HoverFly Question - I'm ready to add the receiver to my new ELEV-8. It says the Gear channel is used for altitude hold. Since I'm not using that, do I really have to plug something into that channel? I'd much rather use my two remaining channels to control other things.
Futaba lost me as a customer after I purchased an early 6EX. What a piece of Smile. Then they wanted a $100 for an "improved" receiver to correct the original piece of Smile. Sorry, I find Spektrum products, while not perfect, to be reasonably reliable and reasonably price. I have noticed that Tower Hobby keeps deeply discounting Futaba products to get their sales up.
HoverFly Question - I'm ready to add the receiver to my new ELEV-8. It says the Gear channel is used for altitude hold. Since I'm not using that, do I really have to plug something into that channel? I'd much rather use my two remaining channels to control other things.
For the HoveryFly code, yes you need something plugged in there. The value it receives on that channel is used as the control gain. It may be possible to fly without it, but there's no mention in the manual of a "default" value or anything.
With my version you'll can hard-code the gain and use that input for something else. I'm currently using it to test different low-pass filter settings.
I have had many Futaba radios over the last few decades. The only problem I had was with my first 12z. It simply stopped working after just a couple hours. Futaba overnighted a new one to me before I sent the first one back.
So tomorrow is the big day! I'm finally done with this thing. Spun it up inside the shop briefly and it seems like the gyro gain is way off. That could just be me. Looking at the gain, I have a question that someone out there can probably answer. It says to set the gains (gear chan-5) to +/-25% on a Spektrum Tx. On Futaba both gain numbers will be positive (there is no negative). I set mine to +/+25% and that's what didn't feel very good. I ran them up to 50% and that seemed a little better, but that seems high. That's where I quit because I was thinking if I kept it up I would probably destroy the place. I did look later at the gains in the Setup program and they were at 88. I figure the last word on this is where the program thinks the gains are so I'm wondering what kind of gains you guys show in Setup? If anyone is running Futaba, I'd be interested in your settings. Oh... One more thing. The right stick seems super touchy. Are you guys running those travels at 100% like the book said? Running expo (I'm not)?
Edit: Should have mentioned that my Tx is a 10CAP and the firmware I loaded in the quad was "Sport". Tnx
I think my gain in 60 or 70, and I've got about 60% expo with full travel, and I have dual rates set to 70%, with about 40% expo there. I also found it really twitchy with their suggested settings. Maybe they're trying to sell crash kits too.
As a note for those following along I am sure Rich and Tiger know this. Futaba vs. JR/Spectrum when setting EXPO Futaba is a negative number (-10%) or what ever number makes it more comfortable to fly less sensitive around center stick, JR/Spektrum is a positive number (+10%) for a similar result. Reason I am posting this is we had a new flyer come out to the feild about a year ago he had just moved to the area and he used Futaba, being a helpful guy he told some new guy's they needed EXPO and offered to dial it in well he went negative on more than one JR/Spektrum radio. We converted him to JR he has a 12X now.
I use a Futaba T6J and the gain end point on mine is set to 20%. I started first with 15% then switched it to 20%. In all honesty I need to spend more time fine tuning it. By the way with a 20% end point set for the gear switch up position the transmitter is outputing 1630us. I have had to much fun flying it and need to take the time to fine tune it. Here is a video flying it with the gain endpoint set to 20%. Still need more practice with my landing!
Edit: I have my dual rates set for 50% - a real beginner!
Got it into the air today! Here is a video of that first flight. It's mostly hovering and gain tweaking with a little forward flight at the end. Pretty cool so far.
I tried to embed this, but couldn't make it work. Anybody know how you do that?
So today I was having a conversation with someone and I mentioned something about discovering that the HoverFly Sport had heading hold capability. This person wasn't very familiar with the heading hold concept so I tried to explain. I don't think I did a very good job so tonight I decided to make a video to demonstrate what I was talking about. This is another one of those cases where I think a picture is worth at least a thousand words. :-) I'm really glad I did this because now I have something to help me explain to my heli friends. I find that only about one in ten of them really understand this. I used the ELEV-8 for the demo so I thought I'd post it here. I hope you guys find it interesting.
Since my ELEV-8 is finished now, I thought I'd post a couple hi res pictures of it. I made a lot of changes to the design during the build process and took a lot of pictures. I hope to share those on here very soon and explain what I did and why. My thanks to Ken and the Parallax team for a great quad!
@Brian: mine weighs 3 lb 5 oz all-up ready to fly. This is 11oz for 4000 mAH Turnigy nanoTech LiPo + 2lb 10oz everything else.
