The Acrobrat

Freestyle flying can be the best FPV experience you can have and sharing those experiences is what this culture’s all about! Currently however, sharing that experience usually involves strapping a $300-500 GoPro to your quad (not to mention building a bigger quad to carry the gopro). We realize that it can be a pretty high barrier to entry for some. And let’s face it, when you have that much money in the air, it can really hold you back with your flying. And if you’re flying conservatively, you’re pilot skills won’t improve. With all that in mind, we set out to make shareable HD freestyle footage more accessible, while knocking down some of the barriers to entry for new pilots as well. We believe that every pilot should be having FUN, not holding back, and sharing their adventures with HD goodness!

“HD quality footage in a compact package, all for less than the price of a GoPro Session 5”

The first question we needed to tackle, was which HD camera to use? Sure there are gopro alternatives, but the weight and form factor still meant a big and heavier quad to build. So we turned to the Runcam split mini. We chose this cam for several reasons:

1 – Runcam is a trusted brand in the industry and is currently on their 3rd version of the model.

2 – The Split Mini functions as an FPV cam AND and HD cam without noticeable latency (helping us reduce space and weight, which in turn allows for a more compact and lighter design).

3 – You can switch the analog output to be 16:9 or 4:3. And it’s low-light performance is actually not bad at all. Flying at night in a parking lot with lamps is totally doable.

4 – I believe that this “kill 2 birds with one stone” solution is the way to go in our industry. We started off having a dedicated FPV cam, and dedicated HD cam… because that’s what technology offered at the time. But tech has gotten better and will only continue to get better. And I believe this solution allows for more refined & elegant frame designs.

While gopro makes awesome ACTION cameras, Runcam makes awesome FPV cameras. I say we help our community by using gear within our industry.

If you are familiar with the original Runcam Split, you know that latency was a real concern. Thankfully, the split mini doesn’t have the latency issues of it’s older siblings. It’s lighter, more compact and easier to use. That said, the current split mini isn’t perfect either.

Here are some of the challenges we’ve found:

• The ribbon cable coming out of the cam (and the connector that ribbon is connected to inside of the cam) is very fragile. It didn’t take much for the camera to give us a blank screen (or something else indicating a failure). We found the fix for this to be actually quite simple, remove the back plate of the cam, and fill up the entire back side with hot glue. Put enough glue so that it squirts out of the cable exit. This will provide some insulation for the cameras board and strain relief for the cable in crashes. Since doing this, the split mini has survived every crash to date!
• Some might find the exposed boards and ribbon cable to be too fragile (some would say it should be contained inside of a case). I’d say that this actually helps us with space and weight reduction. Careful placement of the boards and adequate protection alleviates this problem. We have yet to have a split mini failure from crashing. Just make sure the micro SD catch mechanism is tight, as we have seen SD cards eject. And if you’re feeling extra diligent, you can use some light epoxy (or other light glue) to put a protective coating on the split boards (but don’t get any on the micro SD contacts!)
• The FPV feed will freeze for a few seconds when the memory card is full. Unfortunately, this is out of our hands and the only fix is a firmware update from Runcam. We are working with Runcam to remedy this issue, as well as other ideas for improvement for later versions.
• The other issues are much more manageable. Wide dynamic range (WDR) isn’t up to par with standard FPV cams. And transitions from light to dark aren’t the fastest. These are things that can be improved upon with fw updates and future hardware (but totally flyable IMO). The mic, well it’s terrible, period.

Trivial issues aside, the real problem that we needed to solve was “JELLO” or “rolling shutter”. CMOS sensors are prone to “jello” issues in the HD footage. This occurs when too much vibrations (generated my motors/props) get to the camera. This camera, unlike say a gopro, is simply too light and is quite susceptible to translating much of the vibrations from the motors and props resulting in unwatchable HD footage. When researching other Runcam Split Mini setups, it was very common to see “jello” in other people’s footage. We did, however, find a few exceptions with clean HD footage, so we knew it was possible to get it “jello free”. We decided to dig into this a bit… what’s it take to get clean HD footage?

When we looked into those setups, we concluded that:

• a lot of customization was needed to dampen the camera mount
• You need a motor and prop combo that didn’t create as much vibration
• The lighting might be just right during the flight to make the jello less visible
• Or, a combination of the above

That wasn’t good enough for me. I want a repeatable solution. Someone getting into the hobby shouldn’t have to jerry-rig 8 different pieces of foam to wedge up the camera (they should be flying!). And a specific motor and prop combo? Well… what if you don’t like that motor and prop? Or worse yet, what happens when its no longer a fresh motor and prop? You’re going to crash eventually right? Others we consulted with did say that while they found jello-free success with specific motors and props, that it only lasted until their first crash. And if you’re worried about crashing, then we’re back to one of the concerns we had with flying a gopro!

