Smoothing DC Voltage and inject into RF Cable

[yadiw]

Hi,

I'd like to inject 5V into the antenna in order to drive an external amplifier.
Now I have the problem if I simply connect 5V to it the signal quality will be degraded, can anyone give me some hints how I can improve that?

I guess I need some inductor and diode for that?

 

[FvM]

The function is usually named bias tee. A diode won't help for it. https://en.wikipedia.org/wiki/Bias_tee

 

[yadiw]

The function is usually named bias tee. A diode won't help for it. https://en.wikipedia.org/wiki/Bias_tee

thanks for the answer, the keyword helped a lot to understand it a little bit better!
But there are still some questions..

isn't it like if the the RF signal will be split into the inductor and go towards the line?
Wouldn't it be better to place a diode infront of the inductor?


I looked around and couldn't figure out the topic about the self resonance/capacitance.

I read that inductors are usually used below their self resonance frequency, ok.

So since I'll try to inject some DC voltage the frequency will be 0mhz?
But how about the frequencies on the RF cable they usually range between 100-800 mhz. How will all that affect the signal on the line?

I've taken all the reference values I found on the internet and will do some experiments with it, I also found some lt-spice simulations (I'm not familiar with lt-spice yet but it's time to study it it seems .

 

[FvM]

A diode will act as a differential resistance according to forward current (rd = 26 mV/Id for silicon diodes) and significantly load the RF. A diode parallel capacitance causes further mismatch.

Ideally you'll want a infinite inductance for the bias tee, practically ZL should be large compared to transmission line impedance. The unavoidable inductor parallel capacitance will limit the practical inductance range. Bias tees for the VHF/UHF range are often used with cable supplied antenna amplifiers, even higher frequencies for SAT LNB supply and have good performance.

In RF measurements, you want bias tees with almost ideal behaviour, that's more difficult.

 

[biff44]

a "bias network" is actually a diplexer filter design. The lowpass portion contains the DC and low frequencies (like a slow data link), the highpass portion contains the RF signal. Often there is a specific minimum isolation between the two, and a band of frequencies where the return loss has to be excellent. So it is more like designing a non-commensurate set of filters.

DC thru a cable...make sure the connector is good for DC flow in your environment....do not want it to electroplate itself into a short circuit after a year outside!

 

[davenn]

thanks for the answer, the keyword helped a lot to understand it a little bit better!
But there are still some questions..

isn't it like if the the RF signal will be split into the inductor and go towards the line?
Wouldn't it be better to place a diode infront of the inductor?

no, dont use any diodes anywhere in the bias-tee .... as biff44 said earlier they are not required

I looked around and couldn't figure out the topic about the self resonance/capacitance.
I read that inductors are usually used below their self resonance frequency, ok.

So since I'll try to inject some DC voltage the frequency will be 0mhz?

well if its DC its gotta be 0 Hz, kHz, MHz etc .... doesnt it

But how about the frequencies on the RF cable they usually range between 100-800 mhz. How will all that affect the signal on the line?

it wont, thats the whole purpose of the bias tee ... to inject DC at one end of the coax, and recover it at the other end without affecting the RF signal on the cable

You do realise you need a bias tee at each end of the cable ... dont you ??!!

here's a circuit for one that I have build and used on several systems.....

mine is using 12VDC for relay switching in this case... other times it could be used for power or pre-amps and what ever voltage you need
just be aware that you would be wise to inject a couple of volts higher than what is required at the far end. This is to make up for the voltage drop in the cable

cheers
Dave

 

[yadiw]

no, dont use any diodes anywhere in the bias-tee .... as biff44 said earlier they are not required



well if its DC its gotta be 0 Hz, kHz, MHz etc .... doesnt it




it wont, thats the whole purpose of the bias tee ... to inject DC at one end of the coax, and recover it at the other end without affecting the RF signal on the cable

You do realise you need a bias tee at each end of the cable ... dont you ??!!


Dave

I'm aware that 2 are needed, but I won't need to build a second one by myself
I'd like to use an antenna relay switch that's all.

**broken link removed**
I found this one so far, I read quite many articles and discussions about that topic.

Now I'm so curious about it that I'd like to understand more details about it, one thing I don't understand is the self resonance issue all the inductors are far below the entire frequency range, somewhere else I read that inductors are usually not used over their point of self resonance.

Can anyone recommend a good book about this topic?

- - - Updated - - -

I see you are using a 30uH inductor, is this some special kind of inductor or just some generic one? (I'm just asking because I'm confused by the self resonance value).

 

[FvM]

Self resonance is just another way to look at inductor parallel capacitance. If you assume only 0.1 pF (a very optimistic assumption), the self resonance frequency (SRF) of a 30 µH inductor is about 90 MHz. A technical 30 µH inductor's SRF can be expected somewhere in the 10 to 50 MHz range. So it's surely not well suited for a VHF/UHF bias tee. The example circuit in post #6 is rather targetting to SW band, I guess.

Above SRF, the inductor is working as a capacitor. A small capacitance can be still compensated within a transmission line structure, but not to a large extent. At the SRF, the inductor has highest impedance, in so far it's reasonable to extent the requency range up to the SRF.

 

[yadiw]

Self resonance is just another way to look at inductor parallel capacitance. If you assume only 0.1 pF (a very optimistic assumption), the self resonance frequency (SRF) of a 30 µH inductor is about 90 MHz. A technical 30 µH inductor's SRF can be expected somewhere in the 10 to 50 MHz range. So it's surely not well suited for a VHF/UHF bias tee. The example circuit in post #6 is rather targetting to SW band, I guess.

Above SRF, the inductor is working as a capacitor. A small capacitance can be still compensated within a transmission line structure, but not to a large extent. At the SRF, the inductor has highest impedance, in so far it's reasonable to extent the requency range up to the SRF.

Now I'm even getting more confused about that.
I just had a look at mouser and found following inductors:
Fixed Inductors 39nH SRF=1.8GHz 2.4ohms 120mA
eg. TDK MLG0603P39NJ

SRF=1.8GHz whereas in the other design the SRF was usually between 100-200 Mhz.

Wouldn't such an inductor be perfect for it?

My switch only needs 30mA on the wire anyway.

 

[FvM]

39 nH tranlates to ZL of about 25 ohm at 100 MHz, most likely not sufficient for a 100 MHz bias tee.

 

[pstuckey]

A good rule of thumb is to try to have the bias line impedance 10x the characteristic impedance.

So if your system is 50 ohm, try to design the bias line to have at least 500 ohm impedance at your signal frequency.

If your inductor is below SRF, then you can use this equation for impedance:
Z = 2*pi*F*L

If your signal frequency is 100 MHz, then to achieve 500 ohms at 100 MHz, you will need an inductance of ~800 nH. Now check the data sheets on your inductor to see if an 800 nH inductor is below SRF at 100 MHz. Looking at Coilcraft 0603 inductors, they have a 1000 nH inductor with an SRF of 400 MHz. This inductor would work well at least up to it's SRF. It may also work well somewhat above SRF, but that is a more difficult calculation, the simple equation above does not apply.

 

[FvM]

For ciritical applications, e.g. high performance measurement instruments, there are ultra broad band inductors with pyramid-shape, e.g. AVX GL series
that are free of resonaces up to microwave frequencies.