6:01 So here the responsiveness of all ESCs and firmwares are about the same. In your ESC comparison video a month earlier the BL32 was 2x faster. How can that be ?
Wondering why you haven't included noise as Drone Mesh used to do, I'm still on analogue and noise in my camera feed is probably the main thing that puts me off certain brands/esc's.
It'd definitely be interesting data to see. Looking at the resulting current plots from his other video on this new ESC it's a reasonable assumption that it'd have less noise as it's sudden changes in voltage and current that cause noise.
It would be very interesting to see that different settings on the ESCs will change on this tests. The fettec ESC has probably just more aggressive default settings.
I'm sorry, but.. what exactly does this testing show? That ESCs manufactured by different parties, with a different Amp-rating, running different firmware, at a different age - perform differently? Did you expect different?
I wonder if there is actually any battery for FPV drone that would be able to fully power all 4 motors at 100% throttle as those ESCs are actually capable of?...as I understand those 55A ESCs can deliver 55A of continuous current per motor...for example my Wasp Major datasheet says that it takes about 30A on 100% throttle...that would be 120A in total...I see that those ESCs should handle it without any problem with huge margin but can any battery at standard capacities 1300-1500mAh even deliver this power? Especially considering that the C number is a marketing bullshit and real C number is usually like 1/4 of the advertised? So even Tattu R-line with 150C can delived just about 40A continuously which is only 10A per motor and shortly maybe 100A which is just 25A per motor and it sags quickly...I tested CNHL Black 1300mAh to push it to the limit and it just capped at 84A and didn't go higher, which was barely 20A per motor, which is 2/3 of its max power...
In future testing can you include the flywheel test from a dead stop to full speed? to test the sensorless startup performance. Does it cog at all? does it spin the correct direction on the first try?
I agree, also would have been better to use a good powerful 4s high kv motor like a 2700kv on 6s to see which can actually make the highest rpms before it starts clipping and if it actually desyncs...
Thank you very much for showing this test. I am not using drones, but have started to look for different solutions with BLDC motors, and it have been hard to find specific information on how different ESCs perform regarding speed step response. I do however have a general good understanding of electrical motors. Your information here is the best I have seen so far. You and few other UA-camrs have provided information on how the choice of motor do influence the response time, that you like to be fast for good flying performance. But it is interesting, that the choice of ESC also makes a difference. When I study the breaking performance, I should think that there should be room for improvement for all the ESCs, and especially the new FETTEC ESC with better current sensing should be able to do that. So perhaps they will do it in future firmware versions. With acceleration you got some natural limits, and perhaps it might not be possible to do it much better than shown.
How about adding some measure of efficiency? Perhaps compare current draw at some standard RPM on some standard prop. Or perhaps spin a generator and plot power consumed vs power generated with a sweep over RPM?
I afraid energy in the capacitor is negligible in the context of flywheel. It can only help with peak microsecond current spikes... Supercapacitor bank could have helped with 200A loads :-)
Really interesting work. I know, it's just asking for more - but - I'd be interested to see if the performance remains the same when all 4 outputs are performing the tests at the same time. Do you think that it's worth normalising the results for the last test due to the different current capabilities of the ESCs? The FOC approach seems to generally be advantageous on the T-Motor Alpha ESCs so, I'm not surprised to see that FOC performs well at the FPV scale.
I was wondering something along those lines... what was the current draw on each of those ESC's vs the performance. Could they be pulling at different rates? Maybe the impact on the batteries flight time would be trivial.
It looks like on your chart the SpeedyBee gets the edge on the last test. I do see the new FETTEC overall doing better but how much of a difference will that make in actual flight? Just a thought.
I guess one thing that wasn't mentioned would be are these ESCs just on default settings? Or was there certian things done in the settings to help out as best as possible? Also another ESC that would be interesting to see in the mix would be Advanced Power Drives (APD). They're more common with xclass and heavy lift cinematic rigs. But like FETTEC and Flyduino, APD uses closed source software for it's ESCs.
