Paralleling 2-3 (48v 20Ah) PingPing batteries

We need to learn more about this issue. I am currently using two LiFePO4 48V 20AH packs in parallel on my Evader scooter. So far, things seem to work fine, even though one is a YESA pack and one is a Ping pack, but only time will tell. I don't know what the use of diodes would protect against except maybe some kind of reverse current situation through one of the packs when both are nearly fully discharged. That seems unlikely to me.

There is no problem paralleling these batteries. The diodes are suggested to avoid current charging one battery from the other through the power output. Taking the safety of the BMS charging circuit out of the process. If you understand DC electric circuits it's easy to see that these diodes are not needed. As long as the batteries are alwas paralleled it simply is an impossible scenario that the diodes would be engaged.

I currently run a Perm 132 motor (similar to etek) with a 275 amp limit curtis controller. The bike has a 36 volt set of Oddysey SLA batteries (42ah 10 hour rate) and two 36 volt 20 ah ping LiFePO4 batteries. All three batteries are connected in parallel during discharge and charge. The SLA charge set up gets the LiFePO4 batts 96% - 97% full. The ping battery charging leads are paralled together to a seperate plug in, which is used once every two weeks, or so, to bring the LiFePO4 to a full charge and balance the cells through the BMS boards. During this "top up balancing" charge the LiFePO4 and the SLA are disconnected from each other with a double pole 40 amp breaker (one 40amp single pole for each ping battery Linked so that they both must be in the same position). This is necessary only because the higher voltage for the balancing charge would overcharge the SLA's.

The system works great. Limiting the depth of discharge on the SLA's while using them to provide the igh amp spikes that give the bike great performance. The bike is very quick and very strong on hills. Everyday usable range is 30 miles (not conserving on power during acceleration) with the SLA pulling in at 60% DOD. Max range at deep discharge is near 40 miles. And max range could be around 50 miles with much more conservative acceleration. I've clocked around 2,000 miles with this arrangement so far. Onboard charging is 24 amps (1000 watts out of the wall) yeilding much greater range per day with opportunity charging.

Go ahead and parallel them. Protect them from flexing and vibration and ENJOY.

Aerowhatt

There is no problem paralleling these batteries. The diodes are suggested to avoid current charging one battery from the other through the power output. Taking the safety of the BMS charging circuit out of the process. If you understand DC electric circuits it's easy to see that these diodes are not needed. As long as the batteries are alwas paralleled it simply is an impossible scenario that the diodes would be engaged.

USE DIODES. Electricity flows from a higher voltage potential to a lower and in the scenario where the battery's impedances aren't equal(And they do vary) and/or the battery's "actual capacity's" are different or their "discharge voltage" profiles are different(which happens since cells do become unbalanced), one will discharge into the other if there's not a diode to protect against it.

Don't believe me? Connect a 9V battery in parallel with a 1.5V battery. Tell me what you experience.

(By the ways, I find the emboldened quite hilarious.)

We aren't talking about 9v and 1.5V batteries being connected together. What we are talking about is matched voltage batteries that are both charged and discharged while connected in parallel. In the case where one battery has a different voltage sag under load than the other (i.e unequal impedence). The amperages contributed to the load will vary. Their voltages will always be nearly equal however, preventing any significant (for all practical puposes, any at all) current flowing from one of the batteries to the other.

It's a common misconception that you are operating under. The misconception is that paralleled batteries will contribute equal amperage to a load when paralleled. The amount of current contributed to the load is dictated by the voltage "pushing" the amperage. Not by the load. A 10ah battery paralleled with 100ah battery (same chemistry) and feeding a load will not discharge any faster than the larger battery. It will just contribute less to the total amperage required by the load. At any stage of the discharge the SOC (state of charge) of the two very unequal capacity batteries will be virtually the same as the other one.

The only risk one takes paralleling these batteries to handle a higher amp load, than one unit could alone. Is that if one of the batteries were to have major problems the other one's output would be over amped and that BMS would cut off power. Diodes do nothing to help with the symptoms developed by one of the parallel batteries failing in such a manner. Diodes will offer no protection from a problem that doesn't exist and will simply waste power.

