stunthanger.com
Design => Engineering board => Topic started by: Tim Wescott on October 19, 2014, 11:11:32 PM
-
I got some flights in today with the TUT and aunty servo* in place. Some things I learned are:
- My Tower 40 rises about 800 RPM from start to level flight, and doesn't stay at a very consistent speed**
- My Tower 40 was sagging about 400 RPM at the tops of loops (this may be a needling issue: I think I should have launched a bit richer)
- Rotation rates in the aircraft's z axis (direction of lift) is pretty much proportional to elevator
- Control deflection for square maneuvers isn't all that much greater than control deflection for round maneuvers (this came as a bit of a surprise)
- I never hit maximum elevator deflection -- maximum up is 1.9ms, maximum down in 1.1ms: I just didn't bang into these that I can see
- Profiles vibrate a lot, which makes acceleration data hard to interpret
Here's graphs for the horizontal and square eights; the horizontal starts at about 166 seconds, the square starts at about 187.
* Aunty servo is a servo that's been gutted, and turned into a pulse source, whose pulse width is proportional to input arm position.
** Anyone with a piped engine want to do some experimentation?
-
Good Morning Tim;
Interesting data you've collected, thank you for the effort put forth.
Curious, can you 'cross-reference' elevator throw in degrees to the various pulse train rates?
r/
Dave
-
Curious, can you 'cross-reference' elevator throw in degrees to the various pulse train rates?
I should do that. It's just a matter of hanging a protractor on the tail.
I should note that this model is a flapless plane -- I'm sure that the response of the plane to control input is broadly representative of how a flapped stunter responds, but different in the details.
-
There's something WRONG about waltzing up to a CL plane with a laptop and holding a conversation with it via USB -- yet, here I am, this past Sunday, doing just that.
-
Tim;
You can find solace in the fact that at least you are not having it download telemetry data in real time to a screen held in your hands or to an ear bud a la the latest R/C hardware... ;)
r/
Dave
-
Go wireless? There are wireless data cards and usb devices available. Yes extra weight, but not too much.
OK, they are freaking expensive.
http://www.bestbuy.com/site/sandisk-connect-32gb-usb-2-0-wireless-flash-drive/1542565.p?id=1219054921646&skuId=1542565&ref=06&loc=01&ci_src=14110944&ci_sku=1542565&extensionType=pla:g&s_kwcid=PTC!pla!!!92339115919!g!!58873410679&kpid=1542565&k_clickid=358f2abb-5997-4898-9a1a-df105e00946f&kpid=1542565
http://www.amazon.com/Eye-Fi-Connect-Wireless-Memory-EYE-FI-4CN/dp/B003DV4234
Phil
-
Hi.
Real-time telemetry is easily available, F3D people use such. I think it's from Simprop.
I have a Multiplex/Simprop system (not real-time) that can log up to 16 parametres, but I'm only interested in 3; rpm and 2x temperature.
Now I just have to learn how it works.
L
-
Pretty cool experiment Tim! I would be curious to the degrees of throw per ms as well. If we are truly only using 15-20 degrees of surface deflection, then it would seem we are losing a lot of resolution by building for 30-35 degrees of deflection.
-
Curious, can you 'cross-reference' elevator throw in degrees to the various pulse train rates?
I think the center isn't quite right.
I'm not sure how much the absolute deflection matters that much, given that every airplane will be different, and that this one in particular is flapless. But here you go.
-
I think the center isn't quite right.
Or you need to get your bottoms down.
Very cool, Tim. What's it look like with the time axis stretched out? With pitch rate superimposed?