Hobbywing Xerun AXE

[Ninomaniac]

Looks squarewave to me!

[JohnRobHolmes]

I’ll be doing side by side tests on monolith and rip rap in a few hours, with my video guy taking footage. A storm is coming so I won’t be able to run through full packs, but I’ll start with fresh batteries and see how much they recharge.

[JatoTheRipper]

Originally Posted by Rekreant

They arent claiming that its the same as everything else, they are CLEARLY saying that its the best thing on the market, never before seen! Watch / read all the promotional stuff. I get that you accept lying in marketing as inescapable but in most peoples eyes that is not a valid excuse. If they were saying they were just as good as the competition no one would be complaining. They are claiming to have created the best thing since sliced bread.

Well it would be odd for a company to release a new system and not claim it was the best!

Come on now. Let's be realistic. Most people, in their right mind, don't buy into marketing hype. I know I don't. Marketing, by definition, is to fluff up a product and make it look good for the consumer. And you nor I can prove that HobbyWing is lying about anything in this ESC so...

[Ninomaniac]

There's a difference between calling your new car model "the coolest EVAR" and claiming it's turbocharged when it's not.

[Rekreant]

Just watch the hobbywing official vid side by side with 1967illas video; Its very telling in how differently the system performs.

[JohnRobHolmes]

Looks like I'm getting rained out for vids today, but tomorrow is clear weather and my camera man is available. That is actually good news, as I'll be able to try and drain the battery packs fully and see how the runtime compares. Spent some time tuning the ESC via the bluetooth app this morning, reversing the direction for the TRX, adjusting throttle and drag ramp.

The throttle response is very lacking at the most aggressive ramp setting. Kinda surprising since it is sensor based. It has a slow ramp up compared to even a SW4 sensorless system.

[svt923]

Originally Posted by Ninomaniac

There's a difference between calling your new car model "the coolest EVAR" and claiming it's turbocharged when it's not.

Yeah, that is the main point. FOC is a specific method of motor control, not just marketing BS. If they would have called it "Hobbywing Motion Control" or some other made up name, no one would have batted an eye.

[JatoTheRipper]

Originally Posted by Ninomaniac

There's a difference between calling your new car model "the coolest EVAR" and claiming it's turbocharged when it's not.

What did HobbyWing lie about?

Originally Posted by Rekreant

Just watch the hobbywing official vid side by side with 1967illas video; Its very telling in how differently the system performs.

How so?

[MSH]

Field Oriented Control.

[JohnRobHolmes]

I can say with high confidence (right now, lol) that it is not true FOC since the drive signal is not matched to the motor bEMF. Matching the bEMF is really the entire point of FOC, even "basic" closed loop sensored FOC. Otherwise the rotor and stator flux is not orthogonally aligned and we fail to meet the entire criteria of FOC. The transition from high to low duty is far too linear when compared to the motor bEMF. At half throttle it is exactly 6 cycles low, 3 cycles linear duty rise, 6 cycles high, 3 cycles linear duty fall, 6 cycles low, etc... There simply is no Park transformation going on. It's a little more refined than 6 step commutation with a 14khz A/C carrier frequency overlaid to slap the motor into a synchronous speed, but it falls short of meeting the criteria that every other closed loop FOC system demonstrates.


I'm really trying to stretch the definitions of FOC to make the system fit. The hardware looks completely capable of closed loop FOC (sensored only operation, no active sensorless feedback), which is probably why they are calling it FOC. But somewhere along the path of development they stopped short of fleshing out the commutation table into nice sine waves and simply jammed a trapazoid rise and fall around block commutation. I'm super puzzled about this. Maybe it was "good enough" to run and they didn't think anyone would look into how accurate the attempt was?

[Frank211]

[QUOTE=Frank211;5835983]

Originally Posted by durok

You mean like a Turnigy GliderDrive?

It's probably just the microphone in the recording.

Until someone runs them back-to-back with the same audio equipment, we can't really compare them yet.

https://hobbyking.com/en_us/turnigy-...-5-1400kv.html

"Outrunner in a can"

[/QUOTlike]


I did not know about this motor. Would like to hear that motor ! (I learn something new everyday)

After watching the different vids, IT does not sound like a outrunner motor. I guess the mic made it sound like that.

[Jboucher]

Originally Posted by JohnRobHolmes

I can say with high confidence (right now, lol) that it is not true FOC since the drive signal is not matched to the motor bEMF. Matching the bEMF is really the entire point of FOC, even "basic" closed loop sensored FOC. Otherwise the rotor and stator flux is not orthogonally aligned and we fail to meet the entire criteria of FOC. The transition from high to low duty is far too linear when compared to the motor bEMF. At half throttle it is exactly 6 cycles low, 3 cycles linear duty rise, 6 cycles high, 3 cycles linear duty fall, 6 cycles low, etc... There simply is no Park transformation going on. It's a little more refined than 6 step commutation with a 14khz A/C carrier frequency overlaid to slap the motor into a synchronous speed, but it falls short of meeting the criteria that every other closed loop FOC system demonstrates.


