High Voltage - Low Heat... always?

[fechter]

... doubling voltage quadruples power since doubling voltage also doubles current at any particular rpm. the force(torque) required to overcome air resistance increases with the square of velocity so the power required increases with the cube of velocity. ..

The part about doubling the current would be true if there was no current limiter in the controller. The limiter will keep the current below the set point. When the controller hits the limit, the torque is prevented from going as high as it would without a limiter.

The part about increased air resistance is definitely true, so calculate you power requirements accordingly.

If you increase the voltage (and rpm) of a motor, in most cases the efficiency will go to crap quickly at higher rpms. This is a function of motor design and will vary with the type of motor.

If you want to keep the motor from overheating, you would need to reduce the current as you increase the rpm above the designed operating speed. I don't think the motor simulators reflect this very well. With a hubmotor, there's practically no windage loss and the rpms are so slow even when overvolted that the core loss increase is still fairly negligible.

With a geared motor that runs at a higher rpm to start with, the core and windage losses can get nasty at, say, double the normal rpm. Again, this is largely determined by the motor construction and the type of iron and magnets used. A coreless motor will operate much more efficiently at higher rpm.

I think Safe is just going to have to build a motor side current limiter and test it out. One made with the Allegro hall current sensor would allow moving the sensor to either the battery or motor side.

Personally, I like rocket take-offs and would not be happy with a severely restricted motor. Motor temperature monitoring with feedback to the current limiter seems like an ideal solution to me. Full juice until the motor starts getting hot, then it backs off gradually to prevent overheating. Perhaps the next version of the CA will have this feature.

 

[xyster]

Full juice until the motor starts getting hot, then it backs off gradually to prevent overheating. Perhaps the next version of the CA will have this feature.

Now that sounds like the better way to do current limiting!
Safe, if you built an add-on for normal controllers that could limit the current in accordance with motor temperature, you'd have a marketable product.

 

[safe]

A real world example...

The people that struggle the most with power aren't people like Xyster that have "big iron" or even me with my 750 watt geared machine. The truly suffering folks are the ones that own 250 watt motors which if you tried to overvolt would burn up in a very short period of time.

So I'm going to go through how a typical Unite 250 watt motor can be given a new life that can make even the Europeans pay attention!

 

[safe]

First let's look at the motor as it's sold and the way that it's normally used. The first chart is the motor in the standard 24 Volt setting using a typical 30 amp battery current limited controller like everyone can buy for next to nothing.

Things to notice:

Look at the magenta line at the very bottom. That's the "rated load" heat value which is in this case only 68 watts of heat.

Now look at the light red colored line. This is the "average heat" calculated by taking all the data points across the entire powerband and simply averaging them. We can say that the average heat is something that we want to stay underneath because while the "rated heat" guarantees that the motor stays cool, the "average heat" should mean that the motor will get warm but not burn up. The "average heat" for this motor with a standard 30 amp controller is 317 watts.

 

[safe]

Okay, now we look at two more charts. The first chart is the same motor as before only we've increased the voltage up to 48 volts. When you increase the voltage it also increases the maximum rpm, so the second chart shows how the 24V motor compares to the 48V motor on an rpm by rpm basis.

The thing to notice is that the "average heat" for the 48V motor has jumped from the 317 watts before up to a toasty 560 watts. The fact of the matter is that when you do this your motor will quickly overheat and taken at face value this idea won't work successfully.

But wait.

The motor current limiting controller concept can change everything and in the next posting we'll look at that...

 

[safe]

Motor Current Limiting means that you need to create a feedback loop with your controller so that the controller reduces the throttle whenever the current that makes it's way to the motor rises above the current limit. The standard battery current limited (BCL) controller has no clue what goes on in the motor side because it's watching the battery side current.

Current equals heat. Heat is bad. :wink:

So with this introduction I present the motor current limited (MCL) examples.

The thing to notice is that:

At 24V with a 30 amp current limit the "average heat" is only 203 watts.

At 48V with a 30 amp current limit the "average heat" is only 237 watts. (which is freaking amazing!)

 

[safe]

The Grand Conclusion

The big finale on this is when you compare apples verses oranges and look at a 24V - 30 amp standard battery current limited motor like most people currently own verses the combination of (1) overvoltage to 48V and (2) the use of motor current limiting to squash the current down really low so that the average heat is also low so you don't burn up your motor you get something that is very impressive.

