Thread Tools
This thread is privately moderated by alicecooper, who may elect to delete unwanted replies.
Mar 05, 2021, 08:03 AM
Registered User
Thread OP
Build Log

NOS 1.3gHz RX Module Upgrade to working..

As I am still waiting for several key parts to arrive through the covid curtailed postal system, something I did receive was a new old stock 1.3gHz Nexwave receiver module for my goggles.
I am not planning to be dependent on this alternative to a ground station in any way, but like the idea of having some freedom to fly my 1.3gHz planes and wings in shorter range situations without the hassle of setting up a ground station, which was my only previous option.
I found the early version of the Fatshark/Nexwave modules on AE at a virtual giveaway price, and after doing some research found there were three versions originally released from around six years ago with all of them enjoying a fairly short period of availability. This may have been due the the V1 and V2 versions having some serious image quality problems, and the V3 having sorted these problems out with a complete design overhaul where it then only spent a very short period on the market before getting pulled from sale... very strange as it's just a receiver.
The versions I found on AE were the V1 and V2-(1.6), and after carefully examining some old Youtube footage put up by dismayed users of these modules I deemed the image quality problem to be a termination issue which I was fairly confident I could overcome.

Upon receiving the module and testing it my suspicions were confirmed, the video output signal level was close to double the regular 1V p-p for composite, and only required level correction (matching) to get things right.
Upon closer examination the video processor chip inside the can of the module was a scrubbed 48 pin quad flat package device with the video output running directly from it's assigned pin, across half a millimetre of PCB then dropping into a via (through board connection) then invisibly into a core layer inside the multi layer PCB before exiting just outside the can at the header pin for plugging into the goggles..
This not only posed a big problem for my repair attempt, but also explained the design fault that caused the module to need redesigning.

The processor chip is outputting double the standard composite video level and this with a very low source impedance. This is standard fare for back termination in video, and in circuit it uses a series resistor matching the display device's input impedance to then reproduce half the doubled signal level into the monitor, the other half being divided into the resistor. It seems counter intuitive to do this, but it serves several purposes. One, there is no chance of damaging the driver chip through a shorted video connection, two, the characteristic losses of the video cable are greatly reduced as the cable's total resistance is now in series with a much higher impedance circuit, effectively halving those losses, and three, the effects of interference getting into the video cables become much less significant as they too will be halved by the series division.
So the reason the back termination resistor was not fitted to this module may have been that the Fatshark goggles seem to have a very low (26 ohm) input impedance from the module port. The standard display input is 75 ohms, and I suspect someone looked at this very low value and expected it to have the same reducing effect as using a series termination resistance. This wasn't the case, as the chip has a very low output impedance, somewhere around 2-3 ohms, and the difference into a 26 ohm load compared with 75 ohms is insignificant.
My challenge now was to somehow add a 26 ohm resistor between the chip pin and the header, but with only 0.5mm of PCB track and the pin nestled in a row of twelve just 6mm wide it wasn't going to be easy.
For this problem I used a technique I created several years ago when breaking signals out of installed LSI chips (don't ask...), I take a single strand of bare copper wire, around 0.16mm diameter, carefully bend it about 90 degrees and glue it to the topside of the chip with a thin smear of CA. I do this with about 1mm overhanging the exact pin I wish to tap, then once set in place I then dob a larger drop of CA across it and add a tiny amount of sodium bicarbonate over the top while wet. This makes a ceramic hard shell bonding instantly and effectively makes the wire a solid part of the chip.
Working under a powerful magnifier scope, I then bend the protruding wire down over the IC pin with a jewellers blade, and apply a tiny amount of flux to the bond point.
Soldering this requires just wetting a micro fine soldering iron tip with solder and applying heat at the overlap. Under the microscope it feels like playing Chess with a shovel. I made my own special purpose needle tip for these jobs, taking a regular fine point and grinding it down with a dremel flap wheel rotating in the lathe until it was needle thin, I then polished and coated it with nickel using hydrochloric acid and a low voltage supply. It comes in very handy for SMD work where hot air is not feasible. I run it at 500 degree's
Once the joint was soldered and tested, the next job was to break the original connection to the header and as this was only possible at the via, my best option was to just drill down into the via opening only enough to break the junction. There was a component less than half a millimetre from this via, so I found my number drills and bored it with a 0.4mm hole to just below the surface, I tested it and the connection was broken.. so I was almost there...
I then drilled a small hole in the side of the can, being careful position it to avoid the can lid fouling when closed and also any components either side where I was then able to pass a 27 ohm resistor leg through after covering it with fine heatshrink. I terminated this at the header pin and made a careful connection of the other end to my strand of copper wire exiting from the top of the chip.
When soldering the header pins it is always necessary to have them plugged in to a socket, otherwise they will seriously dislodge when heated, for this task I just used a servo connector with the soldered pin in the middle.
I finally tested the modified module and found the image quality was now perfect, no white colour shift and no white text tearing indicating the video levels were in spec. I confirmed it with the scope and ran a direct AV comparison, despite the goggles AGC correcting ant potential variance.

