Spread Spectrum?

Y'er right. If anything, it makes the problem worse. Spreading the carriers does little to reduce interference. The problem is that such oscillators don't dither the frequency enough to make much difference. For example, the various Maxim spread spectrum oscillators dither up to +/-2% of the nominal clock frequency. The typical FM broadcast receiver has a 200KHz wide IF (HD Radio has

The other problem is that there is fairly limited number of places where a spread spectrum clock oscillator is going to work. Dithering the clock of a sound card is going to make the audio sound like gargling ball bearings. Doing the same to video will smear the image. Anything that has to track a narrowband signal is going to see the dithering as additive noise. A modem constellation diagram and eye pattern are not going to look very pretty with a dithered clock.

My contention is that dithering actually makes things worse.

Agreed. The total power delivered is the same whether dithered or a single carrier. If the dithered signal lands in the middle of the IF bandpass of a receiver, then the total interference potential is the same as it would be a single carrier.

Maxim shows +/-2%. However, this one from Linear claims +/-10%

Yep. My office has about 30 assorted devices floating around. Even the motherboards in their boxes on the shelf radiate RF from battery operated realtime clocks. I also have a collection of 2 way radios and scanners operating when I'm in the office. Every day, a new mystery carrier or noise that has to be tracked down. Usually, it's some customers equipment that's in for repair. I haven't seen any dramatic increase in interference since the introduction of spread spectrum clock oscillators.

That's the problem with one and two sentence questions. Without intent, context, and detail, it's rather difficult to deduce what is being asked. However, since I enjoy a good mystery, it doesn't bother me much.

Incidentally, while OFDM usually involves spread spectrum, there's no requirement that it actually use spread spectrum. What OFDM does is divide the data stream into a number of channels, each with a slow speed part of data stream. If one of the channels gets clobbered by interference, or more likely by frequency selective fading from multipath, then the data continues to flow, although at a slower rate.

Who considers OFDM to be spread spectrum ???

While DVB-T/T2/C2 services (OFDM+subcarrier interleave+time inrerleave+ECC) might be considered spread spectrum, what is the real difference to single carrier time interleave+ECC systems ? All rely on converting burst errors to random errors that can then be corrected by the ECC. It would be more appropriate to talk about spread time communication :-).

Ummm.... 802.11a and g are OFDM, LTE is OFDMA, WiMax is OFDMA, etc. There's a list of acronyms that use OFDM at: Everything that equalizes the power density across the channel bandwidth (as in the multiple carriers used in OFDM) is spread spectrum. Or more crudely, if you don't see any discrete inband carriers, it's probably some form of spread spectrum. Oh, you asked who. Well, I do, as does the FCC, 3GPP, IEEE 802.11 working group, and anyone that has ever seen LTE or 802.11a or g spectra on a spectrum analyzer.