High Voltage - Low Heat... always?

[xyster]

Current Multiplication means Heat Multiplication

Current Multiplication means Heat Multiplication means Torque Multiplication



Conversely, no current multiplication means no torque multiplication, and the motors in ebike applications then won't accelerate worth beans from low rpms, gears or no gears.

 

[safe]

What is the efficiency of the motor running 48V @ 30A?

The first chart shows the standard controller efficiency at 24V and 48V while running at 30 Amps. The second chart shows the standard controller at 24V and the motor current limited controller at 48V also both being at 30 Amps. Notice how the 24V standard controller motor does have a slight advantage in it's best area in both charts, but overall the motor current limited is just better. What more can I say... the numbers point to this theory being a winner. Only actual testing can prove it.

In some ways I think "current multiplication" is mostly a flaw in the design of the standard controller that people have lived with and learned to like for so long that can't think of doing it any other way.

Habits are hard to break... :wink:

 

[safe]

Conversely, no current multiplication means no torque multiplication, and the motors in ebike applications then won't accelerate worth beans from low rpms, gears or no gears.

No, actually if you downshift you gain the peak power again and can therefore out accelerate. It's ONLY the special case of the hub motor / fixed gear bike where there's any advantage to current multiplication.

Power is "mechanically" torque times gearing, but "electrically" it's current times voltage. You can manipulate every aspect of it!

Just recite the mantra:

"Always replace your current with higher voltage."

Power is a combination of current and voltage. It doesn't care if it's "dirty power" (lots of current) or if it's "clean power". (lot's of voltage) So you just want to find every way possible to weed out the high current situations and boost the higher voltage situations. Motor current limiting tends to improve the quality of the power you create. That's all it's doing...

 

[xyster]

If I remember correctly, Fechter measured a current multiplication between input and output of up to 4x. To replace this lost power, you'd need 4x more voltage. I'd do it myself for my bike, though I don't think selling a 200 volt battery pack, 2-5 amp ebike product (like you're thinking about selling) is generally advisable due to safety issues. And do you know of any ebike-sized motors, brushed or brushless, that can handle 200 volts?

 

[safe]

If I remember correctly, Fechter measured a current multiplication between input and output of up to 4x. To replace this lost power, you'd need 4x more voltage. I'd do it myself for my bike, though I don't think selling a 200 volt battery pack, 2-5 amp ebike product (like you're thinking about selling) is generally advisable due to safety issues. And do you know of any ebike-sized motors, brushed or brushless, that can handle 200 volts?

The efficiency at the 4x level (for current) is pretty bad, so you wouldn't be able to stay in that state for long without overheating. I conceed your argument that there is a limit to how much voltage you can use and in fact the other issue is the rpms that get generated because of increasing voltage. It becomes difficult to gear so many rpms down to something that is usable. But for the "practical" example I've presented of going from 24V to 48V that will give you a 2x peak power boost and if you combine that with motor current limiting you don't add any heat in the process. If you think about the "usable" powerband then this approach will win out.

Don't forget that I tried riding around at 4x (current) for a while on my 750 watt motor and it started to smoke and eventually died. So heat is something that you have to consider. You can't "use" that 4x (current) all day long, but with the 24V to 48V + MCL upgrade you can go 2x all day long... :wink:

In the end "power is power". Power is blind to how it's created and used. You can increase the voltage really high and drop the current really low and spin the motor really fast and then gear the thing way down and get the same power as if you did the opposite of high current, low voltage, low rpms, tall gearing.

Power consists of four things:

1. Current
2. Voltage
3. Motor Rpm
4. Final Gearing

Torque is actually only a "partial" value that does not have a time scale to it's dimensions. (there is no time to the torque metrics) Torque becomes "real" when it is combined with gearing and becomes power. The only thing to care about is final horsepower as long as you cover all the gearing range you need to cover... (so that you don't fall out of the powerband)

 

[safe]

Gearing For Our 250 Watt Example

So what kind of gearing results do we get if we take a 250 watt 24 volt motor with a 30 amp (BCL) controller and give it the "upgrade" to 48 volts, the same 30 amp current limit, but with the motor current limiting (MCL) approach so that the motor doesn't overheat?

Here are the ground rules...

First I wanted to set the gearing so that both bikes could climb roughly the same steepness hill, which I chose to be 10%. I'm using an 8 speed that has a gearing range of 305%. From this baseline the gear ratio automatically determines what the potential top speed can be, but the ability to attain it is suspect in that the little 250 watt motor can't pull more than about 16 mph or so. (8th gear more or less goes to waste) So these are the first gear performance charts that would give you an idea of how steep you can climb. At this point we can say that first gear on both bikes is equal as far as torque.

