paralleling LiFePO4 and Lipo

[fivari]

Hallo,
I have this e-scooter with 48V30Ah LiFePO4-battery-pack made with 16s10p cilinderical cells.
The capacity of the pack has come down to 25 Ah. Occasionally, I need more Ah to ensure that I reach my destination.
I was thinking about adding a small (5 Ah) selfmade batterypack made out of 18650 Lipo's from recycled laptop batteries, in parallel.
I don't want to charge the setup in parallel. I realise that i have to add diodes in order to prevent unwanted currents to occur between the two packs.

I have two questions:
1°) What is the better configuration: 13s, 14s (15s)?


2°) I found the next schottky diodes on ebay:

or

I believe the first one is good enough since the voltage difference between both batteries will never exceed 15V.
Thanks for your advice.

 

[dogman dan]

I am doing the same thing with my longtail bike. A 48v 15 ah pingbattery cannot cope with the 40 amps controller I have on the bike.

But it's very very tricky. I am using14s RC lipo with 16s lifepo4 so that fully charged matches good enough so the lipo doesn't try to charge the lifepo4. I'm not using diodes, nor getting anywhere close to a dod that would create a situation where much flow will move from the lipo to the lifepo4 bms. More likely in my case, the lipo will suck wh from the lifepo4.

But on the other end, I have to be very very very very carefull not to leave the lipo connected past it's lowest allowable voltage. The lifepo4 packs lvc is a lot lower than the lipo can go and like it.

Lifepo4 lvc is 44v but the lipo needs to stop by 49v resting. With this disparity between lowest allowable voltages, I simply cannot use all of my capacity.

Since you need all your capacity, I cannot recomend mixing chemistries for your application. You'd be better off to get more lifepo4, and paralell in that. Or, run seperate batteries. When the lifepo4 is used up, unplug that and plug into a lipo pack. But then the lipo pack would have to be bigger.

 

[SamTexas]

I don't see a need for diode at all since you are only discharging them in parallel. 14s would match nicely.

dogman missed the part about 18650 laptop cells. LVC for the laptop cell pack would be the same or lower than the LiFePO4 cell. Laptop cell LVC could be as low as 2.50V.

Bottom line: 14s laptop 18650 and 16s LiFePO4 is a good match. No diode needed for discharging.

 

[dogman dan]

And stay nicely balanced? You could take the lipo I'm using to 2.7v, but it's not going to be recomended by me.

I didn't discuss another approach either, 15s would give you a closer match at low voltage. To match top of charge voltage, you could simply undercharge the laptop cells.

Because my lico is different stuff, and my needs different, 14s worked good for me. Gave me a chance to use some 5s packs converted to 4s I had gathering dust.

 

[fivari]

Dogman,
thanks for your advice. However, the small Lipo-pack will not be able to deliver enough power to my scooter. So, running the two batteries separately is not an option.
Adding a small LiFePO4-pack would be an option, but I was looking for a cheap or free solution to the problem.
But coming back to your combination without diodes. When I go full throttle, the voltage of de life-pack sags down to 48V. That means that de Lipo would contribute depending on the ESR of the Lipo. When applying 14s, this would create a current corresponding to a voltage sag of 57,4V (14*4.1) down to 48V. That would mean a big initial current. After some riding, the Lipos get depleted and the resting voltage has decreased down to 49V. Then the current out of the lipos would become nearly zero. This way the Lipo-pack would be uncharged very soon.
So, maybe 13s is better: both packs would contribute more gradually. Initial voltage of 13s is only 53.3V. Therefore conribution of 13s pack would be lower than in the first case. Furthermore, the LVC would also be closer together. Only just after charging, there is a danger: the Life fully charged is @ 57.6V, which is too high for the 13s Lipo. ==> diode needed?!

For the moment I am using a different approach: in stead of adding a 13s5p-pack in parallel, I add a 2s30p-pack in series. This way the voltage goes up from 52,8V (16*3.3V) nominal to 60.2V (16*3.3V+2*3.7V).
My topspeed is definitely higher, but what about the range? Should be higher too, but I don't get that impression.

 

[SamTexas]

Staying balanced is a matter of matched capacity and internal resistance. You could have an unbalanced pack at 50% DOD if your cells are not well matched.

Anyway, all lithium chemistries are supposed to be balanced charged either manually or automatically via a good balancing charger.