"By the ways" I'll settle for correcting your mistake without trying to belittle you as you felt compelled to try to do.

Aerowhatt

Well, I have waited two days for more on this topic. Especially since I am curious about paralleling SLAs and Pings (same Volt) without diodes.

I can understand Ping's warning not to do it and his advice to add diodes if someone chooses to parallel them. His reputation is important for selling his wonderful, less expensive, "duct tape" LiFePO4 batteries. If someone frys one of his batteries while experimenting with them, they may blame his batteries publically, albeit erroneously.

Cant choose pro or con because I realize that I am unknowledgeable about this and most EV topics. But, my heart is with you Aero. Your argument makes sense to me, but so does swbluto's and Ping's. Okay, I have no sense :)

As for "emboldened," I applaud the fearless and daring. Without them, we would still be living in a cave. Or even worse... extinct. So I see "emboldened" as a compliment. I have been called much worse on here.

As for "hilarious," the arrogant, to put it nicely, used to laugh at Einstein. We dont remember their names now, do we? :)

Im feeling emboldened, so I am going outside now to hook up a 9V and AA in parallel with 10' long 2/0 gauge wires to a Anderson SB350A connector AND see what happens. Is 10' a safe enough distance?

Curiosity sometimes kills the Cat.

Ding! Ding! Round two?

JY

I think there was little further debate because it's hard to argue with 2,000 miles of success. Of course by now it's more than 2,000 miles. I also have three eGO cycles running on this SLA paralleled with Ping batteries set up. The eGO has Regen braking which adds another factor to the mix. No problems and no issues with any of the set ups.

Perhaps the best way to think about it is this, Ping is making up his batteries out of 4AH and 5AH cells. So inside that lovely duct tape batteries of yours are 16 sets of 4 to 5 cells connected in parallel with each other to get 20ah. My "opposition" should be wondering why there aren't 64 or more diodes inside your battery packs. There are not. There is no need for them in like voltage batteries that are parallel connected full time.

It's important to take any information on a forum with a grain of salt I'll completely agree. If it helps at all Physics was my major in college. A little more info on my set ups. I do have the ability to disconnect the LiFePO4 from the SLA's. I do this about once a month to allow the opportunity to fully charge and balance the cells in the LiFePO4 seperately without overcharging the SLA's in the process. The worst case scenario that is possible from user error in my set up is that the LiFePO4 doesn't get reconnected to the SLA's before a ride and the SLA's get drained down. Connecting the LiFePO4 at this time would cause an amperage dump into the SLA's. The LiFePO4 BMS protects against overcurrent if the SLA's draw more than the LiFePO4 can safely provide. Bad things could/would happen if the LiFePO4 were depleted and the SLA were fully charged when the batteries were reconnected. Not much chance of this since the SLA's are hard wired to the drive system. Any discharge is going to include the SLA's.

Make sure to wear safety glasses if you do the 9V to AA parallel "experiment". It's comparing apples to oranges though when compared to two 48 volt Pings in parallel.

Aerowhatt

As I understand it, the risk is to the bms, more than the battery itself. When one bms trips, it is supposed to possibly fry the other. That's why the paralell cells on the other side of the bms are ok. Those that use the diodes have not reported problems to my knowledge. Series connecting may act different, but I think done properly paralleling can be ok.

The most important thing is to have enough ah total so that you don't overdo the c rate, and if one bms trips, you need to know, so you don't hammer the remaining packs for lack of knowing one tripped. Maybe some led's on the dash to show the status of each pack, on or off?

As I understand it, the risk is to the bms, more than the battery itself. When one bms trips, it is supposed to possibly fry the other. That's why the paralell cells on the other side of the bms are ok. Those that use the diodes have not reported problems to my knowledge. Series connecting may act different, but I think done properly paralleling can be ok.

The most important thing is to have enough ah total so that you don't overdo the c rate, and if one bms trips, you need to know, so you don't hammer the remaining packs for lack of knowing one tripped. Maybe some led's on the dash to show the status of each pack, on or off?

