Battery thoughts

[michael.i@me.com]

Theoretical question here

First the setup. I have a bike with a 1000w motor. I currently run a 48 volt 20ah Lifepo pack and get pretty decent performance. If I am going up hills at about 35km/h I am pushing about 1600 watts according to my ca. On the flat I reach around 45-48 km/h. I am using an Infineon controller from grin 48v 40 amp (88 volt max)

What I want to know is this. If I purchased an additional 36v 20ah Lifepo pack, and run it in SERIES to make a 84 volt 20 ah battery, how would my numbers most likely change? What I mean is this, would I still push 1600 watts, or would I push closer to double that? Would my max speed increase, or would I just maintain higher speed up hills? Ie less sag?

I guess what I am asking is this...

All things being the same except voltage, what changes? I mean typically of course.

Also, if I understand electric motors and windings, higher voltages are better for them due to less amperage being required to reach the same wattage. Does this hold true for Ebikes?

I am also assuming that my controller can handle this higher voltage too.

Thanks in advance.

 

[DAND214]

48 + 36= 84v, that is correct. But that is not what you will see charged. Off the charger will be over 100v. 48v is a old term used for lead acid cells. 16s x 3.3 nominal is 52.8 and that is not charged.

Controller might live, but not likely.

Speed will increas so if you go faster you will use more watts. If you go the same speed you will see a increase in range as you have increase the tatal watts. Over 1800 watts if the battery is a full 20Ah.

That is also gonna be a lot of weight to carry.

Here comes the others input coming.

Dan

 

[dnmun]

no, the controller will not live. it will be gone in a few seconds.

 

[wesnewell]

I don't know of any controller with an 88V max. However, 88V max nominal is likely. You need to clarify the specs on your controller. What voltage caps does it use. My guess is it has 100V caps and 100V fets. That being the case. adding another 36V in series would put your very maximum battery charge at 100.8V with cells charged to 3.60V. So, it's very unlikely you will blow the controller if it has 100V caps and fets. I don't use lifepo4, but I run 24s rc lipo (100.8V fully charged) in mine all the time, but usually only charge to slightly over 100V. But, I have run it fully charged on more than one occasion without doing any harm to anything. So if my $35 controller can handle it... well, no need to go into that. Adding the 36V pack in series will increase both speed and acceleration as you have guessed. the bike will have considerably more power. But you will also need to limit the use of such power to short periods of time or risk overheating the motor. Speed wise, only expect maybe 10mph more, due to much more wind resistance the faster you go. But you'll certainly have the potential climb hills faster.

 

[dnmun]

this is great. wes will guarantee your 48V controler will work at 103V DC and he will send you one of his $35 controllers if it doesn't.

what a nice guy.

 

[wesnewell]

If his has 100V cap and fets, there's no reason it should not work with 28s lifepo4 charged to 3.60V per cell. But that's not 103V. Just 100.8V. Don't know here you come up with 103V, as that would have cells over charged. Max charge voltage should not be more than 3.60V.

 

[Jeremy Harris]

Bear in mind that the true voltage seen by the FETs and commutation capacitors will be the battery voltage plus the ripple and spikes. This means that if the battery voltage is just over 100V then with one or two volts of ripple (likely more than this if you draw appreciable current and the commutation capacitors are being run right at their maximum rating) will mean that the FETs and capacitors are seeing over their absolute maximum rated voltage at times. The component voltage rating isn't a recommended maximum, with a bit of leeway so you can run right up to, or over, it reliably, it is what is says on the datasheet, an absolute maximum rating.

The FETs and capacitors may take this for a while, they may not. Even if they do reliability will be lower, as there is a strong correlation between operating voltage and reliability. If you want component reliability then you have to derate the absolute maximum voltage rating, often by a fair bit.

My guess is that the 88V rating for this controller is Justin being smart, and derating a controller with 100V components a little bit, as you can be certain he understands the way derating improves reliability. Bigmoose kindly uploaded an MIT paper on derating components for reliability a while ago ( http://snebulos.mit.edu/projects/crater/docs/engineering/EEE-parts-derating.pdf).

I think the reason folk get away with running close to, or over, component absolute maximum ratings is twofold. First, maybe the bikes don't get run for long enough for the reduced reliability to show itself for many people, secondly, battery voltage seen at the controller often drops quite quickly, both from the surface charge burning off and also because of voltage drops in the wiring.

If you choose to run close to, or over, the absolute maximum voltage rating of any components in the controller then you're playing a lottery when it comes to how long it'll last. You may get lucky and get a year or two out of it, you may get unlucky and blow it on day one. As long as you're happy with the risk, then that's fine, it's your choice. If you want to be sure it'll run reasonably reliably then always run it a bit under the absolute maximum voltage rating.

 

[wesnewell]

If he's a worried about it, he could always just add a 24V 20ah pack. Although it's my understanding that lifepo4 cells bleed off rapidly to about 3.5V or less. That would put it well under 100V. And those 1C pouch cells seem to sag considerably under load lowering the voltage even more. If the controller ever saw more than 100V I'd be surprised. So much so that I will put my money where my mouth is and will offer to exchange his controller for one of mine should it blow within 90 days due to using 100.8V on it.

 

[Jeremy Harris]

All I'm doing here is pointing out the facts to the OP, so that he, and others reading this, can make an informed choice.

Some have got away with running at over the absolute maximum rated voltage of the components in their controllers, some haven't, that's clear from reading posts on this forum. If you, the OP or anyone else wants to run over the absolute maximum rating and are happy to ignore the voltage rating that Grin put on this controller then that's fine by me, as it's then your, or the OPs, informed choice, made with some awareness of the risks involved.

