new dji 800 esc no bldc

sk8brd

Member
maybe someone can explain the benefits of no bldc to me. i asked like 10 times on other forum but no one replied..lol this is pasted from another forum we won't mention the name here from dr turbo..he is the head of the propulsion lab for dji...the bold was the way it was originally typed not trying to point out something. there was also a picture of a sine wave


0. 800g/motor recommended thrust.
It goes withoug saying.

1. Ever Higher Slot Fill Factor.
As many of you may know, slot fill factor describes how much copper do you wind into the slots. The higher the fill factor, the higher the torque constant, the lower the stator armature resistency, the higher effeciency you have. For short, slot fill factor is about all, its one of the few parameters in motor design that is positive to all motor performances.
Since E600/3508, we have adapted the single strand winding, though very difficult to handle, it provided a completely new level of slot fill factor that you can achieve. As a result, DJI 3508 keeps the record of slot fill factor in the makert untill now. In E800/3510, we have refined our manufactering process and achieved even higher slot filling factor. The DJI 3510 motor is more efficient than its little brother 3508. One byproduct of the neat winding is that the airflow path is larger than the multiple strand winding, and thus the motor can be better cooled.


3. 345 mm/13.5 inch new rotors, now comes with two different mounting options, both with quick release mechanism.

If you think 13.5 inch rotor is just a magnified 1242 prop, you are wrong. Yes, many players in this market simply scale up or scale down one aerodynamic design to have a full familiy of props, but we are not one of them. The whole rotor is designed from the sketchs to maximize its perfomamce. New carbon fiber compond is used to produce the 13.5 inch rotor, which allows us to keep the chord length short while expanding its wing span and keep its stength, to minimize its momentum of inertial. We know, the acceleration of thurst is as important as the maximium thurst.

The standard self-tightening hub is compatible with E600 3508 motors, E600 users who want to trade high wind stability with longer indurance could try it, but taking their own risk.

As for the second mounting mechanism, please allow me to keep this secret to the end, giving our IP guys enough time to submit the patent.


4. Programmable ESC

Yes, just as we have discussed in earlier post, DJI propulsion system is haunted by ESC firmware compatibility for a while. Becauce we constantly upgrade our motor and our ESC software was written so to optimize a certain motor. This approach brings higher efficiency as well as troubles for our old custmer, who want to use their old ESC to drive the new motor.

This time, we want to solve this problem once and for all. Starting from E800, all DJI propulsion system ESC will be equiped with a data port. Firmwares could be flashed into the ESC as long as the hardware supports it. For example, E800 ESC, namely ESC 620S will be supported as long as the R&D team is here, new algorithm and control strategy could be enjoyed by early customers. When new motors are released, and the hardware of ESC 620S is capable to drive it, we will release a corresponding firmware for it. You could mount the motor and flash the ESC and good to go.

ESC 620S will be pre-loaded with 3508/3510 firmware at the factory. we will also release firmware for 620S to support 2212 and 2312 motor.

5. Integrated LED drive.

Flying a multirotor in low light is always a challange, especially when you only have one flash LED on your ship. The LEDs on phantom demonstrated the assistant of multiple LEDs can do to pilots. ESC 620S for E800 will feature an intergrated three color LED drive and bright external LEDs with cables as spares. Weights only few grams, the LEDs could be plugged into corresponding ports on ESC 620S and then be placed anywhere you prefer, be it under the motor or at the tip of the arm. What is more interesting is that the color of the LED could be programmed.

6. Anti-Spark circut

Many of us have one or two blackened XT60 connector, a few of us even had a bird or two falling off the sky due to the failuer of a oxidized connector. The sparking during the plug in of the battery oxidize the metal in the connector and put our aircraft in danger. The sparking is actually caused by the charging of the capacitors on the ESC. When the capacitor is empty and the battery just gets plugged in, the circut at the battery end is basically equal to a short circut and the current could be huge and cause the sparking and thus the oxidization.