Could you you please add the settings used on the ESCs to the description? (especially the decreasing resolution of the bluejay-ESC in the rampup/down smells like 48kHz was used)
NeutronRC stuff should be avoided. They are a poor company with no customer support. This team steal open source software as well and don’t respect the licensing agreements used.. if you support them your doing the rest of the hobby a disservice
Great system engineer test, as always Chris. But let me ask the question to Responsive - Acceleration test. Is the FETTEC really the best in this test? It can reach the 90% level faster, but I suppose overall generated thrust is bigger for Speedybee ESC - when you calculate area under the curves.. btw. can you keep same colors for ESCs in charts? :)
The FETTEC rpm data looks concerningly clean. I would take a look and make sure they aren't sending the RPM setpoint over dshot and not the actual measured eRPM.
I would argue that 10ms of acceleration delay is not comparable to 10ms of video link latency. You're talking about 10ms to get from 85% to 90% rpm. It doesn't result in anything close to the pilot seeing where the quad was 10ms ago like you get with video/control latency.
The FOC motor control should be audibly quieter than the traditional trapezoidal control. Probably where the efficiency is gained; less power lost as sound
As expected, these days electronic is on so high level that differences are cosmetic IMO. What you can really compare is EMC performance, like resistance for some electrical shocks while unprofessional user is setting up a drone or, even more interesting, EMC emission (conducted, radiated) to check real impact on other parts of setup...
Funny how the one esc that has over a 15% higher continuous amp rating then all the others has slightly higher acceleration... Why do these tests never seem to have equally matched up test subjects??? I'm sure the fettec fan boys are clapping like seals... Now lets make them cry and do overload and sudden stop testing where you repeatedly do heavy load accelerations a bunch of times but then add a brake to slam it to a sudden stop under load to simulate a prop strike crash scenario to see which one fails first...
Amp rating shows how much Amps it is rated for. A bigger number will only withstand more load, and a smaller rated esc will be damaged more easily, for example.
There is still a larger path of flow do to the bigger mosfets with the same signal input, if it tells the gate to open a certain amount that amount is going to be larger... Why not just test a 65a against a 65a??? His tests seem to always be skewed in some obvious manor
How come the BlueJay ESC shines in motor responsiveness in these graphs, but in other vids Rosser has shown that BlHeli was twice as responsive as Bluejay was?? There seems to be inconsistencies in this guy's testing...
Just installed the speedybee the other day(actually it's the newest 60A version), nice to see how it does considering the price. Thanks again Chris, for all you do!
Sorry Chris but I'm very confused about this testing. What exactly are we comparing here? You have different ESC hardware, different firmwares and even different sizes. You have 3 variables here. Even within the same manufacturer they sometimes change parts due to availability adding even more variables. With BLHELI32 and AM32 it would be more interesting, imho, if you flashed them to the same ESC. Unfortunately with bluejay I understand that's not possible. You also don't describe what settings you are using, they could vary wildly. Lastly, if the goal is to then test different ESC hardware it should be run on the same firmware, no?
Great points. I'm testing all the ESCs on default settings for the FW listed. In general you don't have much of a choice of ESC firmware on a particular board. You can theoretically flash AM32 to some BLHeli32 ESCs but I wouldn't advise this. Better to buy an AM32 ESC in the first place. My hope is to build up a database of ESC performance like I have for motors and batteries.
@@ChrisRosser maybe it would make more sense to test a bl32 ESC vs an old Fettec ESC with similar ratings (as similar as they can get). Then test the old Fettec ESC vs the new prototype and treat your test as a closed "input-output-system" like measuring the motor torque and braking/acceleration capabilities with an external device
Hi Chris. Is it ok for everyone to stop buying BLHeli32? I heard few months ago that it's dead and we cannot patch it anymore. My brother in law bought a $70 Foxeer Reaper few weeks ago and now it's just a paper weight. He cannot install firmware. It was showing PROTO in BLHeli32. The motors were showing letter T or Y in all motors instead of 1,2,3 and 4.
