RPM Based Current Limit?

[safe]

I've been "thinking" again and you know that spells "trouble". :wink:

Decided to run a simulation with the concept being to "adapt" the controller to those nasty low rpm areas that produce high efficiency losses. You can "correct" for that if you slowly REDUCE the current limit downwards as the rpms drop. This is like what people try to do manually with their ammeter, but in this concept the controller would guarantee that you didn't "overdo the throttle" any more than needed.

What I did...

I took the spreadsheet for a 750 Watt "overvolted" 48 Volts Unite MY1020 motor and manually (for now) located the "efficiency peak" at 48 Volts. I then created a variable that subtracted some voltage from the "efficiency peak" downwards in the rpms until you reach zero. I then "played" with that variable a little until a found a "local maximum" for efficiency (the best you can get) and ended up with this chart.

Notice how the voltage AND current decreases with rpms and that the effect is to add a little boost of efficiency in an area that normally suffers a lot. (a regular controller has a flat line for the current limit because it's a constant across the rpm spectrum whatever it might be)

A controller designed like this would be VERY EFFICIENT.

Note: In the lowest rpms I'm sure things would behave a little weirder than the chart actually shows, but you get the idea.

 

[safe]

Just for comparision this is a "normal" (idealized) efficiency curve verses mph/rpm like in a hub motor with a properly configured fixed current limt controller.

 

[xyster]

One problem: what if you needed that extra, albeit inefficient, amperage to go up a hill?

If this less-than-max-available current limit was solely RPM based, the motor would stall on hills it'd otherwise climb.

Your circuit would have to adjust either for slope, or to motor load. How could you adjust for slope? Mercury tilt switches? How could you adjust to wheel load? Some kind of torque sensor?

Seems to me your idea has merit. I've noticed that pinning the ammeter at 35 amps while accelerating on the flats doesn't *seem* to buy me much better acceleration than keeping the meter to 20 amps or less. But on hills, that extra 15 amps really makes a huge difference in acceleration and top speed.

 

[safe]

After redoing the spreadsheet to try to get better accuracy I realized that this DOESN'T WORK very well after all. This chart is more like it... All I'm doing is fiddling with voltage (the only real control you have) and it doesn't really give the results you might hope for.

So this idea doesn't seem to have anything going for it.

(and it places in question how much actual savings you can get with an ammeter)

Gears really are more valuable than people realize at this point...

 

[xyster]

Gears really are more valuable than people realize at this point...

And quite the borderline pathologic obsession for you, as the rest of us realized long before this point. But you're right, this idea of yours', like so many before, sucks. Keep trying! One day you'll hit on something patent-worthy.

 

[safe]

I hope you comprehend that this throws some suspicion on the value of an ammeter too. I guess if you are dealing with an over amped motor controller then maybe it helps to drop the voltage (would make sense) but on an "ideal" motor with the correct controller you can't actually improve on it.

It's impossible to improve on the "ideal" with throttle restriction.

This actually makes sense because the "ideal" aligns the "efficiency peak" and the "power peak" so you can't get any better "by definition".

The LARGER and more PROFOUND siginificance is that any hub motor that is designed to go 50 mph will ALWAYS hover around 50% efficiency if you are below 10-15 mph.

That's a SERIOUS and DISTURBING realization. (given the start and stop nature of actual riding)

You just can't fix this by ANY means...

Wow...

 

[Leeps]

The LARGER and more PROFOUND siginificance is that any hub motor that is designed to go 50 mph will ALWAYS hover around 50% efficiency if you are below 10-15 mph.

wow i guess you forgot our debate on the thread titled low rpm efficiency. If you remembered it you would know that low rpm efficiency does exist. I showed you the dyno charts to prove it.

Joe

 

[safe]

wow i guess you forgot our debate on the thread titled low rpm efficiency. If you remembered it you would know that low rpm efficiency does exist. I showed you the dyno charts to prove it.

