48V in a 36v system

[OneEye]

Running at 48V also means the low voltage cutoff of the 36V controller will put your batteries into deeply cooked territory.  If you don't have another means of monitoring battery voltage you may also want to consider adding a resistor to the low voltage sense line as pdonahue suggests in the Cutoff Voltage thread.  Running the batteries past 80% discharge really kills the cycle life.

[pdonahue]

Another week and the 48V is running fine...  The only thing I've noticed is that my efficiency is quite bad when riding fast into a headwind.  With the old 36V system, I was usually around 12Wh/km.  At 48V I'm riding faster, but the efficiency is down to about 16Wh/km.  Riding home into the wind on Wednesday it was 20Wh/km.  At that rate of drain I can only go about 25km before the battery is drained.  Still, I'm glad I made the switchover. 

Pete   

[myelectricbike]

With what are you measuring your Wh/km?

[trappermike]

A 48v system should be more efficient,given the same watts rating of the motor.Example;

24v-500w:    500W divided by 24V = 20.83 Amps(Current draw)

36v-500w:    500W divided by 36V = 13.88 Amps(Current draw)

48v-500w:    500W divided by 48V = 10.41 Amps(Current draw)

Also,say we add an extra 12v battery(a 12v-12a SLA for example) to achieve 48v, we also have larger battery storage and capacity for more range,extra weight being the only penalty.
This is one reason 24v systems are dinosaurs,and have very poor range.
This is aslo one reason that electric motors in some other applications run on very high voltage,some submersible oilwell pumps are 600v for example. Or another example is motorised equipment in a workshop such as a lathe or air compressor are cheaper to run on 220v than 110v.

[cadstarsucks]

A 48v system should be more efficient,given the same watts rating of the motor.Example;

24v-500w:    500W divided by 24V = 20.83 Amps(Current draw)

36v-500w:    500W divided by 36V = 13.88 Amps(Current draw)

48v-500w:    500W divided by 48V = 10.41 Amps(Current draw)

Also,say we add an extra 12v battery(a 12v-12a SLA for example) to achieve 48v, we also have larger battery storage and capacity for more range,extra weight being the only penalty.
This is one reason 24v systems are dinosaurs,and have very poor range.
This is aslo one reason that electric motors in some other applications run on very high voltage,some submersible oilwell pumps are 600v for example. Or another example is motorised equipment in a workshop such as a lathe or air compressor are cheaper to run on 220v than 110v.

I suspect that is not the way the controller works.  Usually the wattage varies with the voltage and not the other way around.  You are running a 36V 500W spec'ed motor at 48V you will get the wattage of the beast - not less current.

Dan

[cadstarsucks]

Here are the things to consider.
1. 48 v vs 36 is 33% increase in voltage and will cause a similar increase in current draw.
2. You will saturate the magnetic structure of the motor and the additional magnetic force will be wasted, with the waste being turned into heat.
3. Generally a motor can take an overvoltage/current situation for a short time. If the motor is already hot then  the additional heat from too much power will likely start to burn the insulation on the windings.
4. If it is a brushless design you may exceed the semiconductor ratings in the electronic commutator.

This Tim-the-Toolman approach is not a good idea.

Actually a custom controller is just what you need to do this in a brushless.  A reduced output voltage at low speeds prevents saturation, and the extra voltage at "extra speed" ( say 30 instead of 20MPH ) maintains torque at the higher speeds.

Dan

[mustangman]

 You build it , they might buy it.

[cadstarsucks]

You build it , they might buy it.

If I win the lottery so I do not have to worry about putting food on the table I plan to.  ( I am an electronic design engineer )

My thought is 2 of the Golden 500W 36V motors on a 125V bus with a home brew controller for  3000W ... adjusted down 75% for efficiency and divide by 750W give me a 3HP vehicle capable of 60MPH.

Dan

[mustangman]

 I hope you have great medical and life insurance! You might need it if a cellphone distracted driver rolls through a stop sign or you eat the asphalt!!  (which ever comes first). Personally In would opt for a 72 volt , 40 amp comtroller like the other brand "c" has. 

[cadstarsucks]

I hope you have great medical and life insurance! You might need it if a cellphone distracted driver rolls through a stop sign or you eat the asphalt!!  (which ever comes first). Personally In would opt for a 72 volt , 40 amp comtroller like the other brand "c" has. 

Do you have a link?  The lack of specs annoys me. 

I would not recommend that: you might demag the motor.  That is just the sort of thing that torques out the forks.  The controller needs to maintain the rated torque, else risk destroying the structure.  The increased voltage allows you to increase the speed without losing the torque to the frequency or the rotor resistance (which also causes the drop in the speed at high load on the motor curves ).

