Parallel batteries

[WheezyRider]

This is the approach I've always taken. It's no big deal to charge both batteries before connecting. Why would you want to use a half-charged one if you're going on a journey that needs two batteries?

It's whatever you feel most comfortable with. But there are advantages to the diode setup, it gives you flexibility.

You could charge one pack at a time and not have to worry about things. Eg, if you are on a trip, you only have one charger and can spend an hour or so recharging one pack in say a coffee shop. You can take more advantage of opportunistic charging in this way.

 

[Nealh]

For a touring cyclcist is makes sense.
For a day rider who might rides 100 odd miles using a pair of batteries one would surely have planned the ride and made sure before the off they are charged.

 

[Bikes4two]

Parallel, increase in ah and current discharge so easier on the cells, voltage will remain higher initially for staying power. As vfr mentioned make sure both are quite close voltage wise before connecting, 0.025v or better.

A bit of an old post now Neal, but I've been reading through this post and the Schottky Diode Parallel Pack post (link) as I'm wanting the option of paralleling up my 'no name' 36v,10Ah (10s, 4p) bottle battery with my recently built 36v, 7Ah (10s, 2p) using LG M50LT 21700 cells.

I'm looking at going for a simple 'Y' connector (ie no diodes, Shottky, Ideal or otherwise) to keep the arrangement as simple as possible. To connect the two batteries like this then of course as has been posted many times, the batteries need to be at or near the same voltage - the question is how near do the voltages need to be? Herewith are my thoughts/assumptions on this and all feedback is welcome.

1. Where the voltage is different between the two packs, it must not be of a magnitude such that the normal discharge current of one battery nor more importantly, the charge current of the lower voltage of the two batteries, is exceeded.
2. Starting with what is the resistance of the battery packs? I do not know what cells are in the bottle battery but the resistance of the LG M50LT is known, albeit it is not a fixed figure due to many variables such as temperature, SoC and so on, but for this exercise lets take it as 20 milli-ohms per cell. This gives a pack resistance for the 20 cells in a 10s, 2p arrangement, a value of 100 milli-ohms
3. The bottle battery cells at 2,500mAh each I likely to be a higher resistance than this so I'll use the figure for the LG M50LT - ie a pack resistance of 100 milli-ohms
4. Further, let's assume a max charge current of 0.3C so that's nominally 1.6A per cell for the LG M50LT, leading to a max charging current for the 2p pack of 3.2A
5. So what voltage across the 100 milli-ohm battery pack will give a current of 3.2A?
6. Answer from ohms law of V=IxR = 0.320V or 320mV, implying that when connecting my two battery packs together in parallel, I should aim for their voltages to be within 320mV of each other?

Now at this stage that is a hypothesis so I'd be very grateful (Neal or others) for comments on the validity of my approach.

 

[matthewslack]

A bit of an old post now Neal, but I've been reading through this post and the Schottky Diode Parallel Pack post (link) as I'm wanting the option of paralleling up my 'no name' 36v,10Ah (10s, 4p) bottle battery with my recently built 36v, 7Ah (10s, 2p) using LG M50LT 21700 cells.

I'm looking at going for a simple 'Y' connector (ie no diodes, Shottky, Ideal or otherwise) to keep the arrangement as simple as possible. To connect the two batteries like this then of course as has been posted many times, the batteries need to be at or near the same voltage - the question is how near do the voltages need to be? Herewith are my thoughts/assumptions on this and all feedback is welcome.

1. Where the voltage is different between the two packs, it must not be of a magnitude such that the normal discharge current of one battery nor more importantly, the charge current of the lower voltage of the two batteries, is exceeded.
2. Starting with what is the resistance of the battery packs? I do not know what cells are in the bottle battery but the resistance of the LG M50LT is known, albeit it is not a fixed figure due to many variables such as temperature, SoC and so on, but for this exercise lets take it as 20 milli-ohms per cell. This gives a pack resistance for the 20 cells in a 10s, 2p arrangement, a value of 100 milli-ohms
3. The bottle battery cells at 2,500mAh each I likely to be a higher resistance than this so I'll use the figure for the LG M50LT - ie a pack resistance of 100 milli-ohms
4. Further, let's assume a max charge current of 0.3C so that's nominally 1.6A per cell for the LG M50LT, leading to a max charging current for the 2p pack of 3.2A
5. So what voltage across the 100 milli-ohm battery pack will give a current of 3.2A?
6. Answer from ohms law of V=IxR = 0.320V or 320mV, implying that when connecting my two battery packs together in parallel, I should aim for their voltages to be within 320mV of each other?

Now at this stage that is a hypothesis so I'd be very grateful (Neal or others) for comments on the validity of my approach.

I aim for a tenth of a volt when connecting my two solar batteries in parallel, and I have a 5A fuse in the link.

If the same charger is used to fully charge both batteries, and they are checked using the same multimeter just before making the connection, a tenth should be easy to manage.

 

[Bikes4two]

Thanks Matthew.
I've now done two full discharge/charge parallel runs through my test rig and employ a 5A fuse too.

On both occasions the voltage difference was around 300mV and nothing registered amp wise on the Watt Meter, even briefly.

And of course the voltages were measured off load too. I'll try another test run with a more sensitive ammeter next time.

Give the shape of the battery discharge curves where voltage drops off very quickly from the 42v point, I suspect you'd need a closer voltage between the two batteries if connecting up in parallel part way through a discharge.