The "stock" motor produces a peak power output of 416 watts.

The "modified" motor produces an insane power output of 1052 watts.

If I was a European or Australian I'd take my innocent looking 250 watt street legal motor and leave the tag on it that says it's only 250 watts then overvolt it to 48 Volts and at the same time chop the current by using motor current limiting.

If there can be a slogan that you can remember about this whole article remember:

"Always replace your current with higher voltage."

In other words it's always better to get your power with more volts than with more amps... :wink:

(Note: This is why we have people experimenting with 150 volt systems on this messageboard. Any time you can raise the voltage it translates into better performance)

But also remember that when you raise the voltage using a stock controller you also increase the average heat in the motor. The motor current limiting controller is the way to compensate for the extra heat and eliminate it.

 

[The7]

Things to notice:

Look at the magenta line at the very bottom. That's the "rated load" heat value which is in this case only 68 watts of heat.

.... the "rated heat" guarantees that the motor stays cool
[/b]

True.
If the average of the motor heat stays at this "rated load" heat value or less, the motor could run continously with no problem of overheat/burn-up.

Now look at the light red colored line. This is the "average heat" calculated by taking all the data points across the entire powerband and simply averaging them.

.... the "average heat" should mean that the motor will get warm but not burn up. (317W)
[/b]

False.
This "average heat" is 317 W which is about 4.7 times the "rated heat".
If the motor continues to run at this 317 W heat value, the motor will be heated up quickly and will burn up in a short time.

 

[safe]

If you increase the voltage (and rpm) of a motor, in most cases the efficiency will go to crap quickly at higher rpms. This is a function of motor design and will vary with the type of motor.

If you want to keep the motor from overheating, you would need to reduce the current as you increase the rpm above the designed operating speed. I don't think the motor simulators reflect this very well. With a hubmotor, there's practically no windage loss and the rpms are so slow even when overvolted that the core loss increase is still fairly negligible.

Well it's the low rpm current that is the problem. High rpms don't create much heat normally (up near the no load area) so if you're screaming the hell out of a motor it's in a state that is naturally producing little heat so a moderate increase up there should be no problem. (even if air pressure reduces the overall efficiency the heating shouldn't be that bad)

I'm thinking for my #003 Project that I will indeed buy a 250 watt motor and do the overvolting and current limiting I've been talking about and see what happens. I'm hoping to help establish a whole new "cult of the small tricked out motor" that you Fechter have done a good job of initiating.

 

[safe]

False.
This "average heat" is 317 W which is about 4.7 times the "rated heat".
If the motor continues to run at this 317 W heat value, the motor will be heated up quickly and will burn up in a short time.

Well that's reality... (reality is not textbook or "rated loads")

These motors have standard ratings and standard controllers and the standard average heating for the standard 250 watt motor with the standard 30 amp controller is right around 317 watts. The maximum heating is much worse.

This is in fact why a motor can burn up if you go up a long hill and the motor is unable to get out of it's lower rpms. The longer you are at the lower rpms the sooner it is until your motor gets hot.

So it's somewhat out of place to say that "If the motor continues to run at this 317 W heat value, the motor will be heated up quickly and will burn up in a short time" because in real life this is what is going on as one goes up and down the hills and accelerates and decelerates. This is why all our motors get hot... no one stays in the rated load all the time.

The "rated load" is the optimal load for a motor that might have to run for days and days like an industrial application. For an electric bike with a limited range it can heat itself up over the course of a half hour and then when you're done riding it cools off.

You can't argue with the stock reality of the motor... it just "is" this way.

If you are stuck climbing a steep hill you will inflict up to 500 watts of heat to the motor for as long as the hill takes to climb!!!

 

[TylerDurden]

Does acetylene make people stupid?

We're back at the same old roadapples. That's horseshit for you city-dwellers.

These motors have standard ratings and standard controllers and the standard average heating for the standard 250 watt motor with the standard 30 amp controller is right around 317 watts. The maximum heating is much worse.

Standard ratings... standard controllers... standard average heating?? Where do you get that crap??


Kill the current and you kill the torque.

The crippled bike will spend most of its time trying to accelerate. If the bike could possible accelerate, the power needed to accelerate will require current that produces heat and will spend more time there.

Fechter has it right (as did eP in the stupid thread "More power - less heat")... get up to cruising speed fast and stay cool.

Again, the efficiency of the particular motor is being ignored. When the rated load is exceeded the heat will continue to build up until the motor cooks itself.