@others:
I am up and flying for the past week or so. I haven't posted anything because right now I'm just learning how to fly it...focusing on getting to a steady hover 2-3 ft off the ground. 55% gain seems to work well for me. Making good progress, but it is definitely a process of baby steps.
It turns out that a plastic coat hangar can be used to repair a broken landing gear in a pinch.
BR while I was waiting for Ken to ship me a new set (Thank you Ken!) I went down to Lowe's and bought a 6 foot length of 1/2 inch wide by 1/16 inch thick aluminum bar and bent, shaped and drilled them myself. I even added a little height to the existing platform.
Edit: Ok my brother did the work I only bolted them on!
Brian,
That thing is looking awesome! I like how clean it looks, and the functionality you get of being able to fold it. I cant wait to see that thing fly, and to see it in person at the Expo!
I just want to thank everyone at parallax and everyone on this forum for being so "open". Here is my variant of the elev-8 frame made out of 1/8 and 1/4in lexan that i can get down the street at the local glass shop, and some 1/2 in aluminum tube from lowes (i can't stand waiting for materials to ship). In 7 to 21 days (ugh) i will get my parts from HK and be crashing it like a pro! Its been about 6 years since my last R/C heli flight, but hey its like riding a bike right? Maybe i should be working on an auto parachute while i wait, ha!
I just want to thank everyone at parallax and everyone on this forum for being so "open". Here is my variant of the elev-8 frame made out of 1/8 and 1/4in lexan that i can get down the street at the local glass shop, and some 1/2 in aluminum tube from lowes (i can't stand waiting for materials to ship). In 7 to 21 days (ugh) i will get my parts from HK and be crashing it like a pro! Its been about 6 years since my last R/C heli flight, but hey its like riding a bike right? Maybe i should be working on an auto parachute while i wait, ha!
thanks again everyone!
Hey skynugget - welcome to the crash demonstration show! The ELEV-8 design is indeed open and all uses are welcome. Anybody can make these, anytime, for any purpose (personal or commercial). We have only support for such efforts and that's why the drawings are stamped with Creative Commons. Our interest is in the processor mounted on that shiny blue board on the third photo.
You'll be a fine pilot. Nick at Parallax has something around 80 flights logged on his SPORT board, with only a crash here and there. I'm somewhere in the same range. The crashes with these are very forgiving - a broken propeller, snapped landing gear, etc. Ask the other guys for their opininon too.
It's no mistake that the ELEV-8 could be built out of other materials. Up above you've seen Brian Carpenter's thin plywood. And now you've made the clear version out of Lexan. The simple sandwich design can accommodate a bunch of different material choices.
Welcome to the action. We hope to hear of positive results using Jason's code in the very near future!
got up a little courage this evening and attemped to hover. i am impressed with your flight skills those that are flying. hovering is going to take me a bit.
I'm attaching the latest versions of my flight code for the HoverFly Sport / Elev-8 combo, along with the GroundStation software to configure the settings (both the executable and source).
The three attached files are:
1) QuadX - HoverFlySport.zip - This is the package containing the SPIN / PASM flight software that you compile in the Prop tool and upload to your HoverFly board. The top object file is called "Main - HoverflySport.spin"
2) Sport-GroundStation-Executable.zip - contains the .exe file that talks to the running Prop to graph gyro readings, change config settings, and compute gyro drift.
3) Sport-GroundStation-Source.zip - contains the C# source code for the above.
Using the flight software:
To use the flight code, load "Main - HoverFlySport.spin" into the Prop tool. Choose "Run -> Compile Current -> Load EEPROM" (F11) to program the board. I've made the settings comparable to the defaults for the firmware that ships with the Sport board, so your radio settings should be interchangeable. The gain switch does NOT control your gain or altitude hold at this point. Toggling the gain switch flips between "raw" gyro readings and readings that have been additionally low-pass filtered. With the latter setting, there's a little more oscillation when you make rapid control changes, but it will handle a noisy gyro better.
To use the GroundStation program:
Run the executable. There is no installer 'cause there's nothing to really install, though you will need the .NET 2.0 framework installed. Any reasonably modern Windows machine will have this already.
Once your Hoverfly board is running the supplied SPIN program, connect the board to your PC with a USB cable, then run GroundStation.exe. The first tab of the program should start showing gyro readings within a few seconds. Hopefully they'll be a nice flat line, centered very close to zero.
