50Ω Carrier for TV IF SAW filters

Introduction

Some time ago I set out to build a rig to allow the use of TV SAW filters in a 50 ohm environment. I found a few ham resources around the web centered around LC impedance matching arrangements which only match at a specific frequency. Perfectly fine for narrow carriers but I wanted to use them for actual TV carriers which are quite large (7 MHz) when compared with the kind of thing average ham is dealing with; more likely to be measured in tens of KHz.

In my searching I stumbled upon this elaborate circuit in an old Siemens SAW filter catalogue:

It’s the circuit which is used for testing their SAW filters i.e. attaching them to a VNA. And also for generating all of the diagrams which appear in the catalogue, so yes, it’s going to be pretty nice and unsurprisingly it is not LC based. What the actual rig looks like? Who knows. But I want one. Annoyingly the circuit is partially illegible. I found another scan of the same catalogue online, and that one was even worse.

After a quite a bit of squinting and experimentation I managed to recreate the circuit:

Note that in the original some of the components are either unspecified, or vintage/rare thus I did a bit of substitution when coming up with the above. The original used two types of transistor: BF959 and BF241. I found this to be unnecessary in my recreation which uses MMBTH10 for everything.

In the original schematic the text under the transformer appears to say 4:2 (turns ?). Perhaps that was the actual number of turns on the primary/secondary? I drafted in a 2:1 transformer from Mini-Circuits which did the trick.

The value of R2 was altered because I ended up with too little (or too much? I can’t recall) gain on the rig.

Here is the final result:

On this particular board there’s a small oops at the input which is fixed in the project linked below.

I decided to make it all SMD (not sure why). Perhaps it would have been more helpful to others had it been through-hole.

How well does it work?

S21 of an X7251D filter at 0dBm.

Wow. It’s beautiful. 0dB in the passband too! (adjusted by RV1). The filter being tested above is an Epcos X7251D. An exact rendition of the diagram in the datasheet for the filter.

The full schematic can be viewed here.

Check out the KiCad project on GitHub.

Posted in Circuit snippets

Leave a Reply

Your email address will not be published.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