An idea for SAG. Boost pack.

[Jestronix]

Playing around on the motorsim, i was thinking about SAG under full load. If i could have a battery pack kick in as sag kicks in I could keep my controller at max volts all the time.

This would take some electronic trickery though, youd need to add extra cells on the fly and then ramp them off. The unit could watch what the main pack voltage is when under no load, and keep aligned with the capacity. Or if the pack is only being punched on WOT, you could size it properly so your always at peak voltage for your controller. LVD and and such would need to be equated in.

on a crude level you could have it kick in and out at certain amps. so your 100v controller when under load at 90v, could get that extra 10v from the boost pack.

Im sure theres a lot of downsides to this setup, LVD over OVD need to be taking into account.

id say the expense and extra weight would be an issue.

 

[dogman dan]

You can very simply just parallel a small low sag battery with a large high sag battery, provided they both charge and discharge to the same voltage.

I've done this with lead and NiCad, and with lithium and lithium. It works. Mostly I've used this trick with lipo, putting a good newer pack in parallel with the really toasted old pack. Then keeping discharge rate still fairly moderate, the result is doubling the capacity I can squeeze out of the really tired old pack. Keeping that nearly dead old pack from sagging so much, really improves the wh I can wring out of it.

So mostly I was trying to fix low capacity, rather than sag itself.

 

[MrDude_1]

You can do it like dogman dan just said... just be sure the batteries are close in voltage before you connect them. IF you connect a low pack with a charged pack, it will try to instantly pull power from the higher one to bring up the lower one. This is not good for either pack, and depending on your connectors, it can overheat them. It can also damage cells.

but if they're close in charge, it works great.

 

[Ohbse]

By paralleling two different lithium chemistries you can get the best of both worlds to a degree, high peak discharge (AKA low sag) as well as more energy density. For instance a large-ish 18650 pack in parallel with a small very high C rate lipo pack can be a great mix. There's no need to reinvent the wheel.

Sag can be predicted if you know the internal impedance of your battery. This will vary with SOC, heat, age etc so not easy to simulate without actual testing. You can calculate the resulting impedance of a hybrid pack in the same way you would calculate parallel resistance. This is not perfectly accurate, but it will give you a good idea of what behavior to expect.

What sort of use case did you have in mind? Battery type/size/discharge requirements?

 

[Jestronix]

Thanks for the info guys.

Just a question around the smaller higher C packs , would they not always be getting drained before the lower c packs ?

If i go WOT, the higher C packs will be supplying most of the amps right, as their voltage is higher as they havent sagged as much. This would be true when ever draw amps were over the sag differential point between the two different C rated packs. so what happens when the higher c pack drains down? I suspect the lower c main packs charge the high C pack. In a way the higher C pack is really getting slammed as its the work horse in the arrangement. After every WOT, the smaller pack would get an in rush of charge from the lower pack, as the higher pack capacity took a hit. This would happen maybe 20 to 40 times on an average ride. These guys are almost working like capacitors. How long they stand up to this kind of abuse is the question.

 

[neptronix]

If you were to take the same battery and double it's amp hours, the internal resistance would half.
If you triple it, it would be cut by a third, etc..
Therefore, you'd be seeing about half or a third of the sag, depending.

See? battery IR isn't unlike resistance of a resistor or wire.

So, add that into your calculations.

That's why i run a huge 20C lipo pack to get 4kW instead of just get nanotech or whatever. I've got more range and equivalent discharge of a nanotech battery that's 1/2-1/3rd the size.

 

[dogman dan]

That need for the packs to be close to the same voltage when you connect up is a big part of why you connect them at the start of the ride, not later with a boost switch.

Start with pack one at 56v for example, ride awhile draining it to 50v, then flip your boost switch and now you connect a 56v pack to a 50v pack. This in effect makes a short circuit, a huge amps spark trying to charge the lower voltage pack in a huge gush. Bad for the wiring, even worse for both batteries.