Switching to the 2nd tab, you can change the control sensitivity, frame config, and PID values. I don't have a "nice" way to set the PID numbers just yet (like the gain on the supplied Sport firmware). The values provided should work for the Elev-8 platform. If they don't, change one number at a time, and change it by roughly a quarter to half of its present value at a time, either up or down. So, if it was 10,000 to start, change it to 7000, then 5000, then 4000, etc, if you're trying to make it less sensitive, or 13000, 17000, 23000, etc... to make it more sensitive. Tuning the PIDs is an ongoing process, but they're reasonably forgiving when you're in the right neighborhood.
On the 3rd tab you'll see the temperature drift calculator. This isn't a necessary step to use the code, however it will make the code perform better in situations where the quad experiences temperature changes over the course of a flight.
To compute your temperature drift, start with the quad being very still. Point a hair dryer at the quad long enough to raise it's temperature a bit (you should see the dots moving across the display - it usually only takes 5 to 10 seconds). You'll see the lines on the graph move around and fairly quickly settle down and stop moving. When the lines and values aren't changing much, write them down. Open the ITG-3200-Sport.pasm file, and near the end you'll see these six variables:
'' Variables for dealing with drift / division
tempDivisorX long 0
tempDivisorY long 0
tempDivisorZ long 0
tempOffsetX long 0
tempOffsetY long 0
tempOffsetZ long 0
Change the numbers to whatever you've written down and re-program the eeprom on your Sport board with the new values. Once it's reset, if you perform the same routine with the hair dryer, the lines should stay very flat as the temperature rises, meaning the code is doing its job.
If anyone has any questions, you can ask them here, or feel free to PM me. I hope to make the PID settings easier to deal with, as well as making the temperature drift stuff something that can be programmed like the settings, but I wanted to get a first version of the code out there for people to try.
One extra detail:
The gyro display has a couple columns of numbers that aren't labeled. They are Minimum, Maximum, Average, and Variance. They represent the min, max, and average readings of your gyro over the entire display, and the variance value, which is, on average, how far FROM the average each gyro reading is. If your board is stationary, the average value should be close to zero, but the variance value is the most important one - it should be very small. Mine are usually 0.2 or less. Anything less than 3 is probably ok.
Comments
Thanks! ...Tiger
I have channel 5 hooked up, which is the gear channel on the Futaba 6EX, it controls the gain.
Look here: http://www.lynxmotion.com/p-676-dsm2-airmod-war7000.aspx
Futaba lost me as a customer after I purchased an early 6EX. What a piece of Smile. Then they wanted a $100 for an "improved" receiver to correct the original piece of Smile. Sorry, I find Spektrum products, while not perfect, to be reasonably reliable and reasonably price. I have noticed that Tower Hobby keeps deeply discounting Futaba products to get their sales up.
Regards,
TCIII
For the HoveryFly code, yes you need something plugged in there. The value it receives on that channel is used as the control gain. It may be possible to fly without it, but there's no mention in the manual of a "default" value or anything.
With my version you'll can hard-code the gain and use that input for something else. I'm currently using it to test different low-pass filter settings.
J
I share the concern. I'm also not sure they're riding the electric wave.
Ken Gracey
I'm happy with Futaba.
Edit: Should have mentioned that my Tx is a 10CAP and the firmware I loaded in the quad was "Sport". Tnx
Thanks! ...Tiger
Edit: I have my dual rates set for 50% - a real beginner!
http://www.youtube.com/watch?v=auoEB1gqeQs&list=UUtHTG82Nn83T5JSdWdTgkAg&index=1&feature=plcp
1. Start at 50%.
2. Attempt to hover.
3. Quadcopter jitters? Go to #4. Quadcopter flies with stability? Fly!
4. Reduce gain by 5%.
5. Go to #2.
Then dial it in a few percent below the setting that causes jitter.
Ken Gracey
I tried to embed this, but couldn't make it work. Anybody know how you do that?
...Tiger
Thanks Ken!
Can anyone give me an idea?
...Tiger
...Tiger
You are using the DX8. Would you care to share your config file for the elev-8?
@others:
I am up and flying for the past week or so. I haven't posted anything because right now I'm just learning how to fly it...focusing on getting to a steady hover 2-3 ft off the ground. 55% gain seems to work well for me. Making good progress, but it is definitely a process of baby steps.
It turns out that a plastic coat hangar can be used to repair a broken landing gear in a pinch.
Edit: Ok my brother did the work I only bolted them on!
Here are some pictures so far of the build
I came across a small quantity (25) of these. They would fit your DX8 perfectly. PM Me if you would like one for $20. plus shipping.
That thing is looking awesome! I like how clean it looks, and the functionality you get of being able to fold it. I cant wait to see that thing fly, and to see it in person at the Expo!
