A ZMR250 build thread. Like a blackout but on a budget.

[Craftydan]

BLHeli and SimonK fundamentally are just ROMs tuned to work best with multirotors. they shoudl work with ANY flight contorl board . . . for that matter you could toss 'em on a fixed wing.

The feature unique in BLHeli and KISS that would push you to specific contorl boards is "Oneshot" which is a new way for the board to talk to the ESCs. By default, this is disabled in the KISS and BLHeli ESCs, and the "how to turn it on" is detailed enough you'll understand the technology and what board supports what

 

[SnowRocker88]

BLHeli and SimonK fundamentally are just ROMs tuned to work best with multirotors. they shoudl work with ANY flight contorl board . . . for that matter you could toss 'em on a fixed wing.

The feature unique in BLHeli and KISS that would push you to specific contorl boards is "Oneshot" which is a new way for the board to talk to the ESCs. By default, this is disabled in the KISS and BLHeli ESCs, and the "how to turn it on" is detailed enough you'll understand the technology and what board supports what

Okay, so where do I find the 'how-to' on One-Shot?

 

[Snarls]

Does anyone know what GPS would work with the Tau Labs Sparky board. I'm thinking about moving my Flip 1.5 to a mini quad and put the Sparky on my camera platform.

 

[Balu]

I'd not buy a Sparky from RTFQs for now (which has a new URL btw: http://readytoflyquads.com/) until the issues with the baro sensors are cleared.

I always defended Paul for being overwhelmed with work and being a one man show, but now I'm waiting about three weeks for an answer about my problems with two Sparkys. To my first mail he responded after a few days, but after that - nada. And we had a few other people here and on the TauLabs forum who reported the same problems with the baro sensors. :-(

 

[C0d3M0nk3y]

I thought the best GPS FC was the ArduPilot? Or is it but the price makes it less attractive?

It might be challenging to fit an APM on a mini quad.

 

[Tritium]

It might be challenging to fit an APM on a mini quad.

NOT a MINI APM Clone that several vendors sell!

Thurmond

 

[SnowRocker88]

Anyone have more info on "One-Shot"? I can't seem to find much info on it.

 

[Craftydan]

To get up and running, I wouldn't worry about one-shot -- it will allow you to tune your airframe that much tighter, but until you get it tuned simply to fly well, you won't notice the difference for the added requirments.

Otherwise, it's a VERY new feature that's *just* comming into it's own . . . enough so some developers are *still* arguing it doesn't make things any better. Thorough documentation of it is still very sketchy.

Fundamentally, though, instead of sending a constant stream of pulses to the ESC for the "current throttle setting", a one-shot ESC recieves a commanded thottle setting and holds it. The FCB only has to update an ESC as the number needs to change, allowing the board the ability to send changes to the ESCs ever so slightly sooner.

Cute feature overall, but if your running BLHeli on a sloppy ESC, the ESC's inability to respond quickly can overcome any speed advantage in faster Comms, and make the effect worthless . . . while a fast cheap ESC running BLHeli could easily see the difference. KISS ESCs, being consistantly fast hardware, however, can gaurentee the performance . . . for the slightly elevated cost and reduced availability.

Not many of the flight control boards support it, either -- if the FCB is too sloppy it can't take good advantage of it -- but it seems to be a hot add-on item for the nicer boards that are still being improved.

 

[C0d3M0nk3y]

Anyone have more info on "One-Shot"? I can't seem to find much info on it.

https://youtube.com/watch?v=7oX5kX1eBWw

 

[SnowRocker88]

To get up and running, I wouldn't worry about one-shot -- it will allow you to tune your airframe that much tighter, but until you get it tuned simply to fly well, you won't notice the difference for the added requirments.

Otherwise, it's a VERY new feature that's *just* comming into it's own . . . enough so some developers are *still* arguing it doesn't make things any better. Thorough documentation of it is still very sketchy.

