The new forums will be named Coin Return (based on the most recent vote)! You can check on the status and timeline of the transition to the new forums here.
Please vote in the Forum Structure Poll. Polling will close at 2PM EST on January 21, 2025.
Communications/Audio Engineers, alittle help please?
Athlantar
Registered User regular
Registered User regular
Hi all,
I'm a budding Electronics Engineer in my third year. I'm tasked with a design project, and my instructor specifically said he wouldn't assist us. Which is total bullshit, but thats not what I'm having trouble with. The design spec is that we have one single wire connecting two buildings, and we need to construct a communication system that allows full duplex communication. I've got a system that uses lowtech AM, coming out of a microphone, opamped up and a DC bias to the desired modulation index, and mixed with a generic mixer and a VCO. (I hope its all good up to here :P) One of these would sit on each side of the wire, along with a demodulator (demodulating at the frequency which the other sides transmit is mixed with, using another simple mixer).
Now I know its usually okay to tie inputs of systems together so I'm not worried about the demodulators, however, having a modulator on each side is like tieing two outputs together. So what I only recently realised is that this might not actually work. The two modulating mixers will be affecting the wire at 1MHz, and at 3MHz. So, is this okay to do? If not, what alternatives might I have?
Thanks so much for your help H/A. I will gladly provide any additional information if I missed anything important.
tl;dr. can i connect the output of two mixers onto one wire?
I'm a budding Electronics Engineer in my third year. I'm tasked with a design project, and my instructor specifically said he wouldn't assist us. Which is total bullshit, but thats not what I'm having trouble with. The design spec is that we have one single wire connecting two buildings, and we need to construct a communication system that allows full duplex communication. I've got a system that uses lowtech AM, coming out of a microphone, opamped up and a DC bias to the desired modulation index, and mixed with a generic mixer and a VCO. (I hope its all good up to here :P) One of these would sit on each side of the wire, along with a demodulator (demodulating at the frequency which the other sides transmit is mixed with, using another simple mixer).
Now I know its usually okay to tie inputs of systems together so I'm not worried about the demodulators, however, having a modulator on each side is like tieing two outputs together. So what I only recently realised is that this might not actually work. The two modulating mixers will be affecting the wire at 1MHz, and at 3MHz. So, is this okay to do? If not, what alternatives might I have?
Thanks so much for your help H/A. I will gladly provide any additional information if I missed anything important.
tl;dr. can i connect the output of two mixers onto one wire?
10088847214381951643320470475858305731166183151407164751271470824235003318621252307969752086088076499395823874814123350292603347408732347765156628342107995
Athlantar on
0
Posts
Instead, what you might want to do is a dirty little hack involving a voltage divider. I don't guarantee that it'll work, but it seems like it'll work on paper.
Let:
Va be the output of mixer A
Vb be the output of mixer B
R be an appropriately chosen resistor
Vm be the midpoint of a voltage divider organised like so:
Va -- R -- Vm ------ Vm -- R -- Vb ^^ long piece of wire between buildingsFrom here, I can think of two ways you can recover the information, but one is more practical than the other.
Vm = (Vb + Va) / 2
The side that produces Va is interested in recovering Vb, so all it needs to do is measure Vm, and subtract Va/2 from it. This is "easy" since, well, it produces Va. I say "easy", since you're running a subtraction circuit at 3MHz. If you're using badly spec'd opamps with insufficient gain bandwidth, you're going to be doing yourself a world of hurt.
2.) Rely on the demodulator to automatically filter out the other signal, since the two carrier signals are relatively far away from each other.
Also, a couple of things to consider (I realise that you're doing a project, but you might be interested nevertheless):
One thing regarding this I considered was that 1M signal will have harmonic components at the 3M, so I was possibly going to choose 1M and 2.5M (or something random) instead.
The one wire thing has thrown me and made me uncomfortable since day 1 of this nonsense :P
The wire we are using is 10Km long with RG58 coax, so to be honest I'm not sure what all that involves on my part. I'm not great at impedance matching and I havent been able to figure out from the datasheet if the mixer I've chosen is a high impedance output. I know the cable is 50Ohm impedance, so I was going to try to make each circuit on each end appear to be 50Ohms.
With that crazy voltage divider scheme, wouldnt I need something connecting to ground for it to actually work, and avoid the 'outputs duking it out' situation?
Thanks much for the help thus far.
P.S. is the two demodulator thing okay though?
Ah. Odd harmonics. That's a good idea, actually, shifting the carrier over.
Oh wait. One cable, but two wires. Since you're using coaxial, that has a ground shield on the outside. Let me find a picture... here. Either way, at 10km, you have a transmission line, and so the terminology/techniques changes a little.
What I've seen done is:
Tx --> attenuator --> impedance matching <--> transmission line <--> impedance matching <-- attenuator <-- Tx | | Rx <----- (optional filter) <---------+ +-----> (optional filter) ----> RxThe attenuator is so that the output of one transmitter does not:
a.) damage the output of the other transmitter on the other side, and
b.) damage/saturate the input of the receiver closest to it
You will lose a few db of signal, but receivers tend to be much more sensitive than transmitters, so you tend to be okay.
Yup.