ESC 620S of E800 has a build in anti-spark circut, capacitors are not directly soldered on to the power bus but through an active circut. The trick here is that during the initial charging, a resistor is put in the circut to regulates the maximum current. After the initial charging, MCU will open a MOSFET to bypass the resistor and "plug" the capacitor directly onto the power bus to let it play its role.

From E800 onwards, you need never worry about your blackened or melted connectors.

7. ESC: Closed loop control

Most BLDC ESC use open loop control to handle the RPM. It simply convert your input PWM length to a corresponding voltage and let this voltage to affect the RPM. Higher voltage means higher RPM and lower voltage means lower RPM. This is a simple and effective approach, but it does have its limit, because the semi-linear relation between voltage and RPM is only true when the system reaches its static state. During the acceleration between one speed to aother, this linear relation between RPM and voltage is not valid. For example, one motor has a internal resistance of 100 mOhm, and a KV value of 1000RPM/V. Loaded with one propeller, and the voltage is 6V, the current is about 2A and the RPM is 5800RPM. At 7V, the current is around 2.7A and RPM is about 6730RPM. But when you switch between 6V and 7V, the current will rise to 22A at the switching moment and then gradually (an exponential function) drop to 2.7A. Four motors together could generate a current peak of nearly 90A, way above the safe current of a 2200mAH 20C battery. What is more is that the torque of the motor is not smooth but has an impact on the airframe.

Most ESC has a capacity of controling the voltage at a high frequency, say 2KHz. But most morden flight control system can only provide a much lower frequency control signal, around 400 Hz. This means the control input is not smooth for the ESC and motor, and the senario of the example above does happen a lot. From the ESC point of view, the command it receives is a step function, jumping from one point to another. Here is where the close loop control could jump in. The ESC can actually use its own brain to regulate its current at a much finer pace if a proper algorithm is there and the CPU and sensors could handle it.

ESC 620S for E800 has this capacity. The MCU for this ESC is more powerfull than many flight controllers on the market. Helped with current sensors on board, the ESC could control the current and thus the torque in a closed loop and with a much smart strategy. Now the current peak could be carefully smoothed out and the convergence of the PRM from one point to another can be done much faster. Yes, current peak is removed but the overall speed up is accelerated. The battery gets extra protection and thus can run for more charge circles. The motor is now protected from stall and overheat. And many many more possibilities are created.

Cheers,

Dr. Turbo

8. System diagnosis and real time health monitoring--More than a driver

Thanks to the powerful CPU and sensors as well as the corresponding algorithm, E800 ESC620s could not only drive a motor but also diagnose them and monitor them. Motor parameters such as internal resistance, winding temperature, RPM and current are all monitored, in real time.

If you have a demagnetized motor, your E800 will know it and refuse to start.
If you overheat your motor, your E800 will sound the alarm.
If you have a bad connection between ESC and motor, your E800 will refuse to start, If this happened during flight, it will sound the alarm and record this event in the onboard memory.
If you just flashed your ESC620S from 3510 to 2312 motor and hook it to a new motor, before the first start, the ESC will thoroughly diagnose your motor make sure this is a 2312 and it is healthy.

9. Redundant duplex communication

Just as lvale has guessed, E800 will feature a BUS port to communicate with FC. This enable the ESC and FC to communicate in both directions. There is a traditional PWM channel on E800 too, to make it compatible with nowaday flight controllers on the market. Supported with the right flight controller, these two communication channels will hot back up each other to guarantee reliability.

The duplex communication is a real game changer. Together with our real time diagnosis technologies build in the ESC, It enables the flight controller to have a full awareness of the situation in propulsion system. Now FC could know what the temperature every motors has, how fast every rotor is rotating and whether a turbo charge is neccessary and possible at the moment. If a motor fails, the FC will know it at the first second and switch on the preplanned strategy to cope with it.