Fundamentally, though, instead of sending a constant stream of pulses to the ESC for the "current throttle setting", a one-shot ESC recieves a commanded thottle setting and holds it. The FCB only has to update an ESC as the number needs to change, allowing the board the ability to send changes to the ESCs ever so slightly sooner.

Cute feature overall, but if your running BLHeli on a sloppy ESC, the ESC's inability to respond quickly can overcome any speed advantage in faster Comms, and make the effect worthless . . . while a fast cheap ESC running BLHeli could easily see the difference. KISS ESCs, being consistantly fast hardware, however, can gaurentee the performance . . . for the slightly elevated cost and reduced availability.

Not many of the flight control boards support it, either -- if the FCB is too sloppy it can't take good advantage of it -- but it seems to be a hot add-on item for the nicer boards that are still being improved.

Great explanation. Thanks! Yea, seems like it's out of my realm for now. Thanks for the heads up tho. Always happy to learn about interesting things like this!

 

[Balu]

@CraftyDan, you might like Peabody124's take on OneShot: http://buildandcrash.blogspot.de/2015/01/oneshot125-quantitative-testing.html

 

[jhitesma]

@CraftyDan, you might like Peabody124's take on OneShot: http://buildandcrash.blogspot.de/2015/01/oneshot125-quantitative-testing.html

I was glad to see Peabody finally give some good documented proof of the difference Oneshot can make.

But Dan still has a lot of good points:

Oneshot is not a huge improvement by itself. It needs tuning changes and the ESC makes a big difference, until recently KISS were the only ESC's that could do it and they're top notch...but now that BLheli has oneshot just about any ESC can use oneshot - but most of them have slower FET's and many of them can't do damped/regen/comp_pwm so the performance difference is going to be minimal. And the BLHeli version of oneshot still has some issues/quirks that may not be able to be resolved without workarounds (like chirping at low throttle values and slow ramp up from stopped if you like to stop your props in flight.)

Oneshot helps but damped light / regen / comp_pwm makes a much bigger difference in performance that doesn't require retuning to experience. Quite a few ESC's can do it but none of them do it quite as well as KISS.

Oneshot requires a FC and firmware that can handle it. MW can do it - but only on 32u4 based boards and only with a special modified version of the FW. Naze32 can do it but only with Cleanflight. Tau can do it but only if you build your own firmware as it was added too late to make the latest release (it should be in the next release but the last release just hit pre-release so it will be a bit until the next release which will hopefully have oneshot in it.)

There's also a second aspect to oneshot that CraftyDan left out. Yes it does use a shorter signal so it can send changes to the ESC quicker. But in addition it changes when/how the FC sends updates to the ESC by syncing the updates with the PID loop tighter.

Some FC's also use a gyro filter that could impede performance of oneshot - APM for example apparently has their gyro filtering cranked up higher than MW/Naze recommend ever going. And Naze by default uses a higher filter than most MW versions do by default. This is rather important as one of the reasons some people say oneshot can't make a difference is that with oneshot you're getting to the point where the FC can be updating the ESC's faster than the prop can react and faster than the gyro can sense changes. It's something that on paper sounds great at first, but if you look at all the variables you'll convince yourself it won't actually make a difference...but then if you actually qualitatively test it like Peabody did you realize there's more going on than we sometimes think we understand in these systems

So while I'm a big fan of Oneshot I also wouldn't say it by any means necessary or even advisable unless someone really knows what they're doing at this time - which I think is the point Dan was making as well

 

[Craftydan]

So while I'm a big fan of Oneshot I also wouldn't say it by any means necessary or even advisable unless someone really knows what they're doing at this time - which I think is the point Dan was making as well

Indeed. OneShot is a not sharpening a cutting edge, it's honing. Honing a dull knife only polishes a blunt edge but makes it no sharper. Honing a sharp knife makes it unbelievably sharper.

Plenty of other things are needed even to get to sharp, learning to build, fly and tune are the biggest.

 

[SnowRocker88]

Update: Ordered the CPPM receiver. Updating first post...