And things look good on the charts too.... until you reach that "breakdown" level which is the bottom 1/4 of the rpms when the efficiency curve drops off. You CAN'T fix that... there's no way to fix it.

In the examples before we did: (for the sake of argument)

48V - Corresponds to 60 mph
36V - Corresponds to 45 mph
24V - Corresponds to 30 mph
12V - Corresponds to 15 mph

...the area where the efficiency really drops off is below 12V. This is ASSUMING you want to actually join the "50 mph Club". The powerband would have to be set up roughly like the chart shows. (and the efficiency numbers reflect these truths)

It's a saaaaaaaaaaad realization if you are into hub motors. Getting started is a real problem... once up into the rpms it gets better and beyond about 10-15 mph it looks okay, but that means every time you stop at a stop sign you have to pass through a pretty wide band of low efficiency. (gears have the same powerband issues, but it's "narrowed" because of the lowered gearing so you get through it quicker)

There is no way to improve on it either... so the fixed gear options are not very good at start and stop driving situation. And we already went through the PWM issues and resolved that you get extra torque, but AT THE SAME efficiency cost because the efficiency is an attribute of the MOTOR at that rpm and NOT the controller.

I hope by now you "recognize" the "ideal" motor curve. Those are the numbers we played with on that companies website. All motors have that and if you set your "efficiency peak" to equal your "power peak" then at all times you are running in the most efficient manner a motor can run at. All motors "break down" at low rpms like this.

 

[fechter]

But your last graph is basically showing the efficiency of the motor run at FULL THROTTLE and increasing the load until the speed drops. Most people let up on the throttle to go 10mph, which makes the graph look quite different.

 

[xyster]

True, and also the time spent at < 1/4 speed is typically a very small fraction of the total.

 

[safe]

But your last graph is basically showing the efficiency of the motor run at FULL THROTTLE and increasing the load until the speed drops. Most people let up on the throttle to go 10mph, which makes the graph look quite different.

Actually it doesn't change much of anything...

The REASON is that when the "efficiency peak" and the "power peak" are in alignment (that's when your controller matches EXACTLY the current limit to the efficiency peak) then that's "as good as it gets" from an efficiency standpoint.

TOTAL POWER is different... you might back off on the power on a hill and if your controller is "over amped" then it will become more like the "ideal" motor and your efficiency will go up. But it's in effect "correcting" the error of the excessive current allowed in the first place.

The only "hard" fact is that (like the steam engine before it) there are certain "ideal systems" that in reality are usually never achieved. The throttle restriction concept helps to bring closer a dysfunctional system to the "ideal".

But you can NEVER beat the "ideal"... backing off the throttle on an "ideal" motor does not improve efficiency... it just can't when you think about it.

The ONLY way to recapture efficiency is to get out of the "bad rpms" and that's done best with gears. It's becoming my mantra:

Gears, gears, gears... :lol:

 

[safe]

True, and also the time spent at < 1/4 speed is typically a very small fraction of the total.

That depends on where you ride. For me I have lot's of stop signs on suburban streets and so I'm always downshifting to a stop, then accellerating up through the gears only to hold the highest gear for a few seconds and then back down again.

On a race track where you never "stop" then the fixed gear makes more sense because you really only need that low end once on the starting line and then that's it for the race.

It's like freeway verses city driving... cruising on the freeway is totally different than the city streets with stop lights and signs.

Hybrids seek to eliminate "idling" at a light... but from the standpoint of the electric motor (with a high fixed gear) stop and go is not a good thing...

 

[Leeps]

You are right, however what your noticing is a low load condition, i thought you were going to attack low rpm efficiency all over again. The truth is that a motor has a certain constant loss associated to it, just to make it turn, if your load is small in comparison to that, well the percentages dont look good. With gears you would be falling off the efficiency curve on the right hand side, strange not too many people pay attention to that side of the efficiency curve.
Joe

 

[safe]

You are right, however what your noticing is a low load condition, i thought you were going to attack low rpm efficiency all over again.

