How We Test: Water Filters, Air Purifiers and EMF Meters

Clean and Native assesses products on two separate tracks, and every article says which one applies. Track one is first-hand: products I own, measured in my Palm Beach, Queensland home with the instruments listed on this page, with the raw readings published. Track two is everything else: products assessed on certification, manufacturer specification and Australian standards, never described as tested.

This page shows exactly what I own, what I have measured, the instruments I used and their limits, how specification-based assessments are done, how prices are checked, and how mistakes get corrected. If a claim on this site does not match this page, email hello@cleanandnative.com.au and it will be fixed.

7products and tools I own and use at home, listed in the evidence register below
6measuring instruments and apps, each with its stated limits
2evidence tracks, labelled in every article’s methodology box
0paid reviews or gifted units in exchange for coverage

The two evidence tracks

Track 1: measured first hand

Products I own and use at home. The article gives the instrument, the reading, the conditions, and the limits of the instrument. Where I have only used a product, without measuring it, the article says that too.

Track 2: assessed on evidence

Everything else. Assessed on independent certification, the maker’s specification sheet or manual, and Australian standards and regulators. These articles never say “tested”, and the methodology box at the top states what was and was not checked.

The line between the two is strict. A product I have read about is not a product I have tested, however good the specification sheet looks. Most of the products on this site are Track 2, because no one person can own every purifier, filter and meter in Australia. The honest thing is to say so, product by product.

Evidence register: what I own and have measured

This is the complete list of first-hand evidence on the site. If an article claims a reading that is not traceable to this register, it is an error. Filter by category:

ItemWhat I didKey readingsLimits
Water
EcoHero 5-Stage under-sink RO (Pure Water Systems)
In daily use since mid-2025. TDS measured at the tap and after the filter on 14 March 2026, averaged over 3 consecutive tests. First filter service cost $490.69 ppm tap to 3 ppm filtered (95.7% lower)TDS-3 meter, not calibrated. TDS shows dissolved solids, not specific contaminants. Full review
Water
Waterdrop C1S countertop RO
Bought on Amazon AU on 4 Nov 2026 for $379.22. Set up, flushed and used in my kitchen. Timed the 8 oz dispense setting.69 ppm tap to 7 ppm on a handheld TDS and EC meter (at 26.2°C); unit display 3 ppm; setup about 20 minutes; 8 oz in 20.50 sHandheld meter, not calibrated. Short-term use so far. Full review
Water
Earth’s Water shower filter
Owned and used in my shower. Collected a post-filter water sample and read its TDS.Post-filter sample 73 ppm, close to the 69 ppm tap readingShower filters are not designed to cut TDS, so this reading says nothing about chlorine reduction. No chlorine test recorded.
Water
Water test strips
Bought November 2026.No results published yetStrips give bands, not lab values.
Air
Coway Airmega 150
Runs in my 12.5 m² bedroom, in a house with a dog. Pre-filter photographed after a month. Noise measured on Silent at 1 m. Incense smoke clearance protocol run (method below). Clearance time not yet published.32.1 dBA on Silent at 1 m (Decibel X app)Phone app, not a calibrated sound level meter. One room, one unit. Full review
Air
Inkbird air quality monitor
Used for baseline readings in my kitchen and to track PM2.5 during the incense protocol.Kitchen baseline: CO2 762 ppm, PM2.5 10 µg/m³, PM1.0 6 µg/m³, TVOC 0.05 mg/m³, HCHO 0.052 mg/m³, 25°C, 69% humidityConsumer monitor. Optical PM sensors can read high in humid air, and TVOC and HCHO readings on low-cost sensors are indicative only.
EMF
TriField TF2 meter
Room-by-room readings in my home: router at several distances, hallway, pillow, couch, walls, oven, outlets and bedhead. Router now on a timer, off 9pm to 6am.RF: 3.058 to 4.380 mW/m² at the router, 0.521 at 2 to 3 m, 0.124 in the hallway, 0.048 at the pillow (peak 0.150), 0.032 at the couch. Magnetic: 49.8 mG at the oven door, 8.2 and 11.7 mG at walls. Electric: 940 and 997 V/m at outlets, 50 to 105 V/m at the bedhead, 2 V/m at the pillow.One home, spot readings. Electric field readings depend on how the meter is held. Full review

The register is updated when I buy, measure or stop using something. Products sent by brands are not accepted in exchange for coverage.