Yes this is where the confusion comes in. If you were to series connect two ping packs you definately need a properly sized diode in the circuit. One BMS tripping would force an enormous "pressure" for lack of a better word on the Fets. They will blow if the power level was high enough when one BMS dropped out. The diode provides a lower resistence path for the power to pass through sparing the (open) fets on the BMS. A reasonable laymen explanation could compare the tripping of one of the series connected BMS boards to the instant closing of a valve on a pipe carrying a fast flowing fluid. The pressure on that valve instantly attempting to stop the momentum of that flow can break it.

In Parrallel the BMS boards only are subject to the power from the battery they control. If one BMS trips due to hitting current limit then the second BMS will trip in a split second as well. Suddenly the one battery is asked to produce the amperage of both. The BMS quickly opens the circuit. A Diode would do nothing to change this scenario. I've seen this happen multiple times during testing on my motorcycle. The only reason to have a diode between two parallel batteries is to block one from charging or discharging the other. You see this used on small RV's with a starting battery and a deep cycle battery for running accessories while parked. The unit between the two batteries allows the Charging of both batteries from the same charging system but will only allow the deep cycle battery to be discharged running accessories. The starting battery is shielded from providing current by diodes so that it will still be fully charged and able to start the vehicle even though the deep cycle accessory battery is depleated.

Some will argue about voltage surges if one BMS of the two (or more) in parallel trips. None of these possibilities exceed the limits of the BMS capability (not even close). In fact they are transient and significantly smaller than those power levels the BMS handles durring normal operation. If you ask Ping about paralleling his batteries his worry is that one battery charging the other through the power leads and not the protected charging leads of the BMS. Think for a moment of two batteries always connected in parallel. They are "always" at the same potential (voltage). Current always flows from high potential to lower potential. The system continually equalizes. There are extreme cases of mismatched impedence where some small amounts of current will flow from one battery to the other for a few seconds in a parallel arrangement. While the currents flowing from both batteries are in fast transitions by the changing load. But what are we concerned about? Overcharge, overvoltage, Too high of a charging current. These are the only posibilities. None of which are an issue in such a set up. Using a diode in a continually paralleled system will not cause any problems. It also won't solve problems that don't exist either. What they will do is increase voltage drop to the load and waste some power.

Aerowhatt

I agree, a 2000+ mile, longitudinal study is hard to argue with. I wont bring up controls, confounds, bias, false data, etc. :) I do take much of the info and advice given here with a grain of salt, except for everything Link, Reikiman, and Chas say. Sometimes that grain of salt gets in our eyes and burns a little. My major was Psych. Post-doc on Psychometric Properties and Ethnologically Baised Nuerological Implications. Anyone wanna "oppose" me on that topic? :) At least we are playing nice with each other now :)

Okay, If I understand correctly, you guys are saying...
A. Two paralleled Pings in parallel with SLAs
1. The rated voltage of all batteries must be the same.
2. As long as the SLAs are permanently in the circut, paralleled Ping packs can be added in parallel with the SLAs, after the Ping's BMS and without adding additional diodes.
3. Pings of different Ah levels can be paralleled together.
4. The Pings can be charged seperately so they can be "topped off."
5. Do not connect charged Pings in parallel with discharged SLAs.
6. Do not connect charged SLAs in parallel with discharged Pings.
7. The combined Ah total of the Ping packs needs to be large enough to handle high C rates.
8. Each LiFePO4 pack should be monitored to make sure one of the BMS has not tripped.
B. Two seriesed Pings in parallel with SLAs
1. If two Ping packs are put in-series with each other, a properly sized diode should be in the circut of the Pings.

Are these correct?

ps I lied about going to try the 9V/AA experiment. Too chicken! But, I do like licking new 9V batteries though. Must be a Froid-e-in Oral Phase thing?

Aerowatt,

I find this portion of your comment interesting;

The SLA charge set up gets the LiFePO4 batts 96% - 97% full. The ping battery charging leads are paralled together to a seperate plug in, which is used once every two weeks, or so, to bring the LiFePO4 to a full charge and balance the cells through the BMS boards. During this "top up balancing" charge the LiFePO4 and the SLA are disconnected from each other with a double pole 40 amp breaker (one 40amp single pole for each ping battery Linked so that they both must be in the same position). This is necessary only because the higher voltage for the balancing charge would overcharge the SLA's.

