Flite Fest 2017: Bugatti 100P

[quimney]

Wilsonman,

I'm not a quad guy so I am not too familiar with all of those electronics but I have to think that there is some open flight controller that could be hacked and pressed into service... all that you really need is to slave one ESC to another. Rather than trying to control the second motors RPM through it's servo input just force the slave esc to match motor control waveforms with the master.

While I type here it occurs to me that maybe there is an even simpler solution. What if you used one big ESC and hooked both motors to it? You could cross wires on one motor to drive the counter rotation. If the ESC utilized any feedback from the motor this would be a bad idea but if the ESC simply presents waveforms to the motor to cause it to move then this could be the perfect solution.

Just some random thoughts before my brain is fully awake...

 

[cranialrectosis]

The geared setups for F3A would entirely solve this issue but I'm not paying $1100 for one.

Is that $1100 PLUS another $500 for the motor PLUS another $100 to change the color of the spinner?

Looking at the prices of models for sale, I think some people have more dollars than sense.
http://www.f3aunlimited.com/airplanes/f3a-aircraft

Hmmm yeah. Thinking I will stick with my Taranis.
http://www.f3aunlimited.com/radios-accessories/radio-systems

 

[quimney]

Ok, now that I have had some coffee...

It seems that 2 motors on one ESC is not a great idea... there seem to be problems getting both motors to start reliably. I still think that using an open ESC platform you could slave ones frequency to the other but I have never tried it. I know a lot about driving steppers but have little BLDC experience.

 

[F106DeltaDart]

Is that $1100 PLUS another $500 for the motor PLUS another $100 to change the color of the spinner?

Looking at the prices of models for sale, I think some people have more dollars than sense.
http://www.f3aunlimited.com/airplanes/f3a-aircraft

Yeah, those prices are a bit beyond insane... I wish there were more contra-rotating gearbox options out there. I really want to do a Fairey Gannet with contra-rotating 4 blades, but the KV is too high on the Hobbyking system and its out of stock anyway..

As for the RPM difference, I'm not sure that I would worry about it. Most cheaper motors have a questionable tolerance in KV anyways, so two of the same motor can be off by up to 100kv or so. Some of my twins have the same kind of RPM differences between the sides. I believe my P-38 is off by around 250 rpm, and you can't even notice it in flight.

 

[willsonman]

Yeah, it certainly CAN be done on one ESC but the timing issues become more severe when the motors are inline with each other. If timing feedback to the ESC is irregular then the ESC ends up working more to compute timing and intensity of the pulses. It translates to a very inefficient setup that is still prone to failure if angle of attack influences airflow on one prop over another.

The flight control board idea is essentially what I am trying to do with the Taranis system telemetry. One concern I have is that the lag will be felt in-flight as the calculations are done in the Tx and adjustment sent. I'm not entirely sure if it will work well but I am hopeful.

 

[willsonman]

Ok, the latest set of data is in and I think we are on the right track here. I've formatted the data to be within the same relative scale for comparison.

Here is the original 8x8R and 8x6 prop data. You can see how at the top end the RPMs are widely varying due to possible cavitation and turbulence. Also note the relative separation of RPM during the runup, ~3-400 RPM.


Ok, so here is what excites me. The gap has narrowed from 3-400RPM delta to <100RPM delta. In addition, no cavitation and the added bonus... since this was originally a 9x7.5 prop, the blades at the root are a bit more thick and robust (less prone to breaking). I have the motors on separate channels so I think I can do some minor adjustments using offset to try and align the RPM without using a script. One of the risks in using scripts on the Taranis is that if they go wrong they can lock up the radio... not something I want on a model like this... that is why I have been looking to make mechanical changes before I go with the programming changes.


Very promising progress.

 

[PsyBorg]

Maybe not try and do this thru the radio.. (keep in mind this is me thinking out loud as I type this)

If you feed the RPM data into a comparative circuit and have that drive the second motor instead of the radio gear I think you could more closely and automatically match RPM. The only thing is finding an offset to match the differences in the motors, the weight of the prop AND the section of the custom nose cone riding with it. I will assume that the offset would change constantly as well due to air resistance from the other prop or other things like flexing and running flat.

Heck it might even be better to compare tones from the props somehow like you said you can hear when they are "In tune" or out of "tune" with each other. That could be done with two microphones instead of the rpm sensors. Either condenser type in the fuse to pad any air noise or with an inductive mic attached to the fuselage to hear vibration changes or possible a combination of the two.

The more I think on this the more the sound method seems more viable. The rpm data shows one motor lagging thru the entire range yet at some points I am sure you heard them props sync up tonally.

 

[Michael9865]

Wow, and I thought my Dad and I had trouble getting the Veco 19s synced on his Skylark twin. It is incredible that there is not a simple and less expensive way to sync two electric motors in today's tech.

 

[Fyathyrio]

I suspect latency in the Taranis telemetry would make a sensor based programming solution frustrating if not impossible. The telemetry link is not as strong as the control link, and at a location like FF with tons of 2.4GHz noise it could get swamped. While not scientific, I have observed many times having complete control of my plane (and others) while getting constant telemetry lost/telemetry recovered warnings. Plus, I suspect that this solution would require a level of accuracy that the FrSky equipment may not be built to deliver.

