How do I match 75 ohm to 50 ohm and back?

[technix]

With me moving equipment around, the FM receiver got moved to a location that isn’t very convenient, and thus I need to add a length of coax so I can keep my antennae at the original location. I also want to add a LNA in there in hope to improve reception.

The FM antennae, the receiver and the LNA chip all expect 75 ohm, but my existing coax and connectors are all 50 ohm equipment. I don’t have any 75 ohm coax or matching connectors. FM band in China is 87-108MHz. I build my boards using mostly 0603 SMT passives.

Question:
1. Can I just shove the amplified FM signal from the 75 ohm LNA down the 50 ohm coax and SMA connectors and expect it to work?
2. If not, how to build the matching network on both ends of the 50 ohm coax? Bonus if the matching network blocks DC, since that will also allow me to inject power.

[TheSteve]

In this case I wouldn't worry about the impedance mismatch.

[bd139]

If it's receiver I wouldn't worry.

Transmitter / anything that carries power then you tend to use a "minimum loss pad" as described here: https://www.maximintegrated.com/en/app-notes/index.mvp/id/972

[technix]

If it's receiver I wouldn't worry.

Transmitter / anything that carries power then you tend to use a "minimum loss pad" as described here: https://www.maximintegrated.com/en/app-notes/index.mvp/id/972

It is a receiver. The antennae end contains a 50MHz-120MHz BPF and a single-chip wide-band LNA. The receiver end contains a power injector.

[tggzzz]

https://chemandy.com/calculators/matching-pi-attenuator-calculator.htm
https://chemandy.com/calculators/matching-t-attenuator-calculator.htm

Note the loss of power, which might be important depending on the signal and receiver.

If you don't match impedances, then you will get a reduced signal at some frequencies defined by the cable length.

Whether or not the LNA helps reception will depend on the cause of the poor reception; sometimes they can make things worse.

[Howardlong]

If it's receiver I wouldn't worry.

Transmitter / anything that carries power then you tend to use a "minimum loss pad" as described here: https://www.maximintegrated.com/en/app-notes/index.mvp/id/972

Pedant hat on... a minimum loss pad is a very broadband resistive device that would be more lossy (~5.7dB) than a transformer (~0.5dB). A minimum loss pad is much simpler to make though! I'd imagine a simple LC matching network or lamda/4 coaxial transformer would improve on an MLP 88MHz through 108MHz.

Irrespective, forgetting about any matching will almost certainly work well enough.

[bd139]

Fair point! 

[technix]

https://chemandy.com/calculators/matching-pi-attenuator-calculator.htm
https://chemandy.com/calculators/matching-t-attenuator-calculator.htm

Note the loss of power, which might be important depending on the signal and receiver.

If you don't match impedances, then you will get a reduced signal at some frequencies defined by the cable length.

Whether or not the LNA helps reception will depend on the cause of the poor reception; sometimes they can make things worse.

The reception was okay where the antennae was, hence the antenna-end LNA.

[EVS]

Low-Loss Matching Network 75 to 50Ohms (for 50-230MHz):

[Howardlong]

Simple lumped LC with 180nH shunt across the 75 ohm source to ground, 47pF series to 50 ohm port gives 26dB worst case return loss on either port 88MHz through 108MHz. Insertion loss is negligible across the band (0.01dB).

[technix]

Low-Loss Matching Network 75 to 50Ohms (for 50-230MHz):

Simple lumped LC with 180nH shunt across the 75 ohm source to ground, 47pF series to 50 ohm port gives 26dB worst case return loss on either port 88MHz through 108MHz. Insertion loss is negligible across the band (0.01dB).

How do I match it back to 75 ohms then?

[Howardlong]

That's it, that's all you need. The LC match is reciprocal but it is bandwidth limited.

So you put one at each end of your 50 ohm run.

But honestly, I wouldn't even bother to do that!

Do you know you even need an LNA? I ask because sometimes they cause more harm than good.

