1
\$\begingroup\$

I have a Murata Balun available DXP18BN7514T to match an unbalanced 50Ω source to a 300Ω differential load over about 50 - 60 MHz bandwidth. Unfortunately the DXP18BN7514T is specified as a 75 : 300 Ω Balun. I have some options available, some being more convenient than others, so :

  1. I could use a suitable 50 : 75 Ω Balun / transformer and complete the match to 50 Ω this way. Problem is that most such transformers available from Mini-Circuits or Macom are wire-wound, large and fairly high profile components relative to the rest of the design. Otherwise have not been able to source an SMD solution that is readily available so far ...
  2. I could do a discrete component match of the unbalanced 75Ω to the unbalanced 50Ω, but this will be more involved and more time consuming, and also require extra discrete components, calling for a broad-band match due to the 60MHz bandwidth. Extra difficulty is that typical S-parameters are not readily available for the DXP18BN7514T.

There may be a third or fourth option that I've not thought about ... What would one recommend, keeping in mind that simplicity and use of readily available components is prefered, over a more complicated solution (even though the later is not completely excluded...)

thanks in advance for any suggestions...

\$\endgroup\$
2
  • \$\begingroup\$ Have you considered tuning the antenna using a LC filter to reduce the SWR? \$\endgroup\$ Commented Aug 13, 2020 at 22:41
  • \$\begingroup\$ Hi, there is no antenna at this stage of the circuit. It's a broad-band inter-stage match between 2 devices. \$\endgroup\$
    – citizen
    Commented Aug 14, 2020 at 7:31

1 Answer 1

2
\$\begingroup\$

For these frequencies and your particular needs, a discrete LC balun with Z transformation would be the way to go, ditching the Murata part altogether. You really should not have any issues with standard SMD stuff here.

Here is such an example, it consists of two parallel LC filters: enter image description here

The way this works is pretty intuitive, the LC networks are designed to operate at the resonant frequencies, acting as phase shift networks to produce the balanced phase at output. The impedance matching can be viewed as a standard LC section providing an impedance match between 50 ohms and 150 ohms. So this should tell you a bit about the bandwidth you can expect: $$Q=\sqrt{\frac{150}{50}-1}=\sqrt{2}$$ For a centre of 57MHz: $$BW = f_0/Q=40MHz$$

You can also use the calculator here.

\$\endgroup\$
7
  • \$\begingroup\$ Hi Adil, yes thanks this looks like a third way for sure. One thing to note though is that f0 in my case is around 900MHz this makes the BW rather large. Also correct me if I'm wrong, but the Q should probably take the balanced impedance which is double = 300 Ohms, so this would come to √5 instead, which makes the BW at 900MHz = 402.5MHz. I will need to reduce this BW to more like about 65MHz for the application if possible. \$\endgroup\$
    – citizen
    Commented Aug 18, 2020 at 12:13
  • \$\begingroup\$ Well you should have made this a bit more clear. 50-60MHz in your question seemed you were specifying the band edge. For f0 of 900 MHz this is not at all recommended. You need to use transmission line transformers such as the Murata you listed. Check out MACOM and Minicircuits they might have what you need. You could chain the more common 50 Ohm to 75 Ohm balun before this if your OK with the insertion loss. \$\endgroup\$
    – MAM
    Commented Aug 18, 2020 at 12:17
  • \$\begingroup\$ BTW are you aware of a similar phase-shift topology employing more like T or π networks to take back some control of the Q ? \$\endgroup\$
    – citizen
    Commented Aug 18, 2020 at 12:18
  • \$\begingroup\$ Yes MACOM and Minicircuits have them but not as SMD as I'd like them to be ... Anyways thanks for the suggestions ... \$\endgroup\$
    – citizen
    Commented Aug 18, 2020 at 12:21
  • \$\begingroup\$ They are pretty small meant to be SMD, not sure if this is small enough for you? cdn.macom.com/datasheets/MABA-010374-CT4A40.pdf \$\endgroup\$
    – MAM
    Commented Aug 18, 2020 at 12:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.