Synchronize & Compare Two Analog Signals 1

[geomit]

Hello,

I am looking for a circuit design that will allow me to synchronize two incoming analog signals (one reference and one unknown) which would allow me to compare them to see if they are similar or different.
I am not to familiar with appropriate terms and phrases like, "phase lock loop"....etc. This is sort of hindering my search results.
The specifications for the circuit are not really important right now seeing as I am looking for the function first. However, as a reference, I will use +/-5V AC signals with no greater than 100kHz frequencies.
Thank you

 

[IRstuff]

If you have no control over the generation of these signals, you cannot "synchronize" in realtime. You could only digitize them and then manipulate them via software.

TTFN

FAQ731-376

 

[melone]

For what purpose?

 

[geomit]

We are monitoring signals off of a power harness on a vehicle.
We want to compare these signals to pre-recorded signals that pose as a baseline/reference.

 

[geekEE]

What do the waveforms look like and what do you consider when you say they are similar or different? Are you looking at frequency only? Amplitude? Phase? Something else?

It's going to be a much different problem if you're comparing two sine waves than if you're comparing two arbitrary waveforms.

 

[geomit]

We are looking for comparison of freq., amplitude, and phase to the reference signal.
In general, the waveforms will be DC with some transients (~450us).

 

[MacGyverS2000]

Waveforms are not DC... that's not a signal, that's a voltage. Waveforms are sine, square, triangle, mixed, etc.

What exactly are you trying to measure... feeding us info piecemeal isn't going to get you a useful answer and will more than likely just annoy everyone until we give up.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[xray]

I agree with Macgyver, in order to get a solution to the problem the problem has got to be defined, with that said i will offer that a differential pair is such a circuit. it rejects the common mode of signals even dc and apmlifies differential signal. if you design a differential pair and subject the two bases to the reference signal and the other signal you would like to see in your instance a zero voltage gain, there is no difference. The sychronization does not strike me with any kind of solution except within software as was mentioned by IRstuff. One thought if the other signal is flowing through an LTI it might be possible to make frequency selet circuitry to analyze for differenet phase shift through the system.

good luck

 

[jvjtech]

I am just thinking out of the box and throwing out this as an idea. With signals around 100kHz you might want to study RF techniques, specifically superheterodyne signal processing. An incoming signal is mixed with a reference signal at frequency different from that of the known signal by a known amaount. The output of the mixer is a signal at the delta frequency but modulated with the information, e.g. AM or FM or both, on the incoming signal. This signal can then be processed with circuitry designed for the difference frequency. The ARRL (Amateur Radio Relay League) would be a good source for this kind of information. I would also suggest looking at the Communicatoin and Signal Procesing forum. Regards.

 

[Noway2]

If you are working with the signals after they have been digitized, you could run a correlation algorithm on signal versus the reference. The DSP books by Oppenheim, as well as others, will have an explaination of the algorithm. I don't remember too much about it, other than it is similar to performing convolution.

 

[jvjtech]

I will add that it may be possible to combine the superheterodyne and DSP techniques and digitally process difference signal at the difference frequency. Thus the sampling for the DSP analysis need not be as high as what would be required for the "raw" signal but, instead, could be tuned to the difference frequency. Regards.

 

[sreid]

I like Noway2's correlation idea. But the solution to your problem could be a digital filter. This would allow you to look for a "Signature" in your signal.

 

[IRstuff]

This problem has been addressed in many digital scopes. A pass-fail region is defined in voltage and time. The scope is synchronized with some sort of start signal, and the unit under test output is digitized by the scope and passes if the signal never strays outside of the acceptance limits.

TTFN

FAQ731-376

 

[itsmoked]

Perhaps a Lock-in amplifier? One worked well for me doing moon bounce distance experiments.

http://en.wikipedia.org/wiki/Lock-in_amplifier

Note Basic principals.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[VEBill]

"...moon bounce distance experiments."

Amateur EME or professional something-or-other?

(Excuse the thread hijack...)

 

[itsmoked]

Amateur type stuff that occurs when a college student gets access to a large mirror observatory, stacks of various lasers, lock-in amplifiers, and fast expensive scopes.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[IRstuff]

The OP needs to come back with
How similar?
How different?

TTFN

FAQ731-376