Termination Insensitive Amplifier (TIA)

By | November 5, 2017

I have needed some RF/IF amplifiers for my receiver, so I have decided to use the design by Wes Hayward, W7ZOI, and Bob Kopski, K3NHI, originally published in 1999. This amplifier has two main advantages – it has a wide bandwidth ant it lacks any tuned circuits, so it is easy to use in a multiband receiver. All that would need to be switched is the input bandpass filter. The second advantage is the rigid 50 Ohm output impedance on the input, no matter what is the output load thanks to the additional emitter follower at the output. That simplifies the receiver construction.

I have built an SMD version of this amplifier as a module that can be used as both the input RF preamp and the IF amplifier in the receiver. My version is not bidirectional – if you need that, you will need to tie two of these side-by-side.

Termination Insensitive Amplifier Schematics

Populated module serving as an IF amplifier after the first ADE-1 mixer.

More pictures and video of a working receiver using these is available here.

If someone is interested, here are the design files:

All the passives are 0805 sized and the boards use oversized pads so all components are perfectly hand solderable, without any special tools or magnification.

UPDATE: fixed the Kicad project to be openable with Kicad 4.0.7

 

14 thoughts on “Termination Insensitive Amplifier (TIA)

  1. terry lynch

    jan, i just ordered some of these . boards from osh-park thank you for sharing this very useful amp. i also ordered some smd bandpass filters (hf) just search for qrp and read down the list. best wishes, terry, ei3iz.

    Reply
    1. Jan Post author

      Welcome and enjoy! Hopefully the boards will work fine for you. BTW, if you need to mount them into something, I am just using a double sided tape to stick them down while building. Otherwise hot glue works too.

      I didn’t want to add mounting holes because they would make the boards quite a bit larger.

      Thanks for the tip for the filters, will certainly have a look!

      Regards,

      Jan

      Reply
  2. steve NU0P

    I could not load the PCB files into Kicad…Has errors Schematic loaded ok

    Reply
    1. Jan Post author

      Hello,

      Sorry about that! What kinds of errors did you get? It could potentially whine about libraries but only standard ones are used (the ones from Github).

      Also make sure you have a really recent version of Kicad, I am using a version compiled in September this year.

      Reply
    2. Jan Post author

      Hello Steve,

      I have fixed the file – my Kicad is more recent than the last stable 4.0.7 and it was putting some extra keywords in the board file that the stable version didn’t know how to handle.

      Now it should work fine. Sorry about that!

      Jan

      Reply
  3. steve NU0P

    Great…It works now but the Silk Screen layer text is too large but I can fix that I think. I am not an expert on Kicad…Can you tell me if this is a single sided board??? If it is I can make them myself wit this…Steve NU0P

    Reply
    1. Jan Post author

      Yes it is single sided.

      However, the silk screen for the part designators is 0.8mm tall already – if you make it smaller or much thinner it may not be manufacturable.

      Reply
  4. Scott

    Jan, thanks for posting this and especially for posting to Osh Park. I’ve built an RBITX – Receiver only- and going from my painfully build Manhattan TIA modules to SMT will really clean it up and shrink it. Boards are in the mail, can’t wait! 73 and happy holidays!

    Scott Ka9p

    Reply
  5. Scott

    Jan, thanks for posting this and especially for posting to Osh Park. I’ve built an RBITX – Receiver only- and going from my painfully build Manhattan TIA modules to SMT will really clean it up and shrink it. Boards are in the mail, can’t wait! 73 and happy holidays!

    Scott Ka9p

    Reply
    1. Jan Post author

      Hello Scott,

      Glad to see that the work is useful for someone. However, if you want to use my TIAs, you may need to adjust the gain in them by modifying the resistor values – as it is described in the original article. In my own receiver I have replaced the TIAs except for the first pre-amplifier with J310 JFET stages from Pete, N6QW and where I could barely hear a signal before now I have it come booming in. The diode mixers are fairly lossy so a few extra dBs of gain help here. The disadvantage is having to wind transformers.

      Good luck and happy holidays!

      73!

      Jan

      Reply
      1. Scott

        Thanks Jan, Happy Holidays back. You are exactly right. At the moment the breadboard has two sbl1/ 2n5109 blocks that run at a net gain per block of about +8 dB, your 22 dB configuration per block minus mixer loss should net the receiver about 14 dB better – to tell the truth the lower gain is great here on 80 and 40, not loud but very useable with full size inverted vees, but do need the gain above 10 mhz. As soon as the two new 22db smt boards are done, will be experimenting with a 12 mhz Hycas, so may need to lower the gain a bit back. I’m curious how the 2n5109 will compare to the Jfet and 3904s for IMD on an angry contest band.
        Thanks again.

        Reply
        1. Jan Post author

          I have only used this with an indoors magnetic loop on 40m in an apartment so the gain issues are a bit more pronounced than if you have a proper antenna. I don’t have the means to put one up, unfortunately (living in a rented flat and all that …).

          I have been using ADE-1 mixers but those should be comparable with the SBL1.

          Reply
  6. Scott

    Hi Jan-

    I finally completed the first pair of TIAs and the spectrum analyzer shows 20db gain at 45 mhz with up to about -10 dBm input. They work great with the resistors adjusted as you suggested. SMA connectors soldered right to the input and output too! And they are about 15% the size of my Manhattan boards…

    Thanks again for making the boards available, I have 10 more to build.

    73 Scott

    73 Scott

    Reply
    1. Jan Post author

      Hello,

      Happy to hear that it works as advertised. I haven’t measured them apart from using my Mk. I ear because I don’t have a spectrum analyzer and didn’t want to build extra pads for my SSNA – the ones I have are not sufficient to push the gains down in order to not blow the detector. Need to get my lazy backside off the chair and either stock up on attenuators or build some one of these days.

      73!

      Jan

      Reply

Leave a Reply to Jan Cancel reply

Your email address will not be published. Required fields are marked *