Why the Box Number Is Not Your Speed
A powerline adapter that says "2000" on the box has never, in practice, delivered 2,000 megabits per second to anyone's computer. That number is the theoretical PHY rate: the raw signalling speed of the chipset under laboratory conditions, measured across all pairs of your household wiring, in one direction at a time, before any overhead is subtracted. Real throughput — the TCP speed your devices actually see — is a fraction of it.
Three things eat that number before you see a single byte. First, powerline is half-duplex, so the wire carries data in only one direction at a time. Second, error correction and protocol overhead consume a fixed slice of the signalling rate, and on noisy home wiring that slice grows. Third, and most decisive, is the wire itself: your home's electrical circuits were laid to carry 50 or 60 Hz mains current, not high-frequency data. Their quality varies enormously from house to house.
The Four Factors That Shrink Throughput
A realistic estimate starts with the advertised figure and multiplies it down by four reality factors:
- Adapter class. HomePlug AV 600, AV2 1000, AV2 2000, and G.hn 2400 are different generations. Older classes start from a smaller number and many AV 600 units carry only 100 Mbps Ethernet ports, which hard-caps everything downstream of them.
- Wiring age and quality. Modern copper wiring carries the signal far better than decades-old wiring with splices, corrosion, and mixed materials.
- Circuit path. Two adapters plugged into the same room, on the same circuit, lose the least. Different rooms on the same floor lose more. Different floors, with the signal crossing the breaker panel and possibly another circuit breaker, lose the most.
- Electrical noise. Phone chargers, dimmer switches, motors, and cheap power supplies inject noise onto the wiring. In a noisy environment, the adapters spend much of their bandwidth retransmitting corrupted frames.
Worked Example: A 2 Gbps Adapter That Delivers 420 Mbps
Take an AV2 2000 kit — a respectable mid-generation pair — on modern wiring, with both adapters on the same floor, medium electrical noise, and a 500 Mbps internet plan. The estimator applies the four reality factors to the 2,000 Mbps advertised figure and lands at 420.0 Mbps: just 21% of the box number. The verdict is still "good — fast enough", because 420 Mbps uses 84% of the 500 Mbps plan. The adapter is not the bottleneck here; the internet plan is.
The picture changes at the other extreme. An AV 600 kit on older wiring, across different floors, in a high-noise home on a 100 Mbps plan yields about 37.8 Mbps — 6.3% of the advertised 600 Mbps — and the verdict is "not enough". The 100 Mbps Ethernet ports on that old kit are themselves part of the ceiling. And at the top end, a G.hn 2400 kit on modern wiring in the same room with low noise can reach 960.0 Mbps, 40% of its advertised figure, which saturates a gigabit plan almost entirely.
Rule of thumb: expect 10 to 25 percent of the advertised powerline figure in real use. If your internet plan is 100 Mbps or less, even a modest result usually still uses the whole plan.
Powerline vs Mesh vs Ethernet: What to Use Where
No single technology wins everywhere. Use this comparison to pick by room, not by hype:
| Option | Real-world speed | Latency | Best for |
|---|---|---|---|
| Wired Ethernet | About 94% of your plan | Lowest, around 1 ms | Desktops, gaming PCs, TVs that never move |
| Powerline adapters | 10–40% of the box figure | Low to medium, 3–8 ms | Extending the network through thick walls where Wi-Fi dies |
| Mesh Wi-Fi | 50–80% of your plan per hop | Medium, 10–25 ms | Whole-home roaming, rented flats, devices that move |
| Wi-Fi range extender | 25–50% of your plan | High | One stubborn dead zone, last resort |
Practical Tips for Faster Powerline Networking
- Plug adapters directly into wall sockets — never into extension cords or power strips, which halve throughput instantly.
- Keep both adapters on the same electrical circuit when you can; avoid crossing the breaker panel.
- Move phone chargers and other noisy devices to a different outlet, or filter them with a mains noise filter.
- Buy AV2 or G.hn kits with gigabit Ethernet ports; AV 600's 100 Mbps ports are a hard ceiling.
- Keep the Ethernet runs short on both ends — every cable segment adds loss.
- Update the adapter firmware; vendors do ship fixes for chipset bugs.
- If speeds collapse suddenly, re-pair the adapters and reboot them; a noisy device may have knocked them onto a slow modulation.
Try it now: Powerline Adapter Throughput Estimator — enter your adapter class, wiring, noise, and plan speed to see a realistic throughput estimate with a plain verdict.
Frequently Asked Questions
Why does my AV2000 adapter only deliver 150 Mbps?
Because 2000 is a theoretical signalling rate, not throughput. Older wiring, a different-floor circuit path, and electrical noise each take a cut, and typical results land at 10–25% of the box figure. An AV2000 at 150 Mbps is underperforming — check the outlet pairing, noise sources, and firmware before replacing it.
Is powerline networking good enough for online gaming?
Often yes, with a caveat. A healthy powerline link adds only 3–8 ms of latency, which games tolerate well. The failure mode is jitter: noise spikes cause lag bursts. If the estimator verdict is "good" and the link is stable, gaming works; if the verdict is "tight", wire the gaming PC with real Ethernet instead.
Should I choose powerline adapters or mesh Wi-Fi?
Choose powerline when Wi-Fi cannot cross something — thick concrete walls, metal-lined floors — and both devices stay put, such as a TV or a fixed PC. Choose mesh when devices roam and the house is open-plan enough for the signal to travel. When both are viable, real Ethernet still beats both.