:? You were the one that was confused as I recall. (remember you got confused with the title and we spent the whole time debating the wrong question)

My point before was that PWM DOES NOT save someone from low efficiency at low rpms. I thought you had understood that and agreed.

Let's bring that chart back again and ponder it some more. Yes, it's true that going "past peak" is less efficient (with gears you would then "upshift" to a taller gear) And Yes, we can obviously see that at less than 1/4 of the maximum rpm that the efficiency EVEN IN THE BEST of motors will hover in the 50% efficiency range or below. (again with gears you would "downshift" to get back into the "good zone")

I'm not sure if you totally understand the curve yet... tell me you do now... because it's getting tiring trying to explain myself again and again based on your not understanding the question.

The bottom line is that VERY low rpm efficiency is bad, but midrange is not too terrible if you have an "ideal" motor setup. Even at 1/4 of the max rpms the efficiency begins to be okay.

I'm also using as a "baseline" the idea that one wants to join the "50 mph Club". Obviously a hub motor that peaks at 20 mph would begin it's "good zone" at a low mph like around 5 mph, so we need to be talking about the same thing here. 50 mph is the "top desired speed" of this hypothetical fixed gear bike.

Just so you don't try to "pull a fast one" I'll also establish as the "baseline" that the peak efficiency is 80% for the motor at it's top voltage. Other motors can sometimes have better overall performance than this, but it doesn't change the relative performance that is the central focus of the question. So 80% is the efficiency maximum.

And last, we assume a "full load" in all cases to be applied to the motor. I don't want to get into cases of partial load. Assume we are on a hill of sufficient slope to equal the power capabilites of the motor and therefore we are bound by the efficiencies that exist at a given rpm. One can choose to accellerate more slowly on flat land, but they can't bend a hill to make it's slope less steep.

 

[Leeps]

Sheesh i was agreeing with you
I was saying that your right cruising at 10 mph on a bike that can go 40 is inefficient. The power required to maintain 10 mph places you on the right hand side of the efficiency curve at that specific voltage point and simple

ill let you decide if i understand the charts

as far as being confused what was i supposed to think the thread was titled "low rpm efficiency myth" seriously sorry if i thought you were debating low rpm efficiency, my fault

I was actually wondering where you got the idea that pwm magically saved the motor from inefficiency? What i did say was that pwm lets you change the voltage going to the motor. but im sure you remember the debate


Joe

 

[xyster]

Sheesh i was agreeing with you

Safe's always very defensive because he's not used to anybody agreeing with him.

According the 'pro' hubmotor simulator with constants derived from their dynamo, my hubmotor's efficiency at 25% speed is 50%, at 50% speed is 70%. And this isn't taking into account wind resistance, which means 25% speed in the real world may be more efficient than 50% speed. When wind losses are considered, the whole efficiency curve shifts left and down by an amount maybe Safe would like to figure out.

http://ebikes.ca/simulator/

 

[safe]

I posted that 5304 spreadsheet, you can run all the numbers on that you want. (it very closely matches the simulation program and you can run all kinds of "what if" scenarios)

So basically your bottom 25% has low efficiency because of the gearing and the top 25% has low efficiency because of wind resistance. So the only "good spot" is in the middle somewhere which is around 70%. So overall that's not exactly the most efficient system...

 

[xyster]

I agree. It'd be much better for efficiency of any motorized vehicle if we rode in a vacuum, we just wouldn't live to enjoy it without lugging a ton of protective apparatus.

 

[Leeps]

Xyster you got it we have to get safe to design us vacuum bubbles for our bikes.
then somebody has to figure out how to move the vacuum bubbles around efficiently
Joe

 

[safe]

It's called a fairing....

Aerodynamics is the "vacuum bubble" that you need to discover if you want to join the "50 mph Club".