Dan

[pdonahue]

With what are you measuring your Wh/km?

I have a DrainBrain (now CycleAnalyst).  It gives lots of useful info.

[pdonahue]

A 48v system should be more efficient,given the same watts rating of the motor.Example;

24v-500w:    500W divided by 24V = 20.83 Amps(Current draw)

36v-500w:    500W divided by 36V = 13.88 Amps(Current draw)

48v-500w:    500W divided by 48V = 10.41 Amps(Current draw)

Also,say we add an extra 12v battery(a 12v-12a SLA for example) to achieve 48v, we also have larger battery storage and capacity for more range,extra weight being the only penalty.
This is one reason 24v systems are dinosaurs,and have very poor range.
This is aslo one reason that electric motors in some other applications run on very high voltage,some submersible oilwell pumps are 600v for example. Or another example is motorised equipment in a workshop such as a lathe or air compressor are cheaper to run on 220v than 110v.

A lower current only lowers the resistive loss in the wires (Tens of  watts at most).  The higher speed using 48V means much more wind resistance (hundreds of watts) so the net effect is less efficiency.  If we were on the moon though, I would agree that the 48V controller would be efficient. 

Also, moving to 48V does not mean keeping the same output level.  More realistic would be:
24v-500w:    500W divided by 24V = 20.83 Amps(Current draw)
36v-750w:    750W divided by 36V = 20.83 Amps(Current draw)
48v-1000w:    1000W divided by 48V = 20.83 Amps(Current draw)
since the current draw is generally limited by the current feedback loop of the controller and so does not decrease as the voltage is increased. 

Pete

[cadstarsucks]

A lower current only lowers the resistive loss in the wires (Tens of  watts at most).  The higher speed using 48V means much more wind resistance (hundreds of watts) so the net effect is less efficiency.  If we were on the moon though, I would agree that the 48V controller would be efficient. 

Also, moving to 48V does not mean keeping the same output level.  More realistic would be:
24v-500w:    500W divided by 24V = 20.83 Amps(Current draw)
36v-750w:    750W divided by 36V = 20.83 Amps(Current draw)
48v-1000w:    1000W divided by 48V = 20.83 Amps(Current draw)
since the current draw is generally limited by the current feedback loop of the controller and so does not decrease as the voltage is increased. 

Pete

Interesting... this is the first I had heard that they actually have a feedback loop other than the phase sensors.

The Golden graphs are implying almost totally resistive losses of 220W on the 500W model at full load.  I therefore would not expect much more than 220W lost on a 48V controller assuming it is actually current limited.

Dan

[pdonahue]

A lower current only lowers the resistive loss in the wires (Tens of  watts at most).  The higher speed using 48V means much more wind resistance (hundreds of watts) so the net effect is less efficiency.  If we were on the moon though, I would agree that the 48V controller would be efficient. 

Also, moving to 48V does not mean keeping the same output level.  More realistic would be:
24v-500w:    500W divided by 24V = 20.83 Amps(Current draw)
36v-750w:    750W divided by 36V = 20.83 Amps(Current draw)
48v-1000w:    1000W divided by 48V = 20.83 Amps(Current draw)
since the current draw is generally limited by the current feedback loop of the controller and so does not decrease as the voltage is increased. 

Pete

Interesting... this is the first I had heard that they actually have a feedback loop other than the phase sensors.

The Golden graphs are implying almost totally resistive losses of 220W on the 500W model at full load.  I therefore would not expect much more than 220W lost on a 48V controller assuming it is actually current limited.

Dan

Thinking about it, I guess all the losses are resistive (some in the coil, some in the core) except for the loss due to friction which I would assume would be quite small.  So that would blow my "tens of watts at most" out of the water     But I have noticed that the actual current draw at both 36V and 48V is about the same (Peaks at around 30A on startup and then drops to around 20A during operation.) 

There is a metal bridge on the controller board leading to the neg. rail.  I assume that it is being used as a current sense resistor.  I haven't traced out that part of the circuit, but I am guessing that it leads back to an op-amp (I thought there was an LM356 on the board, but it may have been a LM358... It's hard to see the markings.) to cut back the duty cycle when the current gets over 20A.

Pete

[myelectricbike]

Those 2 parallel metal bridges are 17AWG copper (not 16AWG) rated at less than 10 amps each, for 60c insulation. Hence, 20 amp maximum continuous with startup at 30 peak at line voltage; more for 36 DC unswitched supply. On the new controllers an LM317 is used to regulate voltage to the board and the only other LM is a LM339 quad comparator in a PDIP which provides a data signal to port C, it looks like to indicate forward/reverse. If its in a TO-220 then its most likely a voltage regulator like the LM317. If it only has 2 leads then it may be a thermostat like the F11-E06.