**IMPORTANT** Reality check on motor, battery, throttle

[swbluto]

So, it was recently discussing that at a constant power level as would be true under a fixed power level, that a higher motor voltage would result in higher motor currents and a lower bike speed resulted in this higher motor voltage due to the decreased back EMF. This had the effect that motor current would be greater than battery current. I didn't understand what was so shocking about this relevation (Probably because I made the simulator), but it became more understandable when I considered that the most popular simulator used didn't display motor current so I could understand how it was a bit of a mystery.

So today, I will look at comparing full throttle to a lower throttle and see how that affects the motor currents.

At lower throttle, your motor voltage is decreased due to the magic of PWM, which decreases your motor current just like decreasing the voltage to a resistor lowers your current. The following things are true at 50% throttle and full throttle and no current limiting.

-50% throttle will have HALF the motor current at exactly 0 RPM than full throttle
-50% throttle will have its motor current go to zero at HALF the RPM of the full throttle case.



Another factor to consider is, what does a current limit do? It was recently shown in another thread that the battery current is "multiplied" into a higher motor current during current limiting, and it seems some were mislead into believing that phase currents actually were higher than they would be otherwise at full throttle (i.e., no current limit, full throttle) due to the wording. What actually happens is that the current limit restricts the power (keeps it constant) at lower speeds, whereas at full throttle, the power continually increases until reaching a maximum at 0 RPM, so the current limit "bends down" the motor current and results in a lower motor current. See the picture below.



So taking those things into consideration, it seems almost too agreeable that lower throttle = lower motor current and lower current limits = lower motor currents. When you're limiting or decreasing the power, motor power doesn't increase. From those graphs, it also shows why regular ebike controllers include current limits. So that the motor/phase currents are low enough so that the controller doesn't instantly self-destruct.

.....

There's been some controversy (Actually, the only controversy seems to come from one member... *ahem*), so I'll show more "practical graphs".

Here's the 5302, 100v, 20 inch rim, 50 amp current limit, .01 ohm battery resistance, .01 controller resistance. This isn't actually representative of real life because the battery resistance will typically be more than .01 ohms (That's like 2+ kWH of 30C Lipo with short 4 gauge wires), so they'll be less in real life, but the differences will remain comparable.

And here's the same 5302 with a 100 A battery current limit.

As you'll notice, the 0 RPM current for the 50 Amp one is ~260 amps while the 100 amp one is ~360.
At 50 mph, the 50 amp one has a phase current of around ~90 amps while the ~100 amp one has a phase current of around ~160 amps.

So, an obvious conclusion, a lower battery current limit results in less motor current at any given speed (except the no-load speed where they both go to "zero", or technically the no-load current.), but they'll approach each other as it tends towards the no-load speed (The fastest your motor can spin at a given voltage) and will equal each other once both are outside of current limiting.

I reserve the right to edit this post at any time, without notice, to correct any accidental miswordings

(Caveat: the above assumes an "ideal battery", i.e., no internal resistance. Batteries with larger internal resistances will deviate from the ideal results above and reduce the full throttle phase currents because their output voltage will significantly decrease at higher currents.)

 

[johnrobholmes]

Lower average motor current yes. Lower peak currents, no.

 

[swbluto]

Lower average motor current yes. Lower peak currents, no.

That depends on the L/R ratio. For most hub motors, where the peak-to-peak current or "ripple current" is 2 to 8 amps, that isn't true for the vast majority of "practically comparable" situations. (I'm not going to compare 97% throttle and 100% throttle, for example.)

According to Justin's testing, the L/R ratio of typical outrunners is comparable to hub motors. I don't know about inrunners like the astros, but I'm guessing that's why matt/bob suggests setting the PWM frequency to 24 kHz instead of the typical 10 kHz due to the motor's fairly low inductance.

 

[johnrobholmes]

I'm not quite sure what you are getting at with this thread. It sounds like you are saying less battery current = less work, which is true. Less battery current cannot predict the motor currents (other than simple average) without considering duty cycle and commutation frequencies, as well as system inductance.

 

[John in CR]

I'm not quite sure what you are getting at with this thread.

Seems like more trees blocking the forest view to me. Thread title certainly needs modification until a view of the forest is brought into focus.

 

[swbluto]

I'm not quite sure what you are getting at with this thread.

I would quote what john and matt has written to show you what the point of this thread was. In a nutshell, some have the impression less throttle or a battery current limit means more phase current than if that didn't exist (no battery current limit or "no partial throttle"/full-throttle). This isn't true. If you already understand that, than you're on the right page.

Less battery current cannot predict the motor currents (other than simple average) without considering duty cycle and commutation frequencies, as well as system inductance.