What i don't understand, is why do you need a "top up balancing" charge for the LiFePO4 batteries, if they are already charging up to 96+%? I know it must have something to do with the BMS circuit, but I don't understand that part of the process.

Thanks,

Tom

Maybe the bms doesn't start the balancing function till at least one cell reaches a voltage that the sla charger won't acheive but the ping charger will?

I totally agree with the unquestioned posters. Do take all my posts with a grain of salt! I'm often just repeating what Link or somebody I trust said so they don't have to type the noob post over and over. Being an old fart that tires fast, I tend to have a lot of time to waste typing.

If you paralell batteries of different ah capacity, won't the small ones discharge first and then risk overdischarging? Example, a ping 20 ah and a 10 ah paralelled, won't the 10 ah trip when the 20 is still half full? I think the controller is going to pull close to the same amperage out of both packs. If that is the case, then the 10 ah is going to pull twice the c rate as the 20 ah as well. Everything I ever read about ev's says match the batteries in the pack as much as possible.

I know lead and lithium can be paralelled, but I think the real usable ah capacity needs to be similar. Maybe sise the lead so that when the lithium bms trips, you are likely to be at a depth of discharge on the lead that you can live with, like 75% or less?

Thanks Aerowatt, that is the first really understandable to me explanation of what happens when you paralell, or series connect. No suprise it's from a pilot. I only have a small ammount of right stuff, with PP, lighter than air certificate.

I kinda thought that the danger of a paralled bms was a sudden current flow through the bms, going the wrong way, but all reports of fried ones seemed to be series conected to me. If the bms trips on one, there will be a sudden increase in the amps drawn from the other, but that should be ok if the bms trips on the second. If it doesn't, then that pack could experience a higher discharge rate than you would like it to. That is why I thought some kind of dash indicator that a pack tripped would be a nice thing to have. Scooter places have high voltage turn signal bulbs.

Okay, If I understand correctly, you guys are saying...
A. Two paralleled Pings in parallel with SLAs
1. The rated voltage of all batteries must be the same.
2. As long as the SLAs are permanently in the circut, paralleled Ping packs can be added in parallel with the SLAs, after the Ping's BMS and without adding additional diodes.
3. Pings of different Ah levels can be paralleled together.
4. The Pings can be charged seperately so they can be "topped off."
5. Do not connect charged Pings in parallel with discharged SLAs.
6. Do not connect charged SLAs in parallel with discharged Pings.
7. The combined Ah total of the Ping packs needs to be large enough to handle high C rates.
8. Each LiFePO4 pack should be monitored to make sure one of the BMS has not tripped.
B. Two seriesed Pings in parallel with SLAs
1. If two Ping packs are put in-series with each other, a properly sized diode should be in the circut of the Pings.

Are these correct?

ps I lied about going to try the 9V/AA experiment. Too chicken! But, I do like licking new 9V batteries though. Must be a Froid-e-in Oral Phase thing?

A very good synopsis JY. I would only quibble with number 8. While it wouldn't hurt anything at all to monitor each LiFePO4 it would serve much of a purpose. In theory it makes sense but in use when one of the paralleled Ping's drops out due to over current the other drops out in a split second because it's load suddenly doubles (2 batteries) or triples (three batteries). So in practice they all drop out with an overcurrent situation. If one drops out due to overdischarge then you don't have enough battery capacity aboard. If you want them to last you simply don't run them that low.

Aerowhatt

Yeah, Dogman is on the right track with the reason for a periodic balancing charge. The Ping BMS only balances the cells by disapating some power from the ones that get full first while the others continue to charge up to match them. This all happens (and only happens) at a higher voltage than the SLA's finish thier charge at. So if you didn't do a balance charge occationally the battery cells would drift out of balance and cause range/capacity issues. It's this higher voltage required which is why the LiFePO4 must be disconneted from the SLA's during the balancing charge. The voltage needed is too high for them and will cause overcharge damage to them.

