Paralleling batteries of different Ah?

[Crash Machine]

Sorry to be such a newbie. I tried to read as many threads as possible before I posted this. My Fusin battery is dying. Steve will gladly send me another but what for? I need a dependable battery. I already have the Ping 48v 10 Ah and it rocks. I was going to get the 48v 15 Ah Ping.It takes about 20Ah to do all my running around. This would leave me 5Ah to spare. I contacted Ping and he suggested hooking the two up in parallel. He even offered to send me the diodes to do it. Great guy. So I'm wondering if this is a good idea? Has anyone ever done it? Will the difference in Ah be a problem or are there really any advantages to doing this? Is Ping even my best choice? Thank you for any input. Long live Endless Sphere...

 

[grindz145]

The diodes are a workable solution. This is so that you don't end up charging one battery with the other in an uncontrolled way. Especially if they are not ideal diodes and even if they are, there will be losses associated with using diodes. Since you already have the investment and the older set still works, I don't see why this would be a bad setup to have. I doubt the difference in capacity will cause an issue, the current should find an equilibrium and even when the one pack is completely depleted the other pack should still be enough on it's own.

 

[swbluto]

What will happen is they won't actually discharge evenly due to differing internal resistances and capacities (The 15Ah will discharge more current due to a lower internal resistance), but as long as there's a diode, the maximum current any one pack should see would be the same as if only one pack was being used. As long as you don't increase the current limit, you'll be treating both battery packs equal to or better than one alone. The decrease in current per pack will also prolong life as the cells will be at a lower temperature, so it'd actually benefit you. Since the diodes do incur a loss, there'll be a slight loss in voltage performance, but decreasing the internal resistance by doubling the packs should also increase the voltage performance, so you'll probably not notice a difference.

As the 10 Ah decreases, the 15 AH would start picking up the pace until the very end and then they'd extinguish the rest of their respective charges until LVC. I created a simulator for paralleling batteries, once.

 

[Crash Machine]

Wow, a discharge simulator. I feel like a monkey with a math problem when I hear something like that. Thanks you. That was my big concern draining one pack first and damaging it.

 

[dogman dan]

Get the replacement fusin battery for sure. And then paralell it with your ping till you do get another. Use the diodes, of course, and then you will have a 20 ah pack that should make the fusin pack last longer. The guys selling 10 ah, 1c cells packs with a 20 amp controller are making a mistake.

 

[dnmun]

put them in parallel without the diodes.

you do not need diodes.

the lifepo4 packs have a BMS to control the current into the battery.

ping is providing more than an abundance of caution, which i do not think is needed if you have 2 lifepo4 packs in parallel with a BMS on each one.

you would need diodes if you run in parallel with nicads or SLA.

 

[grindz145]

I was thinking about that as well, you really don't need the diodes. The main problem is that you have a pack which has been used and has a different impedance. Otherwise they're all in parallel anyway. As long as the low voltage cutoff and overcurrent protection is sufficient in these packs I wouldn't worry about it. The diodes are more useful with different battery chemistries with different voltages...but then that is a mess....

 

[swbluto]

Ok, all, beware what you're getting into. If you don't parallel with diodes, you will have no control over the cross-currents and the recharging current may exceed the rated charge rate (Especially the 10AH as its charge comes to the end.). You probably won't notice "problems" as it's not exactly immediately obvious, but you'll probably shorten the life of the batteries as you would have the risk of too high of charging current (Resulting in battery over-heating) and the risk of one battery putting out the current for the motor PLUS the current to charge (Unless you're sure the "charging battery" can take these current demands.) which would increase battery heating.

More battery heat = less battery life.

 

[dogman dan]

Yer right, but so is dnmun. Series connections are where diodes are mandatory with a ping bms,( at least the v1 type). An experiment with paralell would be safe enough with no diodes, but a permanent setup would be better with them. Theoretically one battery shouldn't charge the other, they should just discharge at different rates according to the internal resistance as stated before and both end up fully discharged at about the same time. But the real world kinda sneaks up on ya fast, and diodes would be a good idea. If one got discharged to a lower voltage than the other, then it would tend to pull a charge from the other, and you could not controll it, or know it was happening.

