zyx bites back. again!

[RedOctobyr]

paedbo, please excuse my popping into your thread, but it caught my eye. In part because I have been going through a somewhat-opposite experience.

I recently converted my 6S 450 Pro to a Mini VBar, from my ZYX (running FW 2.5, I think). Due to a recent servo failure, I am temporarily running just (2) DS92A+ on my cyclic, along with a KST DS215MG, and another DS215MG on my tail. All at 6V, I'd run with (3) DS92A+ for hundreds of flights off my Ice Lite 50's BEC.

When the servo failed, it bound up, and I believe dragged down my BEC voltage enough that it caused issues in the air, and I actually got a flashing red light on the ZYX after I landed, only time I've seen that.

During setting up the Mini VBar, I was having issues with brownouts on the bench, just from things like moving the collective up/down as fast as I could for 15 seconds or so. I'd get blinking satellites, and 2.5V or 3.3V warnings in the VBar event log. Sometimes it would have trouble even powering up properly, browning out when the servos all jumped at bootup.

I bought a Gryphon Polaris BEC (rated to 5A continuous, 10A peak, on 6S, vs the Ice's 3A continuous, 5A peak), set to 6V, which did much better. But I could still trigger brownouts on the bench in certain conditions.

But here's the interesting thing: I did some testing with my external RX & satellite being powered by a few different BECs. The RX fed the ZYX, which had all the servos attached. I also made a double-ended servo cable, which took power from the RX, to the VBar. But the VBar was simply acting as a voltage monitor, no loads were actually attached to it. Try as I might, the RX and ZYX were very resilient during testing (no weird behavior, no blinking sat on the RX), but I could get the VBar to trigger voltage warnings, and could brownout the satellite connected to it. Really strange, the ZYX wasn't bothered by this at all, despite actually having servo current flowing through it.

I eventually wired in a 2200uF capacitor, as a test, which improved pretty much everything. The marginal BECs, which could barely boot up properly before (they'd get blinking sats, and/or voltage warnings), would now boot up OK. I could still cause VBar brownouts with them using continuous, extended servo movements (the capacitor would just delay the inevitable). It made the Polaris work OK in my testing. I went to a 3300uF capacitor before flying it.

It's been great with the Polaris for about 50 flights, though something may have just changed. Now I got brownouts yesterday and today in-flight. Tonight I added a second capacitor, 6800uF, as a test, to see if that helps.

Long story, short summary: I have actually found the ZYX very tolerant of power-related issues, at least compared to an alternative. I don't know what exactly is causing my "issues". I've gone to a pretty-good BEC (it's not a $90 WR, but I have a hard time justifying that for a 450), and can still see some issues.

I described some of my BEC-related testing/findings in this thread:
https://www.helifreak.com/showthread.php?t=551287

[paedbo]

OK 6s is harder on the bec. What ur experiencing isn't quite the opposite of what i displayed here. I knew my zyx wasn't browning out.

Ill take a look at ur thread tomorrow.

[Daddy]

Yes, very good troubleshooting...gotta luv persistence!

You don't happen to have a scope handy do you? Would luv to see the gain signal while this is happening to see how much it is varing.

I wonder if the gain was set to 100 if it would reset. And if it's also causing small in-flight varying gains on our/some units we're not yet aware of.

edit. I just powered up my cx unit and set the gain at the 50% threshold to see if I could duplicate this. I don't have a corona so I tested with the futaba and couldn't duplicate it. However, I did notice there's a larger deadband on the gain rate to HH switching than I expected. Going down it has to get to 49.0% to switch to rate and going up it had to get to 51.5 to switch to HH, and very repeatable.

[paedbo]

No scope sorry. Next thing to add to my collection! My gain stayed the same with the cap. 49.5 for rate mode. 52% for hh.

I cant get it to happen with a futaba clone or a micro futaba only the corona is doing it. I think I have a spare 929 somewhere ill try that later.

Update - I forgot i have a dynam 3gs. Tried it with corona 919 just to see if problem replicates, it did. I also have a corona 929 so I tried that. At first I couldnt get it to fail on the dynam then I realized if i held the corona 929 in my left hand and wiggled the 3gs in right hand I was able to get it to fail repeatedly.

[RedOctobyr]

Originally Posted by paedbo

OK 6s is harder on the bec. What ur experiencing isn't quite the opposite of what i displayed here. I knew my zyx wasn't browning out.

Ill take a look at ur thread tomorrow.

I didn't mean to imply you were browning out, exactly. More than the discussion of electrical sensitivity, adding capacitors, etc, caught my eye. Because my ZYX has recently shown itself to be my more-robust FBL when it comes to handling input power fluctuations. And that the same setup (essentially, albeit with the swapping of one servo) which has run for a long time off 6S with my Ice BEC, and ZYX, suddenly was unreliable when running a different FBL.

My apologies, I hadn't ready some of the last posts last night (outlining that the gain channel was being influenced), they'd been posted after I'd loaded the page with your thread.