Just as an aside the LiFePO4 battery and BMS should last longer in this kind of set up. It is usually less than 100% charged and that reduces cell stress. Plus the BMS isn't balancing on daily charges reducing the wear and tear on it's balancing circuitry by a factor of ~20

Aerowhatt

If you paralell batteries of different ah capacity, won't the small ones discharge first and then risk overdischarging? Example, a ping 20 ah and a 10 ah paralelled, won't the 10 ah trip when the 20 is still half full? I think the controller is going to pull close to the same amperage out of both packs. If that is the case, then the 10 ah is going to pull twice the c rate as the 20 ah as well. Everything I ever read about ev's says match the batteries in the pack as much as possible.

I know lead and lithium can be paralelled, but I think the real usable ah capacity needs to be similar. Maybe sise the lead so that when the lithium bms trips, you are likely to be at a depth of discharge on the lead that you can live with, like 75% or less?

This is a very common misconception. For paralleled batteries they contribute different amounts to the load. Their SOC (voltage) will be equal or nearly equal at any point in the discharge. For instance if you took two identically constructed batteries. One 20ah and one 10ah battery connected in parallel and then connected to a 60 amp load. The 20ah battery will supply 40 amps and the 10ah will provide 20amps in a perfect world.

Now series connection is a completely different case. In that case the same amperage passes through every battery in the string. A mismatched (ah capacity) string will have the capacity of the smallest battery. Battery capacity and type must be matched in series strings.

Aerowhatt

That is fascinating! A good example of how common sense can be stupid. I would have thought the two parallel batteries would surely contribute the same amps, unless the small one tripped since it's bms was set lower.

Here's one that made me suprised. Today I did the first lifepo4 voltage charge on my ping pack in 100 cycles. I tweaked my soneil charger up to 47v from the sla voltage of 44.5. But the pack is so balanced still that it just sat there and did no balancing. It's a Ping 36v version 1.0 Not bad, 100 cycles and still balanced. 95% of the 120 cycles were to about 75%. Im sure that had something do do with it. I've never caught it balancing with any charger since June.

Here's one that made me suprised. Today I did the first lifepo4 voltage charge on my ping pack in 100 cycles. I tweaked my soneil charger up to 47v from the sla voltage of 44.5. But the pack is so balanced still that it just sat there and did no balancing. It's a Ping 36v version 1.0 Not bad, 100 cycles and still balanced. 95% of the 120 cycles were to about 75%. Im sure that had something do do with it. I've never caught it balancing with any charger since June.

I'm not sure if this applies to you or not but you have to be a little careful when using a switch mode charger (like a soniel) for balance charging these batteries. What happens is if the bulk cycle voltage setting on the charger is 47 volts the float voltage will be at least a volt lower.

The ping BMS works as follows. It lets charger power through until higher cells reach a certain 2nd voltage threashold. The resistor is switched in across that battery at the 1st voltage threashold. But that cells voltage will continue to rise since the resistor can only disapate a fraction of the charging current. Once the cell reaches a 2nd higher voltage threashold the charging circuit is opened (disconnected). A soniel charger will see this as the battery disconnected and switch to a lower float voltage output. Once the BMS resistor has reduced the cells voltage below the top 2nd threashold the BMS will reconnect (close) the charging circuit. The soniel then sees a battery with a higher voltage than the float voltage. Or high enough to not re initiate bulk mode.

The Problem here is that. Very little balancing has taken place and the input voltage is now too low (at the float level) to continue balance cycling. One way around this is to set the Float voltage of the charger up to 3.9 volts per cell. You should have a significant period of balancing cycles after the number of cycles you have had on the battery no matter how matched the cells were at assembly. The ones in the middle run warmer and more efficiently because of that warmth. Check your input voltage from the charger after it switches from bulk voltage. The resistors on the board should get warm or hot and when completely balanced they are evenly hot down the whole line just before the BMS opens the charging circuit for a very long period of time due to high (100% full) cell voltages.

Aerowhatt

Thanks for all your help Aerowhatt.
A few more questions if you dont mind?

If two Ping packs are put in-series with each other, how do I determine the proper size diode that should be in the circut of the Pings, and where does it actually go?

Can this paralleling be somewhat carried over to SLAs and Prius Gen III NiMH ebike packs too?

Thanks again!

