dual voltage for more speed

[James28]

I've been running a bafang hub motor on 36v for a while now and it gives me reasonable low speed performance and tops out around 20mph which is ok but I would like a higher top speed.
If I upgrade my battery to 48v this will give me a higher top speed but I think I'm right in saying this will effectively move the efficiency range to higher rpm and reduce my low speed performance (heating up and straining the motor).
I have considered a mid drive but my commute involves quite a bit of flood water and I'm not great at keeping on top of drivetrain maintenance and i like how reliable my hub motor is.

I've been wondering for a while whether it is possible to put a 12v battery on my bike that I can turn on with a switch when I'm up to the top speed for 36v (20mph) to make it 48v and allow the motor to spin faster. Effectively giving the bike a second gear without straining the motor at low speed.
I'm pretty sure this would work by using a SPDT relay to add the 12v battery to the 36v battery to supply the controller with 48v.

My controller is up to it, its a 25A 36v/48v kt controller with 63v capacitors in it so it will survive the voltage.

I will be treating the 36v and 12 v batteries separately so they will have their own BMS and I will charge them separately.

I'm just after more electronically experienced people to tell me if this is viable or impossible before I waste time and money in trying it out.

 

[saneagle]

You have a good understanding of how it all works. What you want to do will work, but there's a couple of things to look out for. When you connect a battery to the controller with a switch, you get a large inrush current that burns the contacts in most switches, so better to just use a connector. It will spark when you join it, but it's nothing to worry about. That's assuming that you have no switch on your main battery. If you have one, there'll be no spark as long as it's switched off.

Next point is to get the controller to recognise the correct voltage. It will get confused unless you connect a charged 48v battery or a discharged 36v, asuming that it has automatic voltage detection.

Finally, you'll no longer have reliable low voltage cut-off on either battery, so it's best to use two seperate voltage displays that are visible while riding.

Considering everything, it's a lot simpler and more convenient to just get a 48v battery. A pretty good one of 20Ah can be had from Greenlance for about £280, which would be plug and play.

You won't notice any loss of efficiency when going up to 48v because the 30% increase in torque will raise your RPM to a more efficient level. It's only a problem if you already have a high-speed motor.

 

[AntonyC]

If you were to ride the 48V version of your bike alongside the 36V version i.e. used the same way they'd have about the same efficiency and you wouldn't lose low speed performance, the controller and motor would act together like a 48 to 36 volt converter. It'll do what the 36V rig does about equally but can surpass that when asked.

So just wire in a 12V battery, charging separately is good, and as saneagle says save a lot of hassles. But take note of this bit:

Finally, you'll no longer have reliable low voltage cut-off on either battery, so it's best to use two seperate voltage displays that are visible while riding.

If the batteries have the same Ah capacity you'd get 48/36 more range you might think, and ideally when either battery hits LVC the controller should LVC too as it sees the battery as 48V. In practice that's unreliable and when one battery goes LVC there's a risk of damage by reverse charging from the other, so you'd need to keep an eagle eye on the meters and stop well short of fully discharged.

The solution is to put a diode and heatsink in parallel with each battery, each diode rated for the other battery's voltage at max current (>25A). That protects them from reverse charging and gets the full range back.

KBPC3506 Bridge Rectifier Diode 35A 600V

 

[Peter.Bridge]

Is your motor the Bafang G020 (H400b) ? I have a 48V G020 (winding code 12) on a 26 inch wheel. With a full battery, unrestricted no load speed is 28 mph. I (100kg ish ) travel round at legalish speeds and had no problem with it at lower speeds (including a 1 mile hill which has a small section of 28% incline at the end - where I am going 5 mph , with quite a bit of pedalling effort I manage to get up!) Not had any problem with overheating

I think the only thing that @Woosh said was that the dedicated 48v motor has heavier wire gauge than the 36v

 

[James28]

You have a good understanding of how it all works. What you want to do will work, but there's a couple of things to look out for. When you connect a battery to the controller with a switch, you get a large inrush current that burns the contacts in most switches, so better to just use a connector. It will spark when you join it, but it's nothing to worry about. That's assuming that you have no switch on your main battery. If you have one, there'll be no spark as long as it's switched off.