 

[safe]

Here's where it gets interesting. Now look at the two tables that show the various mph figures for each of the eight gears on the two bikes. The table titles are:

Peak - Mph at which the peak power occurs.
Actual - Mph at which the bike can actually go on flat land.
Efficiency - Mph at which the efficiency is the highest.
Maximum - Mph at which the no load limit occurs.

Now what should shock you is that the 48 volt bike not only can climb equally as steep of a hill in first gear, but it can go 15.9 mph in first gear when the 24 volt bike can only go 9.0 mph in first gear.

Why does this happen?

It's because the gear ratio from the motor to the rear wheel is different on the two bikes. In this hypothical bike the gear ratio for the 24 volt bike is 18 (motor) to 60 (rear hub). For the 48 volt bike the gearing is 18 (motor) to 77 (rear hub)... a much lower gear ratio for first gear. (keep in mind that your own gearing might be totally different and this is just an example) But when you expand the voltage on a motor it also expands the width of the powerband, so while on the one hand you need to gear down a little, the higher maximum rpm of the higher voltage motor allows a higher top speed in first and ALL THE GEARS.

So in the end we have two bikes:

24 volt - 10% hill climbing ability - 16.2 mph actual top speed on the flat. (7th gear)

48 volt - 10% hill climbing ability - 42.2 mph actual top speed on the flat. (8th gear)

Top speed is tied closely to aerodynamics and I'm using my own "Road Racer" values so you would need to adjust this if you were riding something like a mountain bike which has poor aerodynamics.

 

[dirty_d]

I think you're really over complicating things especially current multiplication you don't even have to deal with it, its not like it is forcing the motor to draw more current. the amount if current the motor draws depends solely on the RPM, internal resistance, and voltage across the terminals. I think its far better to just consider the controller as a variable voltage power supply with a current limit(which just works by lowering the voltage).

 

[safe]

I think its far better to just consider the controller as a variable voltage power supply with a current limit(which just works by lowering the voltage).

Nooooooooooo... you can't do that. If there was no current multiplication caused because of the buck converter and the motors inductance then there would be no increased low end torque. You can "feel' current multiplication when you pull down low in the rpms.. you get more torque there than you would if there were a simple resistor as a controller. The buck converter is weird... it allows more current to flow than one would normally expect and that's both good and bad in some ways. It's good for the hub motor user because it adds extra torque for hills, but it's bad in that it does this by adding heat.

The approach I'm suggesting, the "MCL", motor current limiting, basically converts a regular "BCL", battery current limited controller back to what you would hope a variable resistor solution would look like.

In a sense I'm returning to the past... 8)

 

[safe]

Duty Cycle Comparison

Ultimately the motor is completely controlled by what the duty cycle curves look like. That's the actual "control" that we have. The chart shows how the duty cycle compares in the overlapped state that the motor (in it's two configurations) is functioning in. The 48 volt has a much higher maximum rpm and yet the duty cycle rises very, very slowly. The 24 volt jumps right up to a high duty cycle from the start of opening the throttle and it's this behavior that gives the current multiplication we see in the BCL controller.

The duty cycle is really everything... it's just a very fast on-off switch, but the consequences of how that switch behaves is very significant in the power and heat charactoristics of the motor.

So this chart is showing an "rpm to rpm" comparision of how the duty cycle functions in either the 24 Volt BCL or the 48 Volt MCL configurations.

 

[The7]

So in the end we have two bikes:

24 volt - 10% hill climbing ability - 16.2 mph actual top speed on the flat. (7th gear)

48 volt - 10% hill climbing ability - 42.2 mph actual top speed on the flat. (8th gear)

What are the motor currents for these twr bikes?

 

[dirty_d]

yea, like i said its a variable voltage power supply with a current limit, it just draws less current from the battery than is flowing through the motor(just think of it as working like an AC transformer with a constant primary voltage but with a variable number of secondary turns) its not magic, its not comparable at all to a resistor. this is a great thing if your batteries are wimpy and cant put out enough current for starting and hill climbing torque. but for your lead acid batteries its not doing anything but giving you an efficient variable voltage power supply as far as you should be concerned. just pretend the battery side doesn't even exist.

 

[safe]

What are the motor currents for these two bikes?

