CarbonCore Cortex X8 - AQ suitability?

[hogster]

Hi Max,

Yeah, I think that's the new DJI flight controller – the AQ is diminutive in comparison!

In that build thread the guy has gone with his own power train, rather than using the ARTF setup provided by CarbonCore. If he's selling it as a tried and tested system, I'm happy to trust him that it works well, and I imagine I could spend (waste) a lot of time and money experimenting with different props and motors to find a setup that works. However his choice of props does create the issue we're talking about.

I'm thinking of using a single 6S 8000mAh pack, so I'm not sure how I could centrally mount both the battery and flight controller ... what would happen if the AQ is not centred?

Thanks again for your help

David

[Max]

Ah the ARTF, gotcha. Makes sense, and you can always try different (matching) props if you need to. There are a number of people using successful X8 AQ configs here.

As for the battery, yeah that's a problem. That's one reason they often end up under the frame (which is not very convenient). I've never tried mounting the AQ (or any FC) off-center... one person on these forums reported trying it with an Ecilops frame and wasn't happy. That's not much to go on. But AQ code assumes it is in the center of rotation and there is no offset parameter to tell it otherwise.

-Max

[Trnquill]

Very problematic frame for AutoQuad indeed! You effectively loose half of the cool if you use that frame with AutoQuad. DJI has magnetometer and GPS neatly separated from the other components and enables you to mount all the rest wherever you want, inside the frame, next to power distribution board, next to ESCs, all is OK! All you need outside of frame mounted on a stick is the MAG/GPS block. Very convenient.

The "AutoQuad way" of tucking all the sensors in one blob makes things considerably harder. I would maybe mount the power distribution board under the frame (if it's possible?) to get it as far away from AQ as possible. Same goes for ESCs, I would try to find some room under the frame for those guys, too. Looks ugly as heck compared to neatly tucking them inside the frame but whatcha gonna do... If you end up putting them inside the frame push them as far out away from the center as possible. The only possible place for AQ board is on top of the frame, on standoffs. Height of the standoffs is highly dependent on current levels and how far from the AQ board you can get the power train parts.

I had quite similar situation minus the ESCs and suffered from major MAG noise. I had to raise AQ as far as 7cm above the top frame plate. And you can just guess what having the gyros on top of a tower that high does to vibration levels... MAG noise there, RATE noise here, beauty of having all the different sensors in one package!

I believe someone from the AQ team responded to my question about having a separate MAG sensor like all the other flight controllers some while back. I remember the answer being it is impossible to align the MAG sensor perfectly with the frame and all the other sensors if it is separated from the FC. I agree, having MAG sensor soldered on FC board makes alignment trivial. But does is really have to be THAT well aligned with the other sensors to make AQ work? Surely few degrees of misalignment could be compensated during the long and thorough dynamic calibrations?

[hogster]

Many thanks for your thoughts guys

Is there any scope to put a grounded metallic film across the whole upper layer of the frame to reduce any EMI getting to the sensors? Just as the GPS antenna gets mounted above its own big ground plane? Or will that prevent / affect the magnetic field readings too much?

I can imagine the further up in the air you mount the AQ the less issues you'll have with EMI, but the more issues you'll have with vibration

Is there any scope to add an offset between the AQ and the vehicle's centre of rotation? I think the DJI flight controllers have a feature for this and it was the first time I'd heard of it.

I was thinking of ordering the Cortex semi-ARTF, as in with all the parts but not mounted yet to give me a chance to experiment with the component placement. That said, other than moving the ESCs out to the edges of the frame (they shouldn't need that much cooling as they're rated to 40A and from tests I've seen each motor only draws 4-5A!), the power distribution board will need to stay in the middle of the frame, and it's already mounted on the bottom plate, so I don't see how it can be mounted any lower.

Thanks for your feedback guys

David

[Trnquill]

[hogster]

Hi Trnquill, many thanks for your reply I think it'll be one of those suck-it-and-see things ... and your mileage may vary ... and many other similar phrases!

David

[hogster]

[hogster]

Hi all

I'm just assembling the X8 now, and with my planned method for mounting the AQ, all 8 ESCs will be about 7-8 cm from the centre of the AQ board (they're arranged in 4 pairs around the central power distribution board). Unfortunately the design of the frame doesn't allow for any other ESC placement (as mentioned above), so apart from raising the AQ even higher above the centre of the top plate (it's currently on 15mm nylon standoffs), I'm not sure what else I can do ...

Does 7cm sound a bit too close? I guess EMI decreases with the cube of distance, so every extra mm further away I can mount the AQ the better? But I don't want extra vibration problems

Many thanks

David

[JussiH]

With the DIMU and recent development versions, you can worry less about EMI in flight affecting the heading. This is because the mag is only used to compute initial heading in disarmed state, once armed and flying heading is maintained by the gyros and corrected by GPS.

And I know....DIMU is somewhere in transit now and I really hope we can get them available to you all soon. Im not particularly proud it has taken this long!

But even with DIMU there is a few "best practice" you should always follow, since EMI affecting mags is not the only reason why you should try and reduce EMI from your power distribution.

I would think that the X8 with AIMU could be made to fly well in your config if you do the following:

Use a round or star shaped configuration power distribution board with all the power leads eminating from a central point and the + layer on the bottom so that the ground layer is in between the current layer and the FC.

Twist all power leads and make sure they go in a straight line from the PDB to the ESC. No hoops or loops. Braid motor wires too.

Is the ESCs in the center of the frame between the plates with the FC on the top plate? If there is 3-4 cm and a CF plate in between the ESCs/PDB and the FC, I think you should be OK.

Youll see some mag noise in flight but I think the filter should be able to correct and maintain heading just fine.

[hogster]

Hi Jussi,

Many thanks for your in-depth reply.

If you look in this image here:



The ESCs are mounted in pairs underneath each of the oval holes in the chassis plates. The power distribution board is screwed to the bottom of the two chassis plates. I'm afraid the power distribution board and the ESCs have already been completely wired up, so I can't really change that now

I've mounted the AQ on ~15mm nylon standoffs in the centre of the frame.

I will continue the build as-is, and hope the EMI effects aren't too intrusive. I will post a log file here for analysis once I've flown it ... which I hope won't be too far away!

Many thanks again

David

PS. When you say braid the motors wires, do you mean plait them like 3 bundles of hair? Or put metal sleeve around them?