Ultra budget quad build.

[jhitesma]

Only had a few minutes before work...but did a few quick tests in the front yard and 2.3 seems to be working. One odd issue was it wouldn't let me arm right away...not sure why - had to disconnect and reconnect the battery before it would let me arm.

Flew pretty well, acro felt a lot more stable to me but that could just be because it was less windy today than the last few test flights Level mode however felt way off today - must not have been very level when I did my calibration last night. No biggie, I'll try recalibrating over lunch and see how it does.

Now for the slow boat with my GPS to get here....

 

[jhitesma]

Oh heck, just checked my tracking info on a whim...the GPS is already in LA and in the hands of the USPS which means I could see it by this weekend. Didn't really expect it for another week. Guess I better get an order into Digikey for the battery it will need!

 

[jhitesma]

GPS left phoenix yesterday...sounds like a decent chance it will be here tomorrow. Fingers crossed!

Added the LED's Wednesday night and gave it a few test flights last night while fighting off a turkey coma:



Think I need to add a few more to the "inside" of the front/rear arms still....or maybe the bottoms. Liked flying with the lights though.

I posted that photo on my facebook last night with the caption "The chances of this ending well are so slim..." due to the combination of my small yard, the fire, the homebrew on tap and topped off with my still developing skills. And why replies back first? A friend who's an actual helicopter pilot and instructor (who also flies RC Helis) and who happens to be very conservative about safety. His comment? "Go for it!"

And speaking of aviation friends on facebook. One of my friends got to go on a fun little heli ride earlier this week:

 

[jhitesma]

Looks like no chance of my GPS arriving tomorrow now

Because the mailman just left it my mailbox about 10 minutes ago

Stuck working a half day today but guess what I'm doing this afternoon

I went with the ublox NEO-6M and ordered it off ebay for $17. The seller mentioned in their listing that it comes without the battery installed and lists the type of battery needed. So I ordered one from digikey since they were <$2. Except it turns out that because it's a rechargeable lithium battery digikey can only ship it ground...which meant $10 for shipping! If I had realized that the $2 battery would cost me almost $15 to get I would have just ordered the GPS module from a US seller since they all come with the battery pre-installed and cost about what I ended up paying for my GPS and battery. But the kicker....my battery hasn't arrived yet. So I was expecting the GPS to not be usable yet...except I just opened it and took a look...it DOES have the battery installed.

Oh well, it's here and looks ready to go. Now to finish my work so I can try to get it installed

 

[xuzme720]

Good deal!

 

[jhitesma]

Well...GPS won't be going on tonight. Doing some rebuilding instead.





Looks like my arduino is toast. Went out to the desert with my girls today and after doing some hiking and exploring I got to do some flying. It felt great. Was too windy for me to work on fine tuning my trims and PID's but it was fun to get to fly in the wind. I told myself the plan was fly one battery just to time how long a battery will go mostly hovering and gentle flying, then try to hoist a camera for a battery, then I'd see about pushing the rates up a bit and having some fun.

Apparently I just can't be trusted with a camera turned my direction though. I setup my camera on a tripod aimed between two bushes so I'd have an area to fly in. Video didn't come out very great though - I didn't do a good job of staying in frame. Level and Level with mag hold were both feeling great. And acro I was actually able to keep a handle on for a bit without it getting away from me. I even got to try altitude hold and it seemed to work very well. But then I got bored flying sedately around in a small area. Started playing in aero mode and felt like it was ready to try a flip again. But again I forgot to set those rates higher and I wasn't quite as high as I thought I was. Landed hard. Really hard. Top down.

Looks like my Arduino board is toast. Every connector on it is smashed or bent, the crystal is crushed, 2 caps near the USB port are broken off...it's ugly.