 

[Crash Machine]

He's already sent the diodes so I might as well use them I guess. Here's the diagram Ping sent me. Is it really that easy. Black to black. Red to red with the diodes inline on the positive side?
This is of three batteries but still the same idea right?

 

[dnmun]

no you do not need them, they just soak up power and lose energy in heat. you do not have to use them if you have BMS on your lifepo4 pack.

 

[Rassy]

There have been previous discussions concerning a problem that can happen when one of the parallel packs shuts down under load, such as LVC, loose connection, blown fuse, or whatever. Ping, has stated that an (inrush?) of current could blow up the BMS. I don't pretend to understand this level of electronics, but currently I use two 48V 20AH batteries in parallel daily. They are charged individually using the BMS's, but the BMS is bypassed during discharge. Since they are normally discharged to only about 25% DOD I don't worry about LVC.

I would prefer to use the BMS's during discharge, but have been convinced that is only wise if there is a diode protecting each battery (BMS) from a potential inrush of current. Until someone fully explains why Ping is incorrect, I will continue using my batteries with the BMS bypass.

 

[grindz145]

If the guys from PING recommends using diodes, it may be that the Fets on the BMS are junk and they blow up during switching when in parallel. I wouldn't imagine this being a problem but to be safe you might as well diode them.

 

[dnmun]

you need to think through what you said.

if the pack that is low has a low cell and drains it down then the LVC function in the comparator for that low cell will shut down the output FETs and isolate the battery from the adjacent battery so the adjacent battery is not dragged down with it. that is the reason for using isolation diodes.

therefore the use of diodes is not necessary if you have lifepo4 packs in parallel with a BMS on each one.

the use of diodes is essential if you have different chemistries in parallel. such as nicad or SLA.

 

[swbluto]

I don't know the inner workings of ping's particular fets, but perhaps his discharge fets pass current through better forwards than in the reverse direction? I don't know if N-channel fets (Which he likely uses) have this "directional preference", but it seems possible.

Or perhaps the negative voltage generated from the shunt would possibly destroy another component?

Looking up an equivalent circuit for a mosfet, it appears I found this at http://www.necel.com/en/faq/n-mos1.gif .



I think the current normally flows from drain to source, and that the body diode would conduct in the case of a reverse current? So, it seems like the mosfet may generate more heat during reverse conduction. I don't know though - this requires confirmation from an expert.

 

[dnmun]

when a FET is off it is like an insulator and does not conduct current.

the breakdown voltage on the irf3205 power mosfets that ping uses is 55V. this will withstand the full voltage of a fully charged 48V lifepo4 battery pack.

you do not need to use diodes, how hard can that be?

 

[swbluto]

you do not need to use diodes, how hard can that be?

You do not need to live, either. However, there might be some reasons why you might want to. :wink:

Those types of reasons some of us are now exploring more fully.

 

[louispower]

according to my engineer diode makes no sense, what you need to do before parallel is to meausre the actual voltage of the 2 packs, and ensure the voltage difference is below 0.5V between the 2 packs, otherwise the current may burn the battery fuse.

Another thing we would be interested to kow is about the Fusin battery, is its capacity lowing down or anything others happen?

thanks,

 

[Russell]

according to my engineer diode makes no sense, what you need to do before parallel is to meausre the actual voltage of the 2 packs, and ensure the voltage difference is below 0.5V between the 2 packs, otherwise the current may burn the battery fuse.

Another thing we would be interested to kow is about the Fusin battery, is its capacity lowing down or anything others happen?

thanks,

If you are putting identical batteries in parallel and they are charged in parallel then no you don't need diodes, for example I have three 36V Bosch tool batteries wired in parallel. It is wise however to add Schottky diodes to the output of each pack if the battery packs are not matched for voltage, if they are different chemistries or if they have a BMS.

-R

Oh yeah, I think many would like to know how the Fusin batteries perform :wink:

 

[grindz145]

you need to think through what you said.

if the pack that is low has a low cell and drains it down then the LVC function in the comparator for that low cell will shut down the output FETs and isolate the battery from the adjacent battery so the adjacent battery is not dragged down with it. that is the reason for using isolation diodes.

therefore the use of diodes is not necessary if you have lifepo4 packs in parallel with a BMS on each one.

the use of diodes is essential if you have different chemistries in parallel. such as nicad or SLA.

