2014 - best ESC at the moment?

[redridinghood]

2 of my major crashes came from Hobbywing esc's. Just because my setup is working doesn't mean someone else is.
All hardware come with imperfections, some are lucky some get a lemon.​

 

[kloner]

ferdinandK,,, is there a high refresh rate hobbywing esc without simonK?

 

[econfly]

I agree. It really seems that the "best ESC" is really just personal opinion about how they have performed for that particular person. Since it doesn't seem that there is a vast difference in components (for the majority of basic ESCs), it really may come down to QC by the manufacturer, and then the compatibility with your specific list of components. I say this because I am using some ESCs on a hex that have proven top-notch for many people, but with my specific set-up, they won't even arm the motors. Ugh! Also, Tundra has had nothing but great things to say about the Castle, but RTRyder had one smoke and drop. Different usage, different builds. Lots of variables.

And then there's always a little luck....

I have been thinking about this a lot lately:

http://www.multirotorforums.com/showthread.php?19923-Product-Review-**-JETI-HiCopter-30A-Opto-ESC

Completely agree with you. There are some objective ways to evaluate ESC quality and performance, but in planning the review above I found very little online where people were offering test results or standardized measurements. There is no lack of opinion of course.

ferdinandK,,, is there a high refresh rate hobbywing esc without simonK?

Which rate are you looking at? Just about any ESC can respond to high input PWM signal frequency. The output modulation frequency of simonk is "~18kHz PWM with a POWER_RANGE of 800 steps". A Castle ESC I tested could come near that with an output PWM rate of 16kHz. I just checked the HobbyWing site and like just about everyone else out there they don't offer clear specifications on the output modulation rate.

 

[SamaraMedia]

I think I messed up and calibrated a set of iPeaka 30 amp with SimonK last week. My TBS Pro was flying nicely after completing the build but for some dumb a** reason I decided maybe I needed to calibrate them. Everything seemed ok after calibrating, throttle response worked and all, but when I took to the field for a test the motors wouldn't start and acted like they needed calibrating. Should have left well enough alone, the quad was flying fine, maybe a little sluggish but that might have been the gain settings of the NAZA. I'm using a Futaba T8FG and R6208SB. I ordered some ZTW Syder 30 amp Opto's with SimonK to replace the iPeaka's and I don't want to make the same mistake twice. Just to confirm my first recollection, you do NOT calibrate esc's with SimonK, correct?

Is there any hope of salvaging the iPeaka's.


Sent from my iPad using Tapatalk

 

[econfly]

I think I messed up and calibrated a set of iPeaka 30 amp with SimonK last week. My TBS Pro was flying nicely after completing the build but for some dumb a** reason I decided maybe I needed to calibrate them. Everything seemed ok after calibrating, throttle response worked and all, but when I took to the field for a test the motors wouldn't start and acted like they needed calibrating. Should have left well enough alone, the quad was flying fine, maybe a little sluggish but that might have been the gain settings of the NAZA. I'm using a Futaba T8FG and R6208SB. I ordered some ZTW Syder 30 amp Octo's with SimonK to replace the iPeaka's and I don't want to make the same mistake twice. Just to confirm my first recollection, you do NOT calibrate esc's with SimonK, correct?

Is there any hope of salvaging the iPeaka's.


Sent from my iPad using Tapatalk

I wouldn't even try to calibrate the ESCs. If they work out of the box, leave them alone.

You should be able to recover those original ESCs. The issue you are having is the result of calibrating to the receiver, when in reality it is your NAZA that is talking to the ESCs. Your radio is just sending signals to the NAZA. The NAZA has its own independent throttle endpoints that have nothing to do with the endpoints on your radio.

One easy potential fix is to set the motor idle speed in the NAZA setup to high (advanced menu, motor tab).

If that doesn't work you can re-calibrate the ESCs. The problem is, it's a challenge to do it the right way. What you want to do is calibrate the ESCs to the NAZA, but to do that you need to have some way to arm the NAZA, set throttle to high, then power up the ESC, then shut down the motors from the NAZA. To get it done you need to be in manual mode and it's awkward as hell to pull it off (definitely pull the props for this). You need some feedback to know that the NAZA is really armed and at high throttle before you power up the ESC being calibrated. And if you have all of your ESCs soldered to the board you can't power them down while leaving the NAZA powered up unless you rig up some way to power the NAZA PMU independently. It can be done, but I would try this solution as a last resort.

