A quick look at the prototype HD FPV system design provided by Fatshark after DJI announced their new HD FPV video system.
Full disclosure I have not had a look at the real device and this is just info from what I can glean from the facebook post and other info on the web.
Here is the recent facebook post from Fat Shark.
It looks like a 30mm X 30mm stackable board for the drone transmitter but no word yet on the goggle receiver, but it sounds like it will plug into any of their goggles that have an HDMI in port.
I am more interested in the technology they are using and want to share what I think they are using to achieve Digital HD FPV transmission with low latency. It is very similar to the HDZERO system on the HDZERO web site here https://www.hd-zero.com/ and this is further backed up by the Partners section at the bottom, where Fat Shark and Analog devices are listed.
From the Fatshark facebook post, I can see that they are using 4 main ICs on the top side of the board. I labeled them on the image.
From the image, the IC that makes this possible is the Divimath DM5680 in Red is a wireless video transceiver that claims to operate with near-zero latency. It transmits uncompressed video up to 1080P30 and boasts smart de-noise which will provide analog-like degradation in low SNR scenarios. The video chip is supported by the Analog Devices AD9364 in Blue which is RF transceiver with integrated 12-bit DACs and ADCs, Tuning range of 70MHz to 6GHz and can support bandwidths up to 56MHz. This AD chip is an SDR quality transceiver which is similar to the one in the ADLM-Pluto SDR.
The chip labeled in green is likely a filter or amplifier to achieve better performance over the FPV bands that are used. The chip labeled in yellow is likely some camera interface controller that will provide the correct video protocol to the DM5680.
It is interesting that Fat Shark has gone with one antenna instead of a MIMO solution that is depicted in the DM5680 Brochure, this is probably to keep the device cheaper but could limit the resilience to multipath. This means that flying in urban environments could be more challenging.
I mocked up a similar block diagram to the one in the brochure from Divimath linked above. This time I substituted in the AD chip that does not use MIMO.
My Thoughts
I really like that Fat Shark is coming out with their own solution to Digital HD FPV video. DJI has tons of resources to solve the problem and their solution looks amazing from what I have seen, and the integration with open-source flight controllers is unprecedented for the company known for its closed, all in one designs, that locks users out. With that being said they are not as embedded in the community as Fat Shark and could easily leave the market and early adopters stranded.
I believe that the solution that will dominate the market for Digital HD video will be from a company that opens its design to interoperability with other vendors and can be integrated with our current rigs. This will also drive down the cost and make switching to Digital HD more viable than using Analog video.
The proprietary nature of all the Digital HD systems so far makes it unlikely that it will capture the market, as FPV pilots value the community aspect of flying together and sharing the experience with others. This means that if the HD technology cannot be used alongside analog and forces pilots to only use a particular system or systems cannot be used together it will not be adopted.
The introduction of digital control transmitters made it easier to fly, reducing the cognitive overhead of managing pilots frequencies, we need this for FPV. HD looks great but not having to think about frequency management and just getting in the air is something that would improve everyone’s experience.
