Sending power and control over single coax

[makidoja]

Hi all

I'm working on a project which will allow a single RG8 cable to have 13.8vdc injected to power the remote circuitry, to receive RF back down the coax (AC) but also control 8 relays.

This is for a remote receiving antenna for the shortwave bands.

Injecting the power volatge and filtering ac / dc both ends i have covered but I have no idea where to start with being able to switch 8 different relays (1 at a time) over coax.

Now I could put a amtel microcontroller at the remote end but could a serial bus be injected along with the dc voltage and RF signal coming back from the preamp?

Sorry if it's a bit heavy but it's making my head hurt

Mak

 

[Harald Kapp]

I see two ways:

1) use a carrier frequency well below the RF range your antenna is receiving, modulate that carrier and filter it on the receiving end using a band-pass filter. Demodulate the filtered signal. Use any protocol you like, e.g. UART serial protocol, to transmit the control information.

2) Use the DC-power to transmit the control information. You will need to decouple and bufffer the DC signal at the receiving end (diode+capacitor), then you can modulate the DC by e.g. a serial UART protocol (idle=13V, active = 0V). You can easily decouple this information using a capacitor (to block DC), scale it to the µC's signal level using a resistive divider and receive the signal with the µC's UART. The µC then can control the relay(s).

 

[makidoja]

Hi Harald

OK well option 2 sounds like a good way

Do you know of any circuit diagrams for this kind of implementation??

I was under the impression that serial needed bi directional communications and 2 wires?

Mak

 

[Harald Kapp]

You don't have to use bidirectional communication, if you only want to send commands without feedback.

I don't have a circuit diagram at hand. Look up 1-wire protocol. this is almost what I described. You need to make adaptations for the 13V power supply.

 

[duke37]

I was looking for a similar circuit and thought that a telephone dial running a uniselector would do the job but I never found a circuit, perhaps someone here knows.

Different DC voltages on the line could switch different outputs. You could use pulses with the peak level to run the circuit and the mean level to set the output.

 

[makidoja]

You don't have to use bidirectional communication, if you only want to send commands without feedback.

I don't have a circuit diagram at hand. Look up 1-wire protocol. this is almost what I described. You need to make adaptations for the 13V power supply.

Ah not this sounds interesting

So i could have an Arduino Mega at the control end and an Arduino Mini at the remote end using 1-wire protocol the data and power could be sent all down one RG6 cable (all DC) with the amplified shortwave signal (AC) coming back the other direction

The arduino mini code would just set outputs high or low to switch the relevant relays through driver transistors with commands sent from the Arduino mega at the manned end.

This all sounds very reasonable and i'm quite excited to have found this informaiton .. better start reading up

Thanks

Mak

 

[duke37]

I have had a cogitate and think the attached circuit may work. You will need a few extra resistors on the LM3914 to set the range.

NOTE that the second diode from the transformer is the wrong way round

 

[makidoja]

I have had a cogitate and think the attached circuit may work. You will need a few extra resistors on the LM3914 to set the range.

NOTE that the second diode from the transformer is the wrong way round

Not sure I get it

this is what I have

I need to feed 13.8vdc from the shed to a remote box with 8 relays and a RF pre-amplifier in, I would also like to control all 8 relays AND get the RF signal back from the remote box all on a single coax cable.

I know I can put 13.8vdc down the coax by injecting in via a rf choke and block the DC with capacitors from getting into the RF receiver and antenna.

What I dont know is how to send signals along the same cable to control the 8 relays individually.

Mak

 

[Harald Kapp]

I was thinking along the lines of a circuit like this:

Component values are arbitrary and the circuit is untested. I think chances are godd to get it up and running. For transmitting a UART signal an additional inverter may be needed at signal_out. Use a Schmitt-trigger type to reduce sensitivity to noise.

 

[duke37]

You said that you knew how to combine two signals on to the co-ax cable so I did not bother to draw that.

I suggested using 12VAC which is rectified to take the positive half cycle. This is rectified again to provide a DC supply for the LM3914 and up to ten relays.

The negative half cycle (turn the diode around) is fed through a resistance so that the voltage varies with the resistor value.

The LM3914 sees part of the positive voltage and part of the negative voltage which varies with the setting of the resistor. It outputs a positive voltage to one of ten outputs depending on the input voltage. I only drew one relay, you can imagine the rest.

The positive supply may be a bit rough but you can always smooth it or add a regulator at the relay end.

You can get four outputs using only diodes with two relays using no energising, positive energising, negative energising and both energising. I think.

 

[makidoja]

Hi Both

2 very interesting directions to try .. i will get building and report back

Mak

 

[makidoja]

I was thinking along the lines of a circuit like this:

Component values are arbitrary and the circuit is untested. I think chances are godd to get it up and running. For transmitting a UART signal an additional inverter may be needed at signal_out. Use a Schmitt-trigger type to reduce sensitivity to noise.

So V1 is something like a 7512 regulator and we are switching the gnd line on and off?

HF filter will be just dc blocking capacitors are we are modulating the DC and not AC?

Do we not need another RF blocking choke at the remote end?

Inverter would be something like a 74HC240
#
What kind of distance do you think this would operate over? Something like 100m?

Mak

 

[Harald Kapp]

V1 is your 13.xV source, whatever it is.
The HF filter need sto be designed to your requirements. I thought you had that covered. A simple capacitor my suffice, a real high pass filter using capacitors and inductors may be better.
This is just a coarse schematic. A HF blocking choke is required at both ends (I haven't shown that).
I would use a 74HV14 or 74VHC14 for inverter. The 240 doesn't have Schmitt-trigger inputs and you don't need the tristate-capability of the 240.

Distance? I don't know. 100m should be possible, but I wouldn't bet on it. You will have considerable attenuation of your RF signal over a disctance of 100m plus the two HF coupling/decoupling filters.

 

[makidoja]

ok so the shield of the coax must be isolated from the 13v supply or the modulation of the power supply wont happen?

I have the HPF covered thats no problem

The RF signal will be amplified at the remote end, bit like a sat LNB but using a wideband pre-amplifier circuit with a HPF to stop any BC breakthrough.

Freq of operation will be 1.6mhz to 14mhz roughly.

Mak

 

[Harald Kapp]

Shield is 0V in my schematic.

 

[makidoja]

Shield is 0V in my schematic.

Hi Harald .. sorry for being hard work

So the - of the power supply is isolated from the 0V line by M1 which is usually "closed" but when a data signal is supplied to m1 it opens and modulates the 13vdc supply.

Mak

 

[Harald Kapp]

I would think so.

You can also use a transistor in the + wire of the power supply. Only that this will be a bit harder to turn on and off than a transistor in the - wire.

 

[makidoja]

Hi again

I'm still working on this project and have the majority of the circuit designed all apart from the 1 wire control

I'm thinking of using the Dallas DS2408 IC for the remote end

I'm thinking your circuit is still relevant for this?

Thanks

Mak

 

[makidoja]

Revisit

Hi all

Revisiting

So to switch the NMOS I just feed with TXD data?

Mak