To solve this issue we looked back to the early roots of multirotors. Way back. Before mini quads, before freestyle, before racing, even before DJI Phantoms. Back then, most multirotors used 8”, 9”, or 10” props and their sole purpose was to carry an HD Camera. These quads generated massive amounts of vibration and had repurposed motors (from airplanes), older gyros, slower processors and simple firmware. It was the perfect recipe for JELLO! To mitigate this, the camera section of the frame had to be mechanically dampened through foam, rubber, or silicone bobbins, separating it from the motors/props. These types of frame layouts were known as “clean & dirty” sections. The “dirty” section of the frame was were the motors were attached, and what caused all the “dirty” noise that caused jello. The “clean” section was where the flight controller, battery and HD cam were mounted (the 2 sections were separated by dampening bobbins). This type of isolation usually worked quite well in reducing the amount of jello seen on a rather “noisy” multirotor.

Unfortunately these solutions were quite fragile, and don’t exactly scale down, so we had to get creative. Those old bobbins would break, or tear in a crash (we didn’t fly acro with multirotors back then so it wasn’t an issue) and just the size of those dampening solutions would have made the micro too big and bulky. And to make things more complicated, they were usually tuned to the weight and severity of the vibrations. Back then we tried using 10 bobbins instead of 6, we tried stiffer ones, softer ones, different shapes & sizes… we actually spent quite a good amount of time “tuning” jello out of our gopros (these days we spend all the time tuning PIDs lol). All that to say, that we couldn’t just take those same bobbins and throw them on the micro. Okay, big problem… we knew what we needed to do (dampen vibrations) but didn’t quite have the actual solution just yet.

With some varying ideas about solving the dampening issue, we decided to switch focus to the flight characteristics. It needed to feel and fly like a 5” while still accommodating the runcam split mini. Why you might ask? Well… I want beginners to have a smooth and easy transition if they wanted to build a bigger quad. And most 3in frames as of current have very distinct flight characteristics (short arm base, low CoG designs with bottom mount batteries). It was very obvious to me when I was flying these 3in quads and wasn’t a fan… at all. Likewise for the pilots who are already accustomed to 5/6in quads (like myself), I wanted them to have a quad that they can fly in places where they otherwise wouldn’t with a bigger quad without having to adjust their flying style. I wanted them to have a kwad that they could practice with, without the risks/consequences of crashing a bigger quad.

All that in mind, we were NOT out to build the lightest 3in frame in the world. This is a freestyle frame. In fact, we are firm believers that a bit of weight really plays well for freestyle flying. The additional mass really helps with the “throwability” of a miniquad and helps with stabilization (especially in windy conditions). Right off the bat, we already knew that “center of mass” (CoM) plays a huge role in how a multirotor flies. Taking everything we learned from developing the remix, we knew that we had to bring the “CoM” as close to the prop line as possible. This meant a bottom mounted design was out the window. Let’s be real though, a bottom mount design never stood a chance with us. While there are certainly some merits to a bottom mounted design, it just didn’t fit the bill with what we were trying to do. The battery needs to be at or above the prop line (and not too far from it). We played with a few designs that flush mounted the electronics into the main plate. This allowed the battery to sit inside the prop line (just like the remix). However, the design proved to be too complex and out of budget.

Originally tried to go for a baby remix…. But we couldn’t fit everything we needed underneath, nor would we be able to keep the MFG process from getting too complex.

We also tried a dual deck design… but same issues with complexity plus it seemed like it would be a complete PITA to build and work on.

Another concept we played with. This would allow us to make the deck height adjustable and provide flexibility with whatever FC stack you chose (while bringing the mass as close to the prop line). But the problem with this, was that it now limited your battery size.

Back to the drawing board, we decided to tackle it a different way. See, one thing we learned while R&D’ing this frame… is that where the CoM is on these micros made a much bigger impact to it’s flight characteristics than it’s bigger 5in counterparts. In other words, I could feel the difference when flying a bottom mounted 5in… but I could REALLY feel the difference when flying a bottom mounted 3in. What we concluded was that motor length, relative to the CoM was a factor.

