We test products that claim to block EMF.
The product is the test. Meters are instruments only. Every verdict comes from paired measurements — same source, same meters, same distance, same mode — judged against thresholds we publish in advance.
Change one thing. Measure twice.
The only fair way to judge an EMF protection product is to make it the single variable. We fix the source, the distance, and the meters — then the product either moves the number or it doesn't.
Baseline first
We measure the source — a router, a phone, a smart meter — with the product removed. Same distance, same operating mode, and every meter that can see the field takes its own baseline. A meter only counts when its baseline clears its resolution floor (TF2 ≥ 1, GQ ≥ 20 µW/m²) — below that, the meter can't see the field, so its pair is logged but not counted. A test runs on one meter or two. At the levels these tests produce (roughly 1.6–5 µW/m²) the GQ EMF-390 is usually under its own floor, so a TriField TF2 test on its own is a complete test — adding a second meter that cannot read the field would only add a pair that cannot count.
Add the product only
The product goes in exactly as its instructions say to use it. Nothing else changes. Each meter's baseline / with-product pair is its own test. Alongside it we run a no-product control — the same source read twice with nothing added — so the ordinary drift of the source becomes a measured noise floor, not a guess.
Verdict from the numbers
Each meter has its own Before / After pair and its own percent change. The headline is the average of those percent changes — never a blend of raw µW across meters, because readings on different scales don't add up. It's judged against the thresholds below and the control run, both set before testing, so no product gets a moved goalpost.
| Verdict | Measured change | What it means |
|---|---|---|
| ✅ Works | Reduction of 80% or more | A large, unambiguous drop. The product does what shielding physics says a working product should do. |
| 🟡 Partial | Reduction of 20–79% | A real, measurable effect — but check the exact number against the claim on the box before paying for it. |
| ❌ No effect | Reduction under 20%, or any change inside the control run | Within what normal measurement variation explains. We can't distinguish this product from doing nothing. An increase with no control run to justify it lands here too. |
| ⚠️ Backfires | The reading rose by more than the control run | It happens — partial shielding can make a phone transmit harder, and a poorly placed shield can reflect energy back. We only call it Backfires when a no-product control run proves the rise is bigger than the source's own noise; otherwise it's No effect. We publish these too. |
Recipes by product type.
Same paired rule every time — the product is the only variable. The setup follows the claim and how the box says to use it. No invented numbers here: these are the setups, not published results.
Router guard / mesh
Router on. You pick the distance and write it down — same mark for cover off and cover on, on the side you actually sit on. The TF2 Before and After at that mark is the test. Leave any box you did not measure blank.
Optional: a speed test from a client in another room, cover off and cover on, same client and same spot, with that client’s signal strength noted each time. A cover that cuts the field cuts the Wi-Fi with it, and that trade belongs in the report.
Phone pouch / case
Phone on a real call, each meter on the body side of the closed flap, at body distance (~1 cm) — the way the thing is actually carried. Baseline is the same call at the same distance without the product.
Shielding a phone is directional and partial, so the position is the test: a flap folded behind the phone instead of between the phone and you is a different test, and we say which one we ran. We also write down the bars and the signal strength on both runs, because a case that costs signal makes the phone transmit harder — a lower reading is not automatically lower exposure.
Paint / film / fabric panel
Source on one side, meters on the other. Baseline is an untreated same-size panel — not empty air — so the comparison is the material, not the gap.
Canopy
Meters at pillow height inside. Open versus closed. Note the seams — that's where a canopy usually leaks.
Match the meter mode to the claim
We only measure the field the product claims to affect. Harmonizers, stickers, and pendants still get a field reading — if the number doesn't move, that's No effect. Not a feelings test.
- RF claim → RF mode on the TriField TF2, and on the GQ EMF-390 when the field clears its floor. Each meter used gets its own pair.
- Magnetic claim → magnetic mode on those same two meters.
- Electric or grounding claim → electric mode on those same two meters.
Every published lab result names the meters used and lists each pair with its own percent change. The headline is the average of those percent changes — meters on different scales are never blended into one µW number. With one meter that average is simply that meter's own change, which is why a single-meter test is reported the same way as a two-meter one.
Document everything, explain the limits.
Every product test documents the full setup: the claim, the source and how it was driven, distance, meters and mode, each meter’s pair and its percent change, the no-product control run, the average headline change, what the product cost the radio, and the verdict. You can check our math against the thresholds above. We do not require RF, magnetic, and electric on every record — only the field the claim is about.
We avoid medical claims. We publish what was measured, how it was measured, and the limits of the setup. Readers can draw their own conclusions with complete context. New to the units? Start with our plain-English guide to what V/m, mG, and µW/m² actually measure.
Required fields for each product test
- → Product name, brand, and the claim on the box
- → Price paid and test date
- → Source and how it was driven (router on, phone on a call, etc.)
- → Meters used — one or both of the TriField TF2 and GQ EMF-390 — and mode, for the field the claim is about
- → Distance, and where the pair was taken (body ~1 cm as carried, 3 ft on the bench, pillow height inside)
- → Per-meter Before and After readings in µW, each meter’s percent change, the average headline change, and the no-product control run
- → What it cost the radio, where one is involved: bars and signal strength, or client throughput
- → Verdict
- → Setup photo when available
Suggest a product.
Name a product that claims to block EMF, include the claim on the box, and we'll buy it and put it through the paired test.
Suggest a product →