Trex 700 heli shaking

Jacob there is these new settings for sag compensated voltage, etc… I don’t know much about that, or how it works. Alan is not measuring current so I don’t think he can use that anyway. But doesn’t experience flying it, and then see what the battery recovers to after it rests, tell you what the voltage sag is under load anyway? Are those figures you gave under load? Or after the battery rests and recovers?

Sorry, yes…apart from the 3.8v/cell = 50%, the numbers are under load.

Huh, so there is
Sory, http://ardupilot.org/copter/docs/parameters.html#batt-fs-voltsrc-failsafe-voltage-source

I cannot find how it calculates the sag?.
Sag will vary greatly depending on battery type and load.

I set up my failsafes based on under load values. If you go below 3.3v/cell, under load, you will wear down the cell quickly, below 3v/cell, you can kill it.

First few flights, I rarely use more than 50% of estimated usuable capacity. I then go home, look at the discharge curve, vs expected and then go 75% next flight.

As a general rule of thumb, I expect usable capacity to be ~80% of rated capacity. Some batteries, like the Multistars, you were lucky to get 50%, even if you stayed well under the batteries claimed C-rating. (The “cheaper” tattus weren’t much better).

You also have to anticipate that as voltage goes down, current goes up and the battery will discharge faster.

I now run li-ion, so my values are quite different, but as a starting rule for lipo, I would have 3.7v/cell as my first failsafe (RTL if you don’t trust the telemetry!). and 3.4v as Land.
Then adjust as needed.

Cool, thanks. That sounds about right. I probably over-worked my batteries when I flew electric because I remember when I hooked them to the charger if it said 3.5V on all the cells I was pretty happy. When a cell went bad I’d start finding one at 3.3, and as it got worse one day that battery would come with the cell at 2.9. And then the charger would refuse to charge it. So I “cheated” by setting the charger for lead-acid and boost that cell above 3, then switch to LiPo and it would charge it. Then I marked that battery as “bad” and used it for short flights.

I did NOT enjoy dealing with batteries. Some of these guys got elaborate balance boards and stuff to mass-charge batteries, and they’re running Honda generators at the RC field to charge their batteries. It’s like dude, just put the gas in the helicopter instead of the generator and get rid of all that :neutral_face:

Well, as you know I’m leaning heavily that direction (Gas), however the li-ion pack has proven quite practical so far. I’m still running it conservatively, but I still gain more flight time than I did with a similar weight Lipo. It’s got an onboard BMS, that I can plug into a PC and check the heath and charging is just plugging it in and forget about it. (Well, you wont forget, as the fan on the charger sounds like a jet engine, but anyways).

However, due to the severe voltage sag, li-ion will be limited to low-to-medium power operations. S, perfect for surveys.

I’m not a fan of batteries but what I can afford to play with for now and crash! I looked into li- ion batteries cannot find a dealer in USA. I suppose I could make my own if i bought spot welding equipment. Iv been looking at 2170 batteries. From Tesla cars or Samsung. The more I have troubles, some of my own doing, and some cuz I don’t know enough yet. Buying a ready built and TUNED heli is looking like I might mortgage the house and get a heli that flies good so I know what that looks like lol.

Early in this thread I posted a log that had hard auto in the title. During that flight. Batteries reached 22.5 volts and dropped straight down below 18 volts. The heli looked sluggish. I had soft cut off on the ice 2. I was 50 ft up. I said here comes my first auto rotate. I dropped collective. I let it come down to 10 ft flared way to early. I watched the blades dam near stop and it dropped 8 ft. Good thing I just put my tall landing gear on cuz it bent and took the blow. So that the reason for the voltage Failsafe’s. When I went into the tree. Voltage was at 22.9 and I was trying to auto it in. Not what the ESC did that time. But heli blades were slow and unresponsive.

I have looked at Milwaukee batteries but voltages are to low for amps unless I could run 60 volt instead of 50 max charge.

