Which wireless meat thermometer keeps its signal? 5 tested from barbecue to sofa

Our back door sticks in wet weather. You have to lift the handle and put a shoulder into it, and the whole house hears the thud. Last summer Rhian timed it on a Saturday night: once every twelve minutes, because that's how often I went out to check the barbecue.

The daft part is that I already owned a wireless meat thermometer. It worked a treat until I closed the lid. Then my phone on the sofa froze on one number for twenty minutes, and out I went again, letting the heat out and the rain in.

So this autumn I tested the question I actually wanted answered. Which smart meat thermometer keeps a steady signal from the oven or the barbecue to the sofa, and hits the target temperature within 1°C? From Wednesday 2 September to Wednesday 21 October, 50 days in all, five of the best-known models went into lamb, chicken, pork and beef at our terrace in Cardiff. One of them didn't drop out once.

A Victorian terrace makes a tough test

Wireless thermometers look great in adverts with glass-walled kitchens. Ours is a red-brick Victorian terrace. The kettle barbecue lives at the bottom of the yard and the sofa is in the front room, 15 metres away through two solid brick walls. From the oven to the sofa it's 9 metres, one wall and the steel box of the oven itself.

A lidded barbecue and a shut oven both behave like a biscuit tin for radio signals. Most of these probes talk over Bluetooth, and the signal has to squeeze out round the lid or the door seal. That's why nearly every model now comes with a booster, base or charging block that catches it near the cooker and passes it on.

The rules: every kit set up as its quick start guide says for the longest range. Anything with a screen came to the sofa with me, and relays without one sat on the kitchen windowsill, halfway between. I wasn't allowed up until something beeped.

On the barbecue, always lid down: lamb, whole chickens, pork shoulder, one brisket and plenty of thick steaks. In the oven: Sunday roasts, pork loin, chicken thighs. Every cook had two or three thermometers in the same joint or in matching pieces, rotated so each model met every other one. That made 31 cooks, 18 on the barbecue and 13 in the oven, and 15 for each thermometer.

Signal was half the question. The other half needed a referee.

How I checked the 1°C

The referee was a Thermapen ONE, the yellow instant-read thermometer made by ETI down in Worthing. It's quoted at ±0.3°C and I've used it for years at the rugby club barbecue. Before any meat went near them, I checked every probe, and the Thermapen, in a jug packed with ice and water, which sits at 0°C.

The Combustion read 0.0°C, the Typhur 0.1, the MEATER and both ThermoMaven probes 0.2, and the TempSpike 0.6. On day 50 nothing had moved by more than 0.2°C. Every probe arrived honest, so any later errors came from the cooking.

Each cook had a target: 52°C for pink beef, 57°C for lamb, 63°C for pork loin, 93°C for pulled pork, 75°C in the thickest part of a chicken. When an alarm went, I pushed the Thermapen in right next to the probe, at the same depth, three times. Within 1°C counted as a hit.

For the signal, I counted a dropout whenever the app or the screen showed a lost connection, or froze on the same reading for more than a minute. A frozen reading is worse than a blank screen, because you believe it.

Fifty days in one table

Each thermometer did 9 cooks on the barbecue and 6 in the oven. The dropout columns count cooks with at least one lost connection, not individual blips.

Two things jump out. The oven was harder than the barbecue for every model with a dropout, and accuracy followed price more closely than I expected. The two dearest thermometers were the only ones to hit 15 out of 15.

Typhur Sync Gold Dual: the one that never blinked

The Typhur Sync Gold Dual is two stainless probes and a slim black base with gold-coloured sides and a small colour screen. The difference is the radio. The probes talk to the base on Sub-1GHz, a lower frequency than Bluetooth that gets through brick and round a barbecue lid far more easily. Typhur claims up to 213 metres indoors. I had 15.

Across 15 cooks, lid down, oven door shut, Nel asleep across the doorway, the base kept ticking up without a single gap. On the first big Saturday, a 2.4 kg leg of Welsh lamb on Saturday 12 September, the TempSpike in the same joint lost its signal twice in the first hour. The Typhur base sat on the sofa arm and just carried on.

"That's the first Saturday this year I haven't heard the back door once." Rhian, Saturday 12 September, with a plate of lamb on her knee

Accuracy was just as calm. On average the Typhur's alarm went 0.3°C away from the Thermapen, and its worst miss in 15 cooks was 0.7°C on a pork loin. Like the MEATER and the Combustion, it has several sensors along the shaft and reports the coolest point it finds, which is the one that matters.

The base joins your Wi-Fi too. On Sunday 4 October I ran the Cardiff Half, and a pork shoulder went into a low oven at half five that morning with Rhian at home, because I'm not that brave. I checked it on my phone in the start pen by the castle. 71°C and climbing, if you're wondering.

The downsides: £199.99, a base that insists on 2.4 GHz Wi-Fi, and 5.6 mm probes, a touch chunkier than the MEATER's 5 mm. Its closest rival is cleverer still, and dearer.

Combustion Predictive: the cleverest, at a price

The Combustion Predictive Thermometer is the one the serious barbecue crowd talk about, and now I see why. Its probe has eight sensors from tip to handle, and the yellow WiFi Display, a chunky kitchen timer to look at, shows the core temperature and a countdown to dinner, no phone needed.

Over 15 cooks it was the most accurate of the five, 0.2°C from the Thermapen on average, with a worst miss of 0.6°C. The prediction is the clever bit. On that leg of lamb, two hours out, it said ready at 8:51 p.m. The lamb hit 57°C at 8:58. In the last hour of big cooks it was within two minutes.