Next point is to get the controller to recognise the correct voltage. It will get confused unless you connect a charged 48v battery or a discharged 36v, asuming that it has automatic voltage detection.

Finally, you'll no longer have reliable low voltage cut-off on either battery, so it's best to use two seperate voltage displays that are visible while riding.

Considering everything, it's a lot simpler and more convenient to just get a 48v battery. A pretty good one of 20Ah can be had from Greenlance for about £280, which would be plug and play.

You won't notice any loss of efficiency when going up to 48v because the 30% increase in torque will raise your RPM to a more efficient level. It's only a problem if you already have a high-speed motor.

I do have a switch with a soft start on my main battery at the moment. It has survived a good few hundred uses even after I stopped using the soft start, although I've never switched it on/off under load. the idea is to use something like a 60v 20a relay so I can switch under load.

The point you make about the controller I will certainly need to test, I'm using a 36V LifePo4 so full charge is slightly higher than li-ion. there would be no point in trying this if the controller has a problem with the sudden voltage change.

I'd never rely on the LVC anyway 2.5v per cell on LifePO4 is just to low for my liking so I watch the voltage while I'm riding anyway.

If 48v wont loose me efficiency at low end then I might as well just run it as 48v. I just don't want to subject my 250w motor to any more undue stress as I'm already putting 650W through it already.

 

[James28]

If you were to ride the 48V version of your bike alongside the 36V version i.e. used the same way they'd have about the same efficiency and you wouldn't lose low speed performance, the controller and motor would act together like a 48 to 36 volt converter. It'll do what the 36V rig does about equally but can surpass that when asked.

So just wire in a 12V battery, charging separately is good, and as saneagle says save a lot of hassles. But take note of this bit:

If the batteries have the same Ah capacity you'd get 48/36 more range you might think, and ideally when either battery hits LVC the controller should LVC too as it sees the battery as 48V. In practice that's unreliable and when one battery goes LVC there's a risk of damage by reverse charging from the other, so you'd need to keep an eagle eye on the meters and stop well short of fully discharged.

The solution is to put a diode and heatsink in parallel with each battery, each diode rated for the other battery's voltage at max current (>25A). That protects them from reverse charging and gets the full range back.

KBPC3506 Bridge Rectifier Diode 35A 600V

I keep a close eye on my voltage already, I was going to wire it so it would effectively just series all the cells together from 12 to 16

Its a good point, I would end up with 12 cells at lower voltage than the extra 4 I'm adding if I only use the extra 4 part time but I though cells would only balance (back charge) if they were connected in parallel and not series.

My battery is 12s1p configuration by the way.

 

[saneagle]

I do have a switch with a soft start on my main battery at the moment. It has survived a good few hundred uses even after I stopped using the soft start, although I've never switched it on/off under load. the idea is to use something like a 60v 20a relay so I can switch under load.

The point you make about the controller I will certainly need to test, I'm using a 36V LifePo4 so full charge is slightly higher than li-ion. there would be no point in trying this if the controller has a problem with the sudden voltage change.

I'd never rely on the LVC anyway 2.5v per cell on LifePO4 is just to low for my liking so I watch the voltage while I'm riding anyway.

If 48v wont loose me efficiency at low end then I might as well just run it as 48v. I just don't want to subject my 250w motor to any more undue stress as I'm already putting 650W through it already.

The soft switch you have on the battery is already like a relay. It switches the BMS on/off, and the BMS switches the power using mosfets.

Unless it's a small motor, it should be able to handle 17 amps at 48v . Your 18A would probably be OK as long as you don't have circumstances that would cause it to slog slowly up a hill, bearing in mind that you'd have 30% more torque.

Which controller do you have?

 

[James28]

Is your motor the Bafang G020 (H400b) ? I have a 48V G020 (winding code 12) on a 26 inch wheel. With a full battery, unrestricted no load speed is 28 mph. I (100kg ish ) travel round at legalish speeds and had no problem with it at lower speeds (including a 1 mile hill which has a small section of 28% incline at the end - where I am going 5 mph , with quite a bit of pedalling effort I manage to get up!) Not had any problem with overheating

I think the only thing that @Woosh said was that the dedicated 48v motor has heavier wire gauge than the 36v

It is a G020 hub winding code 10 so its a little quicker but it is 36 v so 48v might put the rpm up quite a bit more than a 48v code 12.