So we opted to make a low profile (low deck) design that has the battery technically top-mounted, but VERY close to the prop line. And then to better balance the mass, we pushed the motors out farther than what you would typically see for a 3in. This provides the leverage the frame needed to really balance out the flight characteristics.

Once we have a general direction, we print a 3d model to look over. As much as I wanted to keep this low profile deck design, doing so meant that you can’t run a FC/ESC stack, limiting you to the all-in-one single board FC/ESC boards, which as of this time meant one option.

So we decided to look at the most common FC/ESC stacks that we liked (HGLRC and DYS and HobbyWing) and played with short standoffs and M2 nuts/washers to see how short of a stack we could build. We concluded that we could get away with a 14mm deck height and that we would include some M2 hardware to help builders create an FC/ESC stack short enough to work with the frame (we later increased this to 16mm in production to prevent the USB connector from being crushed in hard crashes). All good and gravy…. But we still needed to figure out the jello concern.

And then the idea literally rang in front of us. The computer fan started making a lot of noise and it dawned on us that computer fans have vibration grommets used to quiet them down because they generate vibration. BAM! That was it… we needed to use grommets, not bobbins! And even more of a DUH moment when we remembered that FC’s use grommets to dampen vibrations! Our early prototypes used computer grommets for the litmus test. We just needed to see if there was any improvement to know if we were on the right path.

Now things were starting to shape up. The grommets weren’t perfect, it didn’t get rid of jello 100% but the improvement we saw was enough for to march down this path. As much as we wanted to use off the shelf grommets, the shape of what was available wouldn’t allow us to create the side plates as strong as we wanted while still providing the dampening effects.

We tried different sized tabs to try and make off the shelf grommets work.

So we decided to create our own grommets. And since they weren’t being used as “grommets”… we also decided to call them “bushings”, lol. In addition to our own shape and size, we also made them in varying “shore hardness” (the measurement of how stiff/soft a material is). We knew from previous experience with bobbins, that different builds required different bobbins (there is no one bobbin that works for all). So we made as low as 40 (soft) and as high as 70 (stiffer). So that we can find out which one works best.

The white ones are rated at 40, the red at 50, blue at 60, and black at 70.

Things were starting to shape up nicely. With the bushings in hand, and several ratings to try… I was delighted to find that all colors reduced jello drastically. Mind you, I was testing with less-than-perfect rotorgeek 1407s (which are the notchiest/coggiest motors I have every seen). These motors have NO business on a build where clean HD footage is a requirement! And while all worked, I found that my build worked best with the blue bushings. The blue ones just seem to be much more tolerant to chipped/bent props than the other colors.

The custom bushings were working so well that we even tried mounting the FC to the “clean” section. It worked GREAT! This was it! The clean section not only worked for the jello-free split mini cam, but it also worked well to soft mount the FC! Two birds with one stone! But there was a problem with this approach…. Working on the acrobrat was a complete pain the butt! With the FC attached to the battery plate… working on the acrobrat was way too cumbersome if you needed to take it apart. I felt like I needed 3 hands to work on it. So we decided we would integrate FC soft mounts into the main plate. At the time, only the m2 grommets used for 1.5mm thick PCB boards were available. This would have forced us to mill a pocket into the main plate in order for the m2 grommets to work properly. But this meant more MFG cost/complexity, plus we would weaken the main plate. So… we decided that if we were going out of our way to make custom bushings… what more would it be to make custom FC grommets that would work with the 3mm plate? And that’s what we did and it works quite well.

Pocketing the base to use “off the shelf” FC grommets proved to be more costly to manufacture.

A few revisions in, you can see the general look starting to take shape. The camera was too far forward on this revision (causing CoM to be rear baised), so we decided to bring the camera placement back some.</ span>

The last piece to work on once we proved all this innovation, was to make it look good, and more importantly DURABLE. Crash testing was done with what we think will be the common build (bigger 1407 motors and a heavier 850mah 4s battery). We went through a series of specific tests:

• Crashing onto grass
• Full speed tumbles to one corner
• Hard belly landing on grass
• Belly landing on concrete
• Full speed into a tree
• Full speed into a concrete wall
• Disarm from 50ft up

Each time we broke something, we first would ask the question “should it have broke”? Let’s face it, physics rules us all and there are simply some crashes that would break any frame. But we held the acrobrat up to higher standards. And each time we broke something, we addressed it and stepped up the crash testing. Two revisions later, we were happy with the type of crashes the acrobrat could hold up to. But, we were still curious as to what it took to break something. Even when we got to the point where we seized 2 motors, we continued crash testing by throwing the acrobrat (as hard as we could like a baseball pitch) against a concrete wall. We were delighted to see that the only damage was delamination on the carbon. It was still flyable! Now, this is not to say that the acrobrat is indestructible, but we are confident that if you do break something, it will be because you were doing something epic!