I am just being completely cautious. I have seen battery protection diodes fail on low voltage cutoff when there is poor hysteresis and the fets toggle a bunch of times and die. It's true that this would also be an issue if you did not have another battery in parallel but the hysteresis problem can be worse if there is a voltage on the output and of course there is danger over overcurrent charge from the other battery. Fets do tend to fail safe (open) but I have seen lots and lots that did not! I am not familiar with the Ping BMS and if others feel comfortable that the fets are sized properly and the control circuit is sound then absolutely, dioding the batteries in parallel is not necessary.

 

[Tiberius]

Gentlemen,

You really should use diodes when connecting packs in parallel. This is to stop cross currents flowing if the packs are different voltages.

The cases when they are not necessary are:

1) If there is never any differential voltage, or only a small differential. For instance if the packs are the same chemistry and always charged and discharged together. Beware though that charging them together isn't always possible if they each have their own BMS.

2) If something else stops the cross currents. I have heard it said that some BMS will do this, but I can't confirm that unless I know what's in the BMS circuit. None of the circuits I've seen published will do it, but some people I trust say some of the newer BMS will. The short answer is that some BMS may be ok, but certainly don't assume that every BMS is.

I saw one comment in the thread that a FET is off when its off. Not quite so. When a power FET is off its a diode, so it conducts one way, but with a voltage drop. This is one reason that some of the comments about BMS worry me. If a BMS uses a FET to trip the forward current, either because of low pack voltage or overcurrent, then the FET will still allow reverse current back into the pack.

Suppose then that a discharged pack is connected in parallel to a charged pack. Which BMS will stop the cross current flowing? If its the one in the low pack, then that's a pack that won't take regenerative braking. If its the one in the high pack, then that's equivalent to blowing an overcurrent fuse and the vehicle is disabled.

So, I recommend using diodes.

Yes, I'm not entirely impartial here, I could be accused of bias. The job can be done with Schottky diodes, but I have some Ideal Diodes available.

ES post with links to threads http://www.endless-sphere.com/forums/viewtopic.php?f=9&t=12575
Website with datasheet http://www.re-voltage.eu/electronics.html

Nick

 

[grindz145]

I saw one comment in the thread that a FET is off when its off. Not quite so. When a power FET is off its a diode, so it conducts one way, but with a voltage drop. This is one reason that some of the comments about BMS worry me. If a BMS uses a FET to trip the forward current, either because of low pack voltage or overcurrent, then the FET will still allow reverse current back into the pack.
Nick

Usually in this type of circuit there is a charge fet and a discharge fet biased in opposite directions so as not to allow charging when you don't expect it to but again, I am not familiar with this particular implementation.

 

[Tiberius]

Hi Grindz,

As I said, I'd really like to see a circuit. Then we could see whether the particular BMS really does include this function.

I think its fair to say that many BMS don't, and its telling that Ping seems to be saying use external diodes.

Nick

 

[grindz145]

Exactly Nick,

Which leads me to believe either :

A. The salesman doesn't know what he's doing
or
B. The engineer isn't terribly confident with the BMS setup in this configuration


-Troy

 

[Rassy]

I've been following this discussion with interest. As far as the BMS allowing reverse current when using the discharge leads, IT DOES ALLOW THE CURRENT TO FLOW BACK INTO THE BATTERY. I always check the voltage before hooking my two batteries in parallel. Usually they are very very close or else I will drain the high battery a little before hooking them together. However, I have hooked them together with over 1/2 volt difference, and they always, immediately, balance somewhere between their two voltages. This is not just an average while they are hooked together, because if they are then unhooked, they remain at their new "balanced" voltages. Each battery has a Watts Up in line at the battery.

Edit: Oops, another brain fart from the old guy. When I hook the batteries together in parallel, I BYPASS THE BMS. Because of the warnings, and the fact that I have no Diodes, I have never even tried hooking the batteries together in parallel along with the BMS's. I will try that tomorrow, with a small difference in the voltage, and see if they balance together or remain unbalanced.