The problem with hub motors is that they have just 1 high gear, which means the motor will have low efficiency at low speeds. With high voltage battery, the efficiency peaks at different speed than with low voltage battery. So to get the best efficiency would be changing voltage depending on speed, so that the motor would be always running at good efficiency. The voltage could be changed with some kind of dc transformer or electronically reassembling the battery pack cells. I have been wondering if it would work like that or am I wrong? If it works, why hasn't anyone done it, or has someone? I know changing voltage is more complex than changing current but I think the advantage is very good, no problems on hills anymore etc.
Why not make variable voltage system?
- Thread starter skinkken
- Start date
dmwahl
10 kW
Any efficiency you gain would probably be lost in the conversion, and low speed efficiency isn't really a problem in my opinion. Few people spend much time at low speed, and when they do motor efficiency may be lower, but energy used per mile is better. Ebike controllers are basically a buck converter anyways, so they're effectively lowering voltage into the motor already.
If you use ebike simulator and put there a low voltage battery you can see where the point where efficiency peaks, and when you switch to a higher voltage battery, then at the same point efficiency is much lower, and the peak efficiency is at different place. For example, at very low speed the efficiency may be like 30%, but if you switch to a very low voltage battery then it's more like 70%. Which is much more than the losses in voltage transforming.riba2233 said:
skinkken said:
Lol, just adjust the throttle to ~27% and look at the magic! You have lowered the voltage... That's what your controller does, adjusts the voltage according to motor speed, that's why you have throttle and PWM.
What you are saying would be right if we still used rheostat based motor control.
Ykick
1 GW
The only plausible reason I can see to bother with such complications might be for racing?
Gasoline racing, as fuel burns off vehicles become lighter and generally faster by the end of the race. EV racing, vehicles are as fast as they’re gonna be in the beginning of the race, they don’t lose any weight during the heat and ultimately slowdown significantly due to voltage drop over the course of the race.
So, if racing against gasoline vehicles a person might want a variable voltage system to increase voltage or at least compensate for normal voltage drop so that the vehicle can run just as fast as it did at the beginning of the race.
Gasoline racing, as fuel burns off vehicles become lighter and generally faster by the end of the race. EV racing, vehicles are as fast as they’re gonna be in the beginning of the race, they don’t lose any weight during the heat and ultimately slowdown significantly due to voltage drop over the course of the race.
So, if racing against gasoline vehicles a person might want a variable voltage system to increase voltage or at least compensate for normal voltage drop so that the vehicle can run just as fast as it did at the beginning of the race.
Ykick said:
This can be taken care of in another way, in a way that is already used by EV manufacturers. Use a motor that can spin at max rpm eve when the battery is almost empty, and use smart controller to limit the rpm when the battery is full, or simply use field weakening.
cal3thousand
10 MW
skinkken said:
You're not looking at apples to apples. Without lowering the throttle (on the high voltage setup) to where the speeds are the same as the low voltage speed, you are looking at one that has been bogged down by either a hill or wind to end up at the same speed. What the simulator is saying at that point is if you are going at full throttle and only managing 10kph, something is slowing you down into the inefficient zone.
Exactly!cal3thousand said:
Ignoring the technical viability of the OP's proposed solution, there is no problem to 'solve'.
- Reducing the throttle has essentially the same effect by PWM as reducing the battery voltage, so there is no need to change the battery voltage by other means.
- The often-repeated axiom of 'low speed = low efficiency" is generally overstated.
efficiency and speed as affected by load not by throttle setting.
We aren't interested in some 'peak efficiency' on a load plot - we want to examine the actual efficiency achieved at the exact speed that the bike achieves for a given throttle setting. The rest of the simulator graph is essentially inapplicable in this case and shows 'what-if' situations for different load situations other than the throttle/grade/weight/etc parameters entered into the simulator setup.
Here is a plot on the topic of discussion: efficiency vs speed as achieved by reduced throttle (data drawn from the simulator):
Efficency pretty much doesn't change down to 80% max speed, may fall off some at 60% max, and may for certain motors fall off up to 10% at 30% max speed. That means if the bike can go 25mph it may at most take a 10% efficiency hit if ridden at 7.5mph (without pedaling). It also means that if the bike can go 40mph, it will take about a 5% efficiency hit if ridden at 20mph.
To quote an old commercial: "Where's the beef?"