Again, not trying to derail, or ask you to go off and read through some other long thread. It just seemed to me that maybe I'm being affected by some subtlety of my setup, sort of how you are.

[paedbo]

No problem. I'm not finicky on thread content.

[paedbo]

OK another update. Grounding tail boom fix may be connected to this issue. I pulled the corona 919 off the tail boom. I tested it again on bench I was not getting glitches however if I held it like I did the 929 in my hand it would glitch. Also if u put the servo near a decent size of metal it would do it. I then ran the corona on the heli again. Because of it's proximity to the tail boom it started glitching. The boom is not grounded at the moment. I have decided not to fix this heli as parts for it are scarse. I may buy a clone cheap kit and move the electronics to it.

[Dr. M]

Originally Posted by paedbo

OK another update. Grounding tail boom fix may be connected to this issue. I pulled the corona 919 off the tail boom. I tested it again on bench I was not getting glitches however if I held it like I did the 929 in my hand it would glitch. Also if u put the servo near a decent size of metal it would do it. I then ran the corona on the heli again. Because of it's proximity to the tail boom it started glitching. The boom is not grounded at the moment. I have decided not to fix this heli as parts for it are scarse. I may buy a clone cheap kit and move the electronics to it.

Hmmm. very puzzling. perhaps your servo has some damaged wiring/contacts inside sending anomalous feedback down to the gyro...

[Daddy]

It is puzzling. I recall him saying if he disconnects the gain channel connector it works correct also. I assume that the ZYX goes static high or low and thus doesn't allow the crosstalk on the gain signal to cross the rate/HH threshold point. Or the gain channel leads are acting as an antenna, although I have a problem believing this.

Now one would think the Rx could hold this line stable...so I was thinking something on the Rx bus. Which he also eliminated when he connected the servo power lines directly to the esc (this may not be actually what he did when wiring them together, I don't know).

But that could just be crazy....

What would you want for that wonky corona paedbo?

[paedbo]

I wired all 4 servos power leads directly to an open slot on the RX. The gain , elev , aileron, and pitch leads all went to the zyx as normal including the + and - from rud and aileron.

I also have a logo 10 3D that has a zyx on it with no issues on glitches . All wired to the twin RX that I am using for testing now ( when I say twin I mean its one of a set of two I got in the same kit)

However the logo has a grounded belt pulley and boom because when the heli was flybarred back in the AM radio days I used to get static glitches on the tail so when I moved over to the tarot everything was grounded already.

Another thing that might be an issue is on this particular heli with the glitching is the main blocks are plastic and the frame is fiberglass. The head is not grounded! With the nature of the beast most of these heli's have at least two aileron servos in very close proximity to the spinning main shaft. Possibly another area where some peoples hits or crashes are coming from. I will try to use the futaba clone servo again but hold it in myhand or put metal next to it to see what happens.

Another thing I tried is I just took the make shift tail box ground wire and spun up the heli and would touch the wire to different areas of the heli. The servos would start jittering when I touched the wire to a ball link on the swash.

daddy - you can get the 919 from hobbyking in usa really cheap. I also got the spare 929 servo to do it too but its not as bad.

[paedbo]

ok so alittle more testing today. I took my heli that has 3 corona cyclic servos and an inolab tail servo .

The system was running an msh brain with no issues. even the self leveling worked and I could take off , fly , and land with self leveling on ( which from what I understand is rare to get it that stable)

swapped out the tarot. carefully ran it up on the ground outside to see what would happen. sure enough blue twitching light.

Ok brought it in took the tarot off and put a 1000pf cap on the gain channel. Interestingly the tarot still blinked the blue light and eventually reset. hmmmm.

[ame]http://youtu.be/NSxaCCOz2xI[/ame]


Added another 1000pf cap across the power bus . No blue blinking light but the tail twitches a little when I give aileron input.

I will run it up again outside to see what happens at which point if it still has odd behavior I will go ahead with my full array of filter caps .4.7uf , .1 uf and of course the .001uf that's already on it.

im going to leave the .001 cap on the gain channel to ground .

Another note is like DR M says the tail rotor produces a lot of static and I could see this in my testing. if I left the tail at center I was not getting any blinking lights I would only get it if I gave it moderate to hard right yaw which made the tail rotor blow more air.

I will check the continuity from the tail to the frame. This heli has an aluminum frame. if there's no continuity I might add a temporary wire to see what happens.

im trying to do this slowly so I don't destroy my trainer heli!


ok another update. I checked tail grounding. there is none.

I added my array of caps across the cyclic/tail power connections on the back of the pin headers and retested. I have not grounded the tail and I had it in the air for at least 2 minutes with no glitches. cyclic gain needs to be raised as it wanders all over the place ( but this is just because it was a maiden flight I always set the gain low and raise it slowly).

