Shift Phase of 100 kHz Signal 90 Degrees

[Baluncore]

It might be a good time to consider that, no matter how the 100kHz 4-QAM carrier is generated, the serial data stream will need to be delivered to the modulator at a rate of 200k bits per second.

So the clock that moves the serial data to the modulator will somehow need to be synchronised with the modulator clock. That may be difficult with a 100kHz crystal oscillator. It may need a PLL.

Will that be two separate channels of data, clocked at 100kHz, or one channel clocked at 200kHz, that needs to be split into two staggered bit streams.

 

[sophiecentaur]

It might be a good time to consider that, no matter how the 100kHz 4-QAM carrier is generated, the serial data stream will need to be delivered to the modulator at a rate of 200k bits per second.

So the clock that moves the serial data to the modulator will somehow need to be synchronised with the modulator clock. That may be difficult with a 100kHz crystal oscillator. It may need a PLL.

Will that be two separate channels of data, clocked at 100kHz, or one channel clocked at 200kHz, that needs to be split into two staggered bit streams.

If the modulating signal is going to be binary and all filtering done post modulation then there is no need to have a fast clock for the data. Yes there may be a small 'beat' where the data transitions strobe through the 100kHz but that will be ignored by a demodulator. (I reckon?)
If the data is to be digitally pre-filtered (much healthier) then it will need to be sampled at a high rate. It could be better to choose a clock frequency for the data samples that will strobe well out of the transmitted band. But if the system is all in one box, why not use the same clock source everywhere?
The choice depends on how the system is planned to work. It is good to have an overall plan, involving such choices early on.
If this is a first time project then it may be better to keep everything as simple as possible and then evolve it into something better later.

 

[Fischer777]

With regards to the 4-QAM; I'm not sure if it's an unrelated discussion but the idea behind this circuit was to try and send two audio channels, a left and right, by QAM to a receiver. A 4-QAM device would have to be used at a much higher frequency to send any sort of "good" sound if used digitally, so for now I'd just use the resultant quadrature output as the means by which to obtain a reliable 90 degree shift between the two waves, and do the rest using analog signal processing

Thanks for the ideas/suggestions/advice though, they've been a lot of help.

 

[tfr000]

Have you looked at polyphase networks?

 

[Baluncore]

Polyphase networks, or polyphase filters, are a stack of cross-coupled low-pass and high-pass filters, in other words, integrators and differentiators. Polyphase filters, tend to be insensitive to component variation, but they are quite bulky as they employ multiples of 8 components. They are usually cascaded and applied to broadband signals such as in SSB modulators that employ the audio phase shift method. Used where space is not a problem, they require differential drive and produce differential quadrature output, which makes two inputs and four outputs.


QAM is classed as a spread spectrum technique. I believe it is being used here as an example of how to get a data bit rate that exceeds the carrier frequency by a factor of from two to maybe six times on a quiet channel. The bandwidth required remains twice the carrier frequency.

 

[jim hardy]

Very first post said

For a school project I need to be able to shift the phase of a 100 kHz sine wave by 90 degrees (either positive or negative).

Has this been mentioned ? i wonder if a "bucket brigade" analog delay line device might be a useful mix of analog and digital technology for that task. Clock could be pll to a multiple of the incoming sinewave thereby avoiding frequency dependence of analog RC filters.

http://www.datasheetarchive.com/dl/Scans-091/DSAHI00040423.pdf

 

[Baluncore]

Then in post #7, the final application was specified.

Does you application demand the accuracy & stability of a crystal oscillator?

On that note, what is your application that it needs quadrature signals?

The end product will use a QAM signal to send two audio channels, and the device will need to be very small and low-powered so an LC oscillator would be too bulky. Other oscillators based on ceramic resonators could work, but the principal would be the same as a crystal oscillator I think, which is proving to be very difficult to design.

We do not know the exact wording of the original project specification.

 

[mheslep]

This has been a fun and interesting thread, but I have to confess that I haven't read all the posts in detail, so apologies if this solution has already been covered.

The way I've done this in the past is to use 2 ROMs with sine wave data in them that is offset by 90 degrees (for I and Q demod, but the technique works for encode as well). They are clocked together, and their outputs are driving R-2R ladder DACs with filter buffers. This gives you 2 very nice sine waves that are 90 degrees out of phase.

You can then use those waveforms for your 4QAM modulation circuit.

Works great with I and Q always separated 90deg even in unit number one million. The quantization noise is down 2^2b from the carrier power or, say, 48dB w 8 bits meaning, I think, that an analog osc is still required to realize very high SNR.