 

[willsonman]

That is my suspicion as well. That is why I wanted to pursue the mechanical aspect first before really diving into the telemetry and programming route. At some point here I need to hook up the fish scale and get a value for thrust. Pretty sure we are in excess of 7 pounds but an exact number is unknown. Again, at 13.5K RPM that translates to a prop speed of 102MPH. That is calculated off the 8x8R prop. While we are not going to see that sort of speed I'm confident that the system is robust enough to handle anything other than dumb thumbs.

 

[quimney]

Have you looked at BLHeli ESC code. It has a closed loop mode where the rpm of the motor is regulated to the throttle setting. Most ESCs just try to apply more power as you move the throttle up. The RPM is dictated by the load and the motors response to increased winding voltages. This can lead to RPM variations due to motor and prop differences.

With the BLHeli code a motor with no prop has the same rpm to throttle response as one loaded with a prop.

Perhaps with two such ESCs you would not need to try to regulate one against the other.

 

[willsonman]

The YEP ESCs are based on the YGE design. From my understanding they share the same firmware. These ESCs are some of the most accurate and robust you can buy. They are designed to take a real beating by heli guys and I have to tell you, I am very impressed with the throttle response on them. I use a ratchet on my throttle stick and every click on the stick changes the RPM. Not the case with any other ESC I have used from China.

While the firmware from BLHeli may prove to be something to look into later, I am quite happy with how the setup is now. Without a flight control board to get the TWO ESCs to match up, there is really no point in going to them.

 

[willsonman]

Well, Today will be the first public debut of the Bugatti. My workplace is having a community showcase of what we all do outside of work. The Bugatti will be on display alongside of the Sikorsky S-39 for Flite Fest 2015. I'll grab some shots from the event and post them later. I'm sure there will be some turned heads, even among some scientists.

 

[willsonman]

Reporting back... The event at work was intense! Many talented people but I was constantly talking with people about my building and flying and such. It was a very fun time talking with non-hobby people.

Moving along, I did some final tweaking last night. The wheels were rubbing a bit on the wheel wells when retracting so I adjusted the oleos and did some light sanding to the wells. I took off the tail cone and put some graphite powder on the linkages for the tail surfaces. They operate much smoother now. I also applied some graphite powder to the aileron linkages to keep them free. I did a CG check. The CG ended up in the exact same place as the maiden but without the tail ballast. I added some to the nose under the cockpit tray to get the CG where I want it. If it is too much I can always remove some. The only thing left to do is possibly program some elevator mixed into the flaps. Should not be too hard.

I got some spare props in the mail yesterday so I need to cut them down (9x7.5 to 8x7.5) and paint them. Carl has sent me a sheet of white decal paper that I will print the prop logos on to dress up the front end a bit more as well. Getting super excited to share this project with you all in less than two weeks!

 

[Mid7night]

Great! I'm curious; after your most recent RPM results, are you planning on trying to program in a basic curve or constant offset, or are you just going to run them straight?

 

[willsonman]

Run them straight

 

[wilmracer]

I got some spare props in the mail yesterday so I need to cut them down (9x7.5 to 8x7.5) and paint them. Carl has sent me a sheet of white decal paper that I will print the prop logos on to dress up the front end a bit more as well. Getting super excited to share this project with you all in less than two weeks!

Glad the work event went off well. I found that even people who don't fly RC stuff are fascinated with our "toys".

I'm working on the same thing with the replacement prop for the P-40. She will still have her scars in the paint from previous accidents, but I couldn't bring myself to leave the wooden prop alone for FF. A nice painted prop with scale details adds a ton in terms of scale appearance.

 

[willsonman]

FanTASTIC! Look forward to having a scale companion at the event, Carl... and the decal sheets showed up... DUDE! I asked for one sheet and you send me 6! Seriously, people here are awesome.

 

[willsonman]

Well, I leave in one week. Really looking forward to seeing everyone there.

As an update, I got the decal printing situated and have some results to share. As I will be coordinating the Pylon races there I purchased a $13 hard hat and got a nice decal applied. Pretty sure there is no need for me to put my name on this one.


Spare props are reamed, balanced, painted, and decals applied.


Did a quick conversion of a png file to get the Bugatti properly in the radio. I got my flap mix programmed in and am pleased with how it works. I'll see how it goes at FF and may end up not using it with such a long approach there.


Decals are also applied to the props on the airplane.


I did another proper check of the CG. Without the ballast in the tail that was there for the maiden, the CG was in the same location. So, I added some ballast to the nose area under the cockpit tub. I used a plastic baggie with steel shavings so it is compact but if it ends up being too much, I can remove some weight for subsequent flights.

 

[PsyBorg]

THAT is one beautiful picture. I love how the props are perfectly parallel to the tail and the light reflection off various parts. Di you take time to compose that shot or just set it and snap? If you got lucky that's cool but if you thought that out before the shot and composed it Nice work.!!