"Average Heat"... "Rated Heat" .... WTF? Imaginary specs that do not exist.

Show us where the heat comes from based on the efficiency ratings of the motor.

RPMs are not the issue, efficiency is. Check the efficiency at no-load... 20%.


BTW: The only ebikers I know of @150V are running hubmotors. They don't seem to have heat issues.

 

[dirty_d]

hey safe, i totally misunderstood what you are trying to do, i thought you were saying that by doubling the voltage youre going to get 4X more power with no extra current and heat, that just cant happen. i get what you mean now and it does make a lot of sense, like my motor has a max power of 4430W at 48V with 50% efficiency(at max power which is 1/2 no load speed and 1/2 stall torque efficiency will be 50% for permanent magnet motors). if i increase the voltage to 96V the point on the efficiency curve where 4430W is produced will be at a higher rpm and lower torque closer to the max efficiency region. like you said you do have to be careful with too high a rpm as the brush wear will increase a lot and the bearings might not be rated for that speed as well as increasing iron losses.

 

[safe]

Show us where the heat comes from based on the efficiency ratings of the motor.

TylerDurden you've had difficulty with this stuff all along because you simply haven't studied the math very deeply.

Let's review...

We can estimate the heat by the equation:

Heat = Current Squared Times the Motors Resistance

Heat(Watts) = Current(Amps)^2 * Resistance(Ohms)

So we need to solve for two variables... the current we get from the motor calculations and the resistance we can extrapolate by knowing the rated load and the no load current and other specs. Most of these smaller motors have resistance values somewhere around 0.2 to 0.3 ohms. Unfortunately only the better motors like the Etek and PMG 132 actually list their resistance values. (often a factor of 10 better than the cheaper Unite motors)

From those two variables (current and resistance) we can know at any rpm what the approximate heat generation will be. Fechter points out that at higher rpms the friction and air compression tends to increase the heat above the equations, but for an approximation the equation should give a ballpark estimate.

As you can see the math is there if you want to learn it. You can in fact model the heat behavior of a motor and once you do you gain a mastery over manipulating it. From this level of mastery of the math you can then make the intellectual leaps I've been making and presenting.

Eventually I'll build and test this stuff to prove the theories and then one day it might become old news... 8)

 

[dirty_d]

RPMs are not the issue, efficiency is. Check the efficiency at no-load... 20%.

the efficiency at no load and stall is 0%, at stall you have torque but no rotation, T * 0 = 0, at no load you have rotation but no torque(0 usable torque as all that is being produced simply overcomes friction to maintain that angular velocity) rpms * 0 = 0.

 

[safe]

i totally misunderstood what you are trying to do, i thought you were saying that by doubling the voltage youre going to get 4X more power with no extra current and heat, that just cant happen. i get what you mean now and it does make a lot of sense, like my motor has a max power of 4430W at 48V with 50% efficiency(at max power which is 1/2 no load speed and 1/2 stall torque efficiency will be 50% for permanent magnet motors). if i increase the voltage to 96V the point on the efficiency curve where 4430W is produced will be at a higher rpm and lower torque closer to the max efficiency region.

It's a little more complicated than that.

I'm manipulating a bunch of things all at once. First I'm increasing the voltage which moves all the rpm data points to higher values. Then with the current "diet" that I'm placing the motor on it's running very lean as far as current and that reduces the motors heating tendency. How it's different than just increasing the voltage and lowering the current limit (another approach that will work) is in the way that the current limit is being done. The motor current is always different than the battery current... it's what Fechter calls "current multiplication" and that name seems to have stuck. It's the default behavior of a buck converter. Anyway, the motor current limiting concept is to insert a current meter in between the controller and the motor and measure how much current is actually passing through to the motor. You then create a feedback loop that notifies the controller to lower the throttle ANY TIME the current rises above the fixed rate current limit. On a standard controller it is clueless about how much current it let's through because all the standard controllers look at the battery side current alone and ignore everything else.

I know it sounds complicated, but once you put these two things together the results should be very good. Judging by the comparision chart of the 24 Volt using a regular controller verses the tricked out 48 Volts using the modified controller it does appear on the surface that this could be a really neat performance upgrade.

Imagine 1000 watts from a motor that is rated at 250 watts!!!

And no smoke!!!

The Europeans will love it!!!

 

[dirty_d]

why not just use a 1000W motor since its built for that power and not have to worry about it flying apart.