Wish I wasn't color blind. I see a red, orange, blue, and two dark blue lines. But they are not the same color as the 2 light blue, purple, red, and green lines I see at the bottom of the screen. Leaving me only able to tell that the MAMBA is at the bottom, and the Sky stars is above that, and the other 3 are above that. I cannot tell which are which. Maybe pick more contrast in the colors. And when they are next to each other use white, black, red, blue. And then when one is an outlier at the bottom give it the green or purple that is harder for people to see. 1/12 people have a significant deviation to their color vision. so at least 500 people must share some difficulty like me in viewing these graphs.
Could you please zoom in the interesting parts of the charts? Throttle ramp chart below 50% is totally uninteresting after seeing it's basically the same. The magic happens where we can't see any difference, because it's so tiny
I really do appreciate all of your testing an most of the time I respect your methodology. But here you present this as a test of firmware but your testing is extremely flawed with such a variation in ESCs. Maybe you could repeat these tests with a few ESCs that can be flashed with different firmware so that you can verify that it is indeed the firmware that is consistently responsible for the variation in performance
I agree 💯. This isn't a comparison of FW in isolation. Only of the specific ESCs on test. I fully agree we need much more data to determine if differences are due to FW or ESC design.
Are you serious rn? Since when do you need to be able to fly or "use" the things you are testing and developing? We have 20 drivers in the formula 1, do you think any of them can tight the wheelnut properly? Does a pilot of an aviation aircraft need to know how to build it? Come on. You dont need to be able to use it in order to develop or test it. Using is subjective. Testings are objective. If you chose your parts subjective on your feelings thats totally fine, but dont assume others work and effort and FACTs with your opinion, because your opinion isnt a fact at all.
Bluejay being so cheaper, it is really shining there! 😄
I would recommend measuring current consumption on all tests.
does changing the timing on the ESCs make any difference to these results?
6:01 So here the responsiveness of all ESCs and firmwares are about the same. In your ESC comparison video a month earlier the BL32 was 2x faster. How can that be ?
exactly what 'I'm wondering.
My guess is the ESC is not stock anymore, settings have been changed.
@@Hitman12. Or BLHeli has payed 2x more money then the others ? Hahahahah
Great video. The flywheel testing is very helpful for people in combat robotics.
Wondering why you haven't included noise as Drone Mesh used to do, I'm still on analogue and noise in my camera feed is probably the main thing that puts me off certain brands/esc's.
It'd definitely be interesting data to see. Looking at the resulting current plots from his other video on this new ESC it's a reasonable assumption that it'd have less noise as it's sudden changes in voltage and current that cause noise.
It would be very interesting to see that different settings on the ESCs will change on this tests. The fettec ESC has probably just more aggressive default settings.
I would also include electric noise as a parameter. This is also important for the other sensitive electrical components in the quad.
a parameter dronemesh used which sounded useful indeed!
@@ORBinatusdronemesh is pathetic omg
@@sigterm15 yet he was amongst the first to search for a way to identify how good ESCs would be.
I'm sorry, but.. what exactly does this testing show?
That ESCs manufactured by different parties, with a different Amp-rating, running different firmware, at a different age - perform differently?
Did you expect different?
Think he's just testing the rig out more than anything, if you would like to buy a bunch of brand new escs for him that would be great👍
I wonder if there is actually any battery for FPV drone that would be able to fully power all 4 motors at 100% throttle as those ESCs are actually capable of?...as I understand those 55A ESCs can deliver 55A of continuous current per motor...for example my Wasp Major datasheet says that it takes about 30A on 100% throttle...that would be 120A in total...I see that those ESCs should handle it without any problem with huge margin but can any battery at standard capacities 1300-1500mAh even deliver this power? Especially considering that the C number is a marketing bullshit and real C number is usually like 1/4 of the advertised? So even Tattu R-line with 150C can delived just about 40A continuously which is only 10A per motor and shortly maybe 100A which is just 25A per motor and it sags quickly...I tested CNHL Black 1300mAh to push it to the limit and it just capped at 84A and didn't go higher, which was barely 20A per motor, which is 2/3 of its max power...
In future testing can you include the flywheel test from a dead stop to full speed? to test the sensorless startup performance. Does it cog at all? does it spin the correct direction on the first try?
