Impedance matching of sound blaster with RF mixer

[neazoi]

Hello I design a microwave transceiver using a double balanced 50ohm passive diode mixer. I need to drive this mixer from the sound blaster of a PC.

As far as I know a typical sound blaster output impedance on the green plug is about 32 ohm and the input impedance of the microphone input is about 600 ohm.

For the RX, I have found an amplifier that has a 50ohm input and around 500 ohm output but what about the TX? can I match the 32 ohm to the 50 ohm of the mixer by just connecting a 50ohm resistor in series, between the sound blaster output and the mixer port?

 

[goldsmith]

Dear neazoi
Hi

can I match the 32 ohm to the 50 ohm of the mixer by just connecting a 50ohm resistor in series, between the sound blaster output and the mixer port?

Of course no , you can't ! you should use a matching network what can convert the impedance . such as buffers or RLC networks .
Best Wishes
Goldsmith

 

[vfone]

For matching in this situation you can use a small audio transformer.
Where the Turns Ratio of the transformer = SQRT (Load Resistance / Source Resistance)
in your situation we can consider: TR = SQRT(32/50) = 0.8
As an example, if the transformer primary winding have 100 turns, the secondary winding must have 100 x 0.8 = 80 turns.

Also you can use a resistive matching network, but this introduce much more loss than a transformer, and to get higher level perhaps need to use the speaker output of the sound card, if is available.

 

[neazoi]

For matching in this situation you can use a small audio transformer.
Where the Turns Ratio of the transformer = SQRT (Load Resistance / Source Resistance)
in your situation we can consider: TR = SQRT(32/50) = 0.8
As an example, if the transformer primary winding have 100 turns, the secondary winding must have 100 x 0.8 = 80 turns.

Also you can use a resistive matching network, but this introduce much more loss than a transformer, and to get higher level perhaps need to use the speaker output of the sound card, if is available.

On the RX side everything is matched, since I have used a diplexer design followed by a suitable amplifier to bring the impedance to 500 ohms, which is quite close to 600.

Could you suggest such a network for the TX side? (32 to 50 ohms). I think the line out of the sound blaster is high enough to overcome a not too big loss. An option would be some kind of transformer isolation as you suggest, but appart from the ratio, I do not know the number of turns

 

[jiripolivka]

On the RX side everything is matched, since I have used a diplexer design followed by a suitable amplifier to bring the impedance to 500 ohms, which is quite close to 600.

Could you suggest such a network for the TX side? (32 to 50 ohms). I think the line out of the sound blaster is high enough to overcome a not too big loss. An option would be some kind of transformer isolation as you suggest, but appart from the ratio, I do not know the number of turns

All this matching stuff I think is irrelevant.
If your mixer has the IF port, 50 Ohms, specified for DC to several MHz at least, then the audio frequency spectrum is so narrow that any mismatch cannot affect it. If you want to be sure, use simply a series resistor ~20 Ohms, to match your 32 Ohms to 50 Ohms. Make sure there is no DC component from your audio source (use a 1 uF -10 uF capacitor to block it).

If you do not trust me, take conversion loss data for all suggested matching devices. You will see that none matters except of adding some loss to the IF line.

 

[neazoi]

All this matching stuff I think is irrelevant.
If your mixer has the IF port, 50 Ohms, specified for DC to several MHz at least, then the audio frequency spectrum is so narrow that any mismatch cannot affect it. If you want to be sure, use simply a series resistor ~20 Ohms, to match your 32 Ohms to 50 Ohms. Make sure there is no DC component from your audio source (use a 1 uF -10 uF capacitor to block it).

If you do not trust me, take conversion loss data for all suggested matching devices. You will see that none matters except of adding some loss to the IF line.

Should I have the resistor connected on the mixer side or the dc blocking capacitor?

 

[jiripolivka]

Should I have the resistor connected on the mixer side or the dc blocking capacitor?

I do not know your mixer specifications. Most mixers have two or four diodes connected in a bridge. Introducing AC low-level signals is OK but adding a DC current may kill a diode or at least shift the operation point. Look on the mixer as a 50-Ohm resistor. Your AC signal source is 32-Ohm resistor. For the best match, connect the "live" points by a 20-Ohm resistor. If you are not sure if your audio source has a DC voltage on it, add an DC-blocking capacitor in series with the 20-Ohm resistor.
You can simply play music into a speaker with your audio source, and use a DC voltmeter (recommended is an analog voltmeter) to see if there is any DC voltage.

 

[neazoi]

I do not know your mixer specifications. Most mixers have two or four diodes connected in a bridge. Introducing AC low-level signals is OK but adding a DC current may kill a diode or at least shift the operation point. Look on the mixer as a 50-Ohm resistor. Your AC signal source is 32-Ohm resistor. For the best match, connect the "live" points by a 20-Ohm resistor. If you are not sure if your audio source has a DC voltage on it, add an DC-blocking capacitor in series with the 20-Ohm resistor.
You can simply play music into a speaker with your audio source, and use a DC voltmeter (recommended is an analog voltmeter) to see if there is any DC voltage.

The mixer is a double balanced diode mixer (passive) and all ports are 50 ohm.
I will add a dc blocking capacitor as well. So the correct connection should be sound blaster->resistor->capacitor->mixer OR sound blaster->capacitor->resistor->mixer?

 

[jiripolivka]

The mixer is a double balanced diode mixer (passive) and all ports are 50 ohm.
I will add a dc blocking capacitor as well. So the correct connection should be sound blaster->resistor->capacitor->mixer OR sound blaster->capacitor->resistor->mixer?

Both are correct. You would find that you do not need the 20 Ohm resistor, and if there is no DC voltage from the audio source, the capacitor can also be not used.
The only important thing now is to keep the audio level low, say 0.1 V p-p, to prevent mixer saturation. You can measure the conversion loss if you have a RF power meter; then you can adjust gradually the audio power till the RF output is saturated.
Typical mixers have 6-10 dB conversion loss (IF to RF), and saturate quickly. Check the LO input power, various mixers need certain levels to operate well, starting from +7 dBm.

 

[vfone]

Transformers are more recommended in this situation (myself I used successfully many times), even if there is no perfect source/load matching due to unknown Turns Ratio.
Probably you already have few of them in the junk box

**broken link removed**