Instruments and their limits

Every instrument has an error band, and a reading is only as good as the instrument behind it. These are the tools used for Track 1 readings and what each one can and cannot tell you.

InstrumentMeasuresLimits I state in articles
TriField TF2RF (20 MHz to 6 GHz), AC magnetic and AC electric fieldsA consumer meter, not a lab instrument. Spot readings vary with position and time. Never described as calibrated.
TDS-3 pen meterTotal dissolved solids (ppm)Not calibrated. Shows the change in dissolved solids, not fluoride, lead, PFAS, chlorine or microbes.
Handheld TDS and EC meterTDS, conductivity, temperatureNot calibrated. Used for the Waterdrop C1S test.
Inkbird air quality monitorPM1, PM2.5, PM10, CO2, TVOC, HCHO, temperature, humidityConsumer sensors. Good for trends and before and after comparisons, not for compliance measurements.
Decibel X (phone app)Sound level (dBA)A phone microphone, not a calibrated Class 1 or Class 2 sound level meter. Used for one consistent reading at 1 m.
Water test stripsChemistry bandsSemi-quantitative. A laboratory test is the only way to confirm a contaminant level.
What none of these can measure: fluoride, PFAS, lead, arsenic, bacteria and viruses in water; filter efficiency at specific particle sizes; ozone output from ionisers. For those, the site relies on certification databases and published laboratory data, and says so.

How water filters are assessed

My test water. Palm Beach water is treated by Seqwater and supplied by City of Gold Coast, disinfected with free chlorine, and it is fluoridated. My tap reads 69 ppm TDS. Seqwater’s monthly reports put Gold Coast hardness at about 38 mg/L, which is soft.

First-hand protocol (EcoHero and Waterdrop C1S). Same tap and the same meter within each test: the TDS-3 for the EcoHero test on 14 March 2026, and the handheld TDS and EC meter for the C1S. I read the unfiltered tap first, then the filtered water, and for the EcoHero I averaged three consecutive readings. I record the date, the meter, and for the C1S the water temperature, because conductivity-based meters drift with temperature. Practical measures come next: setup time, how long a dispense takes, noise, and what maintenance costs in real money.

TDS meter reading 69 ppm in unfiltered Palm Beach QLD tap water
Tap: 69 ppm on the TDS-3 meter (not calibrated).
TDS meter reading 3 ppm after the EcoHero 5-stage reverse osmosis filter
After the EcoHero RO: 3 ppm, same tap and TDS-3 meter.
Handheld TDS and EC meter reading 7 ppm in water from the Waterdrop C1S countertop RO
Waterdrop C1S: 7 ppm on the handheld meter at 26.2°C.

What I cannot test, and how it is handled. Contaminant removal claims (fluoride, PFAS, lead, chloramine, microbes) need a laboratory. For those I check whether the claim is certified to the right NSF/ANSI standard for that contaminant (for example 53 for health contaminants including PFOA and PFOS, 58 for reverse osmosis, which also covers PFAS claims, and 55 for UV), and confirm the listing on the certifier’s own database where possible. A “tested to” claim from the maker is reported as the maker’s claim, not as certification. For plumbed products I also check for WaterMark certification and compliance with AS 3497:2021, the current Australian standard for drinking water treatment units. Results are compared against the Australian Drinking Water Guidelines.

How EMF meters and reduction strategies are assessed

First-hand protocol. Readings are taken with the TriField TF2 in the mode named in the article (RF, magnetic or electric), at the positions that matter for exposure time: where you sleep, where you sit, and close to each source. The router was measured at close range, at 2 to 3 metres, from the hallway and at the pillow. Each reading is reported raw, in the meter’s units.