I found this job quite a good challenge but not needing more than just a few hours of careful fine work, and the risk factor was very low as the module cost was only just over twenty dollars.
For anyone contemplating doing this repair themselves, i have included some pictures below showing the technique described above detailing the chip, the pin and the via that needed working on.
The receiver modules are available now on AliExpress, the V1 being even cheaper than the V2, but the V2 apparently had some range improvement mods done back in the day.
It seems that with nearly a thousand in stock from the AE seller I found, they must have recently purchased leftover stock from the original batches, more than likely held from sale due to this inherent problem, but never destroyed.. AE sellers could care less about image quality problems, so they have no scruples about selling them as they are, but at least they are very cheap.
I am hoping to utilise the dual receiver bay option in my next goggle purchase of a set of Skyzone 04x's the only question being the frequency control functions and again, signal level match, although it's now very easy to modify up or down. Time will tell on that one.

I have also just taken delivery of my long awaited Siyi long range 2.4gHz control system, so a few things can get moving there finally.. that will be next on the bench, and hopefully in the field...
Sign up now
to remove ads between posts
Apr 13, 2021, 04:43 PM
Registered User
Thanks Alice for sharing your work. I have a question:

Is this mod of adding 26 ohm resistor valid only while using it with fatshark goggles or it can be used with any display to have perfect 75ohm level?
Apr 13, 2021, 07:34 PM
Registered User
Thread OP
I have tested it with my bench monitor (a high res medical CRT) and the image looks perfect with the 26 ohm resistor fitted, although I didn't try it before the mod I still suspect it would have been worse without it.
As a sidenote, for anyone planning to use this module only as a groundstation receiver, the resistor could more easily be fitted at the breakout connector with the same result.
Apr 14, 2021, 03:11 PM
Registered User
Thanks Alice. Then I will just solder the resistor to the connector as I will using it as a ground station. I am gonna look for the pin outs.

Quick Reply
Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Fatshark FSV2443 1.3ghz rx Module v3 Discontinued? Dr. |2udie FPV Equipment 8 Jun 26, 2019 12:51 AM
Discussion 1.2 - 1.3GHz Video Tx/Rx New Technology availability prjoe FPV Equipment 5 Nov 10, 2016 12:28 PM
Discussion 900mhz-1.3ghz Comtech module w/ SAW filter upgrade DIY jaded FPV Equipment 11 Nov 23, 2015 10:51 PM
Discussion Fat Shark 1.3ghz RX module tchanter FPV Equipment 2 Jun 21, 2014 05:32 PM
Discussion Upgrade SAW filter 1.3Ghz RX gibatron FPV Talk 2 Nov 02, 2012 03:59 PM