Both bikes use 30 amps as a current limit, but with MCL that functions differently. The voltage is double for the 48 volt and since power is voltage times current if you keep the current constant and double the voltage you get double the power. (and the same heat)

The trick is in realizing that if you double the voltage in a BCL you get double the current almost by mistake. (as a result of current multiplication) So it's the controller "error" so to speak that actually produces the heat and not the extra voltage.

 

[safe]

but for your lead acid batteries its not doing anything but giving you an efficient variable voltage power supply as far as you should be concerned. just pretend the battery side doesn't even exist.

Like I said before, this is not true.

Listen carefully... I went for probably three months FIGHTING against the realization of current multiplication until I finally succumbed to it's reality. One wishes that things worked more like the MCL and we could pretend that all this strange extra current doesn't exist for the BCL, but it does. (and if your SLA batteries are small enough they can make the current multiplication muted because they sag, but if you have a large battery pack that can handle the full current draw you will know it does exist as a phenonemon)

 

[TylerDurden]

78% efficiency * 1000W = 780W Power to Driveline + 220W WasteHeat

220W wasted to heat is 3x more than the motor can dissipate. (An arithmatic average of the powerband cannot change this.)

Heat will continue to build: reducing the efficiency rapidly, killing the power and the motor; regardless of RPM.

 

[safe]

220W wasted to heat is 3x more than the motor can dissipate. (An arithmatic average of the powerband cannot change this.)

Someone else was struggling with this before... I'd have to go back through the thread to find out who it was. The "rated heat" (the heat that is produced while at the "rated load") is exceeded on pretty much every bike that's out there. Most average about 3x more than the "rated heat" in their stock configuration. So whatever is "stock" is already exceeding the "rated load" most of the time, the upgrade doesn't interfere with the stock configuration and stock heat production reality.

The average for heat stays about the same for the two configurations (24V and 48V) so whatever was true for the stock bike (as far as heat) will also be true for the upgrade.

People don't realize that when they are riding their 250 watt hub motor up a steep hill that if they aren't doing most of the pedaling that the motor is creating 3x more heat than the "rated load" suggests.

This is reality folks...

Reviewing the stock motor chart below we see that the average heat is right about at the same level as peak power. So if you are climbing a moderately steep hill you will likely be using the average heat or worse.

 

[The7]

Safe says:
High Voltage - Low Heat... always?



I would like to support this safe saying (but not all his claims) on the following assumptions:
1) The motor could work at that high voltage and speed.
2) The power loss in the motor is mainly due to the copper loss (neglecting core loss).

Let us use an example for illustration.
An motor rated at 24Vdc has :
Rated output = 200W:
Full-load current = 10.4A; and
Efficiency at rated output = 80%.

Case 1
At 24V and rated output = 200W
Input power = 200W/80% = 250W
Loss = 250-200= 50W (assume all copper loss)
Current I = 250/24 = 10.4A

Case 2
Using 48V and load the output to 200W
The speed of the motor will be approx 2 times and a gear reduction to use obtain the same speed at drive wheel. The frequency of the back emf is also 2 times.
Current = 10.4A/2 =5.2 A (approx)
Loss = 50W X 1/4 = 12.5 W
The heat generated is only 1/4 of case 1 with the same output power.

Case 3
Using 48V and load the current to 10.4A
Input power = 48X10.4 = 500W
Loss at 10.4A = 50W
Output = 500-50= 450 W.
With the same heat generated, the output power is more than twice.

Comments:
1) For power transmission, high voltage is used to reduce copper loss to obtain high transmission efficiency.
2) But for motors, the true story will be quite different.
3) The iron core loss may not be neglected because the iron loss is approximately proportional to the square of the voltage and to the frequency of the back emf. So the order of increase in iron loss is 8 times.
4) Most BLDC motors have an 3-phase frequecy of 200 to 400Hz at top-rpm. The motors would not work optimally if this frequency is too high due to high iron core loss .
5) The centrifugal force on motor part is proportional to the square of the motor rpm. This means 4 times. There is no problem for direct-drive hub motors. But for gear motors running at 3000 rpm (or so), this is a great problem.
6) Most motor could work at several times full-load current (i. e. overload current) for a short time without overheat. The industrial and domestic ac motor could take 5 to 10 times full-load current in direct-on-line starting. But running at this current (or > full-load current) continuously (or for too long a time) will damage the motors. Guess this would also apply to ebike motors.
6) Direct drive hub motors having larger physical size tend to substain the overload current for a longer period.
7) At present, I am using 36V battery on my AL1020 (24V 200W) motor and there is not heating problem in the motor and the controller.