Less battery current = less motor currents, all else equal. That's predictable. If you don't know what the other parameters are, then yes, it isn't predictable. But, that's pretty much true of everything. One thing that people typically do know, however, is their speed and they can predict where their motor currents are on the basis of that assuming they understand the above relationships (And seen appropriate graphs applicable to their setup). On the basis of those graphs, they also know the limits.

 

[swbluto]

I'm not quite sure what you are getting at with this thread.

Seems like more trees blocking the forest view to me. Thread title certainly needs modification until a view of the forest is brought into focus.

You're one to talk. :lol:

As far as I can tell, you see A forest, but I see the world's entire biosphere. (There's still someone yet who sees the galaxy in its entirety, so I'm not going to be too presumptuous of my relative knoweldge.)

Just a side comment, just about the only major point of that "Enlightening" thread was that motor current > battery current during some parts of motor operation. Whoopdie doo. Glad we got past that first step.

 

[John in CR]

...but I see the world's entire biosphere...

You and donob08.

Please do enlighten the rest of us with the point. Does it explain an issue that results in many blown controllers like Luke's thread? Does it explain how to avoid motor meltdowns or how to make our rigs more efficient, or go faster (other than by turning the throttle, which I have to assume didn't need explanation to anyone here)?

 

[swbluto]

Does it explain an issue that results in many blown controllers like Luke's thread?

Does it explain why the price of tea in China is so often referenced? By golly, it doesn't.

Does it address a misunderstanding that some harbor? Is it aimed at helping people understand the relationships between throttle, battery current and motor current? Yes. How people want to apply that understanding is upto them.

 

[John in CR]

swbluto,

You really need to have this thread deleted, because it's filled with misleading statements, and incorrect statements and misleading graphics that highlight your lack of understanding. eg You still don't get how phase currents can be higher at lower throttle settings than at full throttle (not that they always are, that they can be), and you make reference to the occasions when there is no current limiting, when it's primarily the times of current limiting that are a problematic issue. Like Don you've taken a patchwork of narrowly defined true statements and put them together to paint what is an incorrect picture.

You say you see the whole biosphere, well look at the bike as a complete system. Then take the example of a hill steep enough that a full throttle the bike reaches a constant speed generating the maximum power it is capable of. Now analyze what happens on the same hill at 50% throttle and at 25% throttle. For the real coup de grace and to make it really life-like, while steady at 25%, twist that throttle to 75% to pass that slow-ass truck that was making you cruise at 25% to begin with. Look at what happens to phase currents during these partial throttle conditions, especially after considering that BEMF is much lower as well as efficiency. Look at what goes on in the controller too, and then consider that all this is going on for longer since you're going slower up the hill, so you have less cooling too.

 

[swbluto]

Then take the example of a hill steep enough that a full throttle the bike reaches a constant speed generating the maximum power it is capable of. Now analyze what happens on the same hill at 50% throttle and at 25% throttle. For the real coup de grace and to make it really life-like, while steady at 25%, twist that throttle to 75% to pass that slow-ass truck that was making you cruise at 25% to begin with. Look at what happens to phase currents during these partial throttle conditions, especially after considering that BEMF is much lower as well as efficiency. Look at what goes on in the controller too, and then consider that all this is going on for longer since you're going slower up the hill, so you have less cooling too.

What's your point? Just because the phase currents are going to be "really high" in your examples doesn't disprove a damn thing about the relationships shown in the graphics in the first post. I wasn't suggesting they were low. If you're operating a low-wind motor at 100 volts, you can bet that phase currents can be really high under all throttle conditions, whether that be low partial throttle, high partial throttle and especially full throttle.

Just because you would pull 1000 amps from full throttle doesn't make 500 amps at 50% throttle that much kinder...

 

[John in CR]

Do your homework assignment and see that phase currents are in fact higher at those types of partial throttle situations, not the 50% less that you're still claiming. Until you get that, you're still just stuck in the trees.

 

[swbluto]

Do your homework assignment and see that phase currents are in fact higher at those types of partial throttle situations, not the 50% less that you're still claiming. Until you get that, you're still just stuck in the trees.

You're right, the 50% value I showed was only at 0 RPM, when you're actually moving, the percentage will be lower.

So, let's set an agreement on the details of this "homework" that disproves my original post. What are you comparing? 50% throttle at 15 mph with a 100volt 5302 setup and 100% throttle at 15 mph with a 100volt 5302 setup and no current limit on a 20 inch wheel?

If there is a current limit, and the battery current is exceeded at the requested throttle position at the low throttle and full throttle, then the throttle in effect doesn't matter. The controller "decides" the duty rate needed to achieve the battery current limit desired, and the throttle position doesn't matter. In this case, the phase currents will be equal. This is typically the "dead zone" of throttle because any higher throttle doesn't make a difference, so the "effective throttle" area is a lower portion of the throttle. I'm not referring to the "dead throttle" area in the above graphics, I'm referring to the effective throttle. This dead throttle is part of what sucks of speed-based controllers with a current limit because you only have so much "range" on the throttle to control it, and it can be hard to get exactly the amount of power that you want.

