery in place and the battery holder connector locked over the battery, it'd have to be some deluge for moisture to find its way in.

A note about the BMS construction - it's clearly a bespoke item so no chance of replacement. Further, I've tried to show in the pic below the little metal tags emanating from the BMS PCB that make up one half of the BMS battery balance/sense wires (the other sense wires are the thin blue ones in the pic). I mention the tags because if you wanted to swap out a duff cell or two, it'd be extremely difficult to separate the cell from the BMS and remaining cell cluster.

And then there's all that white mastic holding things together - a nightmare to pull away to free up components.

What I'm saying is, these battery packs are just about impossible to repair in any meaningful way.

@saneagle - {quote] What's that white thing glued to the middle of the yellow plastic side piece? [/quote] - its just a part number label which came off the cell pack because it got stuck to the white mastic on assembly, and definitely no water ingress so maybe the BMS was the problem after all?

Finally, the friend with the duff battery also has a second Swytch machine (and battery) - ordered at the same time and left unatended over winter too (despite the owners manual saying about periodic charging, a manual that he's now read!) but that battery seems to have survived the experience and has been used a few times this summer with no apparent loss of performance, but that is a very subjective assessment of course.

Well thanks @saneagle, @Nealh and @Cisco-man for your comments.

 

[radiatewishbone540]

I have a Gen 2 Eco battery which also died from not charging. What are my options? Ideally want to make a straight replacement of the battery inside the case.