The BUS channel will also make the FC to have a full use of the thi-color LEDs driven by the ESC. System status, alarms or simple decorative colors can be displayed by all the LEDs on the your DIY aircraft, just as the phanom.

10

Byebye BLDC
 
Last edited by a moderator:

R_Lefebvre

Arducopter Developer
He's not talking about getting rid of Brushless DC. He's just talking about more intelligently controlling throttle application, so as not to over-current the system. I believe this will help prevent loss-of-sync. But I do not buy the argument that this would speed up the RPM response. It should slow it down, just simply looking at a few basic laws of physics. Could be wrong, maybe there's more going on than I'm aware, but I await to be convinced.
 

sk8brd

Member
attachment.php
---new



attachment.php
---old


R--i was hoping you would respond i know your technical with these things...the biggest feature he was saying was no more bldc...others implied on the thread there were going with possiible 3 phase induction motors? check out the e300/600 porpulsion thread on that other forum. would like to hear more of your thoughts...i like the integrated led port not gonna lie..lol
 

Attachments

  • thumb-BLDCControl.gif
    thumb-BLDCControl.gif
    9.7 KB · Views: 241
  • 3-phase-1.gif
    3-phase-1.gif
    2.9 KB · Views: 350
Last edited by a moderator:

R_Lefebvre

Arducopter Developer
3 phase induction? So basically, and AC induction motor. Not actually that much different than what we are doing now, though the control system is SO much more complicated. My impression based on my experience with industrial motors, is that induction motors are much, much larger and heavier for a given power rating than a PMAC (permanent magnet AC) motor. And our brushless DC motors are basically PMAC motors.
 

sk8brd

Member
i'm not sure some members said it not dr turbo. could be not true. so if true going ac is there a huge benefit esc wise or is it mostly marketing hype
 

Gary Seven

Rocketman
i'm not sure some members said it not dr turbo. could be not true. so if true going ac is there a huge benefit esc wise or is it mostly marketing hype
[MENTION=12728]sk8brd[/MENTION]; I've been following that thread over at "www.rc******.com/forums/showthread.php?t=2046097&page=192" with great interest (under a different alias, of course). I don't have enough technical (electrical) knowledge to chime in on your OP above, but just wanted to comment that I am shocked by the drooling, pandering fanboys over there. I'm literally stunned by all the (teenage?) members following that thread flirting and brown nosing "Dr. Turbo" like he was the second coming of....

Yes, I'm very interested in new developments in technology not just from DJI but the industry in general, but the amount of hype and pandering taking place there is truly stomach turning. Just my $0.02 worth.
 

sk8brd

Member
Gary-- i feel the same way. the company has a larger presence over there so with new product releases there usually faster at getting some info to the public... i'm much more of a multi rotor forums guy myself. theres def a different vibe on here with better conversations and less hype more real talk. there are industry guys on here as well as arial pro's so the the level of conversation is just more mature on here imho. theres smart guys over there of course but that forum has a more hobby type of feel and less pro feel..if that makes sense.
 
Last edited by a moderator:

Gary Seven

Rocketman
Gary-- i feel the same way. the company has a larger presence over there so with new product releases there usually faster at getting some info to the public... i'm much more of a multi rotor forums guy myself. theres def a different vibe on here with better conversations and less hype more real talk. there are industry guys on here as well as arial pro's so the the level of conversation is just more mature on here imho. theres smart guys over there of course but that forum has a more hobby type of feel and less pro feel..if that makes sense.

