DC blocking RF capacitors

Thread Starter


Joined Feb 18, 2012
I am modifying a DX Engineering RPA-1 preamp to put dc power on the antenna 'input' coax for an upstream active antenna.

I am having difficulty finding a good RF coupling/ DC blocking capacitor for my modification.

For a starting point I examined the components of the DC power injector that came with the preamp. The capacitor in it is a generic .1uF ceramic disc.

By calculation the Z of the cap is 5 ohm at 300KHz. I tested it on a HP LCR meter up to 1MHz and was shocked to see how how poorly it responded. Its Q wasnt too bad at 300KHz but it went down to Q=3 at 1MHz. That tells me it is very lossy. Therefore I want to reduce the losses with a higher value and/or different type cap.

I did some research on RF caps and found MLCC COG, and X7R materials to be recommended. Unfortunately COG caps are .1uF max. I purchased some 1uF X7R caps to test and was disappointed with the results...not much better than the original disc cap. I had some .1uF COG caps on hand which I tested with much better results @ 1MHz the Z= .1 ohm and the Q>20. Paralleling 10 caps for 1uF is a real pain.

Can anyone recommend some other type of cap that isnt lossy from 300KHz to 35MHz?


Joined Oct 26, 2011
It is my guess that the coupling capacitor performance is not an issue because the preamp input impedance may be 50Ω (lossy). The feed inductor is the critical element and it must work across the entire range.

Thread Starter


Joined Feb 18, 2012
Zi and Zo for the preamp is 75 ohms.

You are correct that there is an inductor (wire wound on rod) in the injector ckt but it is not resonant, or tuned. Its purpose is to block the RF from entering the DC source. Its impedance is in the 100's of Ω and is purely inductive up to 1MHz. My guess is that its made with a 61 material core.

Circuit representation:

RF -------] [---I-----To active antenna

I found a better, more lossy core material for the choke...Laird( Steward ) LF - a high perm material tailored for filtering apps from 300KHz to 30MHz. 17 turns onto a LFB095051 core gives me 1000's
Ω impedance with a Q<<1.

I still need to optimize the RF coupling