How to switch something inside an RFI enclosure remotely?

[Philipp F.]

Hey Forum,

I have a question regarding the design of RF enclosures.
I am somewhat of a newbie when it comes to the whole electronic design and especially RF deal. Some time ago I got myself interested in SDR because I love the huge DIY/hackability of it.
My current project is getting the signal from my broadband scanner antenna amplified right after the antenna to deal with the rather long cable that's carrying the signal from my garage roof to my basement where my station is, I know that is sub-optimal but that's the way it is.
(If anyone is interested, I am using an Eurotick-DX scanner antenna and a RTL-SDR Blog V3 R820T2 RTL2832U SDR scanner dongle).

I opted to go for a rather cheap, Chinese made SPF5189 50-4000MHz LNA module, I also got myself an aluminium enclosure, SMA flanges, battery holders and linear regulation. I am going to remove the can to heatsink all the components to the enclosure to keep them as cool as possible.

My only problem now is to get the damn thing turned on and off from my basement (a full metal switch would require me to climb onto my garage every time I want to fire up the scanner). My idea was to use some sort of optical toslink cable arrangement, I even got myself toslink transmitter and receiver sockets only to find out that the receiver does not work like a simple optocoupler like I imagined, it seems to do signal conditioning and only receives pulses. In my mind, any sort of cable going into the box seems like an antenna that might carry crap into the enclosure and mess with the LNA... Or maybe that isn't all that bad? After all, the antenna is receiving that noise aswell...

What ways are there to do such a thing?

Thanks a bunch!

[ogden]

What ways are there to do such a thing?

Put EMI/RFI filters on any non-RF cables.

[Ian.M]

Use a DC power injector to feed the power up the cable, keeping the batteries (or PSU) and power switch next to the receiver, and another power injector to extract it at the LNA, if it isn't already equipped to be line powered.

Do *NOT* remove the screening can from the LNA - its likely to make it unstable.  Instead to keep it as cool as possible, heatsink the bottom of the board which is a ground plane to your box, using a slug of aluminum to take up most of the space, thermal transfer grease between the aluminum and the case, and a thin but resliliant silicone heatsink insulator between the PCB and the top of the slug.

If you need more switching options you can vary the supply voltage and decode it with a ladder of comparators to get various steady state digital outputs to control RF relays etc.   Because there is no clocked logic, there wont be any interference.

[Philipp F.]

Use a DC power injector to feed the power up the cable, keeping the batteries (or PSU) and power switch next to the receiver, and another power injector to extract it at the LNA, if it isn't already equipped to be line powered.

Do *NOT* remove the screening can from the LNA - its likely to make it unstable.  Instead to keep it as cool as possible, heatsink the bottom of the board which is a ground plane to your box, using a slug of aluminum to take up most of the space, thermal transfer grease between the aluminum and the case, and a thin but resliliant silicone heatsink insulator between the PCB and the top of the slug.

If you need more switching options you can vary the supply voltage and decode it with a ladder of comparators to get various steady state digital outputs to control RF relays etc.   Because there is no clocked logic, there wont be any interference.

That sounds super convenient! I didn't think of that, but now I remember reading that the RTL-SDR dongle might even support power injection by bridging 2 pads. The LNA however doesn't seem to support it, but I found someone on Reddit that seems to have done just that:
https://www.reddit.com/r/RTLSDR/comments/6zw06u/on_board_biast_and_voltage_regulator_for_cheap/

I'm going to attach a photo of the circuit and a drawing of what I think he did (I am a real beginner when it comes to analog electronics sorry... I am mainly a digital and programming guy ) maybe you could verify it? I wasn't sure where the cap with the enamel wire on the right goes...

Also, is there a reason why I shouldn't put one of those blue, squishy, silicon heatsink things inside the can before I solder it back on? It would fit really nicely.

[Ian.M]

Also, is there a reason why I shouldn't put one of those blue, squishy, silicon heatsink things inside the can before I solder it back on? It would fit really nicely.

Yes - *DON'T* do it!   A silpad or any other heatsink pad has a dielectric constant that's much greater than that of free space so will introduce a significant extra capacitance shunting the LNA IC.  Its also likely to be very lossy at UHF upwards, so if you want to kill off sensitivity above the low VHF band and maybe blow out your SDR receiver input if the silpad turns the LNA into a VHF or UHF power oscillator, go ahead and make our day!

Never mess with RF stuff inside a screening can (except attempted repair with *EXACT* replacement parts fitted exactly the same as the original in exactly the same position, with before and after photos to confirm that all wires are in exactly the same positions and the same lengths to a fraction of a mm, and even then its a gamble) unless you have the RF test equipment and knowledge to be able to thoroughly test the device against its original specs before putting it back into service.