If you lower the "effective throttle", phase currents will be lower. This is what I'm claiming if your intention is a current-limited situation.

 

[liveforphysics]

SWbluto- I created this thread to increase understanding of what happens IN CURRENT LIMITING!!!

At any point during current limiting, or at partial throttle, the phase current (what the motor gets, the load the FETs feel) will be higher than the battery current. Since all we see when we are riding is battery current, it's important to know that even if battery current says 50amps, and you're thinking you're running safely, you may be at 100-150amps of load on the motor and controller etc.

 

[John in CR]

No changing the assignment. Same bike, same battery, same controller, same motor (call it an X5302 to highlight the effect), same wheel, same hill, just different throttle positions or since you may not be able to simulate based on throttle position you can analyze the 100%-50%-25% based on speed because that's what we use a throttle for. Your "all else equal" analysis is invalid, because they can't be.

Luke's thread helped us understand why partial throttle on a hill can easily lead to a controller failure. It was generally known that slow speed up a hill is hard on a motor due to the lower efficiency, but not why these situations are hard on controllers too. It also helped identify the cause of controller failures at very low throttle positions even without a big load, which will hopefully lead to a small addition to the current limiting section of the controller code that also limits phase currents to a specific maximum ratio of the battery current. This is needed because at very low duty big phase current spikes are possible despite the very low power and no current limiting, and on an uneven surface is nearly impossible to maintain a steady throttle position.

 

[ZapPat]

[...]which will hopefully lead to a small addition to the current limiting section of the controller code that also limits phase currents to a specific maximum ratio of the battery current. This is needed because at very low duty big phase current spikes are possible despite the very low power and no current limiting, and on an uneven surface is nearly impossible to maintain a steady throttle position.

Actually the additionnal code is already there, and has been there for quite a long time at least in the infinion-based and XC116-based controllers (2+ years). It's just that most people didn't understand the importance of setting this parameter correctly for a given controller and/or motor. I for one had been trying to communicate the importance of limiting phase current for almost a year, but it takes time for knowledge to circulate around and be absorbed... and then a well illustrated thread like Luke's helps make the info soak into the common ebike tech knowledge base of the forum.

However, it is important to know that it is *not* a ratio of phase to battery current that is to be looked for, but instead that it's the phase current limit itself that is of foremost importance to be set correctly since it is the parameter that stresses both controller and motor the most. Battery current limiting is primarily only usefull to protect the battery itself if it has limited discharge capabilities (low C-rate).

I do agree about the need to change the way throttle control is implemented in ebike controllers though, because that twitchy response is annoying at the very least, and as you say John it can cause large and non-usefull current spikes that are both not efficient and more stressfull on the controller. A hybrid torque + speed based throttle response would be much better for a number of reasons. This means combined variable RPM + phase current limiting done in the controller's "brain", with the throttle position being the variable limiting parameter. A maximum phase current limit will still have to be selected depending on the controller's and/or the motor's physical capabilities, which would then represent the maximum possible output current of the controller when at full throttle.

Pat

 

[swbluto]

Luke's thread helped us understand why partial throttle on a hill can easily lead to a controller failure. It was generally known that slow speed up a hill is hard on a motor due to the lower efficiency, but not why these situations are hard on controllers too. It also helped identify the cause of controller failures at very low throttle positions even without a big load, which will hopefully lead to a small addition to the current limiting section of the controller code that also limits phase currents to a specific maximum ratio of the battery current. This is needed because at very low duty big phase current spikes are possible despite the very low power and no current limiting, and on an uneven surface is nearly impossible to maintain a steady throttle position.

If the point of the assignment is finding the cause of controller failure at low throttle, it doesn't have to do with phase currents being higher at partial throttle than full throttle, though they still will be high. It has to do mainly with the switching losses and freewheeling losses that are present at partial throttle but not present at full throttle (Assuming you're not in current limiting). If you get phase currents low enough, than it isn't so much of a problem, but normal riding with normal speed controllers typically doesn't get them "low enough" with high current limits.

 

[swbluto]

SWbluto- I created this thread to increase understanding of what happens IN CURRENT LIMITING!!!

At any point during current limiting, or at partial throttle, the phase current (what the motor gets, the load the FETs feel) will be higher than the battery current. Since all we see when we are riding is battery current, it's important to know that even if battery current says 50amps, and you're thinking you're running safely, you may be at 100-150amps of load on the motor and controller etc.

I understood the point of your thread. If I really wanted to dispute anything in that thread, I would've replied to that thread. As far as I'm concerned, this thread is an addition to your current thread and doesn't contradict anything you've stated there.