Okay, so it's just about the amount of throttle and not battery voltage, while higher voltage battery is kind of "allowing" more power? So if there is for example a 48V battery and a 72V battery and needed electronics, would we be able to mod the 72V system to work exactly the same way as the 48V system? Why are people using low voltage applications then? Isn't higher voltage always better, having the potential to go faster?
Another example: I would have a 72V bike and a 36v bike. Is it the same if I use 36V full throttle and if I use 72V half throttle?
EDIT: Tried that with the simulator and they were almost the same.
Another example: I would have a 72V bike and a 36v bike. Is it the same if I use 36V full throttle and if I use 72V half throttle?
EDIT: Tried that with the simulator and they were almost the same.
cal3thousand
10 MW
skinkken said:
Lower voltage systems are safer, have easier BMS requirements (less channels), less complexity, more compatibility (with DC converters and other components), and easier on electrical components (63V and 100V limit on caps that are readily available; 100V limit on the most efficient MOSFETS) . Also, 48V packs are easy to find. 72V+ packs require some figuring out.
zener
1 kW
- Joined
- Jan 16, 2014
- Messages
- 353
cal3thousand said:
Also forthecrazy europeans who have tons of laws for dog doo doo .
everything over 48v (60Vmax.) is considered as High Voltage which require you to put High Voltage sticker all over your ebike also i belief its only a learned electrician allowed to work on such systems.
Which is BS if u ask me i touched a 93V battery accidentially and it hurted a bit. But your 230V AC all around and in your house is much more dangerouse.
liveforphysics
100 TW
skinkken said:
The motor is a current driven device, no amount of voltage creates any magnetic field to generate torque.
The effect you are seeing in efficiency is an artifact of the simulator's limitations, it practice the drive voltage makes no difference in motor efficiency.
If the pack voltage is adequate to overcome the motor's BEMF and be in a current limiting state, than it makes no difference in motor efficiency if the pack was 10v or 10kV, it just results in a different pwm duty cycle % in the controller to maintain the same power.
fellow
1 kW
skinkken said:
I have two 54.6V HOTC* systems, and I'm designing** a 75.6V HOTC system now - se the project Triebfluegel in my signature. I'm expecting about the same system efficiency at same speed, or thereabout. I'll post the real word results as soon my ebike is finished. My guess is that some things will became less efficient (LM317 resistor will waste more energy for example, and laminating losses will get worse at high speeds), some things will get better (if I crank down the current=less heatheat in cables). I hope for better acceleration over 20 km/h than before, because of less back EMF. People are using lower voltage because it's safer. A lot of the things from the engeneering point of view are worse when lowering voltage. If you decrese it to zero volts threre is a risk that it could stop moving
.*HOTC=Hot Of The Charger voltage
**Designing, what a nice word that is. What's wrong with the word "play" those days?
fellow
1 kW
liveforphysics said:
My old swedish university book sugests that the torque is proportional to voltage squared. Here is my translation of it (holding it in my hands):
M=(3*U^2*s)/(Omega_s*R), where:
M represents torque in Nm
U represents voltage in V
s represents difference between magnetic field speed and the actual rotor speed. This is for three phase asynchronous motor, set this value to 1 for three phase PM synchronous motor.
Omega_s represents 4*Pi*f/number_of_poles. f is the controller frequency in Hz.
R=resistance in Ohms
3 phase asynchronous motors do create torque by using the voltage, 400 three phase AC for example. They use more current at start, and less when not needed. This without any controller involved. It's starting current can be very high, several hundred amperes, fast fuses usually do not like this. Close to top speed they use almost nothing. When connected to 1,5V AC they produce less staring current and typically do not move at all. Their torque is zero because of the low votlage, current is not limited in any (other) way. On the other side very low current at high voltage can move things around.
In reality both current and voltage are needed together if we want to move things around, and in Europe they teach us that up to 999 Volts there are no downsides besides the life threat
. Passing 1000V, nasty things start to happen, sparks between the wires that are too close and so on. Typically 10A to 20A is the range we should keep out current at, if we want highest practical efficiency - this according to my university books. This because the wires thicker than 2.5mm^2 are very difficuilt to work with, safety switches are larger, and so on.
liveforphysics
100 TW
fellow said:
If that V-using formula were missing the R there, it wouldn't work, because they are simply using a V=IR substitution to arrive at current for the way that formula is written. It's not a mistake or something, just an alternative expression of current.
fellow
1 kW
liveforphysics said:
I know what you mean, thoused ways to strungle the cat. Things can be expressed differently. The funny thing is that current is almost banned in this book, the few times it is mentioned it is in very negative context. Kind of a evil thing
.John in CR
100 TW
Just to help make simulations clear, that shows the performance (power, efficiency, etc) for a given throttle position from 0 rpm up to the max rpm with that load (friction of all types including wind resistance and the effects of gravity, which takes into account the mass and road grade). Cruising slower using lower throttle will have much better efficiency than during acceleration.