Easier is just to recalibrate to your radio, but with a lower throttle endpoint. Adjust the minimum throttle endpoint on the radio to a lower level (don't change the maximum throttle endpoint). Then calibrate the ESC to the radio receiver again. I don't have a T8FG radio so I can't say how to do this, but there should be a way to change the throttle endpoint setting. Just make it more extreme (i.e., lower) by 10% or so, re-calibrate an ESC, then see if that fixes your problem (test with props off to be safe). Again, make sure you are changing the low throttle endpoint and not the high setting. Be sure to change that endpoint back to where it was before you re-connect the NAZA to your receiver.

 

[SamaraMedia]

Thanks econfly, I'll give it a try.

Default settings on the Futaba throttle are 135, 100, 100, 135. I read somewhere that Kloner suggested (for Spektrum radios if you mistakenly calibrated the Simon K's to lower your endpoint to 110, recalibrate, then return the settings to default. So I'm guessing that means I should lower the first setting from 135 to 110 or so ( even though his recommendation is for Spektrum ), recalibrate then return my default settings before connecting back to NAZA.

What I tried to do was follow Trappy's video on esc calibration. I unplugged all the esc's from the NAZA, plugged them into a 4 in 1 cable, plugged the main lead of the cable into ch 3 on the rx, turned of the radio, put throttle to full, plugged the flight battery in powering both the NAZA and esc's, wait for the tones, bring throttle to off position, wait for confirmation beeps, test the throttle while all plugged in ( all motors fire up ), then unplug the flight battery, turn off radio, unplug esc's from cable, reconnect to NAZA, done.

I will try the motor startup setting first as you suggest and see what happens.


Sent from my iPad using Tapatalk

 

[econfly]

What I tried to do was follow Trappy's video on esc calibration. I unplugged all the esc's from the NAZA, plugged them into a 4 in 1 cable, plugged the main lead of the cable into ch 3 on the rx, turned of the radio, put throttle to full, plugged the flight battery in powering both the NAZA and esc's, wait for the tones, bring throttle to off position, wait for confirmation beeps, test the throttle while all plugged in ( all motors fire up ), then unplug the flight battery, turn off radio, unplug esc's from cable, reconnect to NAZA, done.

I understand. A lot of people do what you did based on what they've learned online. I just don't get it at all. Maybe this all made sense at some point in the past or maybe some flight controllers out there send radio inputs directly to the ESCs without intervention (maybe some very sporty controller in manual mode?). But for most of us, this whole effort to calibrate throttle on ESCs is just a waste of time and can only create problems.

For controllers like the NAZA you have:

(1) The radio is connected to the NAZA. You use the NAZA assistant to calibrate the throttle (and other stick) endpoints. The radio is then all setup to talk to the NAZA.

(2) The NAZA takes all of the receiver input from the radio plus its own stabilization and flight computations and outputs control signals to the ESCs.

(3) The NAZA's output to the ESCs has no relationship to the endpoints in (1). That is to say, once your radio sticks are calibrated in the NAZA assistant it doesn't matter what the endpoints are. The NAZA takes over from there.

So, when people set throttle endpoints on their ESCs with the NAZA out of the loop they get one of two results. Either, the setup moves the ESC endpoints arbitrarily, but not enough to screw things up, or the setup moves the ESC endpoints enough create a problem. Either way, it's pointless at best and destructive at worst.

For background, ESCs are getting a control signal that is a voltage pulse with width ranging from somewhere around 1 to 2 milliseconds. Out of the box, most ESCs will arm motors if the signal is over 1.10 to 1.15ms and will output maximum motor RPM at around 1.90-1.95ms.

A NAZA outputs pulses at 400Hz. It sends a pulse width of 0.94ms for motors off.
Low idle is 1.144ms
Then 1.161ms
Middle idle is 1.178ms
Then 1.194ms
High idle is 1.211ms

Why these are not nice round numbers I don't know -- this is just what I'm reading on my oscilloscope.

