Re-Thinking Governor Mode

[skigolfmike]

In case you haven't been paying attention, I've been experimenting with a Phoenix 35 ESC and flying the heli in Governor mode. I actually have it working pretty well; however, there are some quirks. After talking to Finless about it last night, I'm going to go back to the Stock ESC and try a flat 100% throttle curve for idle-up.

What I thought the governor would do is keep the head speed constant. Well, it does to a point. The problem is the way electric motors make power. Max torque on an electric motor is at 0 rpm. The faster the motor goes, the less torque it makes. Also, max amp draw is at 0 rpm and min is at max rpm. The amp curve and the torque curve look about the same. Peak power is usually somewhere around 2/3s of peak rpm. Since we fly these motors at 75% plus, where we're on the back side of the power curve, we need to use the momentum in the head and the blades for energy in hard maneuvers. The collective management is learning to let the head recover at 0* pitch and get wound up again for the next maneuver.

The problem with the governor is once it gets behind, it takes longer to catch up back up than using the 100% flat curve. Also, the governor uses something like a sine wave to hold the rpm which introduces a wag to the tail. With a Phoenix you can tune the wag with governor gain, but the tail is still not as good as a fixed endpoint curve.

What I am going to do is setup up 3 throttle curves. (I have a DX7) Normal will be 0-60-80-90-100. Idle 1 will be 100-90-80-90-100 and Idle 2 will be 100 flat. I will fly it and let you guys know. I'll have to change the pinion back to a 10T too. I would blow up the head for sure with an 11T.

[sparx-]

Max torque on an electric motor is at 0 rpm. The faster the motor goes, the less torque it makes. Also, max amp draw is at 0 rpm and min is at max rpm. The amp curve and the torque curve look about the same. Peak power is usually somewhere around 2/3s of peak rpm. Since we fly these motors at 75% plus, where we're on the back side of the power curve, we need to use the momentum in the head and the blades for energy in hard maneuvers.

This is not my experience on how BLDC motors operate. You might want to check out this PDF, specifically FIGURE 6: TORQUE/SPEED CHARACTERISTICS

http://www.eetkorea.com/ARTICLES/2004MAY/2004MAY25_MSD_AN02.PDF

SPaRX

[skigolfmike]

Figure 6 is EXACTLY what I am talking about. Max torque is at 0 rpm w/ an electric motor. Min torque is at max rpm.

Assuming the torque curve is a straight line like the graph, which in reality it won't be, Peak Hp is going to occur at 50% of max rpm. You have 0 horsepower at 0 rpm and theoretically 0 horsepower at max rpm.

In reality the torque curve drops off pretty quick and then starts to flatten out. It never quite gets to 0 torque either. Also, electric motors are always trying to reach max rpm.

So, you have your 100-100-100-100-100 throttle curve. You're going to hit max rpm at 0 pitch because that's were there is the least drag on the system. As soon as you add pitch, + or -, the headspeed is going to slow. When you come back to 0* pitch, the headspeed is going to wind back up.

Finless explained to me that the guys who do hard 3D with electrics use this curve to build momentum by getting the head speed really high and then they use that momentum to do some of the crazy maneuvers they do. The throttle management part is sequencing the maneuvers, or pieces of the maneuvers in a way that they hit 0* pitch long enough for the headspeed to recover and the way they do it, it recovers quick.

The problem with a governor on an electric motor is that you are only running the motor at ~80% power. Even if the governor senses you are losing RPM and adds power, it can't keep up the rpm at the collective endpoints even if it's giving the motor 100% power. This is the reason a flat 100% throttle curve is better for hard 3D because you can start a maneuver with more energy in the rotor system. Plus you won't have weird pulses from the ESC kicking the tail all over the place and messin' with the gyro.

With nitros governors work because nitro motor make big power at high rpm, plus a governor prevents over-revving the motor.

Lookit - don't believe me. Ask Finless.

[JustPlaneChris]

Actually, that chart is very interesting to me. It tells me that if you keep the motor within it's rated continuous torque RPM range, it will deliver 100% torque as a flat line at any rpm within that range. The problem is, we don't know how much torque is needed to do the work. So we guess, and choose a gear ratio that we think will put the motor in the "sweet spot", but it doesn't always work. Your 11T pinion probably bumped the motor into its "intermittant torque zone" (above the continuous torque zone), and in that case the torque does go down as you increase throttle, and the result is the headspeed drops. I bet a 9T pinion would change the results dramatically.

As for the governor thing, I definitely see Bob's logic. However, if the motor is geared correctly to keep it in the continuous torque rating area it will have enough reserve to maintain headspeed when demanded. An excellent governor will help this, and (sadly) I cannot rate the Phoenix governor as "excellent". I use it on my King micro, and get a bit of tail "wiggle" pretty much all the time (that sine wave thing you mentioned) and since I had it originally geared for a near wide-open curve it's probably operating outside its continuous torque rating and tends to bog relatively easily. My Dragonus is a different story. I use a JETI Spin 33 in 3D governor mode, and it has none of those issues. The tail is rock solid, the RPM never varies more than 200 rpm (even when I ham-fist the collective) and overall it is simply a joy to fly. When it comes to governors, you do seem to get what you pay for.

[skigolfmike]

I think you may be correct about putting the motor out of it's continuous torque operating zone. I think what I'm seeing has a lot to do with the whole battery, ESC, motor equation.

I never got my stock B400 motor above 2450 headspeed with the Phoenix 35 and a 10T pinion. I can easily get 3000 headspeed with the stock ESC though. I pretty much had to go to an 11T pinion to make the whole governor thing work. I have it working OK; however, I still have some tail wags and it still boggs on punchouts and the tail turns 45*.

I figure the ESC difference is the B400 ESC/motor combination is tuned to each other. The Phoenix has a bunch of options to try to make it work for everything. I just never got it to work for the B400 motor and, trust me, I tried everything.

As long as I'm not stick banging, the governor works great. Altitude control in hover is a piece of cake. However, I don't hover that much anymore.

Governor mode might work better the heli was way overpowered. There would still be the tail oscillations to deal with though.

[sparx-]

That is the key to the whole motor/rpm equation is keeping within the RATED torque range of a motor. As JPC said this is hard to do because we do not know our exact torque requirements.

IMHO, the RPM of the Rotor with a Stock B400 motor and 10T pinion is right at the top end of the motors Rated torque. So going to an 11T pinion will put you into an overgeared situation and operating outside of its rated ranges.

Try a 9T pinion, you might be suprised :). I know I was.

SPaRX

[JustPlaneChris]

Governor mode might work better the heli was way overpowered. There would still be the tail oscillations to deal with though.I found a really interesting thread over on RCG dealing with governors, and there's some good info on the Phoenix setup too. Hope it helps!

http://www.rcgroups.com/forums/showthread.php?t=724252

[skigolfmike]

It's a good thread. I still think the governor modes gotta go. I have a 10T and 9T pinion. I'll probably start with the 10T and see what happens. Will be nice to be able to turn the gyro gain back up to 80% again and have a locked in tail.