Same power, different voltage. Pros, cons, no difference?

[miuan]

....that the 48V version will provide electrical assist for pedaling up to a greater speed.

Right. Which brings us back to the "economy" or "efficiency" thing. If the OP really wishes to preserve as much battery power as possible, it makes little sense to use the motor assist while he can maintain the top speed of 36V setup using solely his leg power. Traveling any faster than 36V no load speed, the efficiency of 48V would sure be good, but aerodynamic drag overcomes any such benefits by a fat margin and makes for a more inefficient ride.

I use this rule when I'm left with low juice in the middle of nowhere:

You basically have to travel a given distance using your power and the power left in batteries. So you just pedal all the time, and should your actual speed drop below your desired speed, you use the assist. This is how you get the farthest.
On a mild downhill, all you need is light/no pedaling to keep the efficient speed. On short steep downhills, you don't pedal at all, instead you activate regen (if you have one) and save your leg stamina for the next section.
On the uphill, of course, you use full leg AND motor power to get close to (but not over) your desired speed. This is the least convenient time to save power, simply because aero drag is still low, and being too slow on uphill section will always result in bigger time penalty than having been "too slow" on light downhills.

I believe this is how the greatest ratio of
RANGE / (Ah * available time)
can be achieved, along with active pedaling.

 

[dogman dan]

Yes, the reason to use 48v would be for better speed used for a short time when needed. Not for efficiency reasons. The kind of coat you wear, or how high you set your handlebars is going to affect efficiency much much more than 36v 250 watts vs 48v 250 watts.

My methods aren't so scientific, but I went out and did the thousands of miles on many kinds of motors. One thing I found out to be a fact, The same low wattage on everything produced nearly the same results . Only when bumping wattage up when going higher volts did performance really improve. The dd motors really do work better on 48v, but the little gear motors want 36v.

With no experience Sam, you are thinking 250 watts will do a lot. It would, coming out of your thighs. But on a hubmotor, 250 watts only equals about 100 watts from your thighs. A lot of your wattage is heating the world. The rest is eaten up by drag riding even slightly faster than usual.

While I was writing Miuan posted, I couldn't agree more with him. Great description of how to hypermile. It's not limiting to 250 watts, it's using 500 when you need it, and 0 to 100 watts when you can. Gearmotors rock for this kind of riding, even if regen is impossible. Riding just using a trickle of power takes you amazing distance when you don't want more speed than normal pedal speeds.

 

[SamTexas]

Thank you all for your views and opinions. When the hub motor arrives, I will use it first at 36V and then at 48V while MAINTAINING THE POWER LEVEL ALL THE TIME and see if there is any measurable difference.

Sam

 

[rscamp]

As liveforphysics says, if the motor wind is changed to exactly compensate for the increased voltage (i.e. more turns of finer wire with the same fill resulting in the same output RPM) then the performance remains pretty much identical. In the ebike realm, you can see this for yourself in the simulator at ebikes.ca. Try a motor with different winds and use voltages that result in the same output speed...

Using a higher voltage alone shifts the efficiency vs power output curve to the right - i.e. maximum efficiency occurs at a higher power output. However, this also occurs at a correspondingly higher RPM. Whether this is what is desired or needed depends on the setup and practical limitations.

 

[SamTexas]

As liveforphysics says, if the motor wind is changed to ...

But it is not. This is the exact same motor, exact same wheel, exact same power consumed, exact same controller, .... The only difference is the voltage it's run on.

 

[jbond]

Using a higher voltage alone shifts the efficiency vs power output curve to the right - i.e. maximum efficiency occurs at a higher power output. However, this also occurs at a correspondingly higher RPM. Whether this is what is desired or needed depends on the setup and practical limitations.

This is true for the motor with no current limiting, but is it true for the motor+controller combination? And that's the core of the OP's question. Set the current limit in the two cases so the max output is the same. Now how do the curves compare for the whole system?

I think the 48v system is going to be a wider curve with a higher max rpm. But almost all of that is going to be in current limiting. In the region where current limiting is in effect in both cases is there really any significant difference in efficiency? I suspect there's a small region corresponding to ~13-15mph where the 36v system is running with Full PWM and max efficiency but the 48v system is chopping the PWM to an average motor voltage of 36v. The only obvious source of a difference is losses in the controller. Below that, both systems will be reducing the motor voltage to below 36v to keep the current inside the limit. The 48v system will just have a lower PWM duty cycle.