 

[John in CR]

[...]which will hopefully lead to a small addition to the current limiting section of the controller code that also limits phase currents to a specific maximum ratio of the battery current. This is needed because at very low duty big phase current spikes are possible despite the very low power and no current limiting, and on an uneven surface is nearly impossible to maintain a steady throttle position.

Actually the additionnal code is already there, and has been there for quite a long time at least in the infinion-based and XC116-based controllers (2+ years). It's just that most people didn't understand the importance of setting this parameter correctly for a given controller and/or motor. I for one had been trying to communicate the importance of limiting phase current for almost a year, but it takes time for knowledge to circulate around and be absorbed... and then a well illustrated thread like Luke's helps make the info soak into the common ebike tech knowledge base of the forum.

However, it is important to know that it is *not* a ratio of phase to battery current that is to be looked for, but instead that it's the phase current limit itself that is of foremost importance to be set correctly since it is the parameter that stresses both controller and motor the most. Battery current limiting is primarily only usefull to protect the battery itself if it has limited discharge capabilities (low C-rate).

I do agree about the need to change the way throttle control is implemented in ebike controllers though, because that twitchy response is annoying at the very least, and as you say John it can cause large and non-usefull current spikes that are both not efficient and more stressfull on the controller. A hybrid torque + speed based throttle response would be much better for a number of reasons. This means combined variable RPM + phase current limiting done in the controller's "brain", with the throttle position being the variable limiting parameter. A maximum phase current limit will still have to be selected depending on the controller's and/or the motor's physical capabilities, which would then represent the maximum possible output current of the controller when at full throttle.

Pat

Pat,
I understand the phase current limit...The current limiting routine checks against the battery current against its limit and the phase current (estimated using battery current and duty cycle) against its limit. What I'm saying is that a 3rd check is also needed, which is the ratio of Iphase to Ibatt against a predetermined limit of that ratio. All the data is already there, so it's a simple program addition, and it will prevent the huge phase current spikes that are possible without it despite being at very low throttle. Such a simple addition to the code should prevent the very low speed controller failures. Large phase currents, whether they reach the Iphase limit or not, serve no benefit at low power.

 

[John in CR]

If the point of the assignment is finding the cause of controller failure at low throttle, it doesn't have to do with phase currents being higher at partial throttle than full throttle, though they still will be high. It has to do mainly with the switching losses and freewheeling losses that are present at partial throttle but not present at full throttle (Assuming you're not in current limiting). If you get phase currents low enough, than it isn't so much of a problem, but normal riding with normal speed controllers typically doesn't get them "low enough" with high current limits.

No the point is to have you retract the statements that phase current is lower just because throttle position is lower. While it can be true, at the times were the system is most at risk the opposite is true. That kind of statement will just mislead people into thinking they reduce stress on their system by reducing throttle going up a hill.

 

[swbluto]

If the point of the assignment is finding the cause of controller failure at low throttle, it doesn't have to do with phase currents being higher at partial throttle than full throttle, though they still will be high. It has to do mainly with the switching losses and freewheeling losses that are present at partial throttle but not present at full throttle (Assuming you're not in current limiting). If you get phase currents low enough, than it isn't so much of a problem, but normal riding with normal speed controllers typically doesn't get them "low enough" with high current limits.

No the point is to have you retract the statements that phase current is lower just because throttle position is lower. While it can be true, at the times were the system is most at risk the opposite is true. That kind of statement will just mislead people into thinking they reduce stress on their system by reducing throttle going up a hill.

I didn't say a thing about the risks to their system and that's not what this thread covers. Whether or not it is more risky depends on their particular system and the particular environment - (In general, highly powered hub motors put everything at risk.). What this thread does cover is how the phase currents actually compare to each other under no-current and current limit, and partial throttle and full throttle, to clear some possible misconceptions. Those relationships are always true - how much "risk" you are at depends on the phase currents, which depends on your bicycle's speed which depends on the environment and your situation. The graphs depict these phase currents as a function of speed.

 

[John in CR]

Making inaccurate statements or partially true statements attached to impossible conditions only leads to more misconceptions. Did you really think that anyone came away from the other thread thinking that phase currents would skyrocket if they decrease throttle while riding on a flat road under light load conditions? Why are you avoiding what happens on a hill at different throttle positions? Do you need help with the analysis?

 

[grindz145]

I for one am glad Luke brought this up in the other thread.

I think SW is just making sure everyone understands how important the throttle is in the equation still.

I'm very familiar with why the phase currents would be higher, but I really wasn't thinking about it when I look at the battery current.

so thanks

 

[swbluto]

Why are you avoiding what happens on a hill at different throttle positions? Do you need help with the analysis?

What are you talking about? That's inside the graphs. Just look at the lower parts of the graph at lower bicycle speeds where phase currents are higher. Voila! There they are.

 

[John in CR]

Nice try. Go ahead, provide the analysis which will contradict the statements you've made.