Supplementary Battery Question
- Thread starter Bikes4two
- Start date
I ALWAYS seperate the batteries for charging because I have one 4A charger and one 2A.
However, even if they were the same amperage and make, there are too many variables.
Terminal voltage setting could be adrift by 0.1-0.2v or more.
As they heat up during charging component drift will differ.
Switch over point may not coincide.
Any charging through the discharge port bypasses the BMS, a recipe for potential horror story.
On a couple of occaisions over the years I have mistakenly left the Y splitter connected
and had a, correct and needed, right royal bollocking from this forum.
So, even with diodes, I would seperate, 'cos things have been known to break down if my past experience with zenor diodes on old British motorbikes is anything to go by. I know it does a different job but point made.
However, even if they were the same amperage and make, there are too many variables.
Terminal voltage setting could be adrift by 0.1-0.2v or more.
As they heat up during charging component drift will differ.
Switch over point may not coincide.
Any charging through the discharge port bypasses the BMS, a recipe for potential horror story.
On a couple of occaisions over the years I have mistakenly left the Y splitter connected
and had a, correct and needed, right royal bollocking from this forum.So, even with diodes, I would seperate, 'cos things have been known to break down if my past experience with zenor diodes on old British motorbikes is anything to go by. I know it does a different job but point made.
BMS
True enough which is why I like to use BMS modules that have an on/off switch to mitigate the risk of doing that.
Each battery output is a female, as is the output of the two into one adapter. But of necessity the two inputs of the adapter have to be male, so when only one is connected the other is exposed.
Just needs care during connection, there is no reason to need to use the adapter with an unused side.
That's why I use Anderson power plugs, both sides are shrouded.
- I have two bikes both with TSDZ2 36v 250w motors and the cells in all the battery packs mentioned are 21700 LG M50lt 5000mAh bought from Fogstar wholesale for £2.99 each.
(n.b. - I am very aware of the current limitations of these cells but I am a low power rider using almost exclusively the ECO power mode which from observation draws less than 5 Amp from the battery packs). - I thought it time that I updated this thread as since raising it I have built another two 10s 1p battery packs to add to the 10s 2p pack I built and have been using for several months now.
- One of the goals of building these battery packs was to make the carrying of the batteries as least obvious as possible - stealth mode if you like (lol) but I just fancied that 'look' on my bikes which are old style (as am I ).
- I built the 10s 1p packs in slightly different arrangements - one being more oblong than the other so that it fits into a bottle cage tool holder whilst the other one being more block shaped fits neatly into the side pockets of my Carradice saddle bags.
- The larger 10s 2p battery fits into a handlebar bag, so between this, the bottle cage tool holder and saddle bag I can swap batteries around to give a number of different capacities (ie riding ranges) as the other main goal of this little project was to have enough battery power for rides upto around 60 miles.
Batteries in parallel
There's endless comment to be found on this topic in various forums so just to add to all of the discussion, here's my finding so far:
- I didn't particulaly want to go down the route of Ideal or Schottky diodes (more points of failure, bulky etc).
- A lot of folks seemed to be fine with a straight 'Y' connector providing at the time of connection each battery was fully charged and within 100mV (0.1v) of each other.
- Looking at the data for the cells and seeing that each cell had an internal DC resistance of around 20 milli-ohms, I did a fag packet estimation of what sort of current would flow between battery packs (of the same nominal voltage) at different charge states (SoC being indicated by the actual voltage).
- To me 100mV was very conservative so I set up a little test jig (see below) to measure the current flow between battery packs at slightly different voltages.
- I started aound 100mV difference and iirc the current flow was under 200mA (or something like that). The biggest voltage difference so far has been 410mV (0.41v) and this gave a current of 700mA which started to decay very quickly and between tea breaks (10 mins?) was down to just a few milli-amps.
- For me then and the battery packs I've tested, I'll not be concerned to parallel up if the packs are withing a few hundred millivolts of each other.
100mV is not much energy for a 10s1p 5AH pack , roughly 1% of the capacity of the pack, under 2WH or 7200 Joules. You are looking at a current less than 0.5A.
The problem will arise when you parallel much larger packs and with more cells in parallel. Then the resistance of the packs will be smaller, the current will be much higher Take for example 10s5p pack with the same cells. The resistance of the pack is thus 20 milliOhms * 10/5 = 40milliOhms. 100mV / 40 milliOhms = 2.5A. The process to arrive at an equilibrium is essentially adiabatic like a swing door, that 2.5A current will oscillate rapidly between the two batteries and can damage the FETs. If the difference is 1V, then the current can reach 25A.
There is an oscillation you get when you parallel a new battery with low internal resistance with an old battery with a higher internal resistance . The batteries may balance initially but what happens under load?
The new battery because of its lower internal resistance will supply proportionally more current under load and its internal voltage (emf) will drop. So when the load eases off, the batteries have to equalize again with the older battery discharging a little back into the newer battery. To some extent, the new battery act as a decoupling capacitor for the old battery (which may be an advantage in some ways).
This happen frequently as when cycling e.g in a town, we are continually going from pedaling to freewheeling, from load to no-load. I don't know to what extent in practice its a problem but using diodes prevent it. I suspect that if the supplementary battery is small compared to the older high resistance battery, it will have to work harder.
The new battery because of its lower internal resistance will supply proportionally more current under load and its internal voltage (emf) will drop. So when the load eases off, the batteries have to equalize again with the older battery discharging a little back into the newer battery. To some extent, the new battery act as a decoupling capacitor for the old battery (which may be an advantage in some ways).
