# Bias-tee for adding multiple frequencies to a DC line

Objective - To add multiple frequencies on sine waves onto a $24V$ DC Line.

Each frequency block ($F_1, F_2, ...$) generates a sine wave of unique frequency within the given range. When the $n^{th}$ frequency block $F_n$ is switched on, it's frequency must be added to the 24V line regardless of any frequency present along that line.

I have used bias-tee circuit to combine the frequencies. Each yellow block has a $100nF$ coupling capacitor.

Sine wave generation method : Square wave to sine wave converter

Frequency Range : $8KHz - 24KHz$

Number of Frequency Blocks $N$ : $10$

Current$/$Frequency Block : $10mA$ at $24V DC$

L1 Inductor Rating :

• Inductance : $470\mu H \pm 10\%$
• DC Current Rating : $420mA$
• Self-Resonant Frequency : $100KHz$

Coupling Capacitor : $100nF, 50V$

Issues -

1. There is a lot of high frequency distortion when detecting the signal over the power-line. (The power-line wire is $2$ core, $1.5mm^2$, $25m$ between each device) How do I resolve this?
2. The sine waveform (from one block) gets distorted when more than one frequency block is connected. How do I isolate each block but still allowing the signal to flow through? I have used a buck-converter to convert $24V - 5V DC$ within each frequency block. It has huge capacitors in its input. Is that the problem? Will a diode at the input do the trick?

3. I don't get any sine-waves (added) at the power-line, for more than 2 blocks. Do I need to change the $L1, C$ values?

4. Do I need an end resistor after the $N$ blocks, connecting the positive and negative?

If this design could be improved, kindly suggest so.

Thank You.

What you need is an inductor going into each frequency generator block.

Each of those block looks like a load for your frequency generator, the same as the 24V power supply does.

Everything that uses DC from the 24V line provides DC to the line needs an inductor.

Everything that puts a signal on the line needs a capacitor.

1. High frequency noise or distortion? I expect you mean noise. Twist the wires to reduce received noise. You might also add ferrite beads to the lines, but I think the answer I gave above will get rid of most of it.
2. You are mixing the various signals on one line. If you put two signals on the line, then you will see a mixture of the two. This may be the "distortion" you are seeing. You might also have problems with the different generators outputs being connected directly together. In this case, a resistor in series with the coupling capacitors may help.
3. See above. There's a very good chance that the power inputs to the frequency block short out the outputs when you get enough in parallel. Again, an inductor in the power feed to each block will help.
4. You shouldn't need a termination resistor at the end of the string. The frequency blocks put a load on the 24 volt line, so for the frequencies you are using you shouldn't need any kind of termination.

I think your inductors are too small. At 8kHz, they have an impedance of only 23Ohm - that doesn't do much for you. 100mH inductors would have an impedance of 5kOhm, which would do much more for you in terms of keeping the signals away from the power consumers/providers.

• I tried that too. I used the same 470uH for each block. But, there wasn't any improvement. Aug 16, 2016 at 12:57
• What should be the current rating of the 100mH inductor? I could find only 20mA rating in the market. I think I require 2A incorporating all the 10 frequency modules. Or will that suffice as each module takes only 10mA? Correct me if I am wrong. Aug 20, 2016 at 7:13
• The inductor at the power supply must be capable of supplying all of the needed current, so 10 modules at 10mA=100mA for the main inductor. The others can use smaller ones.
– JRE
Aug 21, 2016 at 12:08