How does the number of carriers affect the peak to average power ratio and intermodulation levels?
Multi-Carrier Linearity Challenges
When multiple carriers share a common amplifier, their instantaneous voltages add coherently at random phase alignments, creating occasional peaks much higher than the average power. The statistical PAPR depends on the number of carriers and their relative phases. For N carriers with random phases, the CCDF of PAPR follows a Rayleigh-like distribution with peaks reaching 10·log10(N) dB above the average.
| Parameter | Class A | Class AB | Class F/Doherty |
|---|---|---|---|
| Max Efficiency | 50% | 50-78% | 70-90% |
| Linearity | Excellent | Good | Moderate (needs DPD) |
| P1dB Backoff | 0-3 dB | 3-6 dB | 6-10 dB |
| Complexity | Low | Low | High |
| Common Use | Test, small signal | General PA | Base station, broadcast |
Frequently Asked Questions
Does OFDM always have high PAPR?
Yes. OFDM is inherently a multi-carrier system. LTE with 1200 subcarriers has theoretical PAPR of 30.8 dB, but clipping by the DAC and CFR reduce the effective PAPR to 8-10 dB in practice.
How does CFR work?
Crest factor reduction detects signal peaks above a threshold and either hard-clips them, cancels them by subtracting a scaled error signal, or applies peak windowing. CFR can reduce PAPR by 3-5 dB with minimal EVM degradation (0.5-1% EVM increase).
Can I use separate amplifiers per carrier?
Yes. Using one amplifier per carrier avoids IM between carriers entirely, at the cost of more hardware. The outputs are combined using passive combiners. Each amplifier handles only one carrier's PAPR. This approach is common in satellite transponders.