Parallel different C rates- anyone measure what happens?

[John in CR]

My thinking is that with different C rate packs paralleled at the ends the voltage stays the same. Under short high load more current will flow through the lower IR higher C rate portion. When I let off the throttle a slight voltage differential from the different SOC will send energy from the lower C rate portion, but the voltage difference should be low enough to keep these cross pack currents from getting high. Sure it will be a bit more stressful than an homogeneous high C rate pack, but as long as I avoid deep discharges the effect shouldn't be too significant.

This view is supported by my experience over the last year and a half with a 30s pack. 20s was homogenous and charged with one charger. The other 10s was a mixed bag totaling a slightly higher capacity, and charged with a separate charger. That 10s included some 20C 18650's I was testing, along with some older 20C hard packs and 45-90C nanotechs. I very rarely went past 50% DOD, and over that period all of the sub packs remained in such good balance that I never had to balance charge any of them.

 

[docnjoj]

Hey John. My wife runs her trike with paralleled 10 AH old Ping (36 volts) and 6 paralleled Fatpacks ( so 10S 2 P paralleled x6). Both are charged separately and checked pretty regularly and she can do 20+ miles in warm weather with this combo. The Fatpacks gotta be 3-4 years old and the Ping is over 4 years old. We ride almost daily. The paralleling really seems to extend longevity, since on occasion she forgets to connect the parallel Andersons and she runs out of juice at about 16 miles. I think it is the Ping BMS shutting down. She then does a no-no and connects the parallel connectors, but only the Fatpacks have voltage, and the BMS from the ping must prevent the overcharge and shut down. Anyhoo it works fine and she has plenty for juice for the final 4 miles from the untouched Fatpacks. I think those Konions were 1 amp apiece and paralleled, so about 12 ah from the Fatpacks.
I hope this helps a bit.
otherDoc

 

[icerider]

I've run 11s 20C LIPO in parallel to 11s6p ~2C LiMn 18650 for several months now. Daily bulk charge and weekly balance check and re-balance on the LIPOs

So far so good. I agree with your speculation that at high current draw, most of the current comes from the LIPO and the LiMn recharge the LIPO a little when the current draw drops again.

My arrangement would let me check that speculation if I were brave enough to run a couple of long wires up to an amp meter -- sounds dangerous, but perhaps I'll try it just once to check.

 

[amberwolf]

I wish I could find the old CrazyBike2 post(s?) about it, but I haven't been able to locat them in several minutes of searching. And I can't remmember the results.

Back when I first got the 14s1p 20Ah EIG NMC pack, and the RC LiPo packs to make into the 14s2p 10Ah ammocan pack of them, I did a test using two Turnigy Watt Meters, one on each pack, plus the CA monitoring teh whole thing, with the packs paralled on the CA side of the watt meters, so I could see how much current and voltage drop each pack would see, to find out if they were different.

If I can fix one of the TWMs (I have two dead ones and one working) I could repeat the test, though that RC LiPo pack is on it's last legs now so it won't contribute a lot.

 

[Hillhater]

Isn't this just a very mild version of the "Hybrid Battery" pack that was debated last year.
There are several guys over on DIY Ecar doing this very successfully with LiPo " booster" packs in parallel with lead or Lifpo4 main packs.
Makes charging and balancing a little complicated, but no major issues.

 

[John in CR]

Isn't this just a very mild version of the "Hybrid Battery" pack that was debated last year.
There are several guys over on DIY Ecar doing this very successfully with LiPo " booster" packs in parallel with lead or Lifpo4 main packs.
Makes charging and balancing a little complicated, but no major issues.

I've always liked the Lifepo/Lipo combination. Get the number of series cells right and the Lipos deliver the juice early in the discharge, which is mostly low capacity surface charge for the lifepo4, and then you hit the brick wall of lifepo4's discharge voltage (preventing deep discharge of the lipo), but when you hit Lifepo4's cliff at 3V/cell you still have some get to a plug juice left from the lipo. I don't deep discharges, but I do like the energy up top of the lipo and then the flat voltage at lifepo4 nominal.

 

[dnmun]

if you have two series of resistors of the same value in each one of the series and the resistor values are different between the two series, then the expected voltage of each of the nodes, or cells, would be identical to each other.

assuming they are connected at the terminals and there is the same node count in each series string.

 

[John in CR]

if you have two series of resistors of the same value in each one of the series and the resistor values are different between the two series, then the expected voltage of each of the nodes, or cells, would be identical to each other.

assuming they are connected at the terminals and there is the same node count in each series string.

Yes, but it's not quite that simple is it? Since the higher C-rate lower resistance string handles more current, then they end up at a lower SOC, so when you stop drawing current there must be a cross current between the 2 strings to bring them back to equilibrium, essentially to the same SOC. My question is how big are those cross currents?

 

[dnmun]

C rate has nothing to do with it. the only variation which would lead to current in the parallel connection would be because of variation of the resistance from the mean of that element between the nodes in the series.