 

[matthewslack]

I would not connect at mid charge. The flat discharge curve means that a few tenths of volts represents a substantial amount of energy needing to move to equalise. Too much risk for me. If needing to do this, connect temporarily using a high wattage low ohms resistor until voltages are equal, and then connect.

If voltages are within 1 volt, only needs to be 1 ohm to keep current below 1 amp, and at 1 volt drop and 1 amp, only needs to be 1 watt rated. Something like this.

 

[Benjahmin]

When I connect my two batteries together I am generally looking at 0.05v differential, I will go at 0.1v but no more.
My packs are of very similar capacity but one uses a 4A charger, the other a 2A. Of course this means that the 4A one finishes first. I then swap the 4A charger to the other battery (as by now the voltage is sufficient to reduce the charging current) to ensure the same terminal voltage.
Using the same volt meter to measure both is crucial as then any reading error is applied to both.

 

[Bikes4two]

When I connect my two batteries together I am generally looking at 0.05v differential, I will go at 0.1v but no more.

Hi Benjahmin - how did you arrive at the 0.05v to 0.1v values?

 

[Bikes4two]

I would not connect at mid charge. The flat discharge curve means that a few tenths of volts represents a substantial amount of energy needing to move to equalise.

I'm with you all the way on that one.

If needing to do this, connect temporarily using a high wattage low ohms resistor until voltages are equal, and then connect.

Again, I'm with you on that one - I have a dummy load on my test rig which I will disipate any excess voltage/energy through if the battery voltage differentials are not within acceptable (to me) limits.

Since my earlier post, after fully recharging both batteries and letting them rest for a couple of hours, the votage differential was 570mV. I stuck my DVM between them and on paralleling up, I measured a current flow of 360mV which over about 30 mins or so, dropped to zero.

I did out of curiosity order up two Ideal diodes which I may play with if for some reason I have to consider paralleling up when the batteries are at much different charge levels. I can see me using such a rig if for some reason one battery is partially discharge and the other not, and I decide to go on a long ride at short notice and don't have time to re-charge to get the range I'd be needing.

I am curious as to how the Ideal Diodes will share the load. I posted HERE previously about my thoughts on hysteresis and I intend to find out more through bench testing.

 

[Eltommos]

I use 2 batteries in parallel - see photos. I also have the option to only carry 1 - either on the downtube or the pannier rack.

I use a parallel battery adapter which I purchased from Amazon - (choose the the suitable amp rating)



The parallel connector is housed inside the box mounted to the seat post tube.

As mentioned by Woosh - if I only carry 1 battery then the exposed electrical connectors on the other battery holder is not live using this parallel connector (Not possible to short out)

This setup with 2 batteries equates to 1080 watt hours (each battery is 36V 15ah) which allows me to travel much further if needed.

It doesn't matter if the state of the batteries are not equal, e.g. One battery can be fully charged and the other can be say 50% charged. I believe it's good idea that both batteries have the same ah rating

 

[guerney]

If this is the same product, I was hoping for a thorough teardown and detailed explanation as to how it works, but:

I'd rather use a couple of kill switches, which are less likely to fail: deplete one, switch it off, switch on the other.

The 40A switches on this build, are from army radios:

 

[Benjahmin]

Hi Benjahmin - how did you arrive at the 0.05v to 0.1v values?

Common sense and caution. Although an electrician of many years my maths and abilities do not stretch to measuring/calculating the internal resistance of batteries, whilst being aware of their scary power density in short circuit.
Using the same charger to top of both batteries (one at a time) means that they are pretty much the same voltage when connected. Discipline is to always measure before doing so.

 

[Bikes4two]

Common sense and caution. .......

Fair enough, nothing wrong with exercising caution if unsure. I asked in case there was a calculation and explanation behind those figures so I could check out someone else's reasoning against my own.

Thanks for posting.

 

[Sturmey]

I use schottky diodes now but have directly connected batteries in parallel. In practise, the batteries seem to equalize if the are within one volt of one another and this is a rule of thumb and common practise with experienced users. e.g.

Max safe Voltage Difference when connecting two Batteries in parallel

What would you consider the maximum safe voltage difference when connecting two 14S10P 18650 batteries in parallel? Is it possible to calculate the maximum equalizing current? ( maybe U(difference)/(Ri1+Ri2) )

endless-sphere.com

PS Expensive batteries with 'smart bms' or 'coulomb counting' should probably not be paralled except possibly with diodes. There may also be a problem with smart batteries under warranty as the smart bms may keep a log file of 'events' as the case with smart phylion batteries with two year warranty.

 

[Bikes4two]

@guerney - I'd be interested to see/know/where you've mounted your switches.

If I were looking at switches (which has some appeal as in KISS) , I'd think about a change over switch with a centre off position to prevent accidentally nudge one of the 'off' switches back on again with the possibility of paralleling up batteries if dissimilar charge levels.

 

[guerney]

@guerney - I'd be interested to see/know/where you've mounted your switches.

That's someone I met - I didn't ask him enough questions about this, being my pre-ebike days. He was making a small folding electric car, which was a common interest. Perhaps ask via a comment on his Youtube video?

Fiat ‘Smiles per Hour’ Design Competition: the finalists - Car Body Design

Fiat UK has announced the short-list of five finalists for its Smiles Per Hour Design Competition. People can vote until 25 Spetember by visiting the London Fiat Flagship Store.

www.carbodydesign.com