Thanks again for the good info. At least my lower cells will get a bit more charging if nothing else. Before the tweak, the resting pack was at 44v and now it is at 46, so at least the thing is pretty well topped up. Spec for a full charge is 3.8 so I at least have close to that for an average now. Haven't checked cell by cell in about a month, but figured if the total didn't change I was unlikely to be in too much trouble. At the last check,It was as balanced as I could tell with my cheap voltmeter. All the cells were either resting at 3.4 or 3.5 v. Once the pack is run a bit, the voltage settles in at the normal 44v for most of the discharge. I keep saying I need a real charger from Ping, but as long as the pack tests this good, I feel ok. I think the Ping chargers are just tweaked sla chargers anyway. But maybe tweaked more than I did to the soneil. I don't know.

Thanks for all your help Aerowhatt.
A few more questions if you dont mind?

If two Ping packs are put in-series with each other, how do I determine the proper size diode that should be in the circut of the Pings, and where does it actually go?

Can this paralleling be somewhat carried over to SLAs and Prius Gen III NiMH ebike packs too?

Thanks again!

Here is a link to a thread over on endless sphere which should clear up most questions. I scanned through it and it appears to be right on all accounts.

http://endless-sphere.com/forums/viewtopic.php?f=16&t=1312

You can parallel Nimh and Lead for discharge only. They must be charged seperately

It takes some experimenting with the actual cells used to find the right voltage match up for a good even discharge of the two chemistries. For instance when using the ping packs with SLA the pings resting voltage is right around the float charge voltage of the SLA. The SLA will sag less and deliver more amps during peaks so it all works out just fine (how lucky can we get). However if their resting voltage was equal. The SLA would do too much work at the beginning of the discharge and the LiFePO4 left having to carry more of the load toward the end of discharge. If your amp loads are low this wouldn't be that big of a problem. For me and my applications there is no way the LiFePO4 can handle the amp load peaks so this would present a problem.

Another parallel set up I'm running is LiMn and LiFePO4 together in a "24" volt application. To get them to play well together the LiFePO is 9 cells and the LiMn is only 7 cells. They have to be charge seperately, unlike the LiFePO4 SLA combination.

Aerowhatt

Thanks for the link Aerowhatt (and Fechter). I think I am finally understanding series diodes much better.
I plan on charging different types of chemistries seperately to avoid problems with topping off and my lack of ability.
As for getting them to play well together, Im running 3 monster DC SLAs in series 36V/125Ah. I am hoping they will handle the early hi amp loads of my 400A peak Curtis. I will be adding either a couple of 36V Ping packs or a 36V Prius NiMH custom pack to help pick up the slack toward the end of the SLA's discharge. With the SLA Peukert issues I think the SLAs will need lots of help. Leaning towards one Prius 36V/13Ah, 30s2p because of their higher discharge rates versus Pings. Do you think the NiMH pack will play well together with my SLAs? Or, since Pings resting volts are about the same as SLAs float charge, it would be better to go with a couple of Pings? Apple juice and orange juice? :)

Thanks again!

I don't have any experience paralleling Sla's and Nimh. So I can't predict how well they will discharge together. Generally speaking I'm not a big fan of Nimh and I do have lots of experience with them from electric RC airplanes. I know they work well in the hybrids and did fine in the EV1 and Rav4 ev. The Vectrix uses them and people are seeing issues there.

Based on my experiences I would lean towards the LiFePO4. They would be smaller and lighter for the same AH too.

Aerowhatt

I dont have any experience paralleling any batteries. Not even SLAs. Thanks to your great advice and Sir Link's, I am going to parallel a pair of Ping's 36V/20Ah LiFePO4s with my SLAs. A safer and easier option for my novice level, especially since they come with their own BMS and charger. Its just hard to pass up a Prius 14 module pack for $400. A few more pounds really isnt much of an issue at this point, since I have 168# of SLA batteries aready. The bike tips the scales at ~400# (not including the cat). The issue is that I dont have the experience or confidence to hook up NiMHs in a series/parallel setup with several BMS, chargers, and diodes. My emboldened attempt might become dangerous and hilarious to some. ;)

Hi there !