Keep up the good work!
thanks again everyone!
Hey skynugget - welcome to the crash demonstration show! The ELEV-8 design is indeed open and all uses are welcome. Anybody can make these, anytime, for any purpose (personal or commercial). We have only support for such efforts and that's why the drawings are stamped with Creative Commons. Our interest is in the processor mounted on that shiny blue board on the third photo.
You'll be a fine pilot. Nick at Parallax has something around 80 flights logged on his SPORT board, with only a crash here and there. I'm somewhere in the same range. The crashes with these are very forgiving - a broken propeller, snapped landing gear, etc. Ask the other guys for their opininon too.
It's no mistake that the ELEV-8 could be built out of other materials. Up above you've seen Brian Carpenter's thin plywood. And now you've made the clear version out of Lexan. The simple sandwich design can accommodate a bunch of different material choices.
Welcome to the action. We hope to hear of positive results using Jason's code in the very near future!
Sincerely,
Ken Gracey
The three attached files are:
1) QuadX - HoverFlySport.zip - This is the package containing the SPIN / PASM flight software that you compile in the Prop tool and upload to your HoverFly board. The top object file is called "Main - HoverflySport.spin"
2) Sport-GroundStation-Executable.zip - contains the .exe file that talks to the running Prop to graph gyro readings, change config settings, and compute gyro drift.
3) Sport-GroundStation-Source.zip - contains the C# source code for the above.
Using the flight software:
To use the flight code, load "Main - HoverFlySport.spin" into the Prop tool. Choose "Run -> Compile Current -> Load EEPROM" (F11) to program the board. I've made the settings comparable to the defaults for the firmware that ships with the Sport board, so your radio settings should be interchangeable. The gain switch does NOT control your gain or altitude hold at this point. Toggling the gain switch flips between "raw" gyro readings and readings that have been additionally low-pass filtered. With the latter setting, there's a little more oscillation when you make rapid control changes, but it will handle a noisy gyro better.
To use the GroundStation program:
Run the executable. There is no installer 'cause there's nothing to really install, though you will need the .NET 2.0 framework installed. Any reasonably modern Windows machine will have this already.
Once your Hoverfly board is running the supplied SPIN program, connect the board to your PC with a USB cable, then run GroundStation.exe. The first tab of the program should start showing gyro readings within a few seconds. Hopefully they'll be a nice flat line, centered very close to zero.
Switching to the 2nd tab, you can change the control sensitivity, frame config, and PID values. I don't have a "nice" way to set the PID numbers just yet (like the gain on the supplied Sport firmware). The values provided should work for the Elev-8 platform. If they don't, change one number at a time, and change it by roughly a quarter to half of its present value at a time, either up or down. So, if it was 10,000 to start, change it to 7000, then 5000, then 4000, etc, if you're trying to make it less sensitive, or 13000, 17000, 23000, etc... to make it more sensitive. Tuning the PIDs is an ongoing process, but they're reasonably forgiving when you're in the right neighborhood.
On the 3rd tab you'll see the temperature drift calculator. This isn't a necessary step to use the code, however it will make the code perform better in situations where the quad experiences temperature changes over the course of a flight.
To compute your temperature drift, start with the quad being very still. Point a hair dryer at the quad long enough to raise it's temperature a bit (you should see the dots moving across the display - it usually only takes 5 to 10 seconds). You'll see the lines on the graph move around and fairly quickly settle down and stop moving. When the lines and values aren't changing much, write them down. Open the ITG-3200-Sport.pasm file, and near the end you'll see these six variables:
Change the numbers to whatever you've written down and re-program the eeprom on your Sport board with the new values. Once it's reset, if you perform the same routine with the hair dryer, the lines should stay very flat as the temperature rises, meaning the code is doing its job.
If anyone has any questions, you can ask them here, or feel free to PM me. I hope to make the PID settings easier to deal with, as well as making the temperature drift stuff something that can be programmed like the settings, but I wanted to get a first version of the code out there for people to try.
One extra detail:
The gyro display has a couple columns of numbers that aren't labeled. They are Minimum, Maximum, Average, and Variance. They represent the min, max, and average readings of your gyro over the entire display, and the variance value, which is, on average, how far FROM the average each gyro reading is. If your board is stationary, the average value should be close to zero, but the variance value is the most important one - it should be very small. Mine are usually 0.2 or less. Anything less than 3 is probably ok.
Brian - Stick with it. It will come faster than you think.
Jason - Awesome work man. Thanks!
...Tiger