A Futaba receiver (I have a 7008sb to test), with default endpoint settings of 135 100 100 135 will output a throttle range of 1.1ms min to 1.94ms max.

So, if you calibrate the ESCs to the Futaba receiver, the low throttle endpoint will be set to 1.1ms. The ESC won't arm until it gets a signal significantly above that 1.1ms lower bound. Depending on the ESC, the Naza's default arm signal of 1.178ms may not be enough. This is why moving the idle setting to high might get things working again.

To get the ESC endpoint back down to something closer to 1ms you need to change the throttle travel for low throttle. This gets confusing, but because you are using a NAZA with a Futaba radio the throttle is reversed. To make this easy (easier?), don't reverse the throttle in the NAZA assistant (it should be "NORM"). Instead, reverse it on the radio. Set the throttle endpoint settings on the radio to 135 100 120 135. Now re-calibrate your ESCs directly to the Futaba receiver as you did before (put throttle to max, plug in the ESCs, get the response tone, move to min, etc.). This will set the low throttle endpoint on the ESC to almost exactly 1.0ms. Change that 120 to a 140 to set the endpoint at about 0.95ms, but I wouldn't go that far.

Now set the throttle endpoint values back to 135 100 100 135, connect the NAZA back up, and things should work just fine. As always, test with props off.

 

[Motopreserve]

Econ,

thanks for the thorough explanation. Like many here, I was taught to calibrate the ESCs with the RX. I started on multiwii, and although there are end points to be set in the GUI, it's not a calibration like the Naza or SuperX - I guess I figured that the ESC calibration was just that next step in getting it to work together.

The only other question I would have is whether calibrating the ESCs with the RX get all the ESCs working together prior to doing the internal Naza style stick cal? Meaning: can one ESC be slightly off - and then something is written to it after the RX cal - getting all ESCs in line with each other (and the Tx) before doing the Naza cal? Or would this still be pointless (or worse)?

 

[econfly]

Econ,

thanks for the thorough explanation. Like many here, I was taught to calibrate the ESCs with the RX. I started on multiwii, and although there are end points to be set in the GUI, it's not a calibration like the Naza or SuperX - I guess I figured that the ESC calibration was just that next step in getting it to work together.

The only other question I would have is whether calibrating the ESCs with the RX get all the ESCs working together prior to doing the internal Naza style stick cal? Meaning: can one ESC be slightly off - and then something is written to it after the RX cal - getting all ESCs in line with each other (and the Tx) before doing the Naza cal? Or would this still be pointless (or worse)?

I suppose it's possible, but I would think it more likely that ESCs come with some pre-set throttle range that is consistent for a given make/model. The loose standard out there is that a PWM signal should range from 1 to 2 ms. So most ESCs, by default, arm at about 1.15ms and max out around 1.85-1.95ms. That gives plenty of room for throttle to be off (under about 1.15ms) and a little head room to make sure max throttle is achieved.

All we accomplish by calibrating is that we tell the ESC what our controlling equipment is using for minimum and maximum throttle range. So, for a Futaba receiver, calibrating the ESCs is just saying that the travel range is 1.10ms to 1.94ms. That's it. Flight control characteristics should be completely unaffected (apart from little things like the throttle midpoint possibly moving a bit, or power output change vs. stick movement being a little more/less sensitive).

The big two things affected are: (1) at what signal level will the motors start?, and (2) at what level will we have maximum motor speed? Max speed turns out to be irrelevant for all but heavy sport flyers (the controller is almost always limiting us, even when we push the sticks to full). So this really all boils down to setting the point where the motors start.

Even if an ESC is slightly different than others, the controller is going to adjust the signal to that ESC to get the response it desires (e.g., level flight), just as it would adjust to wind, an imbalance in the build, etc.

Having said that, in most cases you can't hurt anything by calibrating your ESCs to the radio receiver. At worst, the controller idle/start setting may need to be adjusted. There may be controllers out there that pass on radio endpoints to ESCs -- don't know of any, but it's possible.

