Homebuilt, both PPG and FLPHG. See my github for the canbus display work I’ve open-sourced for these (as well as cars, and see foils.pro for the (also free) software I’ve published which (among other things) generates perfectly-matched props for specific usage regimes.
Agreed - but seems to be the way of the world.
That is a lot of individual wires going to the controller with their own connection points. And to what end gain in efficiency?
And now they are looking a magnetless motors or hybrids with both magnets and a controlled energized rotor coils, so more wires and even more complex controllers, higher cost to the end user.
Getting ride of magnets I agree with - motor will basically last forever with just bearing changes
Cheers
Ah ok. I thought you show us a completely new eppg that is not copied. I expected that after your 1st posting. my expectation was probably too high.
I’ve flagged this post. It’s an abusive attack on a new community member, and in direct contradiction to the site’s stated policy:
I hope the moderators take it seriously when members, even experienced and esteemed members such as @bratwurst, escalate the tone. It’s distracting and not helpful to our conversations.
There are indeed reasons why ESCs aren’t done the way @gitcnd wonders, but that’s not addressed in a snide attack on a new user who has done nothing more than opine-- correctly I may add-- that the current standard 3-phase approach is not optimally efficient for power conversion.
Yes… and no. As an engineer with a background in optimal control, I love anything which makes things that little bit better. However, IMO, efficiency is not a big deal with motors and controllers, they’re already so efficient that a little bump in efficiency only matters for thermal issues. The big hurdle with electric propulsion is reliability, specifically the reliability of the ESC. They tend to work up until the microsecond when they fail, and then we have very sad pilots who are going to touch down sooner and faster than they’d hoped.
ESC reliability is a function of many things, only one of which is heat. Getting to the topic at hand, easier to manufacture and to support are key factors in reliability, and until we have a variety of systems which work without fail it’s probably early to optimize for efficiency.
Putting the FETs in the motor gets rid of those wires entirely, as well as greatly increasing the cooling options…
I’ve written a few basic ATMEGA328p ESC programs for fun, then invested in a professional dev platform (which I’m yet to find time to play with) - the efficiency improvement that excites me the most is adding “smarts”: because we’re spinning a prop, and usually flying level, that’s the perfect scenario for the ESC itself to make micro-adjustments to PWM freq* and timing on each phase (which, like I suggested, should be more than just 3 - e.g. every individual pole) to optimise for lowing current while maintaining RPM…
- At higher RPM, I have the suspicion that it makes no sense to use PWM at all - there is just one optimal period of time that a coil needs to be energised - spreading that out via PWM to include the less-optimal “shoulders” of that moment is a double-whammy - you’re losing efficiency by switching the FETs for no reason, just to lose more efficiency by energising the coils at suboptimal moments…
“already so efficient” - I’m not sure I agree with that. Just because some marketing department slaps “99.99%” on the box, does not make that true in any way… if it’s not using scientific units that can be measured, it’s not true. Genuine efficiency would be measured in something like Nm/kWh or whatever, and never the unitless and unexplained “%” that we always see…
It sure helps your ROI though. If you want to create something useful, or of value, to more than one person, compromises ensue. Design, production, materials, safety, reliability, time… so many variables…
I think @bratwurst was being blunt, skeptical, maybe cynical. I did not read his post as an attack. Certainly not adhominum. More of a challenge. Tone is very hard for me to judge in a text only communication. 
1 Like
IMHO, it poses some substantial challenges for cooling. The FETs need a really good heat spreader, and the FETs won’t have a ton of places to dump heat when the motor itself is a heat source. Not saying a revolutionary idea couldn’t solve this thermal problem in an elegant way, but AFAICT it’s not an incremental step.
Another important part about ESC layout is keeping the FET input pads extremely short so that there is minimal inductance. This might be doable with some creative design, but it’s again an unproven approach which increases risk at a time when risk is not already under control.
Don’t get me wrong, I like the idea and I see what you’re getting at. If you wanted to hop on the VESC Discord server, VESC development, you are very welcome to come chat with us about it.
What I mean is that adding an extra 1% efficiency hardly directly impacts how high or long the aircraft can fly. If heat is reduced past certain critical thresholds then the thermal impact can be a longer climb or a higher top speed, but those are secondary effects.
You might be interested in DTC (Direct Torque Control). We had a good conversation on it here: Discord
I ask myself these questions: 1) did the post advance the conversation? 2) Does it state overtly or by implication that a user is inferior and thus does not have valid opinions?
