High power ESC challenges

here information about the Hacker company. i have been flying hacker motors since 2016. esc and motor in trike since 2018. hacker also offers systems that work with 2 esc. the motor from the 200 and 240 series has 6 phase cables. there it means 2 winding circles in parallel. ( rotex motors and geiger motors have the same option ) these can be controlled with one esc or with 2 esc. S-Moto works together with hacker and now also offers UN-38.3 certified flight batteries. 14 S (52 V)

https://hacker-industrial-solutions.com/

1 Like

geiger has been the father of e-ppg worldwide since 2006. i have had a lot of contacts with users of the products since 2010.

the specialty is that geiger has a system that communicates internally with all parts. the esc knows how often the battery has been charged and how good its health is. as an an example. Geiger also equipped large aircraft. i was at a big meeting in 2017. a flying wing swift was towed here with e-trike. with more performance than comparable rotax systems.

https://www.geigerengineering.de/en/avionics/products

https://www.geigerengineering.de/avionik/referenzen

another company where i see very good potential in the future. I can’t post any long-term experience here.

1 Like

Hello,new motor bench tester almost ready

2 Likes

Hello, www.onemotor.ltd 15Kw continue 3.8Kg still in progress …
in the mean time take a look on the motor 3D

Gianni

3 Likes

Hello, ONEmotor is back as my core business of machine tools has ended and I have retired to pensione and now I can full dedicate myself to these projects!

I had problems on the server dedicated to the ONEmotor project but since yesterday everything has been restored including email: info@onemotor.ltd
and also WEB site www.onemotor.ltd

Meanwhile we have developed a VESC 100V - 100A for an Australian professional drone company which will be used as an advanced drone fan.


2 Likes

Tried to login – only going to join option and says I am already a member.
Ok – it was the lettering on page not letting me see the “login” – poor colour choice - fire your Web page person😂

I have been playing around with a VESC 100/250 on a 20kW motor, and I have to say I’m very impressed. Someone finally cracked the code for open-source ESCs.

2 Likes

Hi does this version of Vesc handle high pole count motors ? i.e > say 40 poles ,. Def will take a look at the website , also can the powerstage be beefed up to handle 250Amps @ circa 100V Cheers

My guess: Probably not at all.
…and I’d love to be wrong on that one.

I can tell you that we burned 3 original Trampa VESC 100/250 with the MAD M50 (which is a high pole count motor) + 140cm 3-Blade E-Prop. The last one created a serious flame why dying. It did not give any pre-warning and was not hot at all. This was in the 10kW load range with a 20S Battery and far away from the theoretical limits.
Luckily that was only on the test bench and nobody was harmed.

Please be careful!

The VESC appears to be great combination of hardware and software, but it appears to be mainly used with different motors (smaller, lower pole count, higher KV).

@kubark42 How far did you get when “playing around”?
It would be great if you could share your experience, exact setup (motor, prop, config and FW) as well as info on how far you were able to push the combination.
During the first test runs we were also playing around and it looked all super nice. Once your cross the boarder of where you want to go for a safe PPG takeoff (let’s say >10kW) it might be different.

@Daniel-Duesentrieb

There are recent developments on the beta firmware, which might allow us to drive a high pole number motor safely

In particular, the inductance calculation formula has been changed and Benjamin has conducted tests on a T-Motor U15 80 kV (a high pole number brushless motor) without burning the ESC, but processing with caution

A safe procedure is :

  • Set the PWM frequency at 30 khz
  • Set the Slow Abs max current stop to false
  • Set the Abs max current to a low enough value for the first run
  • Do the RL, lambda and gains calculation
  • go to FOC telemetry value page
  • Increase the throttle very gently, until you see oscillations in the phase current
  • once you see oscillations, immediately cut the throttle
  • adjust the parameters. In particular the observer gain and the inductance values
  • increase again the throttle, see if you can go higher, stop when you’re seeing oscillations and so on

I will try myself with a T-Motor U15L 70kv when I’ll have my Li-Ion battery finished

So which ESC do you currently use ?

1 Like

It does! The below setup is a 42-pole Geiger HPD20 driven by a Trampa VESC 100/250, using a MagCAD 48x24 folding prop, running at 60V

As can be seen by the graph, the ESC reached peak 300A (phase) with no instability. I was pushing about 14kW at the time.

The motor’s parameters are below:

It’s a low inductance motor, but not as low as they come.

@NexusG mentions some tuning strategies, I think this is the video (created by Benjamin Vedder himself) which sums them up:

However, in my case, after the math bugs identified in RL measurement incorrect with factor 2/3? | VESC Project were resolved, the motor identified perfectly on the first go and I did no tuning to generate the 300A graphs shown above.