TriField TF2 reading 3.058 mW/m2 RF at close range to the home modem
3.058 mW/m² RF at close range to the modem. At the pillow the same meter reads 0.048 mW/m².

Two reference points, always both. RF readings are shown against ARPANSA’s RPS S-1 limit, and mains-frequency magnetic and electric readings against the ICNIRP 2010 low-frequency guidelines that ARPANSA refers to. Separately, readings are shown against the building biology guidance (SBM-2008) that many readers ask about. They are not the same kind of thing. ARPANSA’s limits are health-based exposure limits. SBM-2008 is a precautionary guide for sleeping areas, not a health standard. For example, my router reading of 3.058 mW/m² is far below the ARPANSA limit but sits in the SBM “extreme” band for a sleeping area, which is why my router now runs on a timer, off from 9pm to 6am. Articles report both, and say which is which.

Shielding products. Where I have measured a shielding product, the article gives before and after readings at the same positions. Where I have not, it says so and reports only the maker’s figures and any independent test. Articles also flag the main trap with shielding a phone: blocking signal can make the phone transmit harder to stay connected.

How air purifiers are assessed

First-hand protocol (Coway Airmega 150). I run it in my 12.5 m² bedroom, check and photograph the pre-filter, and measure noise on Silent at 1 m with the Decibel X app (32.1 dBA). For smoke clearance I use an incense protocol: in a closed room, raise PM2.5 above 200 µg/m³ on the Inkbird monitor, start the purifier, and time how long it takes to fall below 12 µg/m³. That end point sits inside the Queensland Government’s “Good” one-hour PM2.5 category (0 to 12.5 µg/m³). It is a practical comparison, not a laboratory CADR test.

Coway Airmega 150 pre-filter after one month in a Palm Beach QLD bedroom with a dog, covered in dust and pet hair
Coway pre-filter after one month: dust and dog hair caught before the HEPA layer.
Decibel X app reading 32.1 dBA beside the Coway Airmega 150 on Silent mode at one metre
32.1 dBA on Silent at 1 m (phone app, not a calibrated meter).
Jayce Love holding an Inkbird air quality monitor in his Palm Beach QLD kitchen
Kitchen baseline on the Inkbird: CO2 762 ppm, PM2.5 10 µg/m³.

Every other purifier. Sized on smoke CADR using the US EPA sizing chart (about 65 cfm of smoke CADR per 100 sq ft, or 9.3 m², of floor, based on an 8 ft or 2.4 m ceiling). AHAM-certified figures are preferred. CADR measured to other methods, such as China’s GB/T standard, is reported but flagged as not directly comparable. Noise figures are quoted with the maker’s test method where one is given, because makers measure at different distances and speeds. Ionisers and plasma features are checked against the US EPA’s ozone guidance and CARB’s certification list, and articles advise switching them off where possible.

Specification-based assessment (Track 2)

When I have not used a product, the evidence is ranked, and the article uses the strongest source available:

  1. Regulators and standards: NHMRC (Australian Drinking Water Guidelines), ARPANSA, state health departments, Safe Work Australia, the WHO and the US EPA.
  2. Independent certification databases: NSF and IAPMO listings, AHAM Verifide, CARB’s certified air cleaner list, WaterMark.
  3. Manufacturer documents: manuals and specification sheets for the model sold in Australia, not the US or EU version, where these differ.
  4. Retailer listings: used only for price, stock and seller, and labelled as listing claims when nothing better exists.

Customer reviews, influencer posts and press releases are not treated as evidence of performance. Where the only support for a claim is the maker’s own testing, the article says that in plain words.

Prices and availability

Amazon AU prices are pulled from Amazon automatically and refreshed through the day, and each one shows the date and time it was last updated. Prices from other retailers are checked on the day an article is written or updated, and carry that date. Prices change often, so always check the seller before buying. The site does not track or chart prices over time, and does not show Amazon star ratings or review counts. Links are checked on a rotating schedule, and dead or out-of-stock products are replaced or flagged.