So you connect both the weak battery and your stiffening strong battery at the start of your ride. Then, you are right, under the heaviest loads, the lowest resistance pack will provide more of the current. This will temporarily result in the stiff pack being at a lower voltage. Then the weaker pack will recharge it, till the volts are equal again.

This is ok though, because the difference in voltage will never get as bad as 6v different. So the gush is less. But it can be a bit hard on the weaker pack, if it is also small. So this works best with a fairly large weak pack, and any size stronger pack. The main thing is you don't want your weak pack to pull to big a c rate. So the surges when you need that extra amps need to be relatively short, and not every 10 seconds or way too often.

What I mean is if you are riding really hard, pulling huge amps every 10-30 seconds or so, you need all your pack to be strong.

But if you need the big amps less often, like a stop sign every mile or two, then a weak pack with a stiffener pack can work great. You'll tend to get more capacity out of the weak pack, if it doesn't get those big amp spikes whacking it.

 

[Jestronix]

Thanks dogman, makes a lot more sense now. Never though of the inrush effect in boosting voltage with the second pack, your right it would cook the lower battery. But wouldnt the motor controller suck power first before the other batteries, as in the resistance from the batteries would be greater. Controller is gasping for power, batteries resist. I wonder if this could work with a very tricky controller, so the two packs are never connected but are mixed in the controller.

 

[neptronix]

No such controller would exist, nor would anyone bother building one.

The correct solution is permanently paralleled batteries, or a battery that is of the appropriate size to handle the load in the first place.

 

[Jestronix]

Ok lets look at this another way. say you have a 100v controller, people talk about their top speeds and power based on a fully charged pack right. I now when im out and about when the pack is near the bottom of its capacity things get pretty boring i would be around 75v. I could have any C rated battery and this will happen. So maybe we need to start looking at something to boost that voltage when its lower, just like MPPT solar chargers , they grab the incoming volts/amps and throw it to the right voltage and amps. So with the right converter, it could monitor the pack voltage keep it at the controllers peak amps and volts all the time. In our case the solar panels are our batteries and the batteries are the motor/controller. I suspect Tesla and such do this kind of thing. With the right bit of kit this could be done, but price would probably put it outside everyday ebikers. Be nice to hit the throttle and get the same power right through the capacity curve.

 

[Ykick]

Now we’re thinking the really weird shit, LOL….

I always think about this in terms of electric vehicle racing - vastly different than ICE car racing.

In the case of the later, ICE race cars start heavy but become lighter as they burn off fuel. The result is that they run faster towards the end of races.

Electric race cars the weight never really changes as Wh burn off and as we all know voltage under load only goes lower and thus vehicle speed (motor RPM).

By the end of a race heat, electric vehicles are now running much slower than when they began the race.

I’m sure there have been methods devised to “scale” power usage over a race so that this behavior is never really noticed. Basically limit the vehicle speed at top charge to about the same as end of charge pack voltage.

But it’s an interesting conundrum and often comes up in my mind when dipping deep into a battery pack?

Fundamental difference that will surely evolve over time.

 

[MrDude_1]

Ok lets look at this another way. say you have a 100v controller, people talk about their top speeds and power based on a fully charged pack right. I now when im out and about when the pack is near the bottom of its capacity things get pretty boring i would be around 75v. I could have any C rated battery and this will happen. So maybe we need to start looking at something to boost that voltage when its lower, just like MPPT solar chargers , they grab the incoming volts/amps and throw it to the right voltage and amps. So with the right converter, it could monitor the pack voltage keep it at the controllers peak amps and volts all the time. In our case the solar panels are our batteries and the batteries are the motor/controller. I suspect Tesla and such do this kind of thing. With the right bit of kit this could be done, but price would probably put it outside everyday ebikers. Be nice to hit the throttle and get the same power right through the capacity curve.

What you want a DC-DC step-up converter built into the controller/inverter. That would be cool, but expensive and inefficient when the battery is full.

