Paging cmoalem...

AlexLTDLX

AlexLTDLX

Administrator
Staff member
I'm pretty sure you're aware of my spectacular APD ESC failure. I was looking into VESCs and saw you were on a forum over there. You clearly know a lot more about ESCs than I do. What are your thoughts on the APDs (I saw you found a schematic for them too), VESCs and MGMs? Educate me, please.
 
The beauty of the vescs is the Field Oriented Control (FOC) which allows for a smoother motor control ... lower "noise" and much better low speed.
For skateboarders this is da chit!
FOC vs simple trapezoidal phase control is like a sinus signal vs a digital/stepped/square control signal ... therotically you would waste less phase current, produce less heat and run the motor with better efficiency!

UNFORTUNATELY the needed compute power for FOC allows for only up to 150000 erpm (borderline) with their currently used processor.
The TP motors we are using are 6 pole, right? That is juuuust ok. (but I am a fan of "overbuilt" too) (The LMT motors are 2 pole! Those would use less computing power for FOC, but 2 pole rotors are controversial for other reasons !)
There is advancement on the FOC front where the FOC control algorithm is molded in hardware ... but not incorporated into the VESC software and design!

The other aspect I could not find reassuring is their newer design 75/300 VESC does not look beefy enough in the power stage as well. (I have noticed APD is using many more mosfets per phase!). Lots of burnt 75/300s and the higher voltage one high junk stack too.
While I love the FOC principles ... I am not sure yet we get the right stuff from Tampa!

Verdict:
Either get one of the high end MGM controllers (400 or 800 amps, good for 200000 erpms) oooor : talk to Trampa/Vedder if they sponsor the project with a 75/300 VESC (would cause us all to buy!) ... and you don't evaporate more dollars!
Problem with the MGMs especially for you ... max 15s LiIon voltage rating! (for me ... I am sticking with 48V (12s))

There is one newcomer company which is doing crazy high end powerstage VESCs ... currently still starting up ... and too early to be considered I think. (But if you want I can dig their name out again...)
 
Last edited:
uhh ... and I have started even looking at "Golf Cart ESCs" (actually 3 phase controllers) ... they have monster powerstages and are readily available for 48-72V systems ... but all of them expect "sensored" motors ... shhhucks
 
Having said all that: Vedder has a video on utube where he is testing a monster EDF motor/propeller on his bench ... with a VESC 75/300 (software control aspects only are discusses here) bolted to a massive aluminium plate/block ... and his introverted explanations about mosfets in some other video shows he knows his shit! (bigger mosfets vs more in parallel ... etc)
But he seems quite non chalant in hardware design ... slapping all in proximity on a single board.
(Plasma arc lurking?)
Thats why besides Trampa other innovative companies are designing high end high power ESCs based on the VESC software!

I ever since your earlier 3 layer powerstage (4 in total) APD tests have been wondering how tiny those ESCs are in comparison. Where does all the heat go, and what mil spec silicone goo have they poured in for isolation AND thermal conductivity???

(Now we all know ... magic is impossible!)
 
Last edited:
Thanks! That's great input. To the guys' at APD's credit, they acknowledged that there are some improvements they need to make, and even suggested having me test some of these improvements. They really are great guys and I feel bad things worked out the way they did. I've been busy emailing back and forth both with Frank at Trampa and Tomas at MGM. To be honest, Tomas has made me an offer I might not be able to refuse (no, it's not free, not close to free even, but a decent offer none the less). Even with their 63 volt limit, my Lishen battery packs should be able to supply the current needed. I need to do a voltage drop test again on the Lishen cells at 500-550 amps (I'll only use a few cells this time - no more arcs) to get an idea of what the voltage drop would be. This would be for their 800 amp continuous/1000 amp peak controller and an LMT 30100 motor (53 hp!). They're projecting 50 volts at 550 amps and 42,000 rpm (~37 motor hp) - which would put us around 850 hp. Funnily enough, that's still less than 5 ft-lbs of torque - my little 6mm hex drive bolt should be able to handle that. I hope.

Do you understand the 6 wire vs 3 wire connections on the LMT motors? That's one thing I'm still not clear about.

Frank at Trampa told me that with that motor in a 6 wire confugration I could use two 75/300 amp controllers, but frankly I'm worried about the complexity; and all the photos of fried VESCs have me a bit concerned. However, he did also say that they're about to release a unit with double the MOSFETs. That should be interesting. He did also warn me that even though the controller's rated at 75 volts, he'd stay 8 volts under that for spikes.

