Understanding battery & phase amp.. some help..

[steveo]

Hey Everyone,

I wanted to better understand battery amp vs phase amps. Here is what is sort of got me confused :?

I have 2 controllers that I've tested on my x6 MHM 602 hub motor with a 24s4p 20c turnigy lipo pack..

Controller 1
IRFB 4110 Mosfet // Square wave // Infinion // tuned in at 200amps battery & 260 phase amps...

Controller 2
IPP075N15 Mosfet // Sine Wave //096150 sabvoton //tunned in at 161amps battery & 260 phase amps approx also..

If I'm pulling more battery amps should i not be pulling more phase amps? or does is the sine wave controller way more efficent?

Despite controller 2 being alot more smoother but less powerfull .. however.. controller 1 is much more throttle responsive and feels more powerfull based on my experience.... however they both put out the same phase amps.. so isn't that equal power at the motor?

thanks
-steveo

 

[zombiess]

There is a large difference in how the controllers treat the throttle. The Xie Chang uses it as a speed control, so throttle position is a command to get to the new speed. Since the controller is regulating speed, it will use as much current as required to reach the new set point unless there is a limit placed (battery amp, phase amp, speed). It does not care how it gets to the new set point.

With the Sabvoton the throttle is setup for phase current. It does not care about speed, only the amount of current. Pick a new set point and it will raise the current to that point. There is also a throttle ramp rate in place to act as a low pass filter. This value is a few hundred milliseconds and prevents large jumps in throttle set points.

Another reason the Xie Chang controller seems more responsive is that they do not do any phase current monitoring, only battery current. This allows them to experience massive overshoots, the lower the inductance the more it can overshoot. The FOC controller is always monitoring phase current so any overshoots are immediately corrected back to the set point.

You'll only pull as many battery amps as required or until you hit a limit.

Some of what you are experiencing is psychosomatic. The same effect could be accomplished by setting 0-90% of current on the sabvoton to happen between 0-50% of the throttle (try it). It will seem more responsive, but only because you compressed the throttle range into a smaller mechanical throttle range.

BTW, Sabvoton told me do not run 150A (battery) for more than 5s because they had high current PCB failures.

 

[steveo]

There is a large difference in how the controllers treat the throttle. The Xie Chang uses it as a speed control, so throttle position is a command to get to the new speed. Since the controller is regulating speed, it will use as much current as required to reach the new set point unless there is a limit placed (battery amp, phase amp, speed). It does not care how it gets to the new set point.

With the Sabvoton the throttle is setup for phase current. It does not care about speed, only the amount of current. Pick a new set point and it will raise the current to that point. There is also a throttle ramp rate in place to act as a low pass filter. This value is a few hundred milliseconds and prevents large jumps in throttle set points.

Another reason the Xie Chang controller seems more responsive is that they do not do any phase current monitoring, only battery current. This allows them to experience massive overshoots, the lower the inductance the more it can overshoot. The FOC controller is always monitoring phase current so any overshoots are immediately corrected back to the set point.

You'll only pull as many battery amps as required or until you hit a limit.

Some of what you are experiencing is psychosomatic. The same effect could be accomplished by setting 0-90% of current on the sabvoton to happen between 0-50% of the throttle (try it). It will seem more responsive, but only because you compressed the throttle range into a smaller mechanical throttle range.

BTW, Sabvoton told me do not run 150A (battery) for more than 5s because they had high current PCB failures.

Hey Zombiess

thanks for your explanation of how these two controller differ in how to put out there power.

besides throttle feel/power feel

why will the xie chang controller peak at 200amps battery & 260amps phase vs the sabvoton 096150 at 160amp & 260amps phase ?

why is there that difference in the battery amps yet the phase amps are the same yielding the same power at the motor..

BTW.. I did some mad pulls the other week at WOT on the sabvoton with that x6 hub i still have and no pop... 162 battery -- i think 268 phase peak was highest i saw..

FYI.. that Xie Chang EB2 series controller you sold me the matched irfb4110 for is still working!! it goes to show the reliability of it .. I have 2 other eb3 series that i built similar after that one and has also worked great on my 2wd x5303 hub bike

-steveo

 

[Samd]

A ratio well below 1:2 seems to be about the most people want to go if they don't want to make a lot of heat inside the hub.

 

[zombiess]

A ratio of 1:2 seems to be about the most people want to go if they don't want to make a lot of heat inside the hub.

A ratio of 1:2 is poor advice for those looking for performance as it will produce more heat. Search on this topic and you'll find a ton of info. John in CR has done a great deal of posting to dispel this advice meant for low power setups.

 

[zombiess]

SteveO, unless you have measured the phase amps the Xie Chang outputs, the numbers you are calculating are just a guess.

 

[Samd]

A ratio of 1:2 seems to be about the most people want to go if they don't want to make a lot of heat inside the hub.

A ratio of 1:2 is poor advice for those looking for performance as it will produce more heat. Search on this topic and you'll find a ton of info. John in CR has done a great deal of posting to dispel this advice meant for low power setups.

Apologies that should say BELOW - I will amend my post.
Info here if anyone looking.
https://endless-sphere.com/forums/viewtopic.php?f=30&t=73393

 

[steveo]

SteveO, unless you have measured the phase amps the Xie Chang outputs, the numbers you are calculating are just a guess.

oh.. but i did with my fluke amp sir!

thus how i have these numbers

 

[John in CR]

The XieChang's can have quite an interesting phenomenon if turned up to a certain level. That 36fet I bought from you SteveO is still alive and well with a 1.25:1 phase limit to current limit ratio have it set to deliver peak battery current of 415A (measured using an ebikes.ca high current shunt), which puts peak phase current at 520A.

The interesting thing was when I had the controller running at about 20kw peak instead of the 30kw+ max it's set to run now with a 74V nominal pack. The phase current limit comes into play on launch, but at some point as the phase current naturally decreases during acceleration the battery limit will come into play. Normally we don't feel the transition but with this controller as the motor would get up to slightly less than half of peak rpm (at 40-45kph in this case) there was a sudden burst of extra acceleration that you really had to be prepared for. As I continued to turn the controller higher the extra burst of acceleration went away. Maybe Zombiess has some ideas of the cause.

FWIW, if you're running a lower inductance higher Kv hubmotor (those capable of higher power), then your 24s pack is likely to kill your controller due to significant voltage spikes occurring in the controller that will send voltage well above the component limits.

 

[Harold in CR]

John, you have explained why I had such a problem, years ago, with controlling a 12V 3/4HP motor on the sawmill we built. It would blast the controller with Voltage spikes. Thanks for that great explanation. 8) 8)

Luckily, I knew a guy in Wisconsin, on a totally different forum, that worked in Electronics. He drove the company engineers crazy, until one of them suggested wiring an electronic "Chiclet" looking device across both motor terminals. Have no idea what that is, but problem immediately was cured. I still have 2 of them, right here in CR, in case I have that problem with this new Sawmill I am finishing up.