Jump to content

Recommended Posts

Yee-haw, I finally purchased a Quad.

After a long back-order wait, I finally received the Gaui 330X-S kit about a week ago. It was put together late last night. And, today it was hovered in the backyard. Although the internet forums are full of good details, here's my first impressions.

I must say everything went together without any issues. Overall, the kit is very good quality (this is the second generation version with the Scorpion brushless setup). All that is needed is some CA, accelerator, and a metric Allen hex driver set. Parts count is quite low and basic model airplane/helicopter building skills will get the job done. The written instructions rely mostly on well organized drawings, so the usual English translation problems are minimal.

First, here is a file photo of the little fellow from Gaui, the Taiwan manufacturer:

post-2-128279150651_thumb.jpg

Details to the 330X-S are found on the Gaui web site: Click Me!

Share this post


Link to post
Share on other sites

For my initial testing I decided to just use a Spektrum DX6 R/C system that has been sitting idle for a couple years. My first mod was to create a little fiberglass shelf and mount it at the rear of the model. The Rx sits on this perch. See photo.

post-2-12827892218_thumb.jpg

Remote 3-axis gyro gain was configured to use the traditional CH-5 Rx port (typical of model heli's). However, for those that only have a basic 4-CH radio, you can omit the remote gain and use the manual pot setting instead. I prefer remote gain.

Share this post


Link to post
Share on other sites

With the Rx moved to the back and off the normal mounting area, a huge amount of free space was available on the main platform. That is where a FeiYu FY-20A-2 Tech Flight Stabilization module will go. Details to this 3-axis stabilizer are found here: Click Me!

The FY-20A requires a isolation damper, which is a suspended rubber platform with o-ring guys at the four corners. It was a bit bulky, so I pulled the damper off its base, drilled mating holes in the 330X's mounting plate, and rebuilt the isolation system directly on the model. This save some height, but not enough. So, my final plans are to remove the FY-20A from its plastic case and shrink wrap it.

Here is the rebuilt isolation hardware:

post-2-128278978779_thumb.jpg

Here is the FY-20A with the factory installed case. This is just a test fit.

post-2-128278987346_thumb.jpg

As mention, the FY-20A's case needs to come off. I'll get to that in a few days. But for now, I want to continue my initial X330-S test flights without the FeiYu Tech stabilization system. For sure, I'll report back soon on how the FY-20A stabilization gadget works out. Word is that it is very effective, so I am definitely anxious to try it. :)

Share this post


Link to post
Share on other sites

Good report , thanks. I have an FY-20 that I'm going to use for my own design quad and 2 FY-21's for my Twinstars. I think it will need turning down to almost zero to work with the quad but can't wait for your update.

Terry

Share this post


Link to post
Share on other sites

As mention, the FY-20A's case needs to come off. I'll get to that in a few days. But for now, I want to continue my initial X330-S test flights without the FeiYu Tech stabilization system. For sure, I'll report back soon on how the FY-20A stabilization gadget works out. Word is that it is very effective, so I am definitely anxious to try it. :)

Hey, congrats on your new quad. I´m not very familiar with the gaui, does it comes with the FY-20A? or are you doing this mod?

could you please tell us what are the key features of this quad?

How much weight could it lift without loosing the feeling of flying nice?

Share this post


Link to post
Share on other sites

Yup, can't wait to hear what happens. I asked an autopilot developer about using one with a quad last year, and he felt it would be a non-starter, as most of these autopilots are modelled around something that is intrinsically stable in flight i.e. a plane with dihedral and a low C of G. Unlike quads, which tend not to be.

Still, the FY20 looks ambitious: let's wait & see...

Share this post


Link to post
Share on other sites
I´m not very familiar with the gaui, does it comes with the FY-20A? or are you doing this mod?

The FY-20A stabilizer is sold by a different company and can be used on a variety of R/C model airplanes. There have been recent reports of good results on Quads, so I thought I would give it a try.

could you please tell us what are the key features of this quad?

The best source of information is in the mfg's link posted earlier. It has all the features and specs.

I asked an autopilot developer about using one with a quad last year, and he felt it would be a non-starter, as most of these autopilots are modelled around something that is intrinsically stable in flight i.e. a plane with dihedral and a low C of G. Unlike quads, which tend not to be.

Lot's of interesting information here: http://www.rcgroups....ght=fy+20a+330x

Share this post


Link to post
Share on other sites

I've been having a good time getting my sea legs with the 330X-S, so I haven't tried out the FY-20A yet. So far I am delighted with the Gaui X330-S. It is Much easier to fly than a traditional model helicopter.

One nice feature is that the mixer can be set up so you can fly the frame in the + orientation (front is indicated by a motor arm) or the x orientation (front is the center between two motors). The x-orientation seems like it would be a good pick for camera installations.