I'd love to hear your thoughts on all this, since you're clearly better versed on the ESC stuff than I am. (I also looked at golf cart/EV controllers - I spent some time on Kelley Controller's site last night too...)
 
AlexLTDLX said:
Thanks! That's great input. To the guys' at APD's credit, they acknowledged that there are some improvements they need to make, and even suggested having me test some of these improvements. They really are great guys and I feel bad things worked out the way they did. I've been busy emailing back and forth both with Frank at Trampa and Tomas at MGM. To be honest, Tomas has made me an offer I might not be able to refuse (no, it's not free, not close to free even, but a decent offer none the less). Even with their 63 volt limit, my Lishen battery packs should be able to supply the current needed. I need to do a voltage drop test again on the Lishen cells at 500-550 amps (I'll only use a few cells this time - no more arcs) to get an idea of what the voltage drop would be. This would be for their 800 amp continuous/1000 amp peak controller and an LMT 30100 motor (53 hp!). They're projecting 50 volts at 550 amps and 42,000 rpm (~37 motor hp) - which would put us around 850 hp. Funnily enough, that's still less than 5 ft-lbs of torque - my little 6mm hex drive bolt should be able to handle that. I hope.

Do you understand the 6 wire vs 3 wire connections on the LMT motors? That's one thing I'm still not clear about.

Frank at Trampa told me that with that motor in a 6 wire confugration I could use two 75/300 amp controllers, but frankly I'm worried about the complexity; and all the photos of fried VESCs have me a bit concerned. However, he did also say that they're about to release a unit with double the MOSFETs. That should be interesting. He did also warn me that even though the controller's rated at 75 volts, he'd stay 8 volts under that for spikes.

I'd love to hear your thoughts on all this, since you're clearly better versed on the ESC stuff than I am. (I also looked at golf cart/EV controllers - I spent some time on Kelley Controller's site last night too...)
Click to expand...

Great to hear all that!

I like the news about Frank/Trampa working on an uber VESC!

re 6 wire config:
There is on utube some mgm info on how to connect 6 wire motors with 3 wire escs ... and how to connect 3 wire motors to 6 wire mgm controllers. lemme dig that out later tonight.

(it is simple as far as i can remember!)

EDIT: The MGM controllers have 3 phases ... just 2 wires per phase:
 
Last edited:
Thx. I'll look at it either later tonight or tomorrow - I have some, umm, "parenting" to deal with and desperately need to clean my house - between the bathroom remodel and clearing out my dad's house, things have gotten out of hand. I don't function well when my environment's a mess. In an odd way, it's kind of good things turned out the way they did. I've been putting this stuff off for too long.
 
Wow... I'm gonna bet making one of these e turbo things is going come way down in price very soon...
 
Last edited:
some more details re latest hardware here:
endless-sphere.com

Axiom: a 100kW+ motor controller

/***************************************************************************************/ Jump to page 11 for the latest hardware! https://endless-sphere.com/forums/viewtopic.php?f=30&t=89056&start=250#p1458465...
endless-sphere.com endless-sphere.com

reading the thread from the start makes you even more excited! :
endless-sphere.com

Axiom: a 100kW+ motor controller

/***************************************************************************************/ Jump to page 11 for the latest hardware! https://endless-sphere.com/forums/viewtopic.php?f=30&t=89056&start=250#p1458465...
endless-sphere.com endless-sphere.com
 
Last edited:
Ok - I finally had a chance to watch that video about the 6 wires on an LMT motor. At first I was confused, then I saw the diagram and I thought "Oh, that makes sense" then I remembered that the earlier part of the video said to connect the first wire to ANY connector and I got confused again.

I know this probably has a very simple explanation, but I'm asking because I just don't know (and I'm sure I'm not the only one) - I understand that having 6 connections allows for a delta vs wye wind (image found online):

WyeDelta.jpg

And here's on of the best images I found that explains how multiple poles are wired to a wye configuration (this is apparently representative of a CD ROM motor; also found on line):

CDROMMotorWyeWind.gif

But how can you just randomly choose a connection on the back of the motor? Isn't it a 50/50 shot that you'll get between phases? What do they mean by Phase 1a vs Phase 1b, etc from an ESC point of view? I'm sure I'm missing something simple here...
 
AlexLTDLX said:
Ok - I finally had a chance to watch that video about the 6 wires on an LMT motor. At first I was confused, then I saw the diagram and I thought "Oh, that makes sense" then I remembered that the earlier part of the video said to connect the first wire to ANY connector and I got confused again.