We are buying cheap lipo’s. Floureon they seem to work but I don’t know anything about adjusting the kV on the motor for more efficiency or if running the ESC at 70 percent is efficient. We only pull 20 amp at 12 s and max I have seen is 30 amp. My idea was if I put a few more in parallel that will lower the amp draw per battery and be able to run them longer

Yeah, I built my own li-ion pack. It still costs a fair bit and I only did it after I had the heli working as intended on lipos.

The ESC’s are more efficient if you can run them at 90 or above. What many people do is adjust the voltage by running like 8S or 10S instead of 12S, then run higher throttle. I did do this with the 626. It was originally 12S but the motor and ESC got really hot. I dropped it to 10S and it ran much cooler, and less heat means better efficiency.

When I went to 10S I used a 5000 6S 45C in series with a 10000 4S MultiStar. Those MultiStars don’t have anywhere close to the capacity they claim they do unless they are discharged at like 5 amps. That combination actually worked and all 10 cells came in at about the same voltage after a flight with fully charged batteries.

I got close to an hour with that helicopter with 4 6S 5000’s at 12S with two stick packs in parallel. But it got really hot and the ESC burnt the paint on the canopy. So then I stuck a canopy on it with a hole in the front and a muffin fan on the ESC. But it still got hot.

It would only fly for a half hour on the 10S. So I bought it in on a survey flight to change batteries and then sent it back out to finish the flight.

I even tried 696 blades on it once. But those took more power than the 626 configuration and it wouldn’t fly as long and had poor tail authority. I lowered the throttle setting to run the bigger blades at 400 fps, same speed as the 626 blades at higher rpm. And then it was back to overheating the ESC again.

In the end I came to the conclusion that electric is great for 7-8 minute 3D flights for a lightweight, quite powerful machine in a short burst. And for 600 class and smaller where gas engines aren’t really all that practical because of their weight. I don’t remember exactly what the engine weighs, but I know it’s heavier than a couple 6S 5000’s and an electric motor. The upside with piston power is that it can produce full rated output continuous. Electric can’t because the batteries can’t take it. And energy density of the fuel + conversion efficiency to shaft power far outweighs the energy density of batteries at the conversion efficiency to shaft power with electric if the helicopter needs long legs and high continuous power output.

There’s likely a reason you don’t see electric Bell 47’s flying around with a bunch of LiPo’s in it. It’s not practical. And the bigger they get, the less practical it becomes. :grinning:

I just bought a motor suitable for running 12S at non-3D headspeeds :).

https://www.scorpionsystem.com/catalog/helicopter/motors_4/hkiv-40/HKIV_4035_330/

330kv rather than the usual ~500kv. Geared it down as low as I could. Gives me 2200rpm at max voltage. (626 blades)

I bought one of those too! Mine is a 300. It will fit in a 626 with a E7 tail boom and 696 blades with the help of custom water-jet cut aluminum frames, custom machined motor mount, and a custom-built clutch shaft.

100_0165

I bought some new batteries for that helicopter and gave it to Kayla to use for a trainer. She said it only flies for 17-18 minutes and then it takes two hours to charge up the batteries. I told her I’ll fix that.

Alan might be interested in knowing that Trex 700 short flight times are also repairable:

We have been talking about that but I have never seen one for this old turd. The pictures on that link are for the 700N. The swash and head don’t look like mine. Don’t know if that matters. But heard I need some N parts if I want to use a straight E heli. Then again. I’m learning more every say about how to tune these wild machines. I have seen the 766 kits with gas frame conversions for 1400$ but knowing what parts cost blades and radios and motors and electronics unless I canabilize the trex. I’ll have 6 grand in one before I know it and still have to work months tuning it. Trying to still talk partners into splitting cost and buying one bulit.