Our nine-year-old calls the Display "the cheese" and keeps stealing it to time her tablet. On Saturday 10 October she put it on the sofa arm and announced the brisket would be ready at 9:20. It was.

Two cooks had a dropout, one on the barbecue and one in the oven, and the longest was 2 minutes 40 seconds before the Display caught up again. The catch is buying it here. Big Green Egg dealers sell the thermometer alone for about £160, but the range comes from the Display or a booster, and the only Display kit I found in a UK shop is a £330 set with two thermometers. We bought that and used one probe. Brilliant, and dear.

ThermoMaven P2: the bargain with a popcorn problem

The ThermoMaven P2 is the surprise of the test. Two probes, a chunky black base with a red rubber foot, a big clear screen with both temperatures, and Wi-Fi to your phone, for about £65 to £110 on Amazon.co.uk.

Its probes talk to the base over Bluetooth, so the base went on the windowsill. Eleven of 15 cooks went without a hitch. Two dropouts came in the oven, and the other two took me a fortnight to explain. Both were Saturday nights, and both times Rhian had just put popcorn in the microwave. Microwaves leak a little on the same 2.4 GHz band that Wi-Fi uses, and our microwave sits right under that windowsill.

With the base moved to the other end of the worktop, it never happened again. Longest gap 6 minutes, accuracy 0.5°C on average, and one miss of 1.2°C in a thin pork tenderloin. The next thermometer has the best app of the lot, and more trouble with my oven.

MEATER Pro: lovely app, shy in the oven

Most people have heard of MEATER first. The MEATER Pro is one sleek 5 mm probe, rated to 550°C on the outside so you can sear with it in, and a bamboo charging block that doubles as a Bluetooth repeater. The app is the best of the five: it tells you when to take the meat off and how long to rest it, and Rhian used it without asking me once.

On the barbecue it dropped out in two of nine cooks. In the first three oven cooks it went three times, once for 9 minutes. The fix was moving the block from the windowsill to the worktop beside the oven door. After that, three oven cooks without a gap.

Accuracy was very good, 0.4°C on average, with one miss of 1.3°C on a pork chop too thin to get the probe in past its safety line. It has no Wi-Fi, so away from home you need an old phone or tablet left running MEATER Link. At about £99 to £115 in the shops, under its £129 list price, it's still a lovely thing to cook with. The last model is the cheapest, and it showed.

ThermoPro TempSpike Plus: fine on the grill, lost in the oven

ThermoPro is the brand most people find on Amazon, and the TempSpike Plus is one thin probe with a black handle in a long orange booster that also charges it. ThermoPro quotes ±1°C for it, right on the edge of my target.

On the barbecue with the booster on the windowsill, it lost the signal in 4 of 9 cooks. In the oven it lost it in 5 of 6, and the worst gap was 22 minutes in the middle of a Sunday roast chicken. My phone sat on 61°C through a whole episode of Rhian's programme.

Its probe has one sensor at the tip for the meat. That's fine if the tip ends up dead centre, but in a chicken it's easy to go too far, and twice it read the hollow inside the bird, not the thigh. Average gap to the Thermapen 0.8°C, worst 2.1°C, 11 of 15 alarms within 1°C. For steaks on an open grill it's fine and cheap. For my sofa question, it wasn't the answer. Whatever you buy, a few habits help.

Five ways to keep any probe talking

• Put the relay close to the cooker. Near the meat, not near you. Moving the MEATER block 2 metres cured its oven dropouts.
• Mind the microwave. Wi-Fi bases can stutter right next to one.
• Push the probe in past the line. Too shallow and the reading suffers.
• Point the handle towards the house. On a kettle, line it up with the lid vent. In the oven, face it towards the glass.
• Check it in iced water once a season. It should read 0°C. Half a degree out, and you know before the Sunday roast does.

Every score here comes from the same routine and scoring sheet we use for all our kitchen tests, written down on our how we test page. Which brings me to the buying bit.

Which one I'd buy, and for whom

The TempSpike Plus suits steaks on an open grill with you stood next to it. If you're still choosing a barbecue to go with your thermometer, our Weber Spirit E-325 review covers gas, and the Traeger Woodridge Pro review covers pellets in a British spring.

UK prices and where to buy

We bought all five ourselves. Here's what each costs as I write this, and the price I'd wait for:

• Typhur Sync Gold Dual: £199.99 at Typhur's UK shop, also on Amazon.co.uk. Good under £170.
• Combustion Predictive Thermometer: about £160 on its own from Big Green Egg dealers, £330 for the kit with two thermometers and the WiFi Display. Combustion's own shop prices in dollars, so add delivery and import charges.
• ThermoMaven P2: about £65 to £110 on Amazon.co.uk. Good under £75.
• MEATER Pro: £129 list price at MEATER's UK shop, around £99 to £115 at John Lewis and kitchen shops like Borough Kitchen. Good under £100.
• ThermoPro TempSpike Plus: about £65 to £90 on Amazon.co.uk, sold by ThermoPro's UK store. Good under £60.

These often drop for Black Friday on 27 November. The Typhur stays on our sofa arm for the winter, and the Combustion Display is back in the drawer, mostly. My heated jacket test is still running for the dark dog walks, and you can follow it on our tests in progress page. Everything else we've tested for the kitchen is in the home and kitchen section.

Questions I get asked over the tongs