They are very simple as you expected. The design process is straight
forward. It's just a capacitor in series with the coax line, and an
inductor connected to the center line, and on the other end of the
inductor the dc is injected. There is a bypass capacitor to ground at
the point where the dc is injected. You need two of these circuits, one
at each end.
To size the series cap just establish a maximum amount of series
impedance allowable that will keep the SWR low at the lowest frequency
of operation. The power dissipated in this cap won't be the RF current
squared times this impedance, it will be the RF current squared times
the cap's ESR. However the cap must be sized to handle this current.
Typical value may be 0.15 uf for 1 MHz.
The inductor needs have high enough impedance that very little RF flows
through it, maybe 1000 ohms. Then calculate the value for the lowest
frequency of operation. Something like 160 uH for 1 MHz. The tricky
part for the inductor is to avoid resonant and parasitic capacitance
problems. Self resonant point of the inductor actually produces the
highest impedance because it is a parallel resonance, but above this
point the inductor looks like a capacitor and current starts to
increase. Shunt capacitance of the inductor decreases its reactance, and
reactance is the parameter you are interested in. Best way to do this is
measure the inductor at the highest frequency you are using. The
inductor will have to be of significant size to withstand the peak RF
voltage on the line without breakdown. Power loss in the inductor is its
current squared times its resistance. That's its DC resistance plus skin
resistance. You can get the DC resistance with an ohmmeter, and if you
don't have an instrument to measure the skin resistance, you can use
something like K6STI's coil program and calculate it.
Then you need a bypass cap on the DC input line, maybe about 0.1 uf. If
you want to inject low frequency control signals as well as DC, you
would want to change this 0.1 to something higher and add a simple
filter that will pass this low frequency control signal but short out
whatever RF made its way thru the inductor.
Jerry, K4SAV
Jim Lux wrote:
>Does anyone have practical experience with HF bias tees used to send
>DC power up the coax? For instance, MFJ has one that is about
>$50/pair. (MFJ-4116) http://www.mfjenterprises.com/products.php?prodid=MFJ-4116
>
>I'm sure they are just a series capacitor in the RF path and a RF
>choke in the DC path, but what sort of values do they have? and, more
>interestingly, what are the component ratings? The MFJ site says 1
>Amp and 50VDC, which is the DC path (what's the series resistance of
>that choke, though?), but I'm curious about the RF path.
>
>We have some MFJ remote switches at W6VIO which use the same
>technique to send the relay current up the coax, and as I recall,
>there's just a couple of disk caps for the DC block. At 1.5 kW,
>those caps have to handle 6 Amps or so of RF without significant heating.
>
>Other sources than MFJ? (e.g. Tessco?)
>
>What about designs to handle more DC power?
>
>What about interaction with transient suppression devices? What
>order do you put the bias tee and transient suppressor?
>
>Feedline to antenna: transient suppressor: bias tee: rig?
>
>What about at the antenna end?
>
>How could you do this and still have a DC grounded antenna? (the load
>at the antenna end is presumably low Z for DC, and you could bypass
>around it for RF that gets through the choke
>
>
>I'm looking at schemes to run equipment that's a bit more than just a
>bunch of relays at a remote antenna location and sending the the
>power and signalling up the same 1000 ft run of coax.
>
>
>_______________________________________________
>
>
>
>_______________________________________________
>TowerTalk mailing list
>TowerTalk@contesting.com
>http://lists.contesting.com/mailman/listinfo/towertalk
>
>
>
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|