This happen frequently as when cycling e.g in a town, we are continually going from pedaling to freewheeling, from load to no-load. I don't know to what extent in practice its a problem but using diodes prevent it. I suspect that if the supplementary battery is small compared to the older high resistance battery, it will have to work harder.
Last edited:
Good testing. To recap what's been said up-thread, given that BMS malfunctions aren't uncommon the procedure to link up is:
- charge both
- swap chargers and top up both
- measure difference (because BMSs can differ)
- switch off both BMS modules
- fit Y lead working upstream (in case of isolation failure)
- connect to bike first (inrush) then
- switch on both batteries
With a bit of discipline and understanding it can work well... provided you aren't all that tired after a 60 mile ride.
This gets old quickly, with a near lossless combiner pack it's charge, connect, ride:
XT60 Plug Battery Discharge Converter Dual Battery Pack Double Battery Discharge Converter Switch Balancerfor E-bike - AliExpress 26
Smarter Shopping, Better Living! Aliexpress.com
www.aliexpress.com
Plus as clued-up makers we're still fallible and sometimes it's better to be seen to engineer things beyond the standard we feel we need (think vested interests > media influence > indiscriminate regulation > captive market).
I've never heard of a BMS malfunctioning because of batteries connected in parallel. I've paralleled batteries many times and so have others. The problem is too many people inventing theories instead of actually doing things
Sounds over complicated. Charge both batteries until full. Check that they're both over 41.5v (36v battery) and connect. That's all I've ever done. Be careful of the other connector on the Y-lead because it becomes live when you connect the first battery, so don't let it touch anything.
The most important thing is to disconnect them for charging because one will charge the other through the discharge port if you don't, and that by-passes the charge safety controls. Obviously, if you use blocking diodes or one of those proprietary parallel battery devices you don't need to worry about anything other than whether those diodes are still working.
Who said that?
"Sounds over complicated" - fair enough it's a thread summary, sorry I missed the bit about live connectors.
A combiner's simpler than a controller and if it fails you can connect either battery around it and carry on. Description of sorts here:
Double Battery Discharge Converter For E-bike 20A/30A/40A/60A/80A/120A Dual Battery Pack Switch Balancer 24-72VDC [Double Battery Converter] - $16.38 : Zen Cart!, The Art of E-commerce
Zen Cart! Double Battery Discharge Converter For E-bike 20A/30A/40A/60A/80A/120A Dual Battery Pack Switch Balancer 24-72VDC [Double Battery Converter] - ★★★★★ Double Battery Discharge Converter For E-bike 20A/30A/40A/60A/80A/120A 1--Current:20A,30A,40A,60A,80A,120A 2--Voltage:20V-72V(max...
www.topbikekit.com
If you charge battery 1 while the battery 2 is still connected by the link at the output, both will surely be charged at the same time and the only thing that is missing is the balancing phase on battery 2? If you charge next time the other battery then the balancing would be done alternately, no?
Go on then, for the un-anititated, what are they please?
Out of interest, I had a 10s 4p 10Ah bottle battery whose BMS failed early on and rather than change it I bypassed the charge side and charged the battery via the discharge ports for nigh on 12 months / several thousand miles of use.
I would once per month take off the the bottle battery case top (to access the BMS connections) and measure each cell bank with a DVM to check for imbalance and at the last time of measuring the max deviation between the cell banks was just 80mV.
I've subsequently broken the pack down - the cells were just 'no name' 2500mAh chinese ones, so nothing special.
It's very dangerous to do that without knowing how your BMS works. Some cheap batteries/BMSs use the same MOSFETs to switch on/off charging and discharging. With those, you can charge in parallel, but if you have a separate charge control MOSFET and you charge through the discharge wires, there's nothing to switch off the charge when the first cell reaches 4.2v, so any cell can go as high as it wants until it bursts into flame.
when you connect two batteries, you are in effect charging one with the other through the charging port too. We discussed this subject some years back, didn't we?
I find it is difficult to discuss without a reference schematics.
Last edited:
We've been through it before several times. You don't charge the second battery through the charge port unless both batteries have combined charge and discharge MOSFETS, which is only on the cheapo crappy ones.
The only dedicated charging port batteries via the discharge port are the same port BMS , they only have two wires from the BMS and not three.
Several have followed utube or fakebook advice to wake up a battery or charge through the discharge port when their battery refuses to charge via the normal route, fools are easily led when bypassing the safety features of BMS's.
Just beacause it has worked doesn't make it inherently safe , the individual has no idea of what is occurring with the BMS and later on risks issues and fires if critical components on the BMS are damaged irreversibly.
Stick to the recommended charging sequence , if a battery doesn't charge either recycle it or if not competent have an expert check/test it. The latter means taking a few simple voltage checks and if needed opening the battery and taking some more simple voltage checks.
Several have followed utube or fakebook advice to wake up a battery or charge through the discharge port when their battery refuses to charge via the normal route, fools are easily led when bypassing the safety features of BMS's.
Just beacause it has worked doesn't make it inherently safe , the individual has no idea of what is occurring with the BMS and later on risks issues and fires if critical components on the BMS are damaged irreversibly.
Stick to the recommended charging sequence , if a battery doesn't charge either recycle it or if not competent have an expert check/test it. The latter means taking a few simple voltage checks and if needed opening the battery and taking some more simple voltage checks.
Related Articles
-
Swytch announce new conversion kit with ‘pocket-sized’ battery- Started by: Pedelecs
-
New Swytch launches on Indiegogo, raises £100k in first hour- Started by: Pedelecs
-
Swytch to unveil 70% smaller, 50% lighter conversion kit- Started by: Pedelecs