More Data: KS Peak Power hit 2400W

[Jason McNeil]

Today I took the KS800 to Greenwhich park, which has one of the meanest paved inclines around London. 

Objective: Test the maximum limits of the Wheel

Approach: repeated (6) runs up & down a 17° gradient incline, over about 50 meters, at maximum speed, recording the power profile with a data logger. With each successive run, attempt to push the Wheel ever so slightly harder to capture that moment of weakness   

Tools:

• KS800 680Wh 14" eWheel
• GoPro strapped to the front of the Wheel recording the action 
• eLogger set to a capture rate of 10Hz (10 samples/sec) monitoring volts, amps, watts & control-board temperature

Result:

The power profile bears a similar resemblance over each of the runs. It is interesting to see the peak power reading reach into the stratospheric level of 2,400W on the last attempt. It is quite possible that with a higher sample rate, (ie 50Hz) more of these spikes will be recorded.   

    

This graph contains the magnified data view of the last run over the course of 1.3 minutes. Despite the appearance of chaos, the data can be smoothed out a 1 sec avg to reveal a more orderly trend of power distribution (this will be done later using excel).  

 

[esaj]

How about throwing larger versions of the graphs into http://imgur.com/ for example (no need to register, just click "upload images" from the top left corner). Can't see shit from these, captain   Is the amperage curve inverted? Looks like it drops as the power goes up... Or is it the voltage? Can't see the texts or values.

 

[Jason McNeil]

Good call, orange = watts, red = volts, gray = temperature.

[Rotator]

Fantastic!

Would be very interesting to log currents data while braking downhill at diferent intensity, in order to know when regenerative charge occurs, and when adittional power is drained from the battery to brake harder.

[esaj]

Interesting, in the magnified image, if the powerline between 0.18 and 0.28 in timeline is coming down the hill, it actually looks like it's drawing around 250-350 watts for balancing only (assuming fairly steady speed coasting down the hill). That's quite a lot. Also a couple of other interesting points, it looks like the motor does pack a lot of punch, between 0.09 and 0.11, it looks like it's giving about 1400W on average. After that comes the big peak of around 2400W and then the power drops to around 250W for a little while (maybe the BMS or mainboard limited the power for a while after seeing a large current peak, or there was a short regenerative braking involved?)... Also, it seems the temperature goes DOWN for the first part of the climb, maybe the wind or rotating tire cools down the mainboard somehow? In the bigger picture though, it looks like the temperature just keeps going up after every try (looking at the other chart with all the attempts), but probably would steady after a while (unless you keep riding around 1.5-2kW all the time) 

[Mono]

Fantastic!

Would be very interesting to log currents data while braking downhill at diferent intensity, in order to know when regenerative charge occurs, and when adittional power is drained from the battery to brake harder.

and I'd like a speed log as well, always easy to ask, much less so to provide 

@Jason McNeil, what would we see in case of energy recuperation? I reckon that we see recuperation between time 0.19 and 0.28, because the voltage is higher than the voltage without any load? Does the wattage in this case indicate the recuperation energy (i.e. we don't see the "sign" of the energy flow)? 

[Mono]

Fantastic!

Would be very interesting to log currents data while braking downhill at diferent intensity, in order to know when regenerative charge occurs, and when adittional power is drained from the battery to brake harder.

just wondering, isn't the current data a direct consequence from the voltage and wattage data (I [A] = P/V [W/V])? Can anybody remind me: the point of measurement is at the battery or does it even matter? 

[esaj]

just wondering, isn't the current data a direct consequence from the voltage and wattage data (I [A] = P/V [W/V])?

Yes, you can always calculate the third one, if you know the two others, ie:

P = U * I  <=>   U = P / I  <=>  I = P / U

And

U = R * I  <=>  I = U / R   <=>  R = U / I

Where P = power (watts, W), U = voltage (volts, V), I = current (Amperes, A) and R = resistance (ohms, Ω). And that's pretty much everything I remember from my electronics / electricity courses taken a long time ago. And some active/reactive/complex AC power-calculations, which are useless with wheels  

[robca]

Why does the voltage drop to 5V at the end of each run?

[Rotator]

and I'd like a speed log as well, always easy to ask, much less so to provide 

@Jason McNeil, what would we see in case of energy recuperation? I reckon that we see recuperation between time 0.19 and 0.28, because the voltage is higher than the voltage without any load? Does the wattage in this case indicate the recuperation energy (i.e. we don't see the "sign" of the energy flow)? 

Since I do not know how fast is going downhill, I can not know if would be regerative power. The voltage increase could be only a battery recovery from the uphill stage current drain.

Edit: looking again the graphics, I would say that it appears regenerative power. Voltage level is higher than idle voltage. In that case, the Niko's question is a very good question. There are not plus or minus sign for the current?

[Mono]

Voltage level is higher than idle voltage.

That was my point  Still, it would be interesting to know whether the gauge is plugged at battery level or at motor level. 

[Rotator]

That was my point  Still, it would be interesting to know whether the gauge is plugged at battery level or at motor level. 

Battery level is easier. To measure current at the motor would be a nightmare. You would need to measure three diferent wires...

[Robotek]

Hello Jason, could you please tell us what is your weight ?

[Jason McNeil]

Very near 70kg.

[Robotek]

Thank you !  The peak would have been higher if your weight were 90kg for example right ?

[esaj]

I took vee73's MCM2s for a short 10km trip that contained "normal" roads and some off-roading in the hiking paths. The current version of my app doesn't have the record-button visible on such a small screen  (plus it's currently probably broken anyway, as I've done heavy refactoring under the hood to separate the different parts of the program, and I think I haven't "wired" the recording-logics back to the button yet ), so I couldn't record the trip, but kept watching the values and the power-graph a lot during riding. Assuming the values are correct, here are my initial findings:

• On a pretty much level road without bumps, the power needed to balance is around 250-400W, on more bumpy road, more like 350-500W at around 18km/h
• Even on a very slightly declining road, the power is near or below zero when going at steady speed, the forward momentum and gravity gives the wheel enough power that it doesn't need to use much or at all power from the batteries
• Even slightly inclining road causes the power usage to go up when keeping speed steady, something like 400-750W
• Strong acceleration can easily take the peak power to above 1kW even on level road
• On the dirt paths, with more bumps, the power usage is very spiky, maybe around 300-1000W at 15-20km/h, on large bumps it can make very large spikes to way over 1kW, and then drop even down to zero/negative territory as the wheel comes "down" from the bump
• Even fairly slight "power braking" can easily make the regenerative power go over 1kW, and more powerful braking can make it hit around 2kW (2000W) peak
• Climbing a more steeper (10+ degrees) hill uses nearer to 1kW (1000W) of power, stronger acceleration can make it go over 1.5kW
• Climbing a VERY steep (but short) gravel hill the power usage was around 1.5-2kW and hit 2.5kW (2500W) peak, though my speed was pretty much crawling... the temperature also raised around 2-3 degrees in seconds (42.x->45.x degrees), probably the mainboard or motor would burn or the batteries would give up on a longer ascend

I'll see if I have the time to work on the recording, so I could record more precise data, the average power is probably more towards the lower numbers than the peaks

EDIT: Also, the 14" Gotway is much MORE stable on the hiking paths than my 264Wh Firewheel... probably because the 4 battery packs (vee's MCM2s is the 680Wh -version) can give so much more power than the 2 packs on the FW.

This leads me to think that the motor rating isn't that big of a deal (as long as it can take the big power spikes without burning) compared to number of battery packs / max discharge rating of the cells.