I think the best bet for the vast majority of us is to ignore throttle setting completely, and only address it if there a problem (like the motors won't start).

 

[Motopreserve]

Makes sense. Thanks

 

[SamaraMedia]

I understand. A lot of people do what you did based on what they've learned online. I just don't get it at all. Maybe this all made sense at some point in the past or maybe some flight controllers out there send radio inputs directly to the ESCs without intervention (maybe some very sporty controller in manual mode?). But for most of us, this whole effort to calibrate throttle on ESCs is just a waste of time and can only create problems.

For controllers like the NAZA you have:

(1) The radio is connected to the NAZA. You use the NAZA assistant to calibrate the throttle (and other stick) endpoints. The radio is then all setup to talk to the NAZA.

(2) The NAZA takes all of the receiver input from the radio plus its own stabilization and flight computations and outputs control signals to the ESCs.

(3) The NAZA's output to the ESCs has no relationship to the endpoints in (1). That is to say, once your radio sticks are calibrated in the NAZA assistant it doesn't matter what the endpoints are. The NAZA takes over from there.

So, when people set throttle endpoints on their ESCs with the NAZA out of the loop they get one of two results. Either, the setup moves the ESC endpoints arbitrarily, but not enough to screw things up, or the setup moves the ESC endpoints enough create a problem. Either way, it's pointless at best and destructive at worst.

For background, ESCs are getting a control signal that is a voltage pulse with width ranging from somewhere around 1 to 2 milliseconds. Out of the box, most ESCs will arm motors if the signal is over 1.10 to 1.15ms and will output maximum motor RPM at around 1.90-1.95ms.

A NAZA outputs pulses at 400Hz. It sends a pulse width of 0.94ms for motors off.
Low idle is 1.144ms
Then 1.161ms
Middle idle is 1.178ms
Then 1.194ms
High idle is 1.211ms

Why these are not nice round numbers I don't know -- this is just what I'm reading on my oscilloscope.

A Futaba receiver (I have a 7008sb to test), with default endpoint settings of 135 100 100 135 will output a throttle range of 1.1ms min to 1.94ms max.

So, if you calibrate the ESCs to the Futaba receiver, the low throttle endpoint will be set to 1.1ms. The ESC won't arm until it gets a signal significantly above that 1.1ms lower bound. Depending on the ESC, the Naza's default arm signal of 1.178ms may not be enough. This is why moving the idle setting to high might get things working again.

To get the ESC endpoint back down to something closer to 1ms you need to change the throttle travel for low throttle. This gets confusing, but because you are using a NAZA with a Futaba radio the throttle is reversed. To make this easy (easier?), don't reverse the throttle in the NAZA assistant (it should be "NORM"). Instead, reverse it on the radio. Set the throttle endpoint settings on the radio to 135 100 120 135. Now re-calibrate your ESCs directly to the Futaba receiver as you did before (put throttle to max, plug in the ESCs, get the response tone, move to min, etc.). This will set the low throttle endpoint on the ESC to almost exactly 1.0ms. Change that 120 to a 140 to set the endpoint at about 0.95ms, but I wouldn't go that far.

Now set the throttle endpoint values back to 135 100 100 135, connect the NAZA back up, and things should work just fine. As always, test with props off.

Great explanation, much appreciated.

I started my first quad was a Gaui 330X, then I moved to multiwii and then to OP before settling in with NAZA. So many different programs and processes to learn that I get some of the old confused with the new.

John

 

[SamaraMedia]

Thanks econfly, tried the 135, 100, 120, 135 calibration settings then switched it back afterward and now everything seems to be back to normal on startup and I didn't have to go and up the motor start point in the NAZA assistant. I'll give it a try with props tomorrow but I feeling much better about things tonight.

