A 210g Micro-Quad that Flies for More Than 21 Minutes

http://www.proficnc.com/all-products/185-pixhawk2-suite.html

The problem with this is that the interface is not UART, it is CAN, and Kakute F7 don’t have a CAN interface. I asked the developers of ProfiCNC to create one with UART interface, but they were not interested… :confused:

i see. makes sense.
again, hypothetically, to cut a bit more weight off you can use omnibus nano with smaller gps and different lidar, but, even as is - nice results. nano still does not have software inverter solution, so it would need some creative methods to get telemetry working that may end up counterproductive, but, for such a small build it is difficult to say what would be of most importance - F7 rich features or nano small size and weight.
what battery is there, and what is model weight without it?

smallest 3" one i have now is the 160g with no battery, but it is a sport 4S build that uses 650mah lipo, an armattan gecko, FPV with nano and gps puck - no lidar and no telemetry. not even close to 20min range. :slight_smile:

Even with a serial interface the VL51L0X is not working outdoor, so your OK with the TFMINI and the Cheerson CX-OF

Yes, Nano was considered when I chose Kakute F7. It is a good and very lightweight option. The total weight of this micro-quad without batteries is 110g, including CX-OF optical flow sensor. I ordered 2 more new 4-in-1 ESC models weighing 4g, so, the weight can be reduced to 104g if some of these new ESCs does not have LVC at 3V. About the batteries… This is the hardest part. I tested 12 models of battery using Mountain Radio CBA IV Pro and the best results were obtained with Sony VTC6 18650 3000mAh. The problem is that this model have a lot of counterfeit in the market, actually, most of them are counterfeit, even at well known stores. The original one is now made by Murata, that have acquired Sony Energy some years ago. I got original ones with a local reseller in Brazil, but this is really hard to get.

cool!, setup is surprisingly close to my octo.

I have the 4500kv motors but with smaller 2534 props. My esc’s are 40a blheli32 so i could get some weight savings there. I don’t get anything like 20 mins flight time tho, more like 5. Not that I have tried really loading it down with a big battery yet, and using li-po rarther than li-ion. Was hoping for sub 250g but about 100g over. Quad is certainly better than octo for efficientcy.

Yes… Actually, I tried with VL51L1X from Pololu (I2C), that have the double range of its brother. It worked with the latest ArduCopter firmware, but was causing some conflict with telemetry pass-through. I think it is a ChibiOS issue. Now that I’m using TFMini, I will keep the VL51L1X to use as obstacle avoidance sensor for indoors operation, because it is extremely lightweight.

I can confirm that I tested both in parallel indoor and outdoor HereFlow Optical Flow Sensor Testing
and the VL51L1X from Pololu signal get ‘‘washed’’ outdoor

Cool assembly this octo! In my case, as the total weight is 210g, I can use 1106 4500kv motors (114g of lifting each) instead of 6000kv (185g of lifting each) and I still keep at least 2:1 propulsion/weight recommended ratio. The endurance can improve if you can use less motors. That’s why my next try will be a micro-tricopter… :slight_smile:.

Hmmm… I flew only indoors when testing VL53L1X (not VL51L1X). It could keep a steady altitude in AltHold mode flying at 2m height. Now I have 6 units of them to find some usage. Probably I will use for obstacle avoidance… Do you think it will work? :thinking:

@iampete, btw, this GPS+Compass from Matek is too slow to get 3D fix. It took more than 2 minutes to get more than 12 sats, while mRo gets +15 sats in less then 30s. Is yours taking all this time too??

can be slow if you have not used it for a while, mostly its been great tho

I think that @olliw42 build a unit and I dont think he had good results.
Yeah… the VL53L1X :wink:

Hmm… I didn’t find a topic from him about VL53L1X. I hope he answer this comment… :+1:

It was in the same thread as above: HereFlow Optical Flow Sensor Testing

1 Like

matek is very light but if it takes long time to lock - it means you get a lot of interference into it. it is not well shielded, and it is an ongoing issue. i have same problem as i use those matek gps pucks on all builds becouse they are the smallest ones. there is no simple solution for shielding - you need to shield both power and signal lines that feed the gps.
with no interference it locks into 8-12 sats in 20 sec time. with interference it usually locks into 4-6 sats and keeps high hdop for quite a long duration.

Great endurance flight video.

I can confirm this based on my own tests ( https://youtu.be/yueXcnFBmpE ) and a great web site for comparing performance of various Li-Ion batteries ( List of tested batteries )

The reason it is the best battery is because with a level of constant drain around 10 amps, the VTC6 from Sony is the one that enables the most mAh to be used versus any other battery brand/model, even the higher capacity ones (like the Efest 3500mAh for ex).

Yes… This website is classic and I bought the same battery analyzer used by them. There is another battery that is good too, but it is a bit heavier. It will make my quad weights about 240g. I will record a video using it. I think it will fly for more than 25 minutes. Maybe a world record?? :joy:

The mRobotics company have a GPS+Compass with the same size/weight as this small from Matek. Check in their website. The weight is around 7g. I use both mRobotics and Matek. The mRobotics one have faster 3D fix and can “see” satellites even when the drone is inside my laboratory…

Just checked it on mrobotics.io web site. It is a copy version of the Matek M8Q (for once americans copy chinese products): same ceramic antenna, same Ublox GPS chip, same size , same weight. The mag chip is from another supplier but this is insignificant.

I also get satellites inside with a Matek M8Q, so this is not a specificity of the mrobotics one.

An sure improvement we can diy would be to solder a bigger ground plane (must be a conductor material suc as copper) to the negative of the antenna and have this ground plane as close as possible to the negative feed plane ot the ceramic antenna. Of course we cannot get it so close as there is at least the thickness of the PCB. To put a ground plane a few centimeters away will not improve GPS reception sensitivity (and increase overall weight. If it is a few centimeters away, it is also not so useful to block EMI anyway).