 

[Jestronix]

Weight and cost is the big issue, dont know how light you could make a DC-DC with these amps.

 

[Jestronix]

actually i thought more about this today whilst out on a ride. In rush from high C batteries will only happen when they are in parallel, but in series they will just add voltage.

on my 18s pack, i could have a solid state 100amp switch, which in milliseconds could flick the current over to the extra cells, based on throttle position or load. no need for massive dc to dc converters. i guess it would need a little more smarts than just an throttle position, but this could work, im thinking ill get a relay and give it a try when the batteries are half capacity. however finding end of capacity will be harder, but can be calculated

 

[dogman dan]

Yes, you can add another pack in series later to boost voltage.

But the big danger there, is inadvertently discharging the original pack too deep, because you don't notice it's low voltage.

One way is to monitor the main pack separately, like with it's own cell level low voltage alarms. The other way would work with a CA on the bars. Say your first pack is 10 ah 72v. And you know from experience, that discharging it conservatively, you get an actual 9 ah from it. You could ride for 4 or 5 ah, then add your 4v or 8v boost pack that is 5 ah, and then finish your ride with a perky top speed.

How do you know when to stop? You stop at 9 ah. adding the second pack does not change the total ah measured from the first pack in any way. Stop at 9 ah, and your first pack is still not over discharged.

Bear in mind though, in this scenario, the first pack is not helped in any way shape or form by the second pack. It sees the same c rate the whole ride.

It works, I've done this too. I used to ride to Sams club on 48v going down hill for 6 miles. Then add 24v 5ah and ride back up the steep hill, while loaded down heavy, to home. Worked great, I got home at 30 mph, even up the steep hill.

DONT overcomplicate things. No switch, no fancy controller, no converter. Just some Anderson plugs and batteries.

Fun to think about this stuff and discuss it. But keep it simple on your bike. If you are too slow at the end of a run on 72v, run 100v. Problem solved.

 

[MrDude_1]

Fun to think about this stuff and discuss it. But keep it simple on your bike. If you are too slow at the end of a run on 72v, run 100v. Problem solved.

Fun to joke, but if I am honest, I need about 115Kw and atleast 3x that in Kw-h. That would give me almost the same power as my literbike (it makes up for being short by having instant power).. and I need atleast 3 hours range of me beating on it.

until then.. everything is a compromise.. usually due to price. lol

 

[wineboyrider]

ggoodrum did a lipo boost for his gem car and used the lipo to lower the sag of lead acid batteries.

 

[Jestronix]

I like the idea of being able to add extra cells with the click of a switch from the handle bars . Some beefy solid states like these would work. If i work the maths, I should know when to turn it on and when the other packs are low. I dont wanna have to get off and fudge around with connectors. Tricky bit would be how to automate all of this ?

The new Crydom SSRs include:

D1D60 – 60 Amps @ 100 VDC
D1D80 – 80 Amps @ 100 VDC
D1D100 – 100 Amps @ 100 VDC
D2D40 – 40 Amps @ 200 VDC

 

[dogman dan]

You'd have to have some beefy expensive contactors.

I was saying to you Jestronix, just get a 100v controller. Then tone it down anytime you need it less perky and fast with a three speed switch. There you go, you'll have what you wanted on a handlebar switch. Cheap, available, off the shelf shit.

It could still be two 100v batteries, one low c rate affordable cells, the other high rate cells.

For me it worked fine to just anticipate my needs. To the store is downhill all the way. Coming back, I'm off the bike already, so not a burden to just plug in the boost pack when leaving the store for the ride home.

The reason I don't do this anymore, is a bigger motor made boosting above 48v completely un needed. The bike flies up the hill fine on 48v now.

Again, the KISS method.

 

[dogman dan]

Mr Dude, I have no idea how to fill you needs. 115,000 watts? I suspect you meant 11.5 kw. At some point the only solution is a big expensive battery that is entirely made of high c rate cells.