The mean time between failure of components decreases with increased voltage, so components that might have lasted for several years could have their operating life shortened to a few months or less if run at, or over their absolute maximum ratings. It all comes down to what people are happy to accept. Some may accept a shorter operating life in return for higher performance, others may want to sacrifice a bit of performance to get better reliability.

 

[Tommy L]

All I'm doing here is pointing out the facts to the OP, so that he, and others reading this, can make an informed choice.

Some have got away with running at over the absolute maximum rated voltage of the components in their controllers, some haven't, that's clear from reading posts on this forum. If you, the OP or anyone else wants to run over the absolute maximum rating and are happy to ignore the voltage rating that Grin put on this controller then that's fine by me, as it's then your, or the OPs, informed choice, made with some awareness of the risks involved.

The mean time between failure of components decreases with increased voltage, so components that might have lasted for several years could have their operating life shortened to a few months or less if run at, or over their absolute maximum ratings. It all comes down to what people are happy to accept. Some may accept a shorter operating life in return for higher performance, others may want to sacrifice a bit of performance to get better reliability.

This is a great topic.
I have run the Clyte at 36v 72v 90v The 90v nominal 28s LiFePO4 is 100.4 off the charger and settles quickly to 96-98v.

But I have now switched to a Lyen 4115 to handle up to 150v and I've fried the hiside of Bank "C" already. I've only used it a few times too
And used it under it's supposed rated use.

This is where LEARNING is interesting. I like to do long rides at 40kph and pedal to get a work out too. Apparently it's harder on a controller to
hold back. I'm guess that all controllers have a "SWEET SPOT" of longevity.

If so, another lesson learned: Size your ebike for what you need it to do.

Any thoughts here?

Tommy L sends....

 

[dogman dan]

In general, more volts = faster spinning motor, more speed. More amps= same max speed or close to it, but it will get to max speed sooner.

So if you were to increase volts, you do increase watts because volts x amps = watts.

Safe enough to say you'd need a 72v controller, which typically has an operating voltage range of 60-90v, though it likely has 100v caps. Above 90v, you risk other components.

So if your new 72v controller had the same amps as your old one, doubling volts would double watts. If you double volts AND double amps, you reallly increase watts.

If you crave speed, getting the same amps 72v controller will work well. Increasing the amps could harm your battery, depending on the c rate of your battery. Then add a 24v battery to operate at the best voltage for that 72v controller. 58v + 28v , about 86v fully charged.

 

[michael.i@me.com]

I love this forum. You guys are awesome. Thanks for all the talk on this, it has been helpful.

Once again you have argued all sides and have helped me make a decision. I am leaning towards adding 24v to my 48v

Thanks again everyone.

Of course when this all goes down expect a thread.

 

[dnmun]

when you add a battery pack on top of another, the positive lead of the lower battery goes to the P- of the BMS of the upper pack.

to charge the packs, you have to disconnect the lower pack from the upper pack and charge them while they are disconnected.

the current that supplies the motor also now has to travel through two sets of output mosfets, first in the top and then the bottom BMS. this eats up more power, and of course you are gonna hear from all the people who demand you use a diode in series and in parallel which eats up even more power.

i have come up with a hack to allow the two packs to be connected permanently but the chargers will be able to charge each section separately and while the battery is still connected directly from the top of #16 to the bottom of #1 on the top pack.

the output mosfets on the lower BMS will be the only place the main current will run so there will not be as much power loss and there will be no need for any of the stupid diodes.

so when you buy your new pack make sure you go with the same source so it has the same BMS on each.

i already figured out how to do it with the v1 and v2.5 signalabs, and also the headway BMS too, and i have also developed a hack to use less of the BMS than it is sourced at. so a 16S BMS can be hacked down to 15,14, as low as you wanna go, say to 9S.

i will eventually get a chance to build it up, but i don't have a need for that big a battery so if you do this we can create a permanent solution for your second pack. it just has to have the same BMS, or i have a buncha BMSs here too i could hack something up for you from.

 

[NeilP]

But I have now switched to a Lyen 4115 to handle up to 150v and I've fried the hiside of Bank "C" already. I've only used it a few times too
And used it under it's supposed rated use.

What pack where you running to run the Lyen 4115?
I am just re thinking what I was going to run my 24 FET 4115 at. Was going to run at 34 or 35S (142 to 145 off the charger) , but now think I should back off a bit.

Just wonder what voltage you ran yours at?

 

[Tommy L]

But I have now switched to a Lyen 4115 to handle up to 150v and I've fried the hiside of Bank "C" already. I've only used it a few times too
And used it under it's supposed rated use.

What pack where you running to run the Lyen 4115?
I am just re thinking what I was going to run my 24 FET 4115 at. Was going to run at 34 or 35S (142 to 145 off the charger) , but now think I should back off a bit.

Just wonder what voltage you ran yours at?

The equivalent to 40s LiFePO4 3.3v per cell 132v Nominal 146v hot off the charger. My 9C 2810 propelled me to 78kph at that voltage
Lyen believes that I had a Hall or Phase not in correct sequence.


Back to the Topic......
The GRIN C7225 controller is 88V max if you want to use regen. If you go over 88v, no regen, but I have successfully ran 100.8v hot off the charger pack and waited until the residual charge diminished to 96v. No regen available up at that voltage. Right now I'm running a GRIN C7225 at 86v hot off the charger and
it is now regen capable. :wink:

Tommy L sends.... \\m//