I agree, also would have been better to use a good powerful 4s high kv motor like a 2700kv on 6s to see which can actually make the highest rpms before it starts clipping and if it actually desyncs...
@@DigitalCruzer That would probably be more helpful than any of the measurements taken here...
The testing is great. After the stress test with the flywheel, a temp comparison would be nice. The cutoff voltage of each is needed too.
Please include ESCape32 fw in the testing!
Thank you very much for showing this test. I am not using drones, but have started to look for different solutions with BLDC motors, and it have been hard to find specific information on how different ESCs perform regarding speed step response. I do however have a general good understanding of electrical motors.
Your information here is the best I have seen so far. You and few other UA-camrs have provided information on how the choice of motor do influence the response time, that you like to be fast for good flying performance. But it is interesting, that the choice of ESC also makes a difference.
When I study the breaking performance, I should think that there should be room for improvement for all the ESCs, and especially the new FETTEC ESC with better current sensing should be able to do that. So perhaps they will do it in future firmware versions. With acceleration you got some natural limits, and perhaps it might not be possible to do it much better than shown.
How about adding some measure of efficiency? Perhaps compare current draw at some standard RPM on some standard prop. Or perhaps spin a generator and plot power consumed vs power generated with a sweep over RPM?
Seems like the speedy bee esc is kinda the best price to performance
nah theyre junk dont do it lol
@@terpsquirter7107I have on all my Quads Speedybee F7 and F405 FC and 50A (Old ESC) and on two Quads the new 55A ESC. They are awesome!
Yet they seem the noisiest..
I afraid energy in the capacitor is negligible in the context of flywheel. It can only help with peak microsecond current spikes... Supercapacitor bank could have helped with 200A loads :-)
Really interesting work. I know, it's just asking for more - but - I'd be interested to see if the performance remains the same when all 4 outputs are performing the tests at the same time.
Do you think that it's worth normalising the results for the last test due to the different current capabilities of the ESCs?
The FOC approach seems to generally be advantageous on the T-Motor Alpha ESCs so, I'm not surprised to see that FOC performs well at the FPV scale.
I was wondering something along those lines... what was the current draw on each of those ESC's vs the performance. Could they be pulling at different rates? Maybe the impact on the batteries flight time would be trivial.
It looks like on your chart the SpeedyBee gets the edge on the last test. I do see the new FETTEC overall doing better but how much of a difference will that make in actual flight? Just a thought.
Thanks Chris for the test. The result of Bluejay ESC really impressive compared to the price of BL_32.
How did this compare to a vesc using simplefoc?
Excellent testing, Chris! Thanks a bunch! 😃
Stay safe there with your family! 🖖😊
I guess one thing that wasn't mentioned would be are these ESCs just on default settings? Or was there certian things done in the settings to help out as best as possible?
Also another ESC that would be interesting to see in the mix would be Advanced Power Drives (APD). They're more common with xclass and heavy lift cinematic rigs. But like FETTEC and Flyduino, APD uses closed source software for it's ESCs.
Would it be possible to extend your flywheel test starting from 0 RPM? Really curious to see the differences there.
Could you you please add the settings used on the ESCs to the description?
(especially the decreasing resolution of the bluejay-ESC in the rampup/down smells like 48kHz was used)
BlueJay was 24kHz for this testing.
@@ChrisRosser you should test different resolutions of bluejay
Right? WTF?
Please test the FrSky Neuron 2, AtomRC and sequre ESCs. Thank you.
NeutronRC stuff should be avoided. They are a poor company with no customer support. This team steal open source software as well and don’t respect the licensing agreements used.. if you support them your doing the rest of the hobby a disservice
Why is the foxeer 65a reaper not in this test which is considered to be one of the best 65a ESC’s out there?
Great question! I don't have one 😔 If you have one get in touch and I'll test it and return it to you FOR FREE. Open door testing policy!
thank you my friend
Great system engineer test, as always Chris. But let me ask the question to Responsive - Acceleration test. Is the FETTEC really the best in this test? It can reach the 90% level faster, but I suppose overall generated thrust is bigger for Speedybee ESC - when you calculate area under the curves..
btw. can you keep same colors for ESCs in charts? :)
Unless you are reaching the peak RPM this will be compensated by a slightly higher thrust gain in the controller.
what about efficiency?