One motor had all three wires ripped off - but the other 3 appear ok thanks to being able to get knocked off thanks to zip tie mounting. All four props are trashed. One boom got cracked, but it's one I've already glued back together so no big loss

The big question is what to do about the arduino. I have a pro mini I can put in...but that won't work with the GPS without another arduino do do the bulk of the GPS work. A replacement mega is only $15...but I could get a Flip board for $15 with better sensors than I'm currently using and with them integrated into the board for a much cleaner setup. I could order a new KK board...but that won't work with the GPS so not really a replacement (I still want to try one...but I really want to get GPS working right now.) Even if I did a replacement arduino I've been wanting to pick up a better IMU to get away from the Wii parts....oh decisions....so many decisions....

 

[jhitesma]

Here's the video of the final flight:

Like I said not the best video. I'm still not sure what happened at the 1:15 point when the alarm on my phone started going off. That alarm usually means either the bluetooth module lost connection with the arduino or the phone lost contact with the bluetooth module. But I wasn't able to pick up the phone and see which condition was happening. I think the module lost connection with the arduino because the quad almost fell out of the sky at the same time like something reset itself in mid flight.

Oh - and made my decision. I really want to play with a better IMU...but just had an expensive holiday, have an expensive trip next week, and another expensive holiday coming up soon...so need to minimize the budget. Will probably put the pro mini together enough to keep flying...but found a US seller with mega clones for $17 with free shipping estimated arrival Tuesday. I will be looking at a "real" FC board for this quad in January though because I never planned on using the mega board for it full time....this board was earmarked for a 3D printer. And the board may be repairable....I think I even have the correct replacement caps on hand, and may have a replacement crystal...the connectors can be replaced, removed or repaired....so unless the board itself is cracked and damaged it should be repairable.

 

[jhitesma]

Was feeling pretty despondent over this crash ... it was my own fault for doing something I knew I wasn't ready for (again!) But worst of all it damaged the one part I never thought I'd break and eventually planned on using for something else.

But daughter went to bed on time and wife went to bed early...so I pulled up my sleeves....

This is the GPS (now temporarily back burnered) I took this as a reference to remember how I connected the "FTDI" cable:

(Side Note: I was able to get the GPS working with my "FTDI" cable which is actually a "Silicon Labs" cable since it uses a Si chip instead of an FTDI chip. But it was a cable I had on hand and was willing to hack apart. Not sure if it's actually configured correctly or not though.)

Next I figured I'd see if I could use my cable with the pro mini I had - so I just shoved the wires in the holes and clamped them in place with a Popsicle stick and clothespin - it worked! I was able to get MW2.3 uploaded to it:


So I started taking apart the quad to get the mega out for repair and get the pro mini wired in for the next few days. This was after I straightened most of the bent stuff on the mega. Looking at the board I suspect those two damaged caps aren't even used when powering it the way I am...so the crystal is probably the biggest problem with it:


With the quad this far apart I figured I should grab some detail shots of my hacked together IMU and how it's mounted. You can see the baro sensor at the top, the LLC board higher up and there in the middle is the sandwich of WMP and Nunchuck boards that makes up my IMU:


Wiring on the IMU is really ugly. But it's been working so I'm not in a hurry to mess with it:


Got pretty far tonight. Ran out of wood so I'll have to hit the store before I can finish reassembly. Went through the motors and found one broke a winding as well as the one that had all three wires pulled off. Will try to remember to pick up some epoxy when I get the wood tomorrow. Soldered bullets on the two spare motors and have them hooked up now, I'll have to rewind these two damaged ones and turn them into the new spares. I also cleaned up the wiring on the LED's and they're tied into the main harness much nicer now...unfortunately I also took out one of the SMT resistors on one section of lights, will have to see if I have any suitable replacements on hand:


Almost ready to finish physical assembly and test fly again...but for some reason I'm getting no gyro reading with the pro mini setup as it is right now. It's 1AM though and I need to get to bed so I'll have to try and figure that out tomorrow. What's odd is all the other sensors ARE working, and that's odd because the baro and mag are both 3v and I haven't hooked up any 3v yet since the pro mini doesn't have a 3.3v line like the mega did. I'm also not sure how I'm going to hook my bluetooth up on here yet. I'll probably do it the same way I hook up the "FTDI" cable - just four pins and a bit of popsicle stick to hold them in place.

The pro mini shaves a good chunk of weight off compared to the mega. But I think I like the extra options the mega gives...and it's 3.3v regulator is handy. Getting started the mega was definitely easier than the pro mini was tonight. But if I was building a quad without GPS the PM is the way I'd go for sure. Since I'm toying with the idea of doing a hex now that I have 6 motors the mega is still a nice option with it's extra pins.

Oh - I also looked at flashing my ESC's...it won't be easy, but I do have two fine enough tips for my iron that I could possibly pull it off. May wait until I order some easier to flash ESC's though so I've got spares in case I fry something. I really want to try simonk because I'm not at all impressed with the performance of these multistars in several ways - still my biggest regret in choices I've made building this.

 

[cranialrectosis]

I read this and I think, why not just invent Flubber and be done with the cumbersome electronics and mechanisms?

Really, I am just in awe. Talk about pushing the envelope.

Simon K. Worth every penny.
http://www.altitudehobbies.com/brushless-esc-bec/ipeaka-iq-multicopter-esc-simonk

 

[jhitesma]

I'll be going SimonK I've already decided. But I can't bring myself to pay extra to have someone flash it when I have within arms reach about 6 different things that could be used to flash it myself.

These multistart 10a's are just crazy difficult to reflash because they don't have pads and they have an extra small processor with very very small leads to try and solder to. With some fine magnet wire (which I have on hand) and a bit of patience and the right tip on my iron I can do it. But I'll probably order something like the 12a HK Blue's and reflash them instead since they have pads so it would be a lot quicker/easier. Then when I get bored some evening I can try flashing the multistars without having to risk a failure leaving me without a quad

Waiting on paint to dry on the replacement arm...Should have it in shortly. But now to figure out why the gyro isn't responding anymore. I may take out the baro and mag for now just to simplify things. Worried that something happened to the gyro in the crash, but the accel is slaved off the gyro and it's still reading so that rules out a lot of simple obvious things that could be wrong with the gyro. And if the gyro is dead then I'm going to have to order either a new gyro or should have just gotten a real FC board since I don't have any other WMPs to tear apart. $6 and I can have a MPU-6050 breakout here probably some time this week. Should give me a much better gyro/accel setup. But for $15 I can have a full 10DOF IMU board which would put all my sensors on one board and clean the wiring up BIG time. Except the IMU's with the sensor package I'd prefer are all shipping from China so I'd have to wait a few weeks.

After I finish paying this weeks bills tonight I'll probably make my decision

Now to go try and figure out what's up with the gyro and finish physical reassembly before my wife has to go to work.

Oh BTW - Gotta give her thumbs up, she hasn't chewed me out for smashing my toy In fact she ordered me something from HK for Christmas I found out. Don't know what...I have a few items specifically picked out on a list for her if she wanted to go that way. I think she did it just to keep me from buying more stuff this next month though - she knows that I'm not going to buy much of anything until I find out what she got me

 

[jhitesma]

Fixed and in the air again...though with diminished abilities:



Removing the 3.3v sensors and enabling the internal pullups got the gyro and accel working...but I'm getting a lot more I2C errors now for some reason. If I turn the internal pullups off I get too many errors and no signal. On the Mega I didn't need the internals and everything was working...must be a difference between the two boards.

Not flying well yet though. One of my motors did finally get a bent shaft, swapped it for one off of one of the motors down for rewinding and it's doing better...but still has a big of a shake, I think the other surviving motor may have a very very slightly bent shaft...have to take a closer look at it.

PID's are off too, Just went from memory instead of my saved config and I don't have the bluetooth hooked up so it's a little tricky to tune without USB or Bluetooth. Tempting to try and mount everything inside now since this board is so much smaller. But it's only temporary until my replacement mega gets here...just can't say no to the extra abilities the mega brings to the table.

Of course it's also windier than it has been in weeks today so other than a few quick hovers over grass I haven't gotten to really try much. But it's good to know I can recover from such a massive crash in less than 24 hours and be flyable even if not flying well

 

[jhitesma]

And the problem was....PID's. Apparently they hadn't actually saved when I saved them. Hooked it back up to the config program and the PID's were back to defaults. Used the alternative config program to load the ones I saved the other day - but for some reason the switch programming wouldn't save from that program - so I had to go back to the stock config program to setup my switches.