 

[safe]

3) The iron core loss may not be neglected because the iron loss is approximately proportional to the square of the voltage and to the frequency of the back emf. So the order of increase in iron loss is 8 times.

First let me say thank you for writing a very good posting that more or less supports the observations I've made in this thread.

I know very little about iron core loss, I'm going to have to look that up and see how that works. The thing that jumps out at me is that iron core loss is tied to voltage. All the heat related issues I've seen so far have been tied to current, so this sort of takes things in a different direction.

My "guess" is that for a typical motor you can roughly double the voltage and get a doubling of power without suffering too many problems. Beyond that level and I'm pretty sure somethings not going to be right. I'm planning to take a 36 volt motor to 48 volts... so that's as much "risk" as I'm personally willing to take at the moment.

Once I study iron core loss some more I'll see if I can integrate that into the spreadsheets I use to improve the model I'm using.

Do you know any good links to "iron core loss"?

 

[Smoker]

This is way over my head. I must say that this would be a cool feature to add a "low heat mode" or an optional "haul more ass than I'm designed for mode" once the optimal engaging speed was reached. Like I said this is over my head but very interesting.

 

[safe]

This is way over my head. I must say that this would be a cool feature to add a "low heat mode" or an optional "haul more ass than I'm designed for mode" once the optimal engaging speed was reached. Like I said this is over my head but very interesting.

You could build a MCL (motor current limiting) circuit so that your motor would stay cool normally and then simply unplug it (maybe have some switch or a button) that reverted back to the default BCL (battery current limiting) behavior when you needed extra power for a hill that you otherwise could not make.

So most of the time you run "lean" and only in special situations you run "rich".

That would be pretty easy.

Actually that would be good for fast starts too...

 

[The7]

First let me say thank you for writing a very good posting that more or less supports the observations I've made in this thread.

It is great to have an open-mind attitude in searching for knowledge and I admire your hand-on experience in your ebike.

 

[safe]

The Death Cross

This chart points out what I'm calling the "Death Cross". The curve for power and the curve for heat intersect to make a very definite cross shape. As far as heat is concerned the best power you can get is at the peak power point. Everything to the left of the peak power produces more heat which robs your motor of life... thus the term "Death Cross". If you can keep your motor at the peak power or above as far as rpms you will get the best performance. Whenever you drop below your peak power things just get worse and worse. This is the core idea I'm working with in (MCL) motor current limiting.... to defeat the "Death Cross".

Why is power consumed to the left of the peak power point unwise?

Because you could gear down a little and shift the rpms back into the peak power and avoid any loss of power or buildup of heat. Something like a CVT transmission could (in theory) maintain a perfect motor performance all the time. Beagle123 needs to figure out how to match the motor rpm on his new bike to the CVT, but if he can pull that off it will be a huge accomplishment. (I have no idea how you would do that) For me the manual approach of gears is fine. (for now)

 

[The7]

After using 20A X-controller in place of 15A A-controller on my AL1020 ebike with 36V battery, I am reluctant to use the lower current limit one because:
1) need better torque for one ratio gear motor.
2) need better accerelation to reach the cruising speed,
3) there is no heating problem for UP-volt from 24V to 36V (would like to try 48V later)

In fact, "cruising" at current higher than full-load current for all the time, will cause overheating problem.
Don't think MCL will help in such "cruising" unless MCL =< than full-load current.
If the high starting torque of the DC motor is suppressed by MCL, then the motor would loss its most salient point in traction application. Even if multi-speed gearing is used, the high starting torque is still an excellent feature for fast accelleration if need.
If one accellerates slowly, the motor current may not reach the current limit.

 

[Toorbough ULL-Zeveigh]

I don't get it. :?

 

[safe]

Even if multi-speed gearing is used, the high starting torque is still an excellent feature for fast accelleration if need.
If one accellerates slowly, the motor current may not reach the current limit.

That comes down to gearing though. Any time the gearing is lower than the hub motor concept you get more torque when the motor is at it's peak rpm than when it's in the low area of the rpm.

Imagine a really low gear, high rpms and big power.

Run the numbers and you'll see this is true... for the hub motor you're really at a developmental "dead end" because you have essentially no things to "tweak". (gears change everything)

The only way to trick out a hub motor is with raw power... and that means you need 25 lbs of metal to handle the "primal" nature of that power.

The small motor concept is more sophisticated... but harder to do well.

Hub motors are sort of the "caveman" approach of the "bigger hammer" to solve the problem. :wink: