High Voltage - Low Heat... always?

safe · Aug 12, 2007

The equation for a simple and ideal motor for calculating heat is:

Heat = Amps squared times the motors Resistance

From this we see that voltage is not in that equation. And yet power is expressed in watts which is:

Power = Volts * Amps (minus losses to get the actual output)

...so these equations suggest that if you double the voltage you double the power, but you don't increase the heat.

Wondeful! Maybe.... okay... maybe there is more to it...

First let's look at a Battery Current Limited controller that is operating at it's initial voltage... 1X:

safe · Aug 12, 2007

Okay, now let's take that motor and double it's voltage (while keeping the same current limit) and we get:

safe · Aug 12, 2007

Now look what happened. At the high rpms the heat is the same, but as you go lower in the rpms the heat is larger for the 2X than for the 1X. You have now created a machine that can potentially overheat a great deal more rapidly than it's stock configuration.

Why does this happen?

The answer has to do with that terrible phenomenon :wink: we call "current multiplication". Let's look at the duty cycle of this situation before (1X) and after (2X):

safe · Aug 12, 2007

In contrast to the "Battery Current Limited" controller which is the standard let's look at the freakish animal that is the "Motor Current Limited" controller:

safe · Aug 12, 2007

That's amazing!

Wait... what did I just see?

The "Motor Current Limited" controller by definition makes sure that the current that is on the motor side never goes above a fixed value. The standard "Battery Current Limited" controller looks at the battery side voltage and has no idea what the current will be on the motor side. So what's happening is that a standard type controller that is of the same current limit, but of a much higher voltage will allow more current through at the low end of the rpms and this means more heat.

The "Motor Current Limited" controller not only holds heat constant across the entire powerband of the motor, but it even holds constant with increasing voltage. One could (in theory) combine BOTH ideas and have an increased voltage (to get more power) and with the use of a "Motor Current Limited" controller the guarantee that no extra heat could be generated. The only practical limit is the speed of the motor. In my case, with the typical Unite motor, the stock maximum rpm is right around 3000 rpm. If you double the voltage you will need to be able to handle 6000 rpm in your transmission and gearing and also the motor will need to be strong enough to not fly into pieces at that speed. One could only know through trial and error (destruction) just how fast you could spin your motor before it flew apart somehow. (or the bearings might fry)

What's the bottom line?

The bottom line is that to increase the voltage increases the power output and for the most part you get this without an increase in heat. However, due to the nature of the standard controller that monitors only the battery side current of the controller and not the motor side current it allows extra current to pass through at low rpms and since more current produces more heat that means that the motor will likely heat up MORE if you increase the voltage. (not something you want to happen )

The way out of the heat increase is to use a controller that monitors the motor side (the "output" of the controller) and that way you can turn down the duty cycle to keep the current low at low rpms.

It's at the low rpms where the heat is at it's worst.

The down side of this for people using a hub motor situation is that you lose that incredible low end torque. A geared machine can adapt to different torque situations better and would be best matched to this solution. However, I would love to hear about someone with a hub motor and high voltage TRYING this out to see how it works. (it can't hurt to try it) There's plenty of power in the high voltage hub motor and if you could still get around without ANY heating at all you might be surprised that it's a good thing. We will have to hear of an actual test one day. But this does say a lot about the potential of how to get MORE power without creating ANY extra heat... 8)

safe · Aug 12, 2007

The best thing to look at is the "head to head" competition of a 1X using the standard controller verses the 2X and the "Motor Current Limited" controller.

Except at the lowest rpms you have more power and less heat just about everywhere!

The7 · Aug 12, 2007

safe said:
The best thing to look at is the "head to head" competition of a 1X using the standard controller verses the 2X and the "Motor Current Limited" controller.

Except at the lowest rpms you have more power and less heat just about everywhere!

If you use 2X voltage, the speed will be also 2X.
Then the power at the same speed could be compared.

xyster · Aug 12, 2007

No! I disagree with everything you just wrote that I also didn't bother to read because it appears to be the same-old half-baked motor current limiting argument. :roll:

safe said:
The down side of this for people using a hub motor situation is that you lose that incredible low end torque. A geared machine can adapt to different torque situations better and would be best matched to this solution. However, I would love to hear about someone with a hub motor and high voltage TRYING this out to see how it works. (it can't hurt to try it)

Yes, losing torque would suck. I have high voltage. I have a hubmotor. Send me one of these controllers, I'll try it and then post a full report. The risk to you is if it's really that much better, you won't get the controller back. But you will have finally made your point.

safe · Aug 12, 2007

In a hub motor it might not work well because torque is directly tied to motor rpm. On a bike that's has a chain going from the motor to the rear wheel (multiple gears or not) you can compensate for the higher maximum rpm with a lower gear ratio. I kind of mentioned that, but, well, I guess I just mentioned it again.

The big idea here was that increasing the voltage using a standard controller doesn't give the results you might expect. You would expect a direct relationship where the power would increase due to higher voltage and the heat would remain constant. But in reality the heat roughly doubles at low rpms along with the voltage.

This realization came by ACCIDENT!!!

I was running some numbers on different battery configurations and all of a sudden I realized this was true. The "Motor Current Limited" controller has as this weird quirk that you can raise the voltage and there is no corresponding increase in heat production.

It came as a surprise...

The siginificance of this discovery is that there has been a myth that said that you couldn't raise the voltage on a motor beyond a certain point or you would risk burning the motor out. What burns out the motor is the current and NOT the voltage. (Fetcher can get into this if he's around) It's the artifact of "current multiplication" in the standard controller that actually masks the potential of higher voltage.

If you could phase out the low rpm "current multiplication" then you can increase the voltage up as high as is practical with gear reduction.

safe · Aug 12, 2007

xyster said:
...appears to be the same-old half-baked motor current limiting argument.

Geez... the motor current limiting argument has been approved as valid by our resident expert Fechter. Don't tell me you are still pretending in your mind that the theory is invalid? (granted they don't make these products very often, but they do exist)

fechter · Aug 12, 2007

Oh geez, here we go again...

Anyway, the results that Safe is showing will be pretty close to actual.

You can run a gazillion volts, but the current really determines the majority of the heating.

Higher voltage will allow the motor to run at a higher rpm, which will increase core losses, which result in heating also. Too high and the efficiency will suck due to high windage and core losses. Way too high and the motor flies apart. :shock:

The downside is the motor will be really gutless off the line and you will NEED gears to make the thing move. Acceleration, even with gears, will be sort of sucky.

If you don't mind sucky acceleration, but want really good range, it might make sense.

safe · Aug 12, 2007

The7 said:
If you use 2X voltage, the speed will be also 2X.
Then the power at the same speed could be compared.

You have the right idea.

The actual corrected chart that shows both looks like this. For a given amount of heat (they both are about the same) which peak power would you rather have?

How significant is that low end torque compared to the massive high end?

xyster · Aug 12, 2007

safe said:
Geez... the motor current limiting argument has been approved as valid by our resident expert Fechter. Don't tell me you are still pretending in your mind that the theory is invalid?

fechter said:
If you don't mind sucky acceleration, but want really good range, it might make sense.

Indeed -- which was why I described it as only half-baked.

safe · Aug 12, 2007

fechter said:
Higher voltage will allow the motor to run at a higher rpm, which will increase core losses, which result in heating also. Too high and the efficiency will suck due to high windage and core losses. Way too high and the motor flies apart.

I think the efficiency of gearing the motor down from a really high rpm to something manageable would add to the losses too. There's a "sweet spot" somewhere where you go up with the voltage and get more power, but beyond a certain point the other issues take over.

The big idea, or should I say the surrendering of a myth that had been holding me back, was that you couldn't raise the voltage on a motor without risking more heat. This is TRUE for a regular controller because of the current multiplication effects. However, the "freaky" controller that watches it's own "exhaust" (how else does one describe a controller that watches it's own results in a feeback loop?) can allow the voltage to rise with no negative consequences.

This really makes things more complex because now I have to figure out how far up the voltage I want to go (how much risk I take on the high end) while at the same time having to create the circuit myself and also deciding on a survivable current limit.

I can almost bet that I'll blow up another motor before I'm through! :lol:

safe · Aug 12, 2007

xyster said:
Indeed -- which was why I described it as only half-baked.

Yeah, but look at what would happen... just how important is that first 0-10 mph? Once you get above 10 mph then you have more power. Of course you're already overvolting on your bike and have a high current limit anyway, so your "big iron" seems able to dissipate through it's 25 lbs of metal. (my supposedly overweight new frame only weighs 17 lbs)

I'm thinking about these smaller motors that weigh 10 lbs or so. Imagine getting this kind of performance out of a 750 Watt Unite motor. (which is what the chart is modeled after) Can you imagine a peak of 3500 watts from a Unite motor and not that much heat? (not much low end, but the high end is great)

xyster · Aug 12, 2007

Perhaps this idea would rise to fully-baked status if you employed both mechanisms on a single controller, using a sensor to switch from battery-side to motor-side limiting once torque requirements fell below the level of the peak battery-side amperage -- only to switch over again once torque requirements exceeded a pre-specified mark for a pre-specified duration. Have your fully-baked cake and get to enjoy eating it too...?

safe · Aug 12, 2007

xyster said:
Perhaps this idea would rise to fully-baked status if you employed both mechanisms on a single controller, using a sensor to switch from battery-side to motor-side limiting once torque requirements fell below the level of the peak battery-side amperage -- only to switch over again once torque requirements exceeded a pre-specified mark for a pre-specified duration.

On a geared bike that's called a "downshift". :lol:

The hub motor is a special case that is kind of stuck in a rigid set of circumstances. When you can mess with the gears then you can also mess with everything else and come up with new ideas.

On my current geared bike I just smoothly go through the gears and the motor is always running very close to it's ideal all the time. In my experience on my geared bike a "peaky" powerband would suit me just fine. I'd much rather have 3500 watts up top and nothing down low than lot's of low end and no top end.

It's also possible to reverse engineer this stuff and go way down on the currnet limit while going up on the voltage. If I'm going to get into the 750 watt legal limit I might do that. What that means is you get a flatter efficiency curve, less heat even than the stock "rated load".

xyster · Aug 12, 2007

On my current geared bike I just smoothly go through the gears and the motor is always running very close to it's ideal all the time.

And I just twist the throttle a little and a second later, same thing....hills, though, hills are where I'd appreciate the gears because I have insufficient power to just pop into the efficiency zone going uphill, as I do on level ground. So to fix this, since I can't very well add gears, I'll add better FETs and batteries instead.

safe · Aug 12, 2007

xyster said:
So to fix this, since I can't very well add gears, I'll add better FETs and batteries instead.

The hub motor has few options. Basically all you can do is add voltage and hope that your motors large mass can absorb the heat. You can burn up the small motor pretty easily trying to do what the big motors do. That's why you have to essentially "outsmart" the small motor to make it more capable. Fechter has done air cooling, overvolting and controller modifications and it's been a success.

It's what us "little guys" do while you "big iron" guys don't have to work so hard...

dirty_d · Aug 12, 2007

if you want to go faster with the same motor you are always going to need more current, torque is proportional to current, if you want to go faster you need to produce more torque in order to accelerate. so higher voltage = higher speed and more heat. and doubling voltage wont double speed, 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. so to double speed you must increase the power by a factor of 8(sounds crazy but true, experiment with a bike power/speed calculator). so to go twice as fast as you go at 24V would be 24 * 2 * sqrt(2) = 67.8V

safe · Aug 12, 2007

This is what happens when you double the voltage and change nothing else. (keep the same current limit and the same gearing and use standard battery current limited controllers both times)

The heat becomes unbearable for many motors and they fail...

dirty_d · Aug 12, 2007

and about gearing, you always want your motor running at the peak efficiency point for whatever voltage it is running on, so if your motor's peak efficiency rpm at 24V is 3000 rpm and draws 15A then if you double voltage the peak efficiency rpm will be 6000 rpm and drawing 60A. i don't believe there is any getting around it.(basically what you're saying is that by increasing the voltage and power you are increasing its efficiency because its current is staying the same) this cant happen.

safe · Aug 12, 2007

In contrast this is the exact same 750 Watt 36 Volt Unite motor being overvolted to 72 Volts and then given a (MCL) Motor Current Limited controller.

Which overvolted motor is going to burn up?

safe · Aug 12, 2007

dirty_d said:
basically what you're saying is that by increasing the voltage and power you are increasing its efficiency because its current is staying the same) this cant happen.

And yet it does.

The secret is in eliminating the current multiplication. The motor current limiting controller allows a small 750 Watt Unite motor to pump out over 2000 Watts and not burn itself up. The low end isn't great, but who cares? The top end is so good that you ignore the fact that the low end isn't so great.

dirty_d · Aug 13, 2007

you cant change a motors max efficiency without changing it physically .

High Voltage - Low Heat... always?

1 GW

Attachments

1 GW

Attachments

1 GW

Attachments

1 GW

Attachments

1 GW

1 GW

10 kW

10 MW

1 GW

1 GW

Administrator

1 GW

Attachments

10 MW

1 GW

1 GW

10 MW

1 GW

10 MW

1 GW

10 kW

1 GW

10 kW

1 GW

Attachments

1 GW

10 kW

Similar threads