But with my switches reprogrammed and the PID's back where I had them it's flying smooth and stable again! Wonder if I have a 3.3v regulator laying around so I can get the other sensors wired back in....

I could try the GPS...but with only one serial port on the pro mini that means no way to run bluetooth or the PC based program while the GPS is hooked up - and that means no way to debug it. I could try hooking up an LCD...but the ones I have are really too big...and again they need the serial port which is taken by the GPS. Only other option for the GPS would be to use another arduino to handle the GPS and talk to the main arduino with I2C like the sensors...but I don't have any other suitable arduino's on hand ( have some digisparks but they don't have enough pins...)

 

[jhitesma]

Oh - and just in case anyone was wondering. My original center booms aren't that faded. I just switched paints. Originally I used a cheap $.80 walmart acrylic diluted with windex. But a week or two ago I splurged on a bottle of real airbrush paint in hot pink and have been using it instead. It's a bit brighter but if you put it on too thick it suddenly goes from bright pink to a rather dull red.

Pink props still aren't here...2 things I've ordered from Hong Kong after I ordered them have already arrived but no pink props yet

And FWIW - the black front boom and black props do very little for me as far as helping with orientation. They do help a bit but not nearly as much as instinct. The lights on the other hand help a lot.

 

[cranialrectosis]

I find the markings I put on the rear of a small copter to be far more valuable than those I put on the front. On my Flying Purple Squirrel Eater with the 20" diagonal, the yellow front rotors are good enough. On the 14" Bumbler Bee, the rotors are too close together for the different colors to be much help unless I am looking down on the copter. I painted bright yellow stripes on the rear boom and those make orientation possible. I am thinking of adding some lights so I can see the stripes in low light.

 

[Cyberdactyl]

Impressed you're sticking with this jhitesma.

I can't help but notice you have a lot of wiring all over your quad. I also notice you are not using any ferrite rings anywhere, not even on the ESC lowV and control wire. You also said the ESCs are not flashed for multi-rotor. Not sure if it would help in your adventures, but it surely wouldn't hurt.

 

[jhitesma]

Impressed you're sticking with this jhitesma.

Haven't really had anything I'd consider a "failure" yet, just a few hard lessons in humility, though those can be harder to endure

I can't help but notice you have a lot of wiring all over your quad. I also notice you are not using any ferrite rings anywhere, not even on the ESC lowV and control wire. You also said the ESCs are not flashed for multi-rotor. Not sure if it would help in your adventures, but it surely wouldn't hurt.

Yep! Lots of room for improvement still! I wanted to try the pro mini since it's so much smaller, and it works and I could easily see myself using it to build a fun little quad. But I really can't advocate doing things the way I am unless you really do have all the stuff on hand to start with. A real multirotor board would be so much nicer and eliminate so many potential issues. But I like the mad scientist look mine has going on and the fact that despite looking so ugly it can fly really well.

The multistar ESC's are supposedly designed for multirotor use...but they're no simonk and it sounds like even stock plush ESC's are better. When I bought them I knew that the 10a's were theoretically flashable, but had not found anyone who tried it on them yet due to the lack of programming pads and the tiny package on the processor. I really do want to flash them...but there's such a high risk of damage that I don't want to risk loosing one. So until I can order a set of backup ESC's I don't plan on flashing anything just yet. Unless I work out a way to make a small enough jig to work like the ones for larger sized chips.

In some ways I think of mine as being at a v3 point due to the changes I've made...but really it's more like a v0.3 because it's still nowhere near what I plan. I'm hoping in a few months to build a second one based on what this one has taught me and knowing a set of parts that work well together I can order complete for <$100. I may try doing it on an anycopter style frame instead and I'll almost certainly use an off the shelf FC.