 

[cranialrectosis]

Indeed. OneShot is a not sharpening a cutting edge, it's honing. Honing a dull knife only polishes a blunt edge but makes it no sharper. Honing a sharp knife makes it unbelievably sharper.

Plenty of other things are needed even to get to sharp, learning to build, fly and tune are the biggest.

Well stated, Dan.

 

[SnowRocker88]

Okay, parts came in and they all appear to be full carbon. I'll do a resistance test to make sure later but for now I'm assuming all carbon.

Anyways, here are the pics that everyone comes for. Haha





Full Kit. As shipped and laid out.





Next order (my flight controller) doesn't show up until the 19th or so. Following order date is TBD. That order has my motors, ESC & props. The very last order (also TBD date) is my Rx.

 

[SnowRocker88]

Well, it looks right but I think I got the fiberglass sandwich. Resistance test showed no continuity. I'll mess with it some more to see if the surface is carbon or not but it seems like the middle is fiberglass. But, it's pretty when it's all assembled. I'll try to find a gram scale to see what the chassis weight is.

 

[cranialrectosis]

Very pretty frame. Can't wait to see it in the air!

 

[jhitesma]

Hmm, I hadn't heard of ZMR using the hybrid material...though to be honest I haven't been paying a lot of attention to ZMR and there are quite a few variants.

A lot of times it's hard to get a continuity reading on the surface of carbon because of the resin. Try reading off the "edge corners" those tend to be the most conductive. Or just across the cut face - those are almost always very conductive on CF but non conductive on G10/Hybrid.

You may also want to try scratching the surface just a tiny bit and rubbing your probes back and forth to make sure you have a good connection.

Either way it should be a great frame. Having done quite a few experiments with plywood, lexan, G10, ply/CF hybrid, pure CF and CF/G10 hybrid I really think the focus on CF is misplaced. It's great material no doubt, but for most people G10 or hybrid material would be a more economical price with a performance difference they'd never notice.

 

[SnowRocker88]

Hmm, I hadn't heard of ZMR using the hybrid material...though to be honest I haven't been paying a lot of attention to ZMR and there are quite a few variants.

A lot of times it's hard to get a continuity reading on the surface of carbon because of the resin. Try reading off the "edge corners" those tend to be the most conductive. Or just across the cut face - those are almost always very conductive on CF but non conductive on G10/Hybrid.

You may also want to try scratching the surface just a tiny bit and rubbing your probes back and forth to make sure you have a good connection..

Yea, I tested across the edge (from one side to the other) but I didn't try along the side. I knew the coating on the faces would keep it from making a good contact with the carbon (if there were any). I will prod around some more later. Was too busy last night to really spend any amount of time testing. Wanted to get the frame together (to make sure hardware lengths and counts were adequate) and then spent time with the GF and her parents.

Either way it should be a great frame. Having done quite a few experiments with plywood, lexan, G10, ply/CF hybrid, pure CF and CF/G10 hybrid I really think the focus on CF is misplaced. It's great material no doubt, but for most people G10 or hybrid material would be a more economical price with a performance difference they'd never notice.

I like the idea of a carbon frame. But, as you can tell from my actions, I am more about the price point. Not worried about rigidity really as much as weight. Just don't like extra weight if I can avoid it. Was actually looking for fiberglass replacement arms as they seem to have more 'give' to them in crashes and will absorb more energy rather than just be rigid and snap. So if my frame ends up being fiberglass in the arms I would be pleased with that.

Again, I'll try to get a weight measurement on the frame to see just how heavy it is. That's the real test for me. As long as it isn't a brick then I'll be happy with whatever material it is. That and it looks really good so I'm very pleased with the appearance!

Edit: I'm at work and remembered they have a pretty accurate scale. So, I went and weighed it.



That's 158g. Estimated weight of a full carbon frame is 145g I believe so another 13g isn't terrible. This is fully assembled; Both camera mounts and isolators are installed. I'm happy with it!