We had a customer express some concern that the PWM decoder frequency was faster and may burn out an analog servo.
We placed the X8R and the decoder under an oscilloscope for a comparison. The results for Ch 1 were conclusive, no material difference. Myth: Busted
There is a very slight difference in the timing with the PWM decoder a smidge slower. This make sense as it is running on a different chip with its own clock and timing triggers.
Please excuse the poor calibration of my probes
For more info on the PWM decoder, see link here and here.
Thought it would be useful to compare the impact of several variables will have on the stiffness of a mini frame upper plate and standoffs. We conducted the test as when building our quad we notice a material improvement in stiffness when we dropped the standoffs from 35mm to 31mm in the BOLT250 and using longer 12mm bolts compared to 6mm bolts.
The results from the simulations were convincing and consistent with our initial observations. Our standoff structure is around 75% stiffer than common structures used in frames like a ZMR.
It would also appear that aluminum or titanium will make much lighter and stiffer frames at around the 26mm stand off height.
The 6mm SS was our base case with stiffness measured against this as a percentage.
Quiet clearly, if you want to make a stiffer, arguably stronger quad, then reducing height has a considerable impact. Stainless steel is the stiffest but comes at a weight disadvantage. Aluminum is the lightest but not as stiff. Ti sits in the middle being both stiff and light.
What was tested
Our analysis comprised simple stress test simulations in Autodesk. The design model comprised two plates separated by two hex aluminum standoffs. We fixed one plate and applied 20 newtons to the front nose of the top plate. The idea was to simulate a head on crash with a static object at 100km with an overall weight of 600grams. We then measured the displacement of the top of the standoff. Less movement = stiffer frame.
We looked at standoff lengths of 35, 31 and 26 mm. We also considered screw lengths of 6mm and 12mm comprising stainless steel (G308), alumnium and titanium.
FYI, the image below details how our BOLT250 (right) standoffs are mounted compared to the ZMR (left). In addition, in order to stiffen the two bottom plates in the BOLT250, we bolt the two bottom plates together, with a 3mm washer, into the standoff. The ZMR uses 6mm bolts whereas we use 12mm bolts. Also, we used 31mm standoffs as standard
A thank you to the many friends and customers that contributed to this project, the first drone from boltrc.com. Robbie, Pav and many others, you know who you are. Your advice, encouragement and in some cases genuine hard labour have made this possible. Whilst the design to prototype was amazingly quick, there have been countless hours and discussions that have led to this design.
Most importantly. a very special thank you to murzwern. Many of his ideas are reflected in the frame and it is his build that has been detailed in the blog! He has spent countless hours pouring over our designs, cutting and then building our mutual fantasy. Thank you!
This is the final completed quad. It came in at 386 grams with all running gear but without battery. This was a health 65 grams lighter than a similar ZMR build. More pictures and vids to follow.
In this post we highlight the battery lead wiring. The battery has been soldered to the top of the PDB and then drops down between the plates through the hex cavity. Ultimately it exits from the rear of the quad.
The RX bracket just squeezes nicely between the two standoff posts.
In this post we detail how our VTX is attached. We have used a small 200mw VTX attached to a pigtail with 90 degree SMA bulkhead. The top plate has a mount hole at either end for this purpose. The bulk head allows your antenna to be firmly attached whilst protecting VTX from being damaged in a crash. The 90 degrees allows the VTX to be mounted under the top plate without bending the coax. The VTX has been affixed with double sided tape.
Overall, not the lightest solution but it is no fun smashing your antenna connection on the your VTX.
The antenna are good quality Chinese cheapies. Based our testing they perform around 95% as well as the Blue Beam but only cost around 20% of the price.
Important : It is not recommended to use metal bolts to hold down the N32 FC without proper insulation. (eg. a plastic washer). If this was G10, would likely be fine but as this is carbon, it may cause a short. The bolts contained in the picture were just used during the build and were later swapped out for plastic bolts before any power was connected.
In this post we continue to wire up the quad. The image below highlights the installation of the Naze 32 and connecting the camera.
The camera lead has been fed from the PDB, through between the bottom plates, and pops up just behind the camera. Alternatively, the last hex hole could be used for this purpose. As detailed in an earlier post, the lead is custom cut to length and hard soldered to the PDB side. The end result is a tidy fit with no extra cable loops.
For the ESC signal leads, these have also been custom cut and terminated with a crimping tool leaving no untidy wire loops. Three of the ESCs are signal wires only whereas one lead powers the Naze 32. You will also note that the channel pins on the N32 are soldered upside down which saves around 10mm in the FC stack.
All up, a very tidy fit!
The image below present the ESC and camera wiring to the PDB.
A key feature to observe is how the ESC wiring feeds nicely along the arms and then through the dedicated slots into the mid section between the plates where they ultimately solder directly to the PDB without “popping up” again.
The various signal wires are also all channeled to the section between the bottom plates and pop up through the large hex where they can then plug into the Naze32 flight controller. All of the servo lead wires have been custom cut to length and terminated with a crimping tool.
The small servo pigtail has been cut to length for the camera power and signal. This has been direct soldered to the PDB (see hot glue). If required, the camera can be disconnected on the camera plug side. The VTX is to be mounted under the top plate hence this has been left as a 3 pin connector on the pdb where it can be unplugged if the top plate is removed.
If you don’t have access to a PDB, this large cut out section can serve as a major junction for your wiring loom. One of the interesting things about this image is you can clearly see how the arms sweep along the edge of the chassis thereby spreading the load in a crash.
This image demonstrates how we chose to mount the camera. It is one of the new generation mini cams, a Surveilzone 1177.
We love this 1177 camera, besides weighing a feather, it utilizes the famous Sony Super HAD II 600TVL sensor and has a voltage range of 5-22v. This sensor’s low latency, backlight compensation, wide dynamic range and low light performance is legendry.
We are using a PDB with camera port, with the PDB mounted below the upper bottom plate. On this basis we have chosen to mount the camera to the bottom plate using the stand. On this basis, we can run the cable neatly between the plates.
The stand has been cut-down to reduce the profile as low as possible. This stand also provides flex to angle the camera as we wish. By mounting the camera on the bottom plate, through the cavity in the mid plate, we can reduce the profile further.
The above images will are of the current build with 31 mm stand offs. The below image is an analysis to confirm that 26 mm standoffs are feasible but that is another mini project down the line a few weeks time.
In this set of images, we demonstrate the relative size of the ‘bumpers’ and one approach to mount the ESCs on the arms.
The motor prop combo on this build are the Cobra 2204 kv1960 with HQP6045 props running 3S . The ESCs used are the ZTW 12A ‘Blue’ with sK which is perfectly matched to this combo. The kv1960 is reasonably popular motor but not nearly as common as the kv2300. We chose this route as we were looking to see what longer flight times could be achieved.
In this build the ESCs make are secured with zip ties through the dedicated holes in the arms. As this is not a ‘hot’ build, the shrink tube has been applied to the ESC to insulate from the carbon as there is no need for the FETs to be exposed for cooling. An alternative approach would be to place the shrink tube around the arm to insulate the ESC and leave the ESC naked but secured by the zip ties. As cooling was not an issue, we preferred to secure the large capacitor within the shrink tube.
The servo leads has been reduced to just the signal cable on three ESCs in order to simply the build even further.