But for now I want something really flexible that I can swap sensors and other subsystems on quickly and easily. Too windy to fly today (7mph sustained here in town means way too windy out at my flying places) and predicted to be windy through Tuesday. Which is when my new mega board should be here so I can hook up GPS and bluetooth at the same time So looks like I need to get serious about building a better camera mount.

 

[jhitesma]

You know, I do tend to over share on setbacks and issues and forget about some success. Probably partly because success tends to turn to failure as soon as I mix in a camera

So I figured I should try to get a video of a "Good" batteries worth of flying. No cameraman and my attempts at a hat cam just haven't worked out (the old flip is too top heavy) and it's really too windy today (7mph with 10-15 gusts here in town). But having the fixed view of the camera to stay in would give me incentive to stay below the fenceline which would help block the wind.

Still a small area to fly in, I did this almost all in angle/stable mode with just a bit of time spent in Horizon. The wind was really pushing me around though making it hard to stay put in such a small area. I also experimenting with pinching instead of thumbs for a bit but it didn't go well.

Oh and I've learned with my last two attempts at video that the preview on the flip is a joke. Because on both I got way more on the left of the scene than I expected and way less on the right. Need to adjust for that in the future. Sorry, I thought the camera was centered over the chair - so some of my flying is off the side of the screen.

It's hard to see in the video but I flew until 11.4v (3.83v,3.82v,3.83v) I had my beeper set for 3.5 but on everything lipo powered I've owned I almost always stop around 3.8 because I feel a sudden major change in performance at that point. After resting the pack was reading 11.58v (3.86v,3.87v,3.86v) 56% and the 5:00 timer on my throttle% in my TX was still showing 2:47 left. When I push past 3.8v the 3.5v beeper will usually go off just before the TX timer which will go off while I'm coming in for a landing - and after those flights the pack is down to 3.4 and reading just a few percent left. But the flying after 3.8v just isn't as much fun as the power just isn't there.

I have Turnigy, GeForce and Zippy packs in the same size/configuration and similar C ratings that I've tried on the quad and all of them I feel that hit at 3.8v per cell. I also feel it in the SkyLipo's I generally use on my 24g swappable pod and the nanotech, turnigy, Tigers single cells I use in my micro helis and X1. Usually by the time my beeper is going off I'm already planning on landing because I've noticed the drop in power.

 

[jhitesma]

Interesting...plugged the arduino mega in for the first time since the crash since I've straightened the pins and headers on it...and it seems to power up through USB at least and responds correctly. I don't have anything connected to it so can't say if it's really working or not...but it seems to be. With the crushed crystal I don't trust it to try and fly though.

I also gave in and ordered a MPU-6050 on a breakout board so I can use it in place of my WMP/NK sensors. It will also mean all my sensors are on 3.3v which will be nice and should simplify the wiring a bit. Being a higher performance sensor doesn't hurt either for under $6 delivered.

I'll probably take the WMP/NK and package them with the promini to use as a stabilizer on my Versa. Or if I get a few more spare motors and a spare set of ESC's.....

 

[jhitesma]

Wow, the ebay seller I ordered my arduino mega clone from must have beat Amazon to the drone delivery setup because it was waiting for me when I got home for lunch today! Incredibly fast shipping.

Almost didn't fly over lunch due to wind...but told myself to suck it up and learn to fly in the wind already. So flew a pack in the front yard. Really like the way it's controlling after bumping the D up a bit yesterday.

But now with the new mega here I'm torn. Dig right in on setting it up tonight? Should go fairly quick and easy. But once I dig in I'm going to want to set the 3.3v sensors back up and that's a bit more involved because I want to mount them better and fix the wiring harness. And until that mag is hooked back up the GPS won't do me any good and getting the GPS hooked up is the main reason I'm itching to get the mega back in tonight....

But the MPU-6050 I ordered just shipped today and won't be here for a day or two. And while I have things torn down to swap back to the mega I'd like to just go ahead and setup the 6050 as well....but at the same time that's starting to sound like too many changes at once...

Maybe I'll just work on that camera cradle tonight....

 

[cranialrectosis]

You might try some type of foam cover. Should still keep this ultra-budget and the crashes less painful.