Recovering Manchester Code: How is it done?

[DigiTan]

I head from several web sites that it is possible to send Manchester (Biphase-L) encoded bits into systems that are DC-intolerant; including transformers. Exactly what would I need to do if I wanted to recover these signals from a transformer secondary? Would this involve some kind of integrator?

 

[mastertech]

from the engineering point of you, you should know how the waveform
looks, and whether there are artifacts like noise ,ringing etc, before deciding what to do next,also the signal at the secondary will appear as
ac and to convert it to pulsating dc you could use a simple diode network
then you would condition the signal by filtering and limiting before
you apply any decoding on it

 

[checkmate]

On almost all communication systems, transmitted signals should have certain characteristics. AC signals (ie biphase) have zero offset. Offsets merely serve to waste energy over the transmission line, hence signals are usually ac-coupled on the transmission side. Also, offsets could lead to signal drop-off over extended distances.

Most digital demodulation techniques involve correlation. Correlation means you integrate the product of the received signal r(t) and a particular signal pattern s(t), wrt to t, to find the "similarity" between them. So if there are, for eg, 4 different possible signal patterns, then you find the correlation of all 4 patterns, the the one with the highest correlation will be the best estimate of the transmitted signal from the received signal, regardless of whether they are DC or AC.

Suppose there are n possible transmitted signals. So using correlators, you need n integrators and n mixers. Mixers are very "expensive" components. Hence, a substitute to the correlator is a filter, with impulse response h(t) = s(T-t), where s(t) is the possible transmitted signal. It is also very important that the output of the filter is sampled at exactly the right moment, and that s(t) is synchronized exactly with r(t), using PLLs or some other means.

Over at my university here, the entire demodulation process is taught over an entire module on Digital Communications. I'd suggest you grab any text on digital comms if you want to know more.

 

[DigiTan]

Thanks! Thoes sound exactly like what I'm looking for. Do you recommend any integrator/mixer tutorials or equipment in general for low-speed correlation (< 20KHz)?

 

[Roff]

I can't imagine why you would need a correlator unless you have poor signal-to-noise. Have you searched for Manchester decoders? They are commercially available.

 

[DigiTan]

Heh heh. I don't know why, but that idea never occoured to me. :lol: But this signal out of the transformer...I still have to integrate it with an op-amp or something?