I know this probably has a very simple explanation, but I'm asking because I just don't know (and I'm sure I'm not the only one) - I understand that having 6 connections allows for a delta vs wye wind (image found online):

View attachment 150

And here's on of the best images I found that explains how multiple poles are wired to a wye configuration (this is apparently representative of a CD ROM motor; also found on line):

View attachment 151

But how can you just randomly choose a connection on the back of the motor? Isn't it a 50/50 shot that you'll get between phases? What do they mean by Phase 1a vs Phase 1b, etc from an ESC point of view? I'm sure I'm missing something simple here...
Click to expand...
yes ... I get what you're saying ...

ESC wise wires A1 and A2 ore on the same end ... on the motor end if you start wrong you could really mess up.
The important take from the video instruction is "start from one PHASE at the motor" ... so it does not matter A, B or C ... but it should be A1 or B1 or C1

the pic at the end of the video suggests the phases might be marked on the LMT motors?
LMTphases.PNG


EDIT : nope ... the LMT motor poles are NOT marked:
LMTpic.jpg


but I found THIS attached pdf:

note: free choice of star vs triangle config doable on the same motor!? (huh? am I confused? YES)
 

Attachments

  • man_30_series_en_01_2020.pdf
    68.9 KB · Views: 4
Last edited:
after some brain pain (still persisting) I found this thread:

Lots of Delta vs Wye connection questions? - RC-Monster Forums

Lots of Delta vs Wye connection questions? Brushless
www.rc-monster.com

trying to digest that ...

and that :

New Lehner 3080/5 HiAmp LK "Twin Wind-12 connection" - RCU Forums

Marketplace - New Lehner 3080/5 HiAmp LK "Twin Wind-12 connection" - Lehner 3080/5 HiAmp LK motor with twin parallel winds allowing for two controllers to drive the motor. Please note: the Schulzes shown in the pics are not available with the motor. Motor kv is 1050 in Delta and 600 in Wye. The...
www.rcuniverse.com
 
Last edited:
I guess what we don't know is how the 6 connections on the back of the motor are wired to the windings. I suppose once I get my motor, I can take some measurements (I do own a milliohm meter). I don't know enough about the ESCs yet to really understand how their outputs would inter-relate. I do appreciate your poking around to try to figure it out.
 
BTW, I got this description of the VESC from Frank at Trampa:

"Hi Alexander, motor control is quite a tricky thing. The VESC controllers actually run on FW that is backed by the operating system VESC-Tool. This sets the VEAC apart from other controllers. It compares to standard BLDC controllers like cell phone to smart phone. Therefore we can precisely control the power and the RPM and torque at any given time. Regular speed controllers can only control the RPM but not the power output. They jump from one input value to another one with full power output.
That's why you don't want such a controller in a skateboard or on a scooter.

A VESC controller is way more sophisticated. You can choose the mode yourself. Typically we control the amps (currents) and indirectly we ramp up the speed (control the voltage). So we can control current and voltage and can therefore adjust the maximum power output, how smooth or agressive the motor should react, we can protect the battery by defining max amp draw etc. We can also plot up to 50 values to see what actually is going on. This helps you to analyse and tune your system.

A simple BLDC controller will not offer you all those features. You can adjust the voltage (motor speed) while the amp draw is typically not controllable at all. The motor will simply suck as many amps as needed to reach the next RPM level.
It jumps from one input level to the next one. The only way to get some smoothness is adding ramping time, so that the impulse is spread out over more time.

Please note that many BLDC controllers are rated according to their Mosfets. The mosfets of the 75/300 can theoretically push 900+Amps. However, that is a spike value. BLDC controllers can't control the current flow really, so they need to be made in a way that allows more amp flow. Since we can control the current flow we can tame the system and avoid huge spikes in Amp flow.

I will have a look into the Lehner motor you suggested. Let's see how much power it can output realistically and what the controller therefore needs to look like.

//Frank"
 
AlexLTDLX said:
I guess what we don't know is how the 6 connections on the back of the motor are wired to the windings. I suppose once I get my motor, I can take some measurements (I do own a milliohm meter). I don't know enough about the ESCs yet to really understand how their outputs would inter-relate. I do appreciate your poking around to try to figure it out.
Click to expand...
I might have figured it out ... first I thought it is a double delta ... but then I came to another schematic which is kind of "extended delta".

will post a pic later when I have computer access.
 
You must log in or register to reply here.
Top