The FETTEC rpm data looks concerningly clean. I would take a look and make sure they aren't sending the RPM setpoint over dshot and not the actual measured eRPM.
Why do you not do the math to show the torque vs RPM curve (and power), after doing the inertia dyno pull?
ok ok waiting for the new Fettec ESC I hope it will be good for KISS Ultra v2
i wonder how will the new speedybee 60A will perform.
In all the tests is possible that the fettec esc is winning because has more amps rating? It s a fair comparison?
Thanks oracle🙏🏻
wen cine35 best motors review?
Any chance you could verify the ESC's aren't lying in their telemetry data as part of future testing?
Very good job!
Great man you videos are gems 👍
the contrast becomes nigh/day difference the more demanding conditions. I'm humbly surprised at how well the speedy bee esc did
I would argue that 10ms of acceleration delay is not comparable to 10ms of video link latency. You're talking about 10ms to get from 85% to 90% rpm. It doesn't result in anything close to the pilot seeing where the quad was 10ms ago like you get with video/control latency.
The FOC motor control should be audibly quieter than the traditional trapezoidal control. Probably where the efficiency is gained; less power lost as sound
how is the result in this video is so much different from the last sec test video you made?
Firmware versions are newer for all the ESCs and the test is much higher resolution now 👍😁
Would be great to add thermal footage of the ESC as well. :) budt otherwise perfect testing
As expected, these days electronic is on so high level that differences are cosmetic IMO. What you can really compare is EMC performance, like resistance for some electrical shocks while unprofessional user is setting up a drone or, even more interesting, EMC emission (conducted, radiated) to check real impact on other parts of setup...
Put the threshold to 85% and the outcome is different.
Actually only the Blheli32 ESC is significantly slower. The rest is +/- the same in my opinion.
Am32 is at 2.08 firmware now.. it performs a bit different to 1.99. The other esc’s are all running modern firmware
1.99 was the latest when I did the testing but that's why I made a point of detailing the FW version because things can change quickly.
@@ChrisRosser makes sense 2.08 is a solid improvement over 1.99
I would love to see Tests with the new Flywoo Goku F722 Stacks 😊
Funny how the one esc that has over a 15% higher continuous amp rating then all the others has slightly higher acceleration... Why do these tests never seem to have equally matched up test subjects??? I'm sure the fettec fan boys are clapping like seals... Now lets make them cry and do overload and sudden stop testing where you repeatedly do heavy load accelerations a bunch of times but then add a brake to slam it to a sudden stop under load to simulate a prop strike crash scenario to see which one fails first...
I wanna see some durability testing, I remember fettec was also good at going in flames.
Besides, my issue with this methodology is also that we're trusting the ESCs' data by using bidirectional DShot. FETTEC is closed source iirc
Amp rating shows how much Amps it is rated for. A bigger number will only withstand more load, and a smaller rated esc will be damaged more easily, for example.
There is still a larger path of flow do to the bigger mosfets with the same signal input, if it tells the gate to open a certain amount that amount is going to be larger... Why not just test a 65a against a 65a??? His tests seem to always be skewed in some obvious manor
I think you've misunderstood how these components work.
How come the BlueJay ESC shines in motor responsiveness in these graphs, but in other vids Rosser has shown that BlHeli was twice as responsive as Bluejay was??
There seems to be inconsistencies in this guy's testing...
Just installed the speedybee the other day(actually it's the newest 60A version), nice to see how it does considering the price.
Thanks again Chris, for all you do!
Sorry Chris but I'm very confused about this testing. What exactly are we comparing here? You have different ESC hardware, different firmwares and even different sizes. You have 3 variables here. Even within the same manufacturer they sometimes change parts due to availability adding even more variables. With BLHELI32 and AM32 it would be more interesting, imho, if you flashed them to the same ESC. Unfortunately with bluejay I understand that's not possible. You also don't describe what settings you are using, they could vary wildly. Lastly, if the goal is to then test different ESC hardware it should be run on the same firmware, no?
