Intermittent Death Roll

[analog siren]

I am trying to figure out this issue I am having with my very first 5" 6s Drone. I am getting an Intermittent Death Roll here and there that I cannot replicate repeatedly. The first 3 packs went fine at the park and then the next day I tried flying it at work and it started going haywire 2 separate times, luckily I was able to catch it both times but it freaked me out so I didn't fly it until I got to the park again today. It then did the same thing on the first pack and I wasn't able to catch it, you can see it in one of the videos I posted. I have also provided pictures of the build and solder joints.

I have tried the following:
Motor screws aren't touching windings
Everything is tight/torqued
DShot is enabled (600)
Calibrated accelerometer/barometer etc.
Checked over every inch of ESC/FC for solder balls that may have happened when soldering but I made sure to cover everything extremely well so I knew it wasn't gonna be that, checked anyways and nothing.

[long fable]

Looks like a desync

https://youtu.be/jWdkl0EHkyU

[dreamy lava]

Normally there should be a capacitor across the ESC at 6s

[analog siren]

There is, it's just ran under the ESC with some wire extension

[dreamy lava]

Take a photo from the side or something. Wire extension = cap work like 💩

[analog siren]

It's the black thing zip tied to the standoff to the left on this picture

[steel vale]

Yep, so long wires in this case. As stated, cap works like 💩

[analog siren]

How else am I supposed to run it then there isn't any space between the ESC and VTX

This is what was recommended to me by multiple people as well as some guides online

[analog siren]

This is very good information, I appreciate it 😁

[dreamy lava]

cap work like 💩 = electrical noise/drops/spikes, maybe 🔥
Power issues can cause desyncs

You can get smaller capacitors in a different chemistry that are slightly smaller
Unicorn UPL 35V 470mF

for one example

[steel vale]

Get creative. On a leg close to the esc already saves on wire lenght.

[analog siren]

i was just worried about it being exposed and in a crash shorting etc. if the arm broke and ripped cap off or w/e but i think ill just zip tie it to that standoff by positive and use the guards i took off to keep testing cause it blocks my usb c lol

i will try this as well as it won't change much, i ordered some 50v 1000uf caps should be here tmrw

[steel vale]

That is true, maybe more for cruising and less suitable for bando bashing. Just give it a try to see of noise is less and maybe fix your issue.

[analog siren]

How's this?

[steel vale]

Looks great! See if it has a positive effect on a blackbox log. 👌

[analog siren]

@steel vale https://youtu.be/014utmYVv2Q

So far I have changed out the motor that sounded weird, relocated the cap, loaded the headsup racing preset and changed props... the issue didn't happen the other day when i flew my 6 packs after doing this stuff and it happened on the 2nd to last pack today so i flew 11 packs before this issue happened. I was able to "replicate" this issue by doing pitch flips, it seems as though the quad tends to freak out when doing "clean" pitch flips ie. no throttle during flip and throttle when you complete a full rotation so you've basically landed one the same plain you started on

[steel vale]

Try lowering the Master Multiplier slider a bit. Might be the pid tune causing it to freak out.

[polar heart]

You need to look at your esc settings

[grand sapphire]

personally i'd set ESC firmware's demag compensation to high or medium-high, whatever is available.
this is how it was done before RPM filter was a thing.

also: https://hackmd.io/@nerdCopter/BJf9cz2ZT

[analog siren]

I was under the impression that leaving ESC settings stock should be fine and any issues related to desync would be caused by hardware & ESC settings were a way to get around that issue without figuring out the issue with hardware...

[analog siren]

I did give up on that theory and changed rampup power from 50% to 30% and changed demag compensation to high but I still need to test fly it after I fix my VTX antenna issue.

Is there anything the black box is telling anyone? I'd sure like to know what exactly is happening and it's impossible for me to understand the black box log right now. 🥲

[grand sapphire]

only that is happened AFAIK. think of it as a gear slipping, bit digitally. i tend to set my timing to manual 23 as well. i know it's a lot of info, but those Bardwell video with Ryan Harrell are great.

[polar heart]

Desyncs are a motor esc Issue 99% of the time and literally the only way to tame them is esc settings if at all. Higher esc timing is the number one way to handle them

[analog siren]

Okay so I changed the demag compensation to high & ramp up power to 30% and could do a flip with no obvious desyncs, I did have 2 instances in 6 packs where it kinda seemed to "catch" itself randomly but it didn't affect me much. I did try changing ramp up power to 40% and leaving demag where it is today because I felt like flips were slower but it did have a desync on the first flip.

Just had a lens on my 2 months old HDZero goggles go bad I think though so I only got to fly 1 pack, going to try lowering back down to 30% and just leave it for awhile

[grand sapphire]

try a static motor timing as SugarK alluded. 23 is fair for freestyle, higher for racing at maybe minor cost of fly-time -- check the videos on the subject for more informative approach.

[polar heart]

Ohh yeah don’t use auto timing it’s not good

[analog siren]

I believe my motor timing is at 16

So is it okay to use those 470uF caps for 6s?

[steel vale]

uF is not the only thing, check the voltage on the capacitor.

At least 35v capable on 6S.

And 16 on motor timing is low. Go to 23 as already suggested to you.

[analog siren]

@steel vale I'm referring to this one recommended by hyp0

Okay I will try 23