One other ESC question that has always confused me is whether or not to remove the red power wire from the esc signal cable. Maybe this goes back to the OP CC3D but if I recall there all I needed was the white signal wire. Some said it didn't matter one way or the other if you used all three, power, ground and signal. I have also heard the same argument in regards to connecting the NAZA. Currently every ESC I have has a BEC 5V 3A whether it is the Turnigy Plush, Hobbywing, ZTW to name a few. I have always removed the power wire from the servo end and taped it back in case I need it in the future but all four esc cables going to my NAZA only have the ground and signal wires and each setup seems to work OK. So what is the proper setup with this? I'm using the NAZA V1 on this TBS Pro and currently NOT using GPS and flying in atti. I probably will add a GPS for RTH feature in the future but at this time. I always thought that the NAZA was getting power through the LED VU module and there was no need for the power wire from the esc. Does it matter if you leave the power wire or not.

I'm getting my first Opto ESC's, ZTW Spider's with Simon K, this week and need to know the proper procedure for hooking up the servo cable. All three wires, power, ground, signal? Or pull the power wire back like I have been with the BEC esc's.

Thanks.

 

[econfly]

Thanks econfly, tried the 135, 100, 120, 135 calibration settings then switched it back afterward and now everything seems to be back to normal on startup and I didn't have to go and up the motor start point in the NAZA assistant. I'll give it a try with props tomorrow but I feeling much better about things tonight.

One other ESC question that has always confused me is whether or not to remove the red power wire from the esc signal cable. Maybe this goes back to the OP CC3D but if I recall there all I needed was the white signal wire. Some said it didn't matter one way or the other if you used all three, power, ground and signal. I have also heard the same argument in regards to connecting the NAZA. Currently every ESC I have has a BEC 5V 3A whether it is the Turnigy Plush, Hobbywing, ZTW to name a few. I have always removed the power wire from the servo end and taped it back in case I need it in the future but all four esc cables going to my NAZA only have the ground and signal wires and each setup seems to work OK. So what is the proper setup with this? I'm using the NAZA V1 on this TBS Pro and currently NOT using GPS and flying in atti. I probably will add a GPS for RTH feature in the future but at this time. I always thought that the NAZA was getting power through the LED VU module and there was no need for the power wire from the esc. Does it matter if you leave the power wire or not.

I'm getting my first Opto ESC's, ZTW Spider's with Simon K, this week and need to know the proper procedure for hooking up the servo cable. All three wires, power, ground, signal? Or pull the power wire back like I have been with the BEC esc's.

Thanks.

Great! Glad things are working again.

You can just plug in the optos. A true opto ESC has no electrical connection from the servo wire to the main power side of the ESC. These ESCs require power over the servo connection so you need the power and ground connected. Some opto ESCs do not require power (they have an internal BEC, but just don't output power back over the servo connection). Either way, I would just plug in the 3-wire servo connection as is, leaving both power and ground connected.

 

[SamaraMedia]

Me Too!

Shifting gears a little now and working on installing a Zenmuse H3-3D on a DIY quad I built.

 

[sk8brd]

econfly- any experience with timing settings and how much do they matter, if medium works just leave it or go to high for pancakes.?

 

[econfly]

econfly- any experience with timing settings and how much do they matter, if medium works just leave it or go to high for pancakes.?

It's hard to know without testing the motor/ESC combination. I would stick with medium or the default unless you have a high pole count motor or timing issues that are evident. For most ESCs the range is somewhere between 8kHz and 16-18kHz. In principle, faster is better assuming the ESC doesn't lose timing. That's the big unknown. Plan B is to go for the fastest setting available and only choosing slower if you have timing issues (sputtering motors, unstable flight, etc.).

 

[sk8brd]

gotcha thanks econfly, , i will start with high as there high pole count and test.

 

[kloner]

i'm pretty sure you want med or low for high pole count.....

 

[sk8brd]

thanks kloner.

 

[Motopreserve]

i'm pretty sure you want med or low for high pole count.....

I thought the opposite. It's near impossible to get any solid data on this. This might be another question for Chris and the boys at KDE. They were gracious with their explanation and definition of the high pole count/pancake motors.

The dance of "to SimonK or not, Maytek or not, BLHeli or not, HobbyWing not working with SuperX or is it?...." makes my head hurt.

I finally bit the billet and bought a set of KDE motors and ESCs for a new build. Pricey - but if for some reason they don't work together - I know exactly who to call.

HERE is the best (and only) explanation I've really ever found.