ChefChrisYoung

@FlashJackso glad to hear this. And, yes, we're going to try to use that subreddit as a place to share information and answer more general questions about the Predictive Thermometer and how best to use it.

That is strange, and makes me wonder if you have a leaky microwave? If things are connected, and you turn the microwave on for a minute and it causes the disconnect, I would guess that’s the culprit. A somewhat surprising fact about kitchen appliances like microwaves and blenders, is that they are not required to pass an FCC emissions test and can radiate any amount of radio frequency noise. There are standards are how shielded microwaves are *supposed* to be, but in my experience they often aren’t as good as they’re supposed to be. When the motor in a blender comes on, or a microwave leaks, it can dump a ton of high energy noise in the Bluetooth spectrum, drowning out other low-power signals.

Of course, and thank you for being a customer. I think that we’ll ship the last of the pre-orders tomorrow or Wednesday; so if your’s isn’t already on its way, it will be soon.

Glad to hear that True Core did it’s job in the meat balls. I’m a bit surprised that the microwave caused problems. Your microwave does operate at close to the same frequency at Bluetooth, but shielding should keep the RF from leaking out. This said, was the display or probe pretty close to the microwave?

Ceramic Kamado-style grills are reasonably transparent to a Bluetooth signal, and I can typically get 20m or more from my Big Green Egg with line of site to the display or my phone. If the probe is inside a metal grill, the range is reduced a lot and difficult to predict, but more typically 3 to 5m before the signal needs to be repeated by the display. Soon we’ll have some additional accessories to extend the range further by being additional repeaters that expanding our mesh network further. Ultimately we’re trying to ship products with connectivity that “just works” and WiFi isn’t really compatible with that because there are too many variations on how it can be configured by the user. HomeKit probably isn’t something we can really take advantage of to do this since it’s really a framework for integrating the control of a device like a light or speaker, rather than a means of routing data around (including up and down from the cloud servers). That’s where Matter is more useful, and specifically it’s Thread protocol.

@mgaretzthe repeater functionality of the display is deployed. I think you’re using the beta-testing app, which does not support the mesh networking. But the public iOS app does work with the display as a range extender. The Android app should (finally) be released on Monday or Tuesday. There will also be an update to the iOS app, although it’s mostly a bunch of behind the scenes infrastructure needed for the next app update. In February we should be releasing major app updates that will add some critical features: graphing, raw sensor output, data exporting, and a bunch of battery life optimizations. WiFi isn’t on the current roadmap. Based on my experience with Joule and ChefSteps, WiFi is incredibly costly for the company to maintain and provide customer support. Every router is a snowflake, and users often have them configured in ways that don’t reliably work. What is on the roadmap, that gets to a similar end result, is the ability of a mobile device (phone or tablet) to route a signal to the cloud so that you can check-in using another device. We’re also looking to support the Matter protocol, which will allow our thermometer to connect to the internet via WiFi devices that support the protocol, like Alexa speakers or newer AppleTVs. But before we tackle these jobs, we’re going to be focused on dialing in the core functions like the Prediction Engine and better battery life management.

Haven't been by this thread for awhile—and my apologies for that. AFS: I can answer your questions. The Predictive Thermometer's sensor tube is made from stainless steel. The rear charging contact on the handle is stainless steel. The handle itself is aluminum oxide ceramic. The internal seals are a fairly exotic high temperature silicone. There is a screw in the handle that is nickel-platted copper, which is directly connect to a precision thermistor in the handle—this screw acts as a heat pipe to measure the ambient temperature around the boundary layer of the food. And, yes, it's fine on an induction stove top. Regarding the predictions, based on the data customers are sending us from our beta-testing program, they seem to generally working pretty well for higher temperature, or relatively quick cooks (pan-roasting). At 30% of the way to the target temperature, the average error across all cooks in our database is just over 11%, but 50% of the way to done the error averages a around 8%, and by the time we're at 90% of the way to done the average error is usually less than 2% and converges nicely to rarely being off by more than a few seconds. The caveats here are that if you flip the food, change the temperature, move the coals around, etc this changes the speed of cooking. The algorithm does a pretty good job of handling these curve-balls and will just recalculating. That's probably about the best we can hope for since we can't actually anticipate what you're going to do on the fly. There is a known bug that is tripping up the algorithm when the heating rate is particularly slow, such as sous vide and sometime low-temp smoking or reverse sears. This should get a fix in one of the upcoming firmware updates (probably the mid-February update). There is also a bug with the instant read filter on some thermometers causing a bit too much oscillating, or sometimes a bit of an undershoot. This seems to be down to some very small variations with the internal sensor position during manufacturing. We're working on retuning this right now using a much bigger sample size of thermometers. Finally, resting predictions are still a ways off. We've got enough data to have a pretty good idea on how we're going