I have been experimenting with parallel batteries for about 2 weeks on a XB-600 48v electric bicycle. The two batteries pack where from my old luyuan bike (see my blog for details on my conversion), one rated at 48v and 7amp/h and the other one rated at 48v and 14amp/h. I think they are SLA type batteries

I use a switch on the bike to connect the two +48v of each battery together. When I turn on the switch I instantly have more power available for the motor.

But when I come back after a long ride and my batteries are almost empty (slow acceleration, and sometimes controller cutoff because of batteries low voltage), my small battery is still full !! It's like if the 7amp/h battery serve as a condenser and is always feed by the big 14amp/h before going to the motor.

Swbluto said:

USE DIODES. Electricity flows from a higher voltage potential to a lower and in the scenario where the battery's impedances aren't equal(And they do vary) and/or the battery's "actual capacity's" are different or their "discharge voltage" profiles are different(which happens since cells do become unbalanced), one will discharge into the other if there's not a diode to protect against it.

From what I understand....my small battery must have a smaller impedance than the motor....so the current flows from one battery to the other, and then after goes to the motor. Am I wright ?

Measuring battery impedance seem to require lot of equipment that I don't have, but one thing is sure I will try that diode thing. From what I understand this could give me more power, and maybe a 50% increase in range because the small battery will discharge completely !

I found a 40amp diodes at work for 4.25$, and I will be able to put it on the bike this week. I'll come back with the results of my test.

adding diodes will lower your top speed due to the voltage drop across them.
it may even reduce your range due to the added losses (though not significantly)
make sure you put a good sized heat sink on them aswell.

a smaller battery (usually) has a higher impedence.
This is of course not always true as it depends upon the specfic battery.
more likely the smaller battery has a disproportionately higher impedence when seen from the controller relative to the larger battery.

are you wiring the motor across both batteries?
or just the big one, with a couple of wires going out to the smaller battery?
what is the voltage across the batteries?
are the voltages on each battery different?
what guage (or m2) wire are you using?
is there current flowing between the batteries with no load?

Matt

Something is not right in the way you have the batteries wired to the bike or you wouldn't be getting this result. Is it possible that the two batteries are not the same chemistry? What are the fully charged resting voltages of each of the batteries?

In a worst case scenario (if both batts are SLA's) one would read marginally higher than the other at the end of a ride.

Aerowhatt

Thank you for your fast answer !

Here is a schematic of how I plugged my batteries together

I used a dual charger plug that came with the bike (it was used to charge the 2 batteries when they were not in the bike)

The small battery is plugged directly in the "dual charger plug" and the big batterie is plugged in the "dual charger plug" using a 3 feet extension cord coming from under the driver's feet.

Wire gage is something like 14awg
fully charged, the 2 batteries are around 67.5 Volt.
I'm not sure what was the voltage of the 14amp/h pack after my ride(If I remember well it was around 64v) but the small one has topped 70v !

I tought a diode would let the current flow only on one side....I hope it won't cut my performance by limiting the current!

again, thank you !

thats even more surprising.....a 48v battery should never be at 64v, even under charge, and definately nowhere near 70v.

are you sure they are 48v packs and not 60v?
64v for a 60v pack at no load is still fully charged.

Matt

I'm sorry

I gave incorrect numbers for the voltage of the battery pack. My cheap digital multimeter had a near empty battery so this was the problem. Now I can read 47.5v when empty, and about 56v when full(these measures where taken under cold condition, so they maybe a bit high again).

By checking my connection again yesterday I realize that I had not identify which plug of the "dual charger plug" goes to each battery. So maybe that was the problem when the big battery was charging the small one. By connecting the plug to the wrong battery, the +48v of the big battery would go for about 6feets of wires before the junction with the +48v coming from the small batterie....this way the small battery would not even have a foot before the +48v junction !. So maybe my problem was only myself and my modification. lol. Sorry to have bothered you :(

Yesterday I connected each battery pack correctly so that each battery pack has about 3 feet of wires before they get connected together. When I arrived home, they were practically at the same voltage. The small battery alone, heaven full, is not able to push to motor without cutting, so maybe that is part of the problem too.

These are batteries that I resurrected, they are old, and have not been used much in the past years....that is surely part of my problem too....I should rewire the bike with bigger wire and change my battery...maybe next year

So I think there is no need for a diode lol
Thanks you !