His bias T circuit looks sane.  I might do it a little differently - solder a small 1uH SMD choke on end to the output center pin track and then use a 100uH wire ended choke from the top of the 1uH SMD choke off to the circuit he has.   That should improve the wide-band signal rejection of the bias arm of the T.   The 100uH choke should be lossy - a higher series resistance is preferable as long as the voltage drop across it  isn't excessive, compromising the regulator's input voltage margin above its dropout voltage.   I'd also solder the 1nF cap immediately after the wire ended choke to the screening can to keep the mounting hole clear.   The whole assembly *MUST* go in a screened case, or it may pick up local AM radio stations on the choke and cause you massive intermodulation distortion problems!  Some insulating tape over the regulator circuit then some copper foil screening tape on top, soldered down to any bare grounds its edges overlap wouldn't go amiss.

[Philipp F.]

Also, is there a reason why I shouldn't put one of those blue, squishy, silicon heatsink things inside the can before I solder it back on? It would fit really nicely.

Yes - *DON'T* do it!

I'm glad I asked. I thought something like that might be the case, RF is scary

For the circuit, I will build exactly what you said. Are you suggesting to do the copper foil screening in addition to putting it in a housing? I have an aluminium enclosure and SMA flanges.

[Ian.M]

That depends.  If you are near an airport or commercial waterway (radar) or near a broadcast transmitter, or really near a cellphone base station, you'll need every bit of screening you can get.   Also, an aluminum enclosure isn't the best for screening unless it has milled faces, a conductive gasket and a LOT of screws closely spaced along the joint, as aluminum oxide is a fairly good insulator, and if it only has screws at the corners, the lid may not be well enough coupled to the body at frequencies where the lid size is a significant proportion of one wavelength.    I'd try without the copper foil, and add it later if its needed (see below).   N.B. don't cover the chokes with copper foil.

You've got a SDR so have at least basic spectrum analyser capability, even though its uncalibrated.   You'll need a fully screened matched terminator for your coax cable - probably 50 ohm, and a M-M coupler  Start off with a full spectrum scan with the SDR input terminated to get a baseline for the noise floor of the receiver.  Then connect the cable and power injector, with the DC source disconnected and its input terminals shorted out.   Terminate the far end of the cable and do another full spectrum scan.   Investigate any suspicious peaks - they are unwanted pickup that's breaking through the supposedly fully screened coax's shielding.    Now connect the LNA to the cable and terminate its input and do another scan.   It shouldn't be doing anything because its unpowered.    Remove the short on the DC injector and connect its power supply.   Now the LNA is active so the noise floor will go up a bit, but with its input disconnected and terminated it shouldn't be picking up much.  Investigate any suspicious spikes that weren't there with just the cable connected.   Finally hook up the antennae and see what you get.

[vk6zgo]

You could use a pneumatic actuator with a plastic tube to operate your switch!
A 50kW HF transmitter I worked on years ago used pneumatics to switch the inductors in the  PA.

It may be a bit of "overkill" in your application though. 

[Ian.M]

... or a long  length of fishing line with a spring return, run through the old ceramic tubes and round the knobs left over from decomissioned knob & tube wiring. 

More sensibly, a RC servo outside the screened enclosure operating a thin non-conductive pushrod or shaft running through a small bore metal tube with length far greater than its diameter, soldered into the enclosure wall, isn't going to leak significant RF at any wavelength that's large compared to the tube bore.

[Philipp F.]

You've got a SDR so have at least basic spectrum analyser capability, even though its uncalibrated.   You'll need a fully screened matched terminator for your coax cable - probably 50 ohm, and a M-M coupler  Start off with a full spectrum scan with the SDR input terminated to get a baseline for the noise floor of the receiver.  Then connect the cable and power injector, with the DC source disconnected and its input terminals shorted out.   Terminate the far end of the cable and do another full spectrum scan.   Investigate any suspicious peaks - they are unwanted pickup that's breaking through the supposedly fully screened coax's shielding.    Now connect the LNA to the cable and terminate its input and do another scan.   It shouldn't be doing anything because its unpowered.    Remove the short on the DC injector and connect its power supply.   Now the LNA is active so the noise floor will go up a bit, but with its input disconnected and terminated it shouldn't be picking up much.  Investigate any suspicious spikes that weren't there with just the cable connected.   Finally hook up the antennae and see what you get.

That's really good advice, this will be immensely useful for me in the future.

I wish I could do proper spectrum analysis and bode plotting, a DS1054z is all I got. Well, admittedly expensive equipment like that would be a little wasted on me, maybe I will buy a cheap Chinese function gen like the yf6600 that can do sweeps and play around with that 

[EDIT: THIS might be interesting to look into.]

Anyway, I somewhat changed my plan. While researching the topic I found LNA4ALL, which seems to be really popular, nicely designed and supports bias-t. While the board is not that expensive either, the custom CNC enclosure is understandably pricey.

Does anyone have some advice on getting adequate, internationally available enclosures for such a thing that are on the less expensive side?
I searched around on AliExpress where I get most my stuff, but its swamped with the usual aluminium project boxes.

I will probably still get the parts for the bias-t mod on my next Mouser order and compare the two LNAs.
Will post results, but it could take me some time to do all that.