One of the things that I learned is that it is not a good idea to fly until the low battery cutoff occurs. Because the four ESC's LVC are not perfectly matched, one motor will go into the soft cutoff first. This causes the model to enter a death slide without warning. Not a good thing at all. It would be nice if the onboard mixed had a low voltage alarm that blip'd the four motors together as a friendly warning long before the ESC cutoffs kicked in. Or, it would be grand if I could disable the LVC on all four ESC's.

I also learned that you can never get too comfortable with electric models, especially helicopters. I was hovering at about 2-3 meters high when the model suddenly pitched backward and immediately lost all power. It survived the tumble into the ground and surprisingly I couldn't find the reason for the total loss of power. Without an explanation I was a bit nervous, so about all I could do was to set the battery aside and be careful about keeping the model low until I found the problem. It never happened again, so I decided it had to be the 3S/2200mAH Poly Quest battery. I cut it apart and found that the factory did not solder one of the battery tabs. The unsoldered cell tab had some black charred edges too from the high currents. I've repaired it and it is now fine, but things like this really mess with my confidence.

I will try out the FY-20A stabilizer soon, but as I said, I'm having too much fun with the new model. Don't worry, I'll install the stabilizer in a week or so.

BTW, for those of you interested in the FY-20A, I just learned that the mfg (FeiYu) is about ready to release a version (FY90Q) that is specifically for quads. Here is a demo video I found:

http://v.youku.com/v...A0MzgzMzY0.html

Share this post


Link to post
Share on other sites

I've been having a good time getting my sea legs with the 330X-S, so I haven't tried out the FY-20A yet.

Pity - I've fitted my FY20A to my X500D, and have hit an interesting problem - the power pulses (briefly) OFF to all 4 motors every 3 seconds or so, regardless of the throttle setting!

Now, the throttle channel doesn't go through the FY20A, but the X500D's onboard IMU/CPU varies power to all four motors continuously to keep level... BUT what exactly is the FY20 doing to the ail/ele/rudd inputs that is persuading the X500D's brain to cut power to all 4 motors every 3 seconds?!

I've tried hooking it up in a conventional ail/ele/rud 'plane, just to check, and it seems to do what you'd expect. Red & blue LED's don't suggest there is a problem. Flies fine without the FY20A.

So, some sort of strange interaction between the FY20A and the onboard IMU/CPU?

I would be interested in your experience with it on the GAUI - different IMU/CPU/firmware, of course...

Share this post


Link to post
Share on other sites
I've tried hooking it up in a conventional ail/ele/rud 'plane, just to check, and it seems to do what you'd expect.

For a sanity check, how about moving the Rx from the conventional plane to the X500D?

Share this post


Link to post
Share on other sites

Good idea. I've been using a Multiplex Synth IPD 7Ch, the same receiver type I used in the 'plane. Still, it could be that receiver.

So I just tried:

- another Multiplex IPD 7Ch - same effect.

- Multiplex DS IPD 9 Ch Rx - same effect.

- Corona RP8D1 - same effect.

So it's probably not the receiver, but worth a try!

Share this post


Link to post
Share on other sites
I've been using a Multiplex Synth IPD 7Ch, the same receiver type I used in the 'plane.

I was hoping it was a different Rx in the plane. Because some Rx designs have peculiar servo pulse synchronization, using a different model Rx might have offered a remedy. Too bad a different one didn't help out.

Share this post


Link to post
Share on other sites

One of the things that I learned is that it is not a good idea to fly until the low battery cutoff occurs. Because the four ESC's LVC are not perfectly matched, one motor will go into the soft cutoff first. This causes the model to enter a death slide without warning. Not a good thing at all. It would be nice if the onboard mixed had a low voltage alarm that blip'd the four motors together as a friendly warning long before the ESC cutoffs kicked in. Or, it would be grand if I could disable the LVC on all four ESC's.

I also learned that you can never get too comfortable with electric models, especially helicopters. I was hovering at about 2-3 meters high when the model suddenly pitched backward and immediately lost all power. It survived the tumble into the ground and surprisingly I couldn't find the reason for the total loss of power. Without an explanation I was a bit nervous, so about all I could do was to set the battery aside and be careful about keeping the model low until I found the problem. It never happened again, so I decided it had to be the 3S/2200mAH Poly Quest battery. I cut it apart and found that the factory did not solder one of the battery tabs. The unsoldered cell tab had some black charred edges too from the high currents. I've repaired it and it is now fine, but things like this really mess with my confidence.

Yes, confidence is an issue with quads, as unless all the active stuff is working fine, they fall out of the sky. I've had a fall from about 100ft - presumed Rx glitch. I've had the "sliding uncontrollably at height" problem, which isn't completely consistent. I've had a 'partial loss of power on one ESC through overheating' problem, although after much soak testing I think I've solved this (fit little IC-type heatsinks on the ESC chips). Thankfully the X500D board does have a 'universal' LBV alarm, although it cuts in a bit late, as if to explain why the quad is sagging out of the sky i.e. it's set a bit low. But at least it sounds before the ESC LVC's kick in!