"Why not make a variable voltage system?" The answer is that's what our controllers do via the throttle, ie give us 0 volts up to pack voltage.
"Why not make a variable voltage system?" The answer is that's what our controllers do via the throttle, ie give us 0 volts up to pack voltage.
regmeister
100 W
- Joined
- May 9, 2009
- Messages
- 263
I have been planning to build a rectactor circuit to use with an Agni. Because of the brushed motor, it is a least worthwhile to experiment.
The range of voltages is pretty limited though. I settled on dividing the battery into 4, so it would be 3 speeds.
The battery/controller will be 20s a123, therefore 5s in each group. At discharge load of 2.7/cell, the voltages would be 16.2, 32.4, or 64.8. I think the sag, and cabling resistance will no doubt affect my results so I may need to go to 6s or change strategies altogether.
btw, this is to drive a propeller, so it will need to work correctly with the redrive ratio. I'm afraid it will be too much shock (electrical and mechanical) changing speeds, but I plan to try anyways.
Designed for full speed with all batteries in series, 1/2 speed with battery split in two and paralleled, 1/4 speed with battery split in four and paralleled. May need some starting resistance from zero, so maybe something similar to a pre-charge resistor circuit on the start button.
Anyways, here is a link you may find interesting re: some history of contactor controllers, and a link to the rectactor just in case someone wants to make sense of the 3 speed I've touched on above. Some day it'll get it built...
http://www.mneaa.com/frame_sunrise_ev2_project.htm
http://www.evdl.org/docs/rectactor.pdf
The range of voltages is pretty limited though. I settled on dividing the battery into 4, so it would be 3 speeds.
The battery/controller will be 20s a123, therefore 5s in each group. At discharge load of 2.7/cell, the voltages would be 16.2, 32.4, or 64.8. I think the sag, and cabling resistance will no doubt affect my results so I may need to go to 6s or change strategies altogether.
btw, this is to drive a propeller, so it will need to work correctly with the redrive ratio. I'm afraid it will be too much shock (electrical and mechanical) changing speeds, but I plan to try anyways.
Designed for full speed with all batteries in series, 1/2 speed with battery split in two and paralleled, 1/4 speed with battery split in four and paralleled. May need some starting resistance from zero, so maybe something similar to a pre-charge resistor circuit on the start button.
Anyways, here is a link you may find interesting re: some history of contactor controllers, and a link to the rectactor just in case someone wants to make sense of the 3 speed I've touched on above. Some day it'll get it built...
http://www.mneaa.com/frame_sunrise_ev2_project.htm
http://www.evdl.org/docs/rectactor.pdf
regmeister
100 W
- Joined
- May 9, 2009
- Messages
- 263
Punx0r said:
Won't argue with you. Nearly purchased a controller several times.
Don't see why it would unbalance the pack, unless caused by differing voltage drops across subsections? Would this be massive?
dmwahl
10 kW
Wouldn't unbalance if you keep the batteries always wired in series or parallel, but any combination of the 2 would cause imbalance since half the pack would be 2 parallel batteries and the other would be 2 series. Why bother though when motor controllers are so cheap and give you much better control? You'll generate a lot of heat in those diodes. If you want three speeds, use a 3 speed switch (~$10 or less) and set the controller to 35/70/100% or whatever levels you want. I certainly would not want only 3 fixed speeds in my boat though, seems like asking for trouble. Mechanical/electrical shock shouldn't be a problem though, especially not in a boat.regmeister said:
I met a guy who used a similar setup in his S10 EV conversion, he had some elaborate method of switching contactors around to get variable speeds. Said it was a pain to use and just about peeled out from a stop if you hit the wrong combination since there was no speed control other than changing the battery configuration. Tangle of wires too, I don't know how he drove the thing.
Similar threads