I'm really trying to stretch the definitions of FOC to make the system fit. The hardware looks completely capable of closed loop FOC (sensored only operation, no active sensorless feedback), which is probably why they are calling it FOC. But somewhere along the path of development they stopped short of fleshing out the commutation table into nice sine waves and simply jammed a trapazoid rise and fall around block commutation. I'm super puzzled about this. Maybe it was "good enough" to run and they didn't think anyone would look into how accurate the attempt was?

I know this must be wayyy more complex than i see it but since hobbywing is already big in the racing and bashing markets, they should offer a partnership with you.
Holmes Hobbies designs the most badass and complete ESC / Motor combo then hobbywing uses it's manufacturing power in China to produce it.
Boom! Market cornered and profits!


Maybe you have an engineering batch AXE system and they will tune it with your reviews

Sent from my Nexus 6 using Tapatalk

[JohnRobHolmes]

If they want to tune it to my liking, it won't be pretty. I'm really trying to like this throttle. I'm trying so hard to not tell yall how much I hate the "velocity loop". They have done something unique, and that has a lot of merit. But 16 years of crawling on "regular" PWM voltage throttle response has engrained so much into my head and finger. If the crawling isn't tough, it feels great. But the moment there is any sort of hard bind, its effing annoying. No control over torque delivery at even the smallest amounts of throttle. It's simply trying to break my rig without my consent, and I'm having a really hard time getting past this.


Enough chit chat for now, back to work for me

[JatoTheRipper]

Originally Posted by JohnRobHolmes

I can say with high confidence (right now, lol) that it is not true FOC since the drive signal is not matched to the motor bEMF. Matching the bEMF is really the entire point of FOC, even "basic" closed loop sensored FOC. Otherwise the rotor and stator flux is not orthogonally aligned and we fail to meet the entire criteria of FOC. The transition from high to low duty is far too linear when compared to the motor bEMF. At half throttle it is exactly 6 cycles low, 3 cycles linear duty rise, 6 cycles high, 3 cycles linear duty fall, 6 cycles low, etc... There simply is no Park transformation going on. It's a little more refined than 6 step commutation with a 14khz A/C carrier frequency overlaid to slap the motor into a synchronous speed, but it falls short of meeting the criteria that every other closed loop FOC system demonstrates.


I'm really trying to stretch the definitions of FOC to make the system fit. The hardware looks completely capable of closed loop FOC (sensored only operation, no active sensorless feedback), which is probably why they are calling it FOC. But somewhere along the path of development they stopped short of fleshing out the commutation table into nice sine waves and simply jammed a trapazoid rise and fall around block commutation. I'm super puzzled about this. Maybe it was "good enough" to run and they didn't think anyone would look into how accurate the attempt was?

So can this be changed with software or is there "missing" hardware to accomplish what you want?

[redemon]

After reading John's take on it I wont be stocking them in my store. Nope. Just aint gonna do it.
Thanks John

[JohnRobHolmes]

I would assume they could get it close with just software changes. A modification to the lookup table would do it, but each motor kV will have a unique table to adapt to the resistance and inductance. Assuming real FOC is the end goal... Not a big deal to store that information in the controller, it's a brute force method that will limit how many motors will be compatible. Since this is closed loop and already not matching the bEMF, it can be assumed they also skipped over phase amp monitoring, which steers us away from any active formula based commutation being possible. In a nutshell, they have a system that works with one generic waveform and don't have much incentive to go the last mile. By all tests, this still fails to meet the criteria of a properly implemented closed loop FOC.

[JohnRobHolmes]

Originally Posted by redemon

After reading John's take on it I wont be stocking them in my store. Nope. Just aint gonna do it.
Thanks John

I wouldn't go so far as to tell people to stay away, as this is simply my opinion on the system. There may be a lot of folks that like the way the throttle feels, even though I don't. I would ask that you wait for the videos before making a decision, as I will be able to put all this information into a succinct summary.

[redemon]

But in my mind for that amount of money it just isnt worth it for me to sell it to someone else, when I can sell your motors to them and then a mamba x or something like that. Idk. Just my opinion anyway. If someone wants one I will for sure get them one. I just wont push it.

[JatoTheRipper]

Originally Posted by redemon

But in my mind for that amount of money it just isnt worth it for me to sell it to someone else, when I can sell your motors to them and then a mamba x or something like that. Idk. Just my opinion anyway. If someone wants one I will for sure get them one. I just wont push it.

One thread on the internet has swayed you enough to not carry a product? I mean you don't even wait for a hands on test?

[durok]

RPP Hobbies has the Axe ESC and the 3 lower-kv motor options, along with combos up for pre-order now.

I still find it interesting that the higher kv version of the motor is not part of the initial release...