1 hour flight time with 3 pound payload and gimbal to aim straight down at 20 kts is the goal. In the wind at 400 feet. Gas is the best bet. But will take a lot of weekend surveys to pay it off. The current drone hexa copter is not the ticket. But is working so far. As long as not much wind. Ground Speed below 14 mph. 4 battery changes a gen to charge battery’s. 5 sets of 6 lipo battery’s. And at least 4 hours on site min. But I am not currently flying it if I was I’d done bought a gas heli. That’s the reason for a trainer to prove concept. But struggling with this thing sets it back a bit.

If I can get 40 min out of this heli it will be a huge advantage over the hexa drone.

Well Alan, you have several things you can look at that have been suggested sources of problems, including that it looks like you are not working your battery hard enough. It appears you did get a successful auto flight out of it where it flew fairly decent. But still not 100% certain about what caused the first really bad one.

Still not sure why your RC telemetry is delayed.

Starting out with helicopters is very difficult. They are not beginner-friendly machines. But you are making some progress.

Edit: also wanted to mention something about your two somewhat rough autorotation landings. You cannot autorotate a helicopter with ArduPilot in anything but stabilize or acro. Any of the altitude modes - Alt Hold, Loiter, Pos Hold, Auto, will stall the main rotor. You might already be aware of this, but have your flight modes set up so one click of a switch gets you into one of those modes if it quits in flight. Stabilize is probably the best because after the flare you just have let the cyclic go and it will push the nose over and level the helicopter by itself and all you have to do is manage the collective.

If you try to autorotate it in any of the altitude modes you won’t have enough collective control to make it work.

Thanks to all the help from you and bill. I agree the flight look lots better. I got 4 battery’s mounted , wired up and charged. I’m going to try a flight with the radio hooked again. Then I am going to revisit the rat p d gains. Then on to the blade tracking. I also thanks to today’s conversation I’m looking into 4s battery’s. To try out 10s as an option. Iv exhausted the li-ion battery in my mind. Their is to much overhead when looking to gas in the future. New charger kicked butt. 4 batterys in 4 hours. 800% better.

Ya I had hopes the radio was causing the delay in telemetry. That or I had the wrong param set in the 4/5 serial port and that it was only getting some if the telemetry. Since the wiki settings did not work. I need to remember to see if artificial horizon is transmitting or if all telemetry is locked up.

Truly flying the heli in normal flying is fairly easy once you get your brain retrained to the orientation vs stick inputs. The mechanics and electronics take a bit longer. But I can say I learned a lot from the crash on how the heli operates.

I agree I have throttle hold on a switch I have memorized I have stabilize on a switch towards me and acro on a switch away from me. I start the heli in stabilize. One click gets loiter next click gets auto then rtl then land and all else fails final click is acro.

I saw that your serial settings have been set to 57 on all ports. Normally, at least SERIAL0 is 115200 because that is the USB. The others it won’t make much difference. The GPS driver sets the serial baud rate on the GPS port. But you are using SERIAL5 instead of SERIAL4 for your RC telemetry. I don’t THINK that makes a difference. On the V3x SERIAL4 and SERIAL5 are on the same plug. You would have to pull Rx/Tx wires out of the plug and switch them to the SERIAL4 pins (the two middle ones), then set SERIAL4 for the passthru instead of SERIAL5 (disable SERIAL5 by setting it to -1).

In the pinout diagram for the V3x they call SERIAL5 “GPS2” and SERIAL4 “Debug”. But most people use SERIAL4 for the passthru telemetry.

I don’t know if that will make a difference or not. But you could try it.

Will look at that this weekend see if it changes any thing. Heli ready to fly after work with gc radio. Not sure how I am going to do gains testing with no GC radio to adjust them. Might gave to try the variable switch again.