There's something in here which is confusing me. Take an overall efficiency (motor+controller) of 80%. For 250w of mechanical power at the rim we need ~300w of battery power. That's ~36v-8A or ~48v-6.2A Set those limits in the controller. Now what average voltage-current does the motor see during 10mph hillclimbing in each case? It's the same isn't it? or is it?

I think this is really a question of appropriate controller settings for a typical 250w hub motor at 36v and 48v to avoid burning out anything or stripping the gears. And then once you've done that, comparing the overall system curves. I've seen people running a 36v LiOn 10AHr, Bafang QWSX, e-Crazyman 6FET controller with upgraded FETs and current settings of 18A-45A (battery-Phase) with no over-heating or reliability issues. So what should the battery-phase settings be for the exact same system but a 48v-10AHr battery?

 

[cell_man]

As liveforphysics says, if the motor wind is changed to ...

But it is not. This is the exact same motor, exact same wheel, exact same power consumed, exact same controller, .... The only difference is the voltage it's run on.

Which is why I responded to your original post stating that sticking to the rated Voltage is the way to go. The motor was configured for a certain rpm with a certain wheel size with a certain voltage. If all you do is increase the voltage it wants to go faster and the motor is not rated for that higher load it will experience at those higher speeds. If you went up to 48V and fitted that same motor into a 20" wheel and reduced the controller current to about 13A from the standard 17A at 36V you would have basically the same performance and may possibly have some small gains in efficiency, but they would be marginal at best. Will it break the motor upping it from 36 to 48V? probably no, but it will not improve the efficiency and that low cost controller is now supplying significantly more power at a higher voltage in the hot Texas sun.

 

[dogman dan]

Cell man is correct, with bells on. So is LFP.

The only reason to up the voltage would be to up the current. Then you recieve a faster take off, higher top speed and better climbing,,,,, till the smoke comes out.

Back in the real world folks, ya know, riding the bike down the street? If you limit current to 250 watts, you are going to limit speed at both voltages. Regardless of the voltages, aerodynamic drag is going to slow you down to about 15 mph or less. So you'll never see the faster speed of 48v anyway.

If you limit current to 500 watts, then you might see a slight increase in speed at the 48v setting. That's because at 36v you'll max your cruising speed at about 400 watts.

 

[grindz145]

Cell man is correct, with bells on. So is LFP.

The only reason to up the voltage would be to up the current. Then you recieve a faster take off, higher top speed and better climbing,,,,, till the smoke comes out.

Back in the real world folks, ya know, riding the bike down the street? If you limit current to 250 watts, you are going to limit speed at both voltages. Regardless of the voltages, aerodynamic drag is going to slow you down to about 15 mph or less. So you'll never see the faster speed of 48v anyway.

If you limit current to 500 watts, then you might see a slight increase in speed at the 48v setting. That's because at 36v you'll max your cruising speed at about 400 watts.

agreed or a higher top speed (which can be achieved other ways as well).

 

[miuan]

There's something in here which is confusing me. Take an overall efficiency (motor+controller) of 80%. For 250w of mechanical power at the rim we need ~300w of battery power. That's ~36v-8A or ~48v-6.2A Set those limits in the controller. Now what average voltage-current does the motor see during 10mph hillclimbing in each case? It's the same isn't it? or is it?

The motor will see the same regardless of 36/8 or 48/6 due to both ctrlrs modulating the voltages.
Again, the best way to increase eff. is go higher voltage, lower current AND a slower motor.

 

[SamTexas]

Take an overall efficiency (motor+controller) of 80%. For 250w of mechanical power at the rim we need ~300w of battery power. That's ~36v-8A or ~48v-6.2A Set those limits in the controller. Now what average voltage-current does the motor see during 10mph hillclimbing in each case? It's the same isn't it? or is it?

My GUESS would be the same voltage-current, but the bike will move a little faster at 48V (less heat loss). However, the difference might be too small to notice. But if the distance tested is very long (10+ miles) along with proper measuring tools (CA for energy consumption, GPS for speed), maybe (hopefully) the gain would be noticeable.

 

[dogman dan]

Now you are getting what I've been saying. The difference is very small. What kind of coat you are wearing will make more difference.

The big difference if you didn't limit the watts, which under some use patterns is worth it. Maybe you have one short, but very steep hill.