I'm glad to here overheating is not a problem but would a motor with a lower winding speed have an easier time getting up a hill as apposed to a higher rpm motor. This is essentially what I'm trying to achieve by being able to switch the motors efficient rpm by changing the voltage on the fly.

 

[James28]

The soft switch you have on the battery is already like a relay. It switches the BMS on/off, and the BMS switches the power using mosfets.

Unless it's a small motor, it should be able to handle 17 amps at 48v . Your 18A would probably be OK as long as you don't have circumstances that would cause it to slog slowly up a hill, bearing in mind that you'd have 30% more torque.

Which controller do you have?

All my switch does is disconnect the battery from the controller.
LifePo4 ebike full build | Pedelecs - Electric Bike Community
There's my build if you were interested.

this is sounding more and more like I should series in another 4 cells and see how 48v runs compared to the 36v at slower speeds

 

[saneagle]

All my switch does is disconnect the battery from the controller.
LifePo4 ebike full build | Pedelecs - Electric Bike Community
There's my build if you were interested.

this is sounding more and more like I should series in another 4 cells and see how 48v runs compared to the 36v at slower speeds

Try it out first. Code 10 is a little fast for 48v but it should be OK as long as you're not a heavyweight. Maybe only use up to level 4 until the speed builds up and make sure you have your controller set to current control (P3=1).

 

[Peter.Bridge]

It is a G020 hub winding code 10 so its a little quicker but it is 36 v so 48v might put the rpm up quite a bit more than a 48v code 12.

I'm glad to here overheating is not a problem but would a motor with a lower winding speed have an easier time getting up a hill as apposed to a higher rpm motor. This is essentially what I'm trying to achieve by being able to switch the motors efficient rpm by changing the voltage on the fly.

Yes, I think a code 12 motor would be more 'in the power band" up a hill than a code 10 because it has a lower max speed.

Your build is 700c wheel too ? Which will make it faster no load max speed than mine (26 inch) but up a steep hill won't be as good.

I don't know whether running at 36v it would be better up hills than running at 48v - I understand your logic that the motor will be at a more efficient rpm. My thought experiment is I can go up hills much better with a fully charged battery than a nearly empty battery , so a higher voltage helps going up hills ?

Are there 43v batteries available ?

 

[saneagle]

Something doesn't add up. A code 10 is 310 rpm at 36v, which would give a no-load speed of 25 mph with a 700c wheel, and 10% faster if the battery is fully charged. I'd expect to see more than 20 mph with a fully charged battery - more like 23 mph.

 

[AntonyC]

I don't know whether running at 36v it would be better up hills than running at 48v - I understand your logic that the motor will be at a more efficient rpm.

It needn't be, if he backs off to climb a modest hill at the same speed as before the efficiency will stay the same, because when the hill levels out both top speeds will be the same. This may not hold true if the PAS levels don't allow such fine control or if climbing near the motor's gradient limit, but 48V can potentially be used almost as efficiently as 36V.

My thought experiment is I can go up hills much better with a fully charged battery than a nearly empty battery , so a higher voltage helps going up hills ?

It does, because it results in more current and torque, unless climbing steeply where the controller's current limit steps in.

 

[Peter.Bridge]

If my calculations are correct :

(so potentially there is quite a big difference on the max no load speed on a 48V code 12 on 26 inch wheel and 48V code 10 on 700C)

 

[Sturmey]

There is a simple if perhaps unrefined way to implement the uninterrupted switching in of extra cells as shown below that can cheaply be made. It does have the advantage that if the extra cells(or 12V battery) cuts out or is removed, the 36V still has a circuit to the motor. The disadvantage is the half volt loss in the diode, which incidentally needs to be mounted on heat sink and rated above 50 v and perhaps 30 or more amps for safety.
As regards the controller LVC, I think you may have to lock the controller to 36v with the 'L' setting. I am not sure how well that will work as the circuit below was only tested with a 36v KT controller and two extra cell. i.e 36 to 44V or 10s to 12s switch.