Clock Recovery Using Tuned Circuit

S

stevieboy01

Guest
I am hoping to design a fairly simple clock recovery circuit using a
Full Wave Rectifier, a tuned circuit and a Schmitt trigger. Currently,
the system uses 4 level ASK signalling over a wireless channel model.

I am having trouble designing a full wave rectifier. I would like the
circuit to be active and have come across a circuit which half wave
rectifies, doubles this signal, and adds it to the original waveform.
This may cause problems with component drift as the gain will need to
be fairly exact. ANy other ideas?

Any possible circuit ideas would be greatly appreciated.
 
On 1 Mar 2005 08:04:53 -0800, stephen_cochrane@hotmail.com
(stevieboy01) wrote:

I am hoping to design a fairly simple clock recovery circuit using a
Full Wave Rectifier, a tuned circuit and a Schmitt trigger. Currently,
the system uses 4 level ASK signalling over a wireless channel model.
Click to expand...
Is the ASK four levels at the same phase, or are two states flipped?


John
 
Sorry, didnt explain correctly. It doubles the amplitude of the half
wave rectified circuit and then adds it to the non rectified signal.
This is a fairly long-winded way of doing it, I admit, but is the only
one I have come across.
 
It is 4 levels at the same phase. I want to use the diode square law to
produce a squared version of the rectified signal and then filter the
baseband signal out, using the tuned circuit or some other type of
filter. Hopefully the clock will be derived from this.
 
By the way, the bandwidth of the channel is approximately 15kHz
 
On 1 Mar 2005 08:35:46 -0800, "stevieboy01"
<stephen_cochrane@hotmail.com> wrote:

It is 4 levels at the same phase. I want to use the diode square law to
produce a squared version of the rectified signal and then filter the
baseband signal out, using the tuned circuit or some other type of
filter. Hopefully the clock will be derived from this.
Click to expand...
Then why not just bandpass filter at the carrier frequency (or
equivalently, lock a PLL to it) and that's your recovered carrier.
Then mix (ie, linear multiply) that with the incoming signal, lowpass,
and that's your 4-level dibit data. I assume data and carrier are
synchronous somehow.

Or even do a simple AM detector!

Why rectify? That's only needed if the data is biphase and has zero
average carrier content, which would force you to look for the 2F
thing.

What's the carrier frequency? Is it synchronous to the data?

This is potentially confusing to do in text. You could post a block
diagram sketch to a.b.s.e. for discussion.


John
 
stevieboy01 wrote:

It is 4 levels at the same phase. I want to use the diode square law to
produce a squared version of the rectified signal and then filter the
baseband signal out, using the tuned circuit or some other type of
filter. Hopefully the clock will be derived from this.

Why not use a mixer chip like the NE612, or just use a windowed
Click to expand...
comparator since you were going to square it up with a schmitt trigger
anyway?

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com
 
Tim Wescott wrote:
stevieboy01 wrote:

I am hoping to design a fairly simple clock recovery circuit using a
Full Wave Rectifier, a tuned circuit and a Schmitt trigger. Currently,
the system uses 4 level ASK signalling over a wireless channel model.

I am having trouble designing a full wave rectifier. I would like the
circuit to be active and have come across a circuit which half wave
rectifies, doubles this signal, and adds it to the original waveform.
This may cause problems with component drift as the gain will need to
be fairly exact. ANy other ideas?

Any possible circuit ideas would be greatly appreciated.


This only works with a rail-rail op-amp, it won't be terrifically
high-speed, the output impedance varies tremendously _and_ it's been
ages since I've used one, but it does work:


___
.---|___|---------.
| |
| VCC |
| + |
| | |
___ | |\| |
--------|___|---o----|-\ |
| >----->|--o--o---------
.----|+/ |
| |/| |
| | .-.
=== === | |
GND GND | |
'-'
|
|
===
GND
created by Andy´s ASCII-Circuit v1.24.140803 Beta www.tech-chat.de

On the negative half-wave the thing acts like a regular op-amp circuit,
with the diode forward-drop being compensated for by the op-amp. On the
forward half-wave the diode is reverse-biased and the signal just passes
through the feedback network into the output (clever, huh?). You may
want to buffer the output with a voltage follower to speed it up a bit.
Click to expand...
You're going to do that at 433MHz?
 
The operating frequency will be approximately 20kHz
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top