 

[jbond]

Take an overall efficiency (motor+controller) of 80%. For 250w of mechanical power at the rim we need ~300w of battery power. That's ~36v-8A or ~48v-6.2A Set those limits in the controller. Now what average voltage-current does the motor see during 10mph hillclimbing in each case? It's the same isn't it? or is it?

My GUESS would be the same voltage-current, but the bike will move a little faster at 48V (less heat loss). However, the difference might be too small to notice. But if the distance tested is very long (10+ miles) along with proper measuring tools (CA for energy consumption, GPS for speed), maybe (hopefully) the gain would be noticeable.

I think that scenario is a trade off for slightly higher efficiency in the motor when the controller is not current limiting for slightly worse efficiency in the controller when it is. We've moved the non-current limiting range up to a higher speed so you're probably less likely to use it. And to answer my own question above. If the 48v controller has the current setting lower then when in current limiting, I think the motor voltage will also be lower. And with less current, the motor will produce less torque/power.

Re the comments about speed. On a good running bike on the flat and hunched down, it's not hard to reach 20mph. For a while anyway. At that speed the 36v motor will have stopped assisting you at all.Give a gentle slope or a tailwind and a good cyclist can push that up above 20mph. The one advantage of the 48v system is that it will still be adding it's 250w (or whatever) to the mix. If you were struggling to maintain 20mph just pedalling, the motor should make it relatively easy. What it probably won't do is maintain that speed on it's own. On an EU legal bike or setting, with an effective cut off at 25Kmph it happens often enough that you are pedalling above the assist cut off to be annoying. Even the guys using the Panasonic system where power is directly related to pedal torque start messing with the gearing to try and get assist at higher speeds. People have tried using a 20" wind in a 26" wheel for the same reasons. You get assist at a higher speed before cut off but you also lose torque at lower speeds.

With all due respect to Cellman, who's got way more experience than me, this question arises because these baby motors are designed for regulations with an annoyingly slow assist max speed and they produce just enough power to be useful. The question is whether the motors can be pushed enough to get a higher max assist speed without reducing their already limited hill climbing ability. It's also a result of the relatively crude controllers. Their current limit is programmable but fixed. Phase current is programmable but fixed and over-current limiting is a guess. And their mode control is a simple speed limiter.
So
1) 48v instead of 36v and the same max current. Push that too far and the motor heats up, gears break. Maybe 48v-20A is ok in cold climates and in bursts.
2) 36v smaller rim motor in a bigger rim (eg 20" in 26") with max current increased. Again, push that too far and the motor heats up, gears break. Maybe 36v-25A is ok in cold climates and in bursts.
3) As designed and just accept it. Maybe push the current up a bit for a bit more acceleration/hill climbing.
4) Push the boundaries with 1) or 2) But build a 2 mode control external to the controller. Economy with a Full power over-ride to discourage you from using Full power and encouraging you to put plenty of effort into the pedals.

 

[motomech]

The limits of the sm. 250 watt geared motors has been well explored here by GaryKard, Russell and others.
Russell melted the phase wires on a 24V GM motor run at 48V 20A.
He also ran a 36V Bafang at 48V 20A w/ no problems.
Broken gears rarely seems to be the weak link.

 

[jbond]

The limits of the sm. 250 watt geared motors has been well explored here by GaryKard, Russell and others.
He also ran a 36V Bafang at 48V 20A w/ no problems.
Broken gears rarely seems to be the weak link.

Really good to know.Thanks. A related question, I guess. How big a 48v LiFePo battery do you need to get 20A? Is a Ping 48v-10AHr too marginal? 48v-15AHr or 48v-20AHr is a *lot* of battery. The other alternative is an A123 based pack from cellman. I've heard of one setup of 16s4p of A123 from him for a total of 48v 9.2ah.

 

[dogman dan]

In pings, or similar lifepo4 batteries, I like to recomend the 15 ah size minimum, unless they are using 15 amp controllers. It's not that a 10 won't work, it's just that the larger the better, for overall lifespan and for simply having a more usefull bike if the range is longer. The idea is to keep max c rate under 2, and average c rate closer to 1. They should last much longer used that way, mine have for sure.

48v 15 ah, and 36v 20 ah are good sizes because they aren't too heavy to carry well on rear racks, may even still fit in the triangle, and fit in some rear rack bags. Once you go 48v 20 ah, the size gets cumbersome.