And like the previous advice, figure out what your speed needs to be even when the battery is empty, and choose a voltage that works for that speed. Then enjoy being even faster when your battery is fuller.

Technologically, so much easier to tone down overkill, ( amps limiting) than to make the battery voltage actually increase during the ride. Overkill it, then if you want a turbo mode switch, just ride in the toned down mode most of the time. Toned down is relative, the slow mode can still be faster than a scalded cat.

 

[gogo]

And like the previous advice, figure out what your speed needs to be even when the battery is empty, and choose a voltage that works for that speed. Then enjoy being even faster when your battery is fuller.

Technologically, so much easier to tone down overkill, ( amps limiting) than to make the battery voltage actually increase during the ride. Overkill it, then if you want a turbo mode switch, just ride in the toned down mode most of the time. Toned down is relative, the slow mode can still be faster than a scalded cat.

Amen.

 

[Jestronix]

Getting a 100v controller and winding it back is a nice idea, I just know in practice I'd leave it on full power and want that extra cell pack at some point, I'm a sucker for more power , don't worry it's an ego thing, it's like money and sex , more is better or so it seems I should just be content, but my eye keeps looking over at that 14kw adaptto. Trouble with the adaptto is I'd constantly hit the heat limit.

 

[Ohbse]

I like the idea of being able to add extra cells with the click of a switch from the handle bars . Some beefy solid states like these would work. If i work the maths, I should know when to turn it on and when the other packs are low. I dont wanna have to get off and fudge around with connectors. Tricky bit would be how to automate all of this ?

Carrying around extra cells and adding them back in at some point makes zero sense. You're adding a lot of complexity, many failure points, extra weight and getting what? The ability to go slow, then go slightly faster. Or - you could just have an adequately sized pack, a motor with the correct winding and a controller that can achieve your desired speed across the full range of SOC/voltage. This is cheaper, more reliable, delivers more performance, better efficiency and less complexity.

More speed does not necessitate more voltage. More speed requires more power. Power is simply volts x amps. Motor KV determines maximum speed when presented with a voltage. There is no difference between a motor delivering 1000w @ 48v/21a with a kv of 12 with a motor delivering 1000w @ 96v/10.5a with a kv of 6. One requires twice the amps, but half the volts, speed and outcome are the same.

A controller takes battery voltage and bucks this down to drive your motor. It's pointless to add another voltage converter in series with your existing voltage converter in order to overcome an arbitrary limitation you have built into your system. If you want to go faster than the natural speed of your battery voltage x the KV of your motor then you either increase the voltage or you purchase the correct motor. Selectively adding voltage can *only* be worse than having these basic variables correct because it is more complex and less efficient with no upside compared with the alternative.

A good controller can overcome an incorrect voltage or motor kv by increase speed at the expense of torque. Once again however, this is less efficient than simply having the voltage/kv relationship correct in the first place.

 

[Jestronix]

wheres the fun in that i like to tinker, plus i buy a controller, i want max amps max volts all the way through, i just cant fork out for the 14kw adaptto right now. But you are right, do it once, do it right.

 

[dogman dan]

Well, it's true, once you adjust to 100v, you won't want to ride much with the 3 speed switch set to 2.

But for sure, if nearly empty 72v is not enough rpm, then you do need to just up the ante.

When I was doing this, it was more the situation the thread started with. I had a 2000w motor on the bike, and the 48v 15 ah ping battery was far too weak to run it back up the hill. Adding 24v to the pack allowed me to run half throttle up the steep part back, thus the ping only saw 20 amps, but still had a full 1500w to get up the steep part. Then the milder part of the hill could cruise 30 mph, only pulling lightly on the ping, also using only 1500w.

That situation was only temporary, and now I can put two not so strong batteries together, and pull the 2000w at 48v wherever I may be riding. 2c discharge rate from 30 ah is still 60 amps. So no problem with my 40 amps rig.