"You're KILLING your ESC by running that low"

[DominicD]

Keep trying Snap . Current still flows in the motor windings during the off part of the duty cycle. The winding has a large enough inductance to filter out the switching frequency so the winding current is effectively constant. Remember your basics, an inductance will resist changes in current ! This filtering is what allows the use of a high switching frequency and duty cycle to create an average voltage. The MOS loses are greatest during the switching not when they are high or low, when they are high and low they are in saturation in which case the Vds is very low and even high currents wont result in large power dissipation.

[OnTheSnap]

HFG, I think you're overlooking the generator aspect of the motor, and the voltage it presents to the ESC. The potential difference between the motor back EMF voltage and the ESC output is what matters wrt stress on the ESC. You get higher currents, and when the FET switches, it sits in the linear region for a short time producing lots of heat. The total aggregate amount of FET switching at 99% is the same as 5%. It just spends more time in the on or off state for the switching period.

So, my point is that at any throttle setting we have the same amount of FET switching. But at lower RPMs, more current (vs. high RPMs) flows through the FET for very short on intervals (Vbatt - Vemf). This puts more stress on the FETs.

[DominicD]

I'm somewhat familar with that a duty cycle is, i dont need it explained to me.

[OnTheSnap]

Sorry, no insult intended. Just trying to politely disagree with your disagreement of my attempted explanation of the thread topic.

[DominicD]

I shouldnt have taken it so personal, you know how it is some times. My understand of it is this. The back emf increases with RPM (we both agree there), which cause a reduction in the voltage potential between the H-bridge voltage and the back EMF, this voltage difference is effectively the winding voltage. As you increase RPM the back EMF increase which results in the ability of the winding to draw current (Its not widely know but motor KV isnt constant). With current being proportional to torque means that electric motors produce the most torque at lower RPM (which they do).

But the original arguement is the best place to run the esc with all else being equal. So for example a 11T at 100% curve or 12T with 80% (just random numbers, i didnt do any math), you have to assume that both sets would be running the same HS which requires the same power regardless of the esc settings. So what it comes down to is the esc has a lower ripple content at 100% than it does at 80%.

[OnTheSnap]

I certainly don't disagree with your assessment of ripple. But I'm not convinced on how this plays into the life or heat generation of the ESC. We both agree that lower throttle settings on the ESC produces a lot more heat. I believe this heat is generated by the potential difference between the back EMF voltage and FET output voltage. I don't see how ripple matters here? There are no components on the output of the ESC that care about ripple?

[dauste]

Can I ask you guys a question:

But the original arguement is the best place to run the esc with all else being equal. So for example a 11T at 100% curve or 12T with 80% (just random numbers, i didnt do any math), you have to assume that both sets would be running the same HS which requires the same power regardless of the esc settings.

Looking at this example, with the same load on the rotor (say hovering) and same HS, would you get more flight time out of the 100% throttle with lower gearing as the motor is running more efficiently?

[DominicD]

A motor's max efficiency is based on RPM not throttle position.

[dauste]

Ah, I think you know what I was getting at though! Do you have an answer? My electronics knowledge stops at 'O' level.

[OnTheSnap]

The power system is most efficient at 100% throttle. This minimizes losses in the ESC. This why everyone always recommends gearing down and increasing the throttle curve to achieve the desired RPM.

Note that there are other problems with running 100% throttle. Mainly, there is no head room for the governor to operate. So while less efficient, actual performance might be better at 85% vs. 100% throttle at the same head speed.

[dauste]

Thanks for that, I will try dropping a tooth. Not had much luck so far with the Scorpion governor as it seems to settle at different speeds when spooling up.

[DominicD]

The gov on the scop esc is useless, the best way to get performance out of the scorp esc is to leave everything default but enable the SS. Then set for 100% flat TC and gear the pinion to get the desired HS.

[dauste]

Cheers Capi! That's what I'll aim to do now.

[Danal Estes]

For what it's worth, I've been through this exact same debate many times, to the tune of hundreds (or thousands) of posts... and I'm too burned out to repeat all of that. I've also spent a LOT of time with everything from motors to O-Scopes to the code (firmware) that runs inside the ESC.

IMHO, HFG has nailed it with everything he's said in this thread.

For what one more opinion is worth...

[Danal Estes]

Originally Posted by OnTheSnap

The power system is most efficient at 100% throttle.

The ESC is most efficient. The motor may or may not be... depends on many, many things

This minimizes losses in the ESC. This why everyone always recommends gearing down and increasing the throttle curve to achieve the desired RPM.

Agreed, again assuming the motor is reasonably happy at the resulting motor RPM

Note that there are other problems with running 100% throttle. Mainly, there is no head room for the governor to operate. So while less efficient, actual performance might be better at 85% vs. 100% throttle at the same head speed.

The last paragraph is misleading. 100% in Governor mode is NOT, repeat NOT 100% throttle. The % sent from the TX in Gov mode is an RPM request. The gov may be able to reach that requested speed... or not... many other facts must be considered.

That's why the general guidance for setting up a gov is:

1) Slowly increase %. Use a tach and note the % where the headspeed no longer increases.
2) Whatever number comes from step (1), no matter how far above or below 100 that number is, multiply it by 0.8. Or perhaps 0.85, depending on how much "headroom" you wish the gov to have, and how flat your batteries hold their voltage. Fly with that number, again no matter what number that turns out to be.

Phrased another way... %s in gov mode just don't tell you ANYTHING. Want further proof? Castle's "Gov Hi" and "Gov Lo" overlap (on purpose). So, on a given setup you might be able to fly the EXACT same governed headspeed at 92% in Lo and 23% in Hi. Now, what did that % mean again?



Actually... one of the reasons I've begun to stay away from Gov mode is because there is SO MUCH misleading information in forums regarding Gov. The "100% leaves no headroom" (or its corollary "80% gives perfect headroom) myth is one of the most common.

[OnTheSnap]

Originally Posted by Danal Estes

For what it's worth, I've been through this exact same debate many times, to the tune of hundreds (or thousands) of posts... and I'm too burned out to repeat all of that. I've also spent a LOT of time with everything from motors to O-Scopes to the code (firmware) that runs inside the ESC.

IMHO, HFG has nailed it with everything he's said in this thread.

For what one more opinion is worth...

Ok, please explain to me how ripple plays into this?

Well, just received the latest RC Heli Mag and it seems to agree with me. Since you've developed ESC's can you please enlighten us?

[DominicD]

You are going to trust a magazine over two EEs ?

[OnTheSnap]

Uh, well I'm an EE also. Although I've never dealt with motors in my designs. But did play with them in school.

[DominicD]

My school work deals with MOS and RFIC so my knowledge on the topic is limited as well. But i did my fair share of research before i wrote my esc 101.

[OnTheSnap]

Go to www.rchelimag.com and you can view 3 pages of the mag. Go to the esc article, which is co-authored by a guy from castle. Let me know if you agree.