@Daniel-Duesentrieb that sucks that you guys burned the controllers, doubly so if it was because of a math bug. Fundamentally, there’s no reason why these controllers can’t be used with larger motors. It’s taking a little bit of time to get enough eyeballs and testers, but once all these issues are ironed out the problems are solved for forever.

I’m not saying they’re all gone now, but the before and after difference of solving the math bug was startling on my motor. Before, I could only push a few amps before issues arose. After, it was like magic.

3 Likes

Cheers Guys some really useful info and good feedback , really appreciate the time taken to provide this and the sharing of knowledge
@kubark42 wow wow just read how this bug was found on Vesc Forum, you guys are on another level , abs great to observe tech guys bouncing ideas off each other and coming to conclusion to fix a overlooked bug, and hear the diffence this make to product capability …damm that was a good read Cheers

Of course we have been following the Beta Firmware closely. I was in contact with Frank and he suggested using them, but that was before the math bug was discovered. We also got a LCR meter to get a more precise measurement of the motor inductance. Before the last one burned we did not had any (more) oscillations in the phase current.

That said I might be better now with the latest VESC beta firmware, but I am not putting any more 500€ chips on that table. The cost of being a beta tester is way to high here.

I would have love this solution to work, because I absolutely like the Trampe WAND remote. (I have one on sale now :wink: ) The finish and the integration with the VESC is great. I also belief that once their VESC BMS is finally released and stable it would be a very powerful combination, especially if the VESC 100/500 is good to go, too.
The BMS is the SP140 is way too basic. No bluetooth, no info, no settings => no clue about the state of the cells. It is black box and I’d rather spend 30 USD more and get the version with Bluetooth to at least get a chance of viewing the cell voltage difference.

@NexusG and @kubark42 please share your results on the T-Motor U15L and the Geiger motor!

  • Have you been able to reach anything close to the SP140 setup, which is able to push around 20kW?
    Now that I have flown the SP140 in chill and in sport mode I can say that I will never use chill mode again. I got around up to 10 kW in chill mode and that was only good for up to +1m/s at a 31sqm wing on a light trike. 10kW sounds more than it feels like. I can keep altitude at 4-5 kW, but there are situations when you definitely want more than that.
  • What battery setup are you running on (in particular anything close to a 100V)?
    I am asking because strictly speaking we don’t care for Amp throughput, but for thrust created at the prop.

I’d like to take a moment to appreciate the work Paul + team have been done by putting to together a ready to use system and taking care of all the test loops necessary to get it into the air.

3 Likes

Is 14kW close? It’s 75% of the way there, so suitable for some applications. My threshold is 16kW, with a target of 18kW. I don’t think I’ll get all the way to 20kW, but it would be cool if it happened.

I’m running two honking big Lambda/TDK 10kW DC lab power supplies, no batteries as of yet.

The nice part about doing it this way is that I don’t have to build a battery pack and hope it’s okay. I can just test the voltage and the prop and see what comes out of it.

1 Like

16KW = 21.4HP
18=24
My Simonini 202 is suppose to put out 24HP so 18 would be nice.
20 would be really nice – better to have the extra power and not need it, than to need it and not have it😉
Cheers

I’ve speculated before about splitting the poles of a motor in 4 sections to continue to use the hobby controllers and have redundancy. As someone noted above the four X4 controllers are less than the one SP140. I have had zero luck of finding documentation where that motor configuration has been tried before though.

It amuses me that inside a motor there are dozens of individual poles, which are strung together rather inefficiently into 3 groups, and then there is an ESC with a huge number of FETs, which have to be strung together into 3 groups to handle the load of the group of poles… when it makes heaps more sense (efficiency) to have just a small number of FETs dedicated to individually driving each pole, rather than all that grouping and unnecessary wire routing… and if you took the time to mount the FETs into the motor itself, you further increase efficiency and reduce cooling issues as well…

It’s strange how “innovative” everyone thinks the world is, when in reality, 99.99% of everyone is just copying everyone else’s ideas, including all the dumb parts…

1 Like

Yes, it is true that a lot is being copied. But there is also a lot that many do not know … You can introduce us to your Eppg that you are flying.

https://plettenberg-motoren.net/en/products/motor-solutions/motors

Possible, but maybe not practical. You are adding the complexity to the controller side. Each pole needs to be turned off and on, doing this in groups is easier and provides more strength – ie, 3 poles vs 1 pole. Would you have enough torque to control/hold the prop - is that the KV?
Why are pancake motors becoming more popular? More poles?
Cheers

Poles are obviously energised at the correct times of course, so there’s never any diminishment of strength, simply the reduction of “waste” in the form of the interpole wiring losses that are now gone, and if you take the time to smarten up the timing per-pole, you also gain some efficiency by matching the ESC to the motors physical characteristics with much more precision possible.

“Easier” is not a goal anyone should optimise for, by the way…