Fact-checking and corrections

Before a new or rebuilt article is published, it goes through an independent fact-check: a separate checking pass that did not write the article verifies every number, specification, price, quote and study against its primary source, and reports anything wrong, overstated or unsupported. Every problem is fixed or the claim is cut before publishing. AI tools are used for research, drafting and checking, under the rules in the AI and Editorial Policy. A person decides what is published, and first-hand readings come only from my own measurements.

If you find an error, email hello@cleanandnative.com.au with the page and the claim. Factual corrections are made to the article, and significant ones are noted on the page.

The approach comes from my Navy Clearance Diver training, where you trust the instrument over the sales pitch, state exactly what you know and what you do not, and check your equipment before you rely on it. That is the standard this page holds the site to.

Scores

Where an article gives a score, products are rated out of 10 on performance, value, suitability for Australian conditions and ease of use, with performance weighted most. For a product in the evidence register, performance reflects my readings. For everything else it reflects certification, specification and Australian standards, and the article states which basis applies.

ScoreWhat it means
9.0 to 10Best in class for Australian conditions, with no significant compromises.
7.5 to 8.9Excellent, with one or two caveats that are clearly stated.
6.0 to 7.4Good for some buyers and specific uses.
4.0 to 5.9Below average. Only worth it if nothing better fits.
0 to 3.9Do not buy. Performance does not justify the price.

Independence and affiliate disclosure

I buy the products I test with my own money or the site’s commission income. No brand has paid for a review, supplied a gifted unit in exchange for coverage, or had input into a rating or conclusion. Products are chosen for coverage based on what Australians can actually buy, what they search for, and whether the product is genuinely different from what is already covered, not because a brand asked.

Clean and Native earns commissions when you buy through some links, at no extra cost to you. Programmes in use include Amazon Associates (Amazon AU), Pure Water Systems, Waterdrop Australia, SaferEMF, Earthing Oz, Premium Grounding, HolyH2O, Waters Co Australia and TAPP Water, plus programmes through the Commission Factory network. Affiliate links are marked as sponsored, and every article with affiliate links carries a disclosure. The rule is simple: the best product for the reader gets recommended whether or not it pays a commission, and where the best option has no affiliate programme, it is still the recommendation.

When a product is discontinued, delisted or replaced by a new model, the article is updated rather than left recommending something you cannot buy.

Frequently asked questions

Do you test every product you recommend?

No. Only the products in the evidence register on this page have been used or measured in my home. Everything else is assessed on certification, the maker’s specification and Australian standards, and each article’s methodology box says which applies.

Are your meters calibrated?

No. The TriField TF2, the TDS meters, the Inkbird monitor and the Decibel X app are consumer instruments. They are used for consistent before and after comparisons in one home, and articles state their limits rather than presenting them as laboratory results.

Can a TDS meter tell me if my filter removes fluoride or PFAS?

No. A TDS meter shows the total of dissolved solids. A large drop, like 69 to 3 ppm after reverse osmosis, shows the membrane is working, but specific contaminants need a laboratory test or a certification to the relevant NSF/ANSI standard.

Do brands pay for reviews or send free products?

No brand has paid for a review or supplied a unit in exchange for coverage. The site earns affiliate commissions on some links, which are disclosed, and recommendations do not depend on whether a product pays a commission.

How do you compare EMF readings with safety limits?

Every RF reading is shown against ARPANSA’s RPS S-1 limit, and mains-frequency magnetic and electric readings against the ICNIRP 2010 low-frequency guidelines that ARPANSA refers to. Both are separate from the SBM-2008 building biology guidance, which is a precautionary guide for sleeping areas. Articles say which reference each comparison uses.

How current are the prices?

Each price is checked on the day the article is written or updated and shown with that date. Prices change, so check the seller before you buy.

How do I report an error?

Email hello@cleanandnative.com.au with the page and the claim. Factual errors are corrected in the article.