A head on with the tree broke the side plate here. Funny thing about it… was that Ryan and I both said “we should probably put a standoff there, it looks like it’ll break there” BEFORE the testing. Turns out we were right. The picture to the right is the damage done to the tree LOL… we made it bleed! Two revisions later, we were able to repeat this test without anything breaking.

With everything shaping up nicely, the last thing to do was get this frame into the hands of testers. We chose testers of all backgrounds and climates (because the climate plays a role in how well the bushings work as they are sensitive to temperature). What we found is that the softer bushings worked better for those in colder climates, while the stiffer bushings worked better for this in hotter climates. The other big question for us, was how well the bushings worked with varying builds. Our testers all had their own favorite motor and prop combos (thus having varying vibration profiles), and this would give us further insight into what color bushings worked best. Blue worked best for me, but I got the feeling that it wouldn’t be blue across the board. After weeks of testing, all testers were able to get great HD footage with little to no jello. BUT, what we also found was that there was no clear winner as to which color bushings worked best. Some swore by the white, others went with the black, some the blue. One person said the white doesn’t work at all, and one said all of them worked equally well. Again, we have to keep in mind that climate plays a role here, as does the condition of your motors/props. The big win IMO though,  was that all were happy to find that they could crash and bend props and still have little to no jello in their HD footage once they found the right color. To me this was huge. There are other builds out there that give jello free split mini footage, but it takes a very clean build that only lasts as long as you don’t crash. The acrobrat now raises that threshold.

We had one more dilemma though. See, I was hoping that the tests would show the blue (or the white) as being the color that works for most/everything. Reason being that I intended to package up the acrobrats with said color so that people don’t have to figure it out for themselves. Instead, the test data told me that I either don’t have a big enough sample size to find an answer. OR, my theory of there is no “one-size-fits-all” is true. I was already well past our budgets and already had to increase the cost of the frame twice. But i WANTED to know the answer to the “which color is best” question. So as an introductory offer, and in the name of collecting data… we decided that we would include ALL of the bushings on the first batch of acrobrats. This way, users can give us feedback on what motor/prop they’re using, the climate of where they’re at, and of course which color bushings worked for them. My goal, is that we will have enough information to decide on which color bushings will become default with the frame kit. As well as help others choose a color to try based on this information. This helps keep the frame cost down, gets information sharing, it’s just an all-around great approach. So if you already have an acrobrat and have found what color works for you, share it with us! Click here! What color are you!?

With the testing phase complete, we took the feedback and made some last minute changes before going into production. Out of 13 testers, only 1 managed to break their main plate. Mind you, I specifically asked everyone to TRY and break their frames. And I screened out all those who applied to be a tester and said “no” to the question of “are you willing to crash/break the prototype frame intentionally?”.  One out of 13, varying degrees of crashes, onto different surfaces, multiple crashes… only one… I’d say we have a winnar! Shawn crushed his USB on his FC on a crash, so we increased the deck height by 2mm. Reggie wanted to be able to squeeze his RX underneath the vtx so we increased the space to the side plates in the rear. One person broke a battery plate, so we beefed that up. I was wearing out the “fins” of the side plate underneath from sliding on the ground and doing wall taps, so I decided to get rid of the “fins” and instead have 3d printed “skids” take that abuse. Then we decided on the hardware (charcoal grey standoffs with stainless steel bolts), and it was off to production! But the development isn’t over! Now we turn to you, the community to give us feedback on the bushings. And as the product gets into more peoples’ hands, as technology continues to evolve… we’ll be here… listening and continuously improving the frame.

The final frame kit is the result of lots of research, thoughtful development, and a large sample pool of testing. I hate to admit it, but I find myself reaching for the acrobrat over my 5in quads now. The control and feeling is as good, if not better, than my 5in quads. Because of its size and not so intimidating noise, I’m able to fly it in places that I wouldn’t otherwise with a 5in. People’s reactions are to stare in awe instead of run out of fear. The fly spots that have been “played out” and “boring” to me has now become something new again, unlocking new levels of flying. Getting ready to fly now takes a fraction of a time (faster to charge smaller batteries than bigger ones). And when I do reach for the 5in, the transition is so seamless and easy. The acrobrat makes so much sense for both the new guy and the seasoned pilot. I’m excited for this one!