 

[safe]

why not just use a 1000W motor since its built for that power and not have to worry about it flying apart.

In Europe they are not allowed to have more than 250 watt motors. In America it's 750 watts. Canada 500 watts. No matter what size motor you start with if you double the voltage and cut the current down using motor current limiting you get a similiar result which looks very positive... though I've never build this idea yet. (I see it in the numbers, but it's still theory until a prototype is built)

In a "perfect world" we would use larger motors and then run them lean.

But "perfect worlds" tend to not interact with reality very often so we have to find ways to adapt to the specifics of our environment.

So I'm actually kind of tweaking the motor upside down from what would be the natural inclination if rated loads were never an issue. If people could run any sized motor then everyone should go out and buy an Etek or PMG 132 because they are by far better than the smaller motors we are forced to deal with because of bicycle laws.

If a policeman sees a PMG 132 he's simply not going to believe your story that it obeys the power limits in the law...

 

[dirty_d]

if you get a 250W motor to put out 1000W then its a 1000W motor, same with a 1000W motor, if its only powered by enough voltage to put out 250W its a 250W motor

 

[safe]

if you get a 250W motor to put out 1000W then its a 1000W motor, same with a 1000W motor, if its only powered by enough voltage to put out 250W its a 250W motor

True.

But this is an "aftermarket upgrade". The idea is that you buy the bike in it's legal form at 250, 500, or 750 watts and then do this little bit of magic and increase the performance of what you bought significantly. To the naked eye it will appear to be a standard "rated load" vehicle, but only an expert will know otherwise... :lol:

The hub motor guys are already doing it, but they have 25 lbs of steel to dissipate the heat. Small motors can't do that so you need to be smarter to make it work. This is that "smarts thing" you need to have... (that's a President Bush-ism :wink: )

 

[The7]

In Europe they are not allowed to have more than 250 watt motors. In America it's 750 watts. Canada 500 watts. No matter what size motor you start with if you double the voltage and cut the current down using motor current limiting you get a similiar result which looks very positive... though I've never build this idea yet. (I see it in the numbers, but it's still theory until a prototype is built)

Would like to see your dream becomes real!?

 

[safe]

Would like to see your dream become real!?

If by that you mean that someone has already done this already and you're going to show me then sure go ahead and show me!

 

[safe]

There's a guy on the YouTube that claims he has modified his Kollmorgen motor so that it puts out 1500 watts with a Crystalyte controller and it's held up very well after a year of abuse. At around five pounds in weight that's very good. One wonders if there might be any benefit to motor current limiting with a brushless motor. It might be harder to do, but then maybe not... if you simply measure the current going into the motor (an external current sensor would naturally average the current... or would it? :? ) and then set up the feedback loop to the controller to turn down the throttle it should work the same.

Just how far can this idea be taken?

If this guy says the Kollmorgen can handle 1500 watts without overheating using the regular controller then I wonder how far you could go with the cooler motor current limiting controller technique?

At what point do the bearings break?

The motor fly apart?

 

[TylerDurden]

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

20%?

30%??

in order to know how much heat is generated, it is necessary to know the motor efficiency at all operating points.

http://endless-sphere.com/forums/viewtopic.php?t=930&postdays=0&postorder=asc&start=285

 

[dirty_d]

that is the easiest way to calculate all of the heat generated including heat from iron losses and friction, if you wanted to know how much heat is dissipated while the motor is putting out 1200W at 78% efficiency then its just 1200 / .78 - 1200, don't even need to know about the resistance or current. safe, plug that into your graphing program and see how it looks compared to just calculating copper losses.

 

[safe]

that is the easiest way to calculate all of the heat generated including heat from iron losses and friction, if you wanted to know how much heat is dissipated while the motor is putting out 1200W at 78% efficiency then its just 1200 / .78 - 1200, don't even need to know about the resistance or current. safe, plug that into your graphing program and see how it looks compared to just calculating copper losses.

Yes, actually you can do it that way too! The results would be identical, but in my spreadsheet I've added the efficiency losses of the controller (it's only 95% efficient) so there's a slight difference. So the extra heat is radiating from the controller. The overall efficiency is lower because the controller takes some energy away from the motor and adds to the heat total. (it's only a 5% difference in results)

That's the great thing about math... everything adds up! :wink:

The bottom line is still the fact that current produces heat:

Current Multiplication means Heat Multiplication