How long are the original and new cable runs, and what sort of cable are you using?

If you do use an LNA, its gain is likely to override any losses in the following 50 ohm run unless it's a very long run.

[technix]

That's it, that's all you need. The LC match is reciprocal but it is bandwidth limited.

So you put one at each end of your 50 ohm run.

But honestly, I wouldn't even bother to do that!

Do you know you even need an LNA? I ask because sometimes they cause more harm than good.

How long are the original and new cable runs, and what sort of cable are you using?

If you do use an LNA, its gain is likely to override any losses in the following 50 ohm run unless it's a very long run.

The original antenna is one meter long. It is one of those simple wire-style FM antenna commonly found with A/V receivers.

I am inserting up to 6 meters of 50 ohm coax. That is why I suspect I will need an LNA at antenna end to boost the signal a bit so it can survive the likely badly matched transmission line.

[tggzzz]

That's it, that's all you need. The LC match is reciprocal but it is bandwidth limited.

So you put one at each end of your 50 ohm run.

But honestly, I wouldn't even bother to do that!

Do you know you even need an LNA? I ask because sometimes they cause more harm than good.

How long are the original and new cable runs, and what sort of cable are you using?

If you do use an LNA, its gain is likely to override any losses in the following 50 ohm run unless it's a very long run.

The original antenna is one meter long. It is one of those simple wire-style FM antenna commonly found with A/V receivers.

I am inserting up to 6 meters of 50 ohm coax. That is why I suspect I will need an LNA at antenna end to boost the signal a bit so it can survive the likely badly matched transmission line.

Why on earth don't you simply use 75ohm coax, and avoid all the problems?

Look at the attenuation specifications of cheap 75ohm coax: at 100MHz it is 6-16dB for 100m. You only need 6m so the cable loss might be 1dB!

[technix]

That's it, that's all you need. The LC match is reciprocal but it is bandwidth limited.

So you put one at each end of your 50 ohm run.

But honestly, I wouldn't even bother to do that!

Do you know you even need an LNA? I ask because sometimes they cause more harm than good.

How long are the original and new cable runs, and what sort of cable are you using?

If you do use an LNA, its gain is likely to override any losses in the following 50 ohm run unless it's a very long run.

The original antenna is one meter long. It is one of those simple wire-style FM antenna commonly found with A/V receivers.

I am inserting up to 6 meters of 50 ohm coax. That is why I suspect I will need an LNA at antenna end to boost the signal a bit so it can survive the likely badly matched transmission line.

Why on earth don't you simply use 75ohm coax, and avoid all the problems?

Look at the attenuation specifications of cheap 75ohm coax: at 100MHz it is 6-16dB for 100m. You only need 6m so the cable loss might be 1dB!

I just happen to have some 50 ohm coax, and it is somewhat hard to find 75 ohm connectors.

[ZS6GST]

Build a 1/12th wave transformer using short pieces of 75 and 50 Ohm cable.

Google is your friend and will show you all the detail.

73 Pine, ZS6GST

[radiolistener]

By using 50 ohm cable with 75 ohm input you will get SWR=1.5.
It's about 14 dB loss, I don't think it will an issue for listening FM broadcast stations.

[Kalvin]

By using 50 ohm cable with 75 ohm input you will get SWR=1.5.
It's about 14 dB loss, I don't think it will an issue for listening FM broadcast stations.

VSWR 1.5:1 means 14 dB return loss and 0.18 dB transmission loss. The transmission loss is relevant in this case, so the 0.18 dB transmission loss is insignificant.

[fourfathom]

VSWR 1.5:1 means 14 dB return loss and 0.18 dB transmission loss. The transmission loss is relevant in this case, so the 0.18 dB transmission loss is insignificant.

There's going to be a bit more loss due to the cable loss (matched and SWR-related), but I looked at the numbers for 6 meters of RG-58 (likely/typical 50 Ohm coax), and it still all added up to less than 1dB at the top of the FM broadcast band.  Truly nothing to worry about.