Great points. I'm testing all the ESCs on default settings for the FW listed. In general you don't have much of a choice of ESC firmware on a particular board. You can theoretically flash AM32 to some BLHeli32 ESCs but I wouldn't advise this. Better to buy an AM32 ESC in the first place. My hope is to build up a database of ESC performance like I have for motors and batteries.
@@ChrisRosser maybe it would make more sense to test a bl32 ESC vs an old Fettec ESC with similar ratings (as similar as they can get). Then test the old Fettec ESC vs the new prototype and treat your test as a closed "input-output-system" like measuring the motor torque and braking/acceleration capabilities with an external device
Hi Chris. Is it ok for everyone to stop buying BLHeli32? I heard few months ago that it's dead and we cannot patch it anymore. My brother in law bought a $70 Foxeer Reaper few weeks ago and now it's just a paper weight. He cannot install firmware. It was showing PROTO in BLHeli32. The motors were showing letter T or Y in all motors instead of 1,2,3 and 4.
Three cheers for Bluejay!
Xrotor are the best ESCs in terms of durability - I would recommend testing these.
Analytics!
I hope it works for Betaflight, gonna negate all that performance having to use KISS LOL
Works for Betaflight. Now supports bidirectional dshot!
Wish I wasn't color blind. I see a red, orange, blue, and two dark blue lines. But they are not the same color as the 2 light blue, purple, red, and green lines I see at the bottom of the screen. Leaving me only able to tell that the MAMBA is at the bottom, and the Sky stars is above that, and the other 3 are above that. I cannot tell which are which. Maybe pick more contrast in the colors. And when they are next to each other use white, black, red, blue. And then when one is an outlier at the bottom give it the green or purple that is harder for people to see. 1/12 people have a significant deviation to their color vision. so at least 500 people must share some difficulty like me in viewing these graphs.
You might want to update AM32, that's quite an old version.
This was the latest on ESC configurator when I did the testing a couple of weeks back. Future testing will be on the latest version! 😁
Could you please zoom in the interesting parts of the charts? Throttle ramp chart below 50% is totally uninteresting after seeing it's basically the same. The magic happens where we can't see any difference, because it's so tiny
Mr. Rosser Im so enjoying my AOS EVO 7 Sir so Thank You!!!! I also have the 5.5 and love it equally as well!
My question is, why wouldn't he compare it to the existing 65 amp ESC from fettec?
Bidirectional dshot is needed for the testing. Existing FETTEC doesn't have this (betaflight doesn't support onewire). ☹️
Felix is a genius
This is similar to the promotion of FETtec equipment.
Great work Chris ! Finally someone that is working hard to go deep in testing fpv hardware and share the results. This is gold for the fpv community ❤
I really do appreciate all of your testing an most of the time I respect your methodology. But here you present this as a test of firmware but your testing is extremely flawed with such a variation in ESCs. Maybe you could repeat these tests with a few ESCs that can be flashed with different firmware so that you can verify that it is indeed the firmware that is consistently responsible for the variation in performance
I agree 💯. This isn't a comparison of FW in isolation. Only of the specific ESCs on test. I fully agree we need much more data to determine if differences are due to FW or ESC design.
FPV JEDI!
❤
The only thing I want is cheaper kiss stuff I’m not impressed with fettec feel
🧐👍✨
Motors xnova
Unquestionably the best channel at indept research 👌.
I don’t trust this dude can’t even fly just another bard well getting ur view money
Are you serious rn?
Since when do you need to be able to fly or "use" the things you are testing and developing?
We have 20 drivers in the formula 1, do you think any of them can tight the wheelnut properly?
Does a pilot of an aviation aircraft need to know how to build it?
Come on. You dont need to be able to use it in order to develop or test it.
Using is subjective. Testings are objective. If you chose your parts subjective on your feelings thats totally fine,
but dont assume others work and effort and FACTs with your opinion, because your opinion isnt a fact at all.