You framed it perfectly. I hang around rc******* mostly to pick up gems of information from the likes of R. Scott Page (I love his Taranis info) and others. But yeah, the "feel" of the place is definitely more rinky-dink than here. Without a doubt I feel much more at home here at mulitrotorforums and almost never post at the other site. Maybe it's an age thing....I'm uh....50ish. :friendly_wink:
 

R_Lefebvre

Arducopter Developer
Most ESC has a capacity of controling the voltage at a high frequency, say 2KHz. But most morden flight control system can only provide a much lower frequency control signal, around 400 Hz. This means the control input is not smooth for the ESC and motor, and the senario of the example above does happen a lot. From the ESC point of view, the command it receives is a step function, jumping from one point to another. Here is where the close loop control could jump in. The ESC can actually use its own brain to regulate its current at a much finer pace if a proper algorithm is there and the CPU and sensors could handle it.

This here strikes me as being a bunch of malarkey. First of all, the typical switching frequency of an ESC is at least 8kHz, even 16kHz, sometimes even more. But that doesn't really even matter. Yes, flight controllers can only provide 400Hz input signals. It's surprising he even says this, because it contradicts DJI's previous statements about 800Hz control frequency, which is physically impossible on a 2mS signal. But anyway...

What's funny, is that what he's talking about is using the ESC's processor, to slow down the rate of change of the throttle pulse in order to prevent current spikes which lead to loss of sync. This is a great idea. But what's funny is that ESC's used to do this in the past. That is what happens with the default firmware! But, with SimonK etc. which up until now have been "must have" upgrades, the filter is gone. This is the point. It was done in the search for faster response. But it has led to all these problems.

But now, DJI is adding a filter back in, and calling it a radical new improvement?

Maybe their filter is not just a dumb low-pass-filter. Maybe it really is closed loop. Not sure. But it's not that much of a radical concept.

Similarly, the other thing they are talking about is basically having the ESC generate a sine-wave, rather than a square-wave. This is a similar idea to true sine-wave inverters that you may be familiar with. Using the switching FET's to build a sine-wave, instead of just turning on/off. Again, this may be a good idea, but it's really not that radical.

Why hasn't it been done before? Well, it is true that AC Induction motors are more efficient when driven with a true sine wave instead of a square wave. But I don't know that a PMAC motor we are using is actually more efficient with sine-wave.
 

sk8brd

Member
thanks for your input rob...making more sense to me. when i was looking for generators for charging batts a while back people were saying to get one with an inverter cause of the clean sine wave was better for electronics. thats all the experience i had with the term sine wave ..i guess we will see if there any difference that you can actually feel while flying. you could feel simon k no doubt... i think i like the led port most..lol
 

R_Lefebvre

Arducopter Developer
Actually, it's funny because you always hear about that with sine wave inverters being better for electronics. But fact is, I don't believe that's true at all. Most electronics run off DC. So the first thing they do with the AC power, is rectify it. The shape of the wave really should not matter. Your AC motor powered saw cares about the shape of the wave more than your computer does.
 

sk8brd

Member
interesting...i saw this nikola telsa documentary 2 days ago. i remember they were saying something about ac being easier when traveling long distances ...him and edison were fighting about what was better..eddison had live demonstrations trying to prove that ac was more dangerous then dc by electrocuting cats i think..lol wouldn't it not matter going ac because of how short the energy needs to travel because of our esc wire length..just speculating..i'm going to learn more about this..thanks for your input.
 

R_Lefebvre

Arducopter Developer
I don't really think any of that has anything to do with this.

I think back in the day, the argument was about voltage vs. amperage. AC makes it really easy to step up and step down voltages (using a transformer). So it made sense to use very high voltage, to lower the amperage, for long distance transport. DC is not easy to change voltage levels (at least not back in the day, they didn't have switching regulators). So you had to transport at the lower voltage, which meant huge amps needed to be carried. That had a lot of thermal loss due to resistance in the wires.

I dunno, was Edison electrocuting cats to show that high voltage was dangerous? Well duh. But you know, I don't want to touch the 10V 5000 Amp DC bus bars on my wire annealer at work either.

There's also an effect where, with 3-phase AC, that third wire makes the system much more efficient, you can transport twice as much current with only 50% more copper. But that's complicated to explain.
 


Top