No matter the tone that was intended, I think we can all agree on the answers to these two questions. I think we can also agree we shouldn’t find ourselves in these situations. Moderation doesn’t have to be heavy-handed, it can just be a private word from a moderator to ask someone to think twice before venting, esp. at new users. But, IMHO, it does need to happen in this case so that the community can hold itself to the highest standards.
In particular, we want to be careful about judging commenters by their PPG experience (which bratwurst seeming does in the flagged post). I personally am not here to fly a PPG, because I fly fixed-wing gliders. I am electrifying my self-launching AC-5M and as a result have very similar problems to you guys. Indeed, gliders and PPG use identical powerplants (mine right now has a Zanzoterra MZ35 and I’m swapping it out for a Geiger HPD20), where the only difference is the prop pitch (PPG fly more slowly so the optimal pitch is flatter). So it’s really useful to participate on this forum because we use the same controllers, motors, batteries, BMS, and propeller manufacturers.
1 Like
yes it was my mistake. I have to learn to talk to pilots who actively fly with electric motors. The people in the forum who post their legitimate opinion on technology and science here will not be bothered by my comments in the future.
1 Like
If you’re not going to take advantage of the individually FET-ed poles this wouldn’t necessarily be the case. You would simply have the controller firing 3 phases in whole blocks as normal. “What’s the point?” you might ask. If one FET block lets out the magic smoke - it’s only the loss of one pole and can’t cascade to an entire phase. But with FET-per-pole you could also run multiple controller redundancy without the expense of multiple FET blocks.
If you are going to take advantage of individually FET-ed poles, then you could dive deep into the weeds on how best to fire the poles since you’d want to find the balance between the lowest frequency of switching (running fewer poles at a higher duty cycle) but then you’d have to consider heat buildup - so building in some kind of skip-fire on the poles and finding the optimal point where the crossover is for introducing new poles, etc. It would be interesting to figure out how much partial-throttle efficiency could be gained. It could probably be partially simulated with a 3 phase controller by only firing a single phase per rotation once the prop is moving, though it might come at the expense of vibration.
Hang gliders and foot-launch sailplanes are superior to paragliders in almost every measure - but there’s at least a 2:1 disparity between PG and HG numbers. In the early days of the sport - the VJ23, Easy Riser, Eipper Quicksilver, et.al were far superior to the Rogallo in performance and safety. What won? The Rogallo - it was lighter and it could pack down in minutes and be tossed on a car roof for transport. Fast forward a few decades and the PG has had some development…the speed range, penetration, L/D, and sink rate isn’t as good, but it’s even lighter and easier to transport and ground handle than an HG. So easier sometimes is what to optimize for.
1 Like
I see it has been discussed about this APD HV 24s 200A(300 burst) ESC (see photo) at the begining of the topic. I understand it has been modified by openppg by adding larger case for heat dissipation, what was it used in?
Would it handle the SP140 M50 34kv motor with throttle management like only chill mode (max about 70%) throttle?
APD website original Esc link:
Esc specs:
Operating within a 32-bit ARM architecture, the HV_Pro ESCs are designed to deliver fast throttle response that is linear over the complete output range of up to 3****0 kW . APD’s proprietary developed firmware ensures unmatched low-throttle and starting characteristics. The ramp characteristics can be configured for fast response or drive-train protection. Advanced features include PWM-frequency dithering to avoid PWM-commutation synchronisation, adaptive motor timing advance and proprietary developed Active Phase-Current Demagnetisation (APCD). APCD along with Synchronous rectification ensure maximum ESC efficiency is achieved.
…
General Features
- Plug and play firmware designed in Australia
- Up to 24S (100 V)
- Up to 200 A continuous* current with airflow
- 300 A burst current
- True 12-bit throttle resolution
- Feed-forward compensator
- Up to 500 Hz input update rate (standard PWM)
- Over 1,000,000 eRPM
- Motor agnostic
- Auto PWM output frequency
- Auto timing advance
- Onboard 16 Mbit flash memory for datalogging
- Firmware updates through the micro-USB port
- Isolated Input and output (can be configured for RPM)
- Fully configurable through the APD ESC Config Tool (micro-USB connection)
- Optional external capacitor board for >1m input lead lengths: Cap Bank
Weight: 0.26kg
Dimensions: 96x52x26mm
On the MAD website the ESC recomandations for the M50 motor with an 160cm prop is surprisingly an 200A esc! Although we know that at full throttle it gows beyond that like 256A with the 140cm prop (Pauls video)
Botom line, if one would stay at max 70% throttle, and got an offer for this 200A esc, would it be worth consider it for the SP140 motor?