Didnt have time to post log yet. Flight with radio. Drunken again. So radio definitely causing some if it. I even moved it clear to the bottom of my tall landing gear. So at least 10 inches away. Landed. Unplugged Radio. Whoops had a 10 min. Delay. Would not find GPS and I had to upload new auto plan cuz old half of one was in and wouldn’t arm. After that. Flew auto flight. Looked fairly straight with spline waypoints tail was yawing off center 6 inches each way. Had SSE wind flying north and south. When it to as flying into the wind it it kept I would call it hopping. The nose would jumping up and down like it was running cyclic rearward. Forward flight with the wind did not notice that. I’m going to replace my main shaft it might be ever so slightly bent. Worth a try I have 2 extra. Ran out of time to do p d gains. Wind was 13 mph and couldn’t get behind anything so gave up on that till weekend. Tommorow is forecasted as 20 mph plus. Wind dies at night if had enough light to see.

ok trying to learn the logs. so the one where ground control was not hooked up. speed was at a max of 6 mph. or kts. i had it set at 22 mph. my vibe chart x is the new high y went to the bottom of the three axis.
height is only showing 22 something. i had it set at 70 ft i bet it was close to that watching it. servo 1 was still going nuts. if you look in the rate file. on the att file they look better. i had it flying 650 ish feet. then 60ft apart on way points for turns. it did pretty tight turns no much overrun.

Is there a way to tell the auto pilot what the wind is and where it is coming from. on the ground control it has a spot that says wind direction and speed,

Telemtry to yaapu was about 2 to 3 waypoints behind. battery voltage looke to keep up a little better. since more batterys means it was gonna change slower…

I changed the defaut in the ice 2 to 8khz instead of 12 to see what it does. i read that most of the 700mx motors were 8 khz. i read that if i speed it up the motor might be more effiecent but the ice might get hot. both were warm but not hot.

another one . i read the wiki on tuning p and d gains. it said to tune vff first. baseline was .015 and mine is .017. do i change that to adjust p and d gains again then change it back.

I would recommend selling that radio set to somebody on RC Groups and get a different one. I have seen these problems before with some of those telemetry radios.

Going to grab some supper and then I’ll look over your log from that latest flight.

With that telemetry radio operating it’s back to the same deviations in the position controller. This is with the radio on.

Despite the wind, this with the radio disconnected:

In the one without the radio I see the deviations during acceleration and deceleration, auto is not real smooth for helicopters under some situations. I have modified code for helicopters that will fix that but we’ll cover that later via private email.

You have a definite RFI problem with that telemetry radio. I don’t know what it’s affecting, looks like maybe the controller itself. But it’s not good.

What did you do different with the machine this time? Your vibe Y is lower than the vibe x, and they’re both mostly below 10. Did you hang some more batteries on it and change the CG?

That’s because the logging is in meters. 22 meters is 72 feet.

The helicopter is not tracking attitude in auto very good at all. It’s on both sides of where it should be and it’s bouncing back and forth. I think that servo1 is causing it.

So, get rid of that telemetry radio and put it someplace safe where it can’t cause problems. Take it completely off the machine. Didn’t you get a little CUAV PW-Link WiFi telemetry radio with the controller? Use that, plug it into the telemetry port, mount the antenna someplace underneath. It doesn’t have much range, but it’s good enough to make settings with your GCS and load flight plans and that stuff.

So when you fire up with that WiFi radio on there, your computer or tablet will detect a WiFi network that says something about CUAV whatever. The password to connect to it is “cuavwlink”. QGroundControl will automatically detect the UDP link and connect to it.

Servo1 is wiggling back and forth, but SERVO2 is also wiggling back and forth at the same time. The helicopter is shaking at rotor speed when you spool it up

This is a mechanical problem, either rotor out of balance or not tracking. Until you fix this problem the helicopter is not going to fly decent. There is no tuning that can fix it, because you can’t put lipstick on a pig and turn it into a beauty queen. You can put all the filters on it you want but the helicopter is still gonna shake, and anything you hang on it will shake. If you don’t want to shake test your payload, this needs to be addressed. THEN you can worry about tuning it.