Further FY20A tests. I've tried the reset procedure - no difference. I've also tried pinging it up to about 30 ft, so when the power pulses off it doesn't hit the ground before it comes back on again; amusing to watch, and strongly suggesting that apart from the power glitches it is working! But not a sensible way to fly a quad (boing boing boing....).

I can't think of an attitude control input combination that would cut all 4 motors, so I'm wondering if it is something to do with the 3 attitude controls having slightly different timing relative to the throttle control input - since the throttle bypasses FY20A. But if this were a problem, I wouldn't be alone, and these days if I google a bit and no-one else is having the problem, it's invariably not the problem I think it is! Having said that, the X500D isn't that common...

Share this post


Link to post
Share on other sites

I temporarily installed the FY-20A in the Quad and test hovered it today in the back yard. I picked the worst day to try it for the first time because winds were gusting to 25mph+.

At first I didn't think the stabilizer was doing much since the wind was pushing the 330X-S around a bit. But, as soon as I remotely turned off the stabilizer it was immediately obvious it had been helping a lot. Once the storm moves out, I will try again to see how things feel in better weather. I expect it will work as advertised.

I did not experience the throttle pulsing problem reported by Doofer. That's an odd one!

Share this post


Link to post
Share on other sites

(Well, consistent - no-one else using FY20As with the Gaui board are reporting problems!)

Share this post


Link to post
Share on other sites

Doofer - My hunch is that it is a servo sync phasing issue on your X500D's mixer board. Because the other channels are processed through the FY-20A, the throttle channel's timing is probably superimposed over another channel, causing occasional missed throttle pulses.

As a test, remove the props and setup the model on the bench. Move the X500D's throttle channel input to the Rx's highest unused channel available (CH8). If this channel is just a simple switch, then during your testing use your Tx's mix menu to vary the throttle speed. If the problem goes away with this setup, then a final solution would be to use your Tx's CH-CH mix and have a high order Rx channel supply the throttle signal (or build something that delays the native throttle channel's servo pulse).

Share this post


Link to post
Share on other sites

Close!... but no cigar. I've found that using a different Tx (MPX Cockpit) I get a different effect - the power still pulses down, but the pulses are much briefer, and more random, occuring every ~0.5-0.75 seconds or so apart. And still no discernable difference between different control settings. Interestingly, your suggestion - I've mapped the throttle to Channel 9 - changes the Royal Evo 9 performance to be pretty much exactly the same as this. So yes, something about frame length, frame desynchrony, something like that. Irritating.

I'll dig out an old jap 5Ch Tx, and see what that does. Must get it on the bench and look at the FY20A outputs vs Throttle output on my o'scope...

Share this post


Link to post
Share on other sites

With the slight improvement you now see, there is a chance you may have actually fixed it. Huh, you may ask? Read on.

Last night I made the permanent FY-20A installation. Without removing any cables, I unscrewed its blue plastic case and added clear heatshrink protection to it. The supplied gyro tape was used to mount the FY-20A on the rubber isolation tray. With the case gone, the 330X's plastic canopy now fits with room to spare.

post-2-128405965733_thumb.jpg

With gentle breezes outside this morning, I eagerly tried it out: I was in for a real surprise! It flew horribly, with irregular throttle (throttle was very bi-stable, pulsing, and jumpy like a frog) and controlling pitch/bank was difficult. Overall, almost impossible to fly. I reset power (unplugged LiPO) several times trying to see if it was an initialization issue, but that did not help. Even with the FY-20A remotely disabled, it still did not fly right. At this point it seemed my Quad had caught the Doofer disease.

Since I hadn't unplugged any cables, I figured either the reduced mass from the missing plastic case had changed the performance of the isolation/damp tray, or the case-less circuit board had flexed/twisted and changed the gyro/accelerometer bias. On a lark, I decided to perform the gyro reset procedure. Thankfully, that restored the model back to sanity. I didn't have much time to really play with it, but at least I confirmed that the reset trick cured the disease.

Long story short, I suggest you look closely at your isolation tray installation and optimize it, then perform the gyro reset trick. With the CH-8 throttle fix and the gyro reset, you might find success. Fingers are crossed.

EDIT: You will read later on about how the real problem was that the FY-20A adversely affected the ESC's throttle curve and confused the 330X's mixer. The permanent fix is to recalibrate the ESC's (with the FY-20A installed) using the standard procedure discussed in the Gaui instructions.

Edited by Mr.RC-Cam
Added comment about ESC throttle curve

Share this post


Link to post
Share on other sites

Hi Rc-Cam

I am a photographer based in Zambia and fairly new to R/C but excited to see your progress with installing the FY-20A on the 330xs. After flying a cp3 for a few weeks for fun I came across the 330xs and ordered it to see what I could do with it photographically. I finally received it last week (after being nailed 80% customs and duty!!!!)

I finished building it over the weekend and used the aero-conversion kit to provide a camera platform. Its been flying great so far (after some initial difficulty understanding how to calibrate the ESCs). Once I add the fairly substantial weight of the camera and battery beneath the baseplate though, achieving a level hover is rather more difficult, especially in any sort of wind. The other day I was hovering at about 100ft when it suddenly flipped on its back and came down like a stone. I managed to right it (by pure luck) just before it hit the grass and the cf legs took the impact and bounced it upside down. Amazingly no damage at all to the 330 but I had to sit down and have a smoke. I still have no idea what caused it as it seems to be flying fine again now...

Anyway, would love to hear more of your experience with the FY-20A as it seems it would be ideal for what I want to achieve with the 330.

Here's a couple pics of it with the aero-conversion kit mounted with GoProHD cam and cf legs.

cheers

post-6136-128411382875_thumb.jpg

post-6136-128411387642_thumb.jpg

Edited by remoteafrican

Share this post


Link to post
Share on other sites

Regarding backflip mentioned in previous post: Just downloaded footage. I had the GoPro taking pics every 2secs. Made for an "amusing" sequence. See below:

1. Hovering nicely

2. Mid-backflip

3. HARD landing right way up

4. Bounced and lay upside down

Cheers

post-6136-128412062386_thumb.jpg

post-6136-128412062825_thumb.jpg

post-6136-128412063298_thumb.jpg

post-6136-128412063786_thumb.jpg

Share this post


Link to post
Share on other sites

Long story short, I suggest you look closely at your isolation tray installation and optimize it, then perform the gyro reset trick. With the CH-8 throttle fix and the gyro reset, you might find success. Fingers are crossed.

Okay, no, re-setting doesn't help. Looking at the outputs on a 'scope, Rx outputs vs. FY90A outputs are two combs slowly going past each other i.e. the FY90A is clearly imposing it's own frame rate on the control outputs, and isn't synching to the Rx inputs. So if the X500D is expecting inputs in order, yes, the throttle will seem to keep going out of sync.

Hmmm... the question is, whose problem is this? At least the FY20A folk reply to emails, I'll see what they say - Why are they selling something you put in some signal paths & not others that doesn't sync to the original input frame? Seems a simple question, should have a simple solution, if only the FY20A were flashable....

Share this post


Link to post
Share on other sites

Perhaps this is the time I finally use a PIC solution - anyone know of widget that takes 4 PPM inputs and synchs them to a single frame rate?

Such a pain - manufacturers really should mention it somewhere if their inline device doesn't preserve sync...

Share this post


Link to post
Share on other sites
The other day I was hovering at about 100ft when it suddenly flipped on its back and came down like a stone.

That sounds like the battery voltage was low enough to trigger one of the ESC's to shut down due to the Low Voltage Cutoff. With all the weight you are carrying, low voltage cut off will happen much sooner than usual.

Looking at the outputs on a 'scope, Rx outputs vs. FY90A outputs are two combs slowly going past each other i.e. the FY90A is clearly imposing it's own frame rate on the control outputs, and isn't synching to the Rx inputs. So if the X500D is expecting inputs in order, yes, the throttle will seem to keep going out of sync.

Assuming that the three servo channel pulses (ele, ail, or rud) are still sequential out of the FY-20A, my gut feeling is that the only important thing to prevent is the throttle pulse overlapping any of them. That said, which of those channels does throttle walk on? How about posting an o-scope screen shot?

Perhaps this is the time I finally use a PIC solution - anyone know of widget that takes 4 PPM inputs and synchs them to a single frame rate?

I would think that all you have to do is ensure that the throttle channel does not overlap Ele, Ail, or Rud. So the big questions is how many mS delay is needed to move the throttle past those? Worst case is that the throttle slowly crawls over them all, which will make a solution a bit more tricky.

Such a pain - manufacturers really should mention it somewhere if their inline device doesn't preserve sync...

Given all the problems in the past about servo signal phasing issues on R/C products, one could say that the mixer on your Quad should have been designed to handle this scenario. If it is flash upgradeable, try asking the developer to help out.

Share this post


Link to post
Share on other sites

Yup, it's a slightly different frame length, so straight-through channels slowly slide over all the channels output by the FY20A. I tried capturing the screen output, but that didn't work, so I went looking for a camcorder then got distracted, made some tea etc. Just imagine two combs slowly moving past each other, and that's what it looks like.

I've never had a reply from the chaps who make the X500-D (over their overheating ESCs), so I don't have much hope that they'll update the firmware...

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×