[ame]http://youtu.be/1oj38wZSxlQ[/ame]

[paedbo]

several more flights have been performed in my yard. It's very close quarters so I can't do much more than circuits. figure 8's and some piros.

Took a few flights to adjust head gain and tail stops.

No issue's so far and the wind is moderate today.

Applied my fix to my second tarot zyx unit. I swapped out the zys's and the second one flew fine without incident.

If anyone exhibits the same issues wants to try this fix im willing to do a few at cost + return shipping. Probably only $2 for the capacitors ( they're 6 cents a piece but I had to pay shipping) . I didn't buy many but I have enough left over to do a few more tarot zyx's.

This would also be a good test to see if it will fix other people's similar issues.

Now I just wonder if anyone has reset issues when using sats on zyx-s as the signals never go through the pin header?

[RedOctobyr]

Wow, awesome, glad you've made headway.

For those of us (ahem, me) uncertain of which things actually did it: which cap made the difference? Gain->ground, or power->ground?

Power->ground would also help smooth out BEC-output spikes/dips, I think.

From my decidedly-rudimentary understanding of how caps affect this stuff, I'm curious about putting one on the gain signal line.

Since I'm a mechanical person, I tend to make physical/mechanical analogies for myself for these things. My understanding of caps is they almost act like a flyweel for voltage, helping feed back into the system when the input voltage sags briefly, etc, and generally smooth things out.

If that's accurate (probably a stretch), wouldn't the capacitor impact the gain signal? The gain signal, I presume, looks like a typical servo signal, alternating between high and low voltages, at controlled intervals. So wouldn't the cap kind of smooth out the square wave, making it tougher to interpret?

There's a pretty good chance I'm thinking about this wrong But I'm willing to ask anyhow, how else am I going to learn?

[paedbo]

although I left the .001uf caps on the gain channel. The value is way too low to interfere with the signal coming from the receiver. Yes it smooths out high frequency RF. But is probably not needed. This cap is so small of a value I believe its 1/1,000,000 the size of a 1000uf capacitor. It wouldn't be large enough to smooth out the signal from the receiver.

However it was the caps I put across the power bus + and - that stopped the resets.

They are not like adding a 1000uf or higher caps to smooth out voltage for brownouts etc. If you read the whole thread one of the very first things I did was use a large value electrolytic cap on the zyx which did nothing for the resets. They are such a low value that they will just filter out high frequency variations in the power bus. Each value of capacitor I put on the power bus targets a different frequency range. They are called bypass capacitors or decoupling caps you can look it up for more info.

Could it be the bec? well it worked fine with a MSH brain before.

It has been stated corona servos are noisy electrically. So maybe my fix has done nothing for static issues but then I saw the video I posted a link to earlier where he was getting the same flickering blue light as I was. When he would apply a temporary ground wire from the tail to main frame the problem would go away.

[paedbo]

Update- I've been flying this heli everyday for testing. However the humidity level has gone way up.

I still have a dynam unit that I took off my 250 because it kept resetting mid flight. I will apply the caps some time this week and try it on the above 450.

[Dr. M]

I'm glad it's finally working properly, paedbo. Great work! And thanks for sharing.

[Daddy]

Your tenacity is admired paedbo

Would you try 1 more thing? Have you tried a battery as a bec? Probaly won't matter since I think we're all in agreement of emi and crosstalk.

Umm, nwm, canx that...take a break and go enjoy some well desired flying

[paedbo]

daddy b4 I do anything to the dynam 3gs. I will put it on the heli and spin it up without blades to see if I get the blue hits and reset. If it does then I will plug in a battery pack instead of the bec just to find out.

However don't forget that I started this endeavor with a Blackhawk 500 and moved it to a hk450. Both using different esc's and electronics. The only thing they had in common was one was using a corona tail servo and the other had corona cyclic servos.

The only thing I used the same was the 2200mah batteries and the transmitter.

I'm still curious to see if anyone else has tried or had success using corona 929's or 919's with the tarot.

Another thought is if you setup for bank switching instead of yaw gain on the conditions, does it still reset the tarot if you flick the switch 4 times real fast? I haven't tried yet but eventually i'll get there.

my logo with high end electronics has always run good with the tarot. On the other hand the cheap heli's I have never gotten the tail or head to feel locked in until recently.

[Dr. M]

Originally Posted by paedbo

snip...
Another thought is if you setup for bank switching instead of yaw gain on the conditions, does it still reset the tarot if you flick the switch 4 times real fast? I haven't tried yet but eventually i'll get there.

Yes, it still triggers a reset even if you have set up the Gain channel to switch flight banks (aka "conditions").

[paedbo]

I have been switching between the two tarot units and this morning I had a tail servo jitter. I realized after some fiddling that I was getting crosstalk between the wires going to tarot. I had in the process of moving the other unit over accidently running corona servo wires too close to signal/ tail. I seperated the wires and its fine again. No resets tho.

Corona and tarot = bad news! Maybe ill open coronas and apply decoupling caps