 

[dnmun]

Hallo,
I have this e-scooter with 48V30Ah LiFePO4-battery-pack made with 16s10p cilinderical cells.
The capacity of the pack has come down to 25 Ah. Occasionally, I need more Ah to ensure that I reach my destination.
I was thinking about adding a small (5 Ah) selfmade batterypack made out of 18650 Lipo's from recycled laptop batteries, in parallel.
I don't want to charge the setup in parallel. I realise that i have to add diodes in order to prevent unwanted currents to occur between the two packs.

I have two questions:
1°) What is the better configuration: 13s, 14s (15s)?


2°) I found the next schottky diodes on ebay:

or

I believe the first one is good enough since the voltage difference between both batteries will never exceed 15V.
Thanks for your advice.

how much do they want for the IR 100A shottky? i found the same one on a solar panel ebay shop. i tried to talk doc into buying some since he has to combine two different packs for his zero. he thinks the diode would be exposed to 58V if the BMS opens but of course it won't be exposed to any more voltage than the difference between the two packs. so the 15V is more than adequate.

http://www.ebay.com/itm/400244740084?ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1438.l2648

since you are using the dinky little laptop lipo you could use some of the 10A axial diodes that they sell there too. connect two or three in parallel:

http://www.ebay.com/itm/20PCS-10AMP-Bypass-Blocking-Diode-DIY-Solar-Cells-Panel-10SQ045-Schottky-/130687274953?pt=LH_DefaultDomain_0&hash=item1e6d918fc9

or you can use these 15A axial schottky diodes. i bot these for combining pack because it was the best deal for diodes: two would be enuff on your pack.

http://www.ebay.com/itm/20-PCS-15A-45V-Schottky-Diode-SCHOTTKY-BARRIER-RECTIFIER-solar-panel-DIY-/271150583889?pt=LH_DefaultDomain_0&hash=item3f21d59051

i bot mine for $9.99 free shipping, but i slept through a $4.99, free shipping auction for these when he offered them at bid.

 

[fivari]

how much do they want for the IR 100A shottky? i found the same one on a solar panel ebay shop. i tried to talk doc into buying some since he has to combine two different packs for his zero. he thinks the diode would be exposed to 58V if the BMS opens but of course it won't be exposed to any more voltage than the difference between the two packs. so the 15V is more than adequate.

http://www.ebay.com/itm/400244740084?ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1438.l2648

I found them at the same supplier you are refering to: 6$ each.
Anyway, the other I selected comes down to the same price and would be more convenient to use. Or are there contra indications?

 

[dnmun]

like i said, i think you could use the axial diodes and it would be cheaper and more compact. i doubt if you could ever get more than 30A so just few of the axial diodes would be sufficient imo.

 

[cam3087]

So I'm kind of doing the same thing by adding to 48v 10ah lifepo4 in parallel but they discharge at different rates . Now those 15A axial diodes seem like they would do the job but would i need to add two in parallel on the positive side of the battery so I can get up to 30amps on the controller? Thanks- Cam

 

[dnmun]

i would add 4 of the 15A diodes in parallel if you wanna carry 30A. maybe 5. they are cheap. the more the current is spread out the lower the forward bias too.

i plan to do something similar and was thinking i would put an S shaped dog leg in the end of each lead on the diode, so they end would be offset by a little bit from the axis of each diode and then they could all be collected and soldered to the wire on each end with the diodes arrayed in circle in between, equally spaced to allow max cooling.

 

[cam3087]

So I got those 45v 15 amp axial diodes in from ebay and did a test on them from a 9v battery. Foward flow was normal volts but when I reversed the diode expecting zero I still got about 6v. Im I missing something here I don't want to put 5 in parallel in a 48v system and fry my bms because the resistors don't work.

 

[dnmun]

what did they measure on the diode tester? it should be open circuit if reversed. just measured one of mine and got 163 mV forward bias.

that is at 1A from the diode tester so as the current climbs you can expect the forward bias to climb also.

 

[cam3087]

Regular voltage, foward voltage across diode, foward voltage on reversed diode. Shouldn't the reversed diode be showing 0v?

 

[cam3087]

No diode voltage

 

[cam3087]

So if I put 4-5 diodes in parallel on the positive wire of each 48v 10 ah battery should I be ok and not worry of one battery charging the other?

 

[dnmun]

i guess i do not understand why you did that. you have the diode tester right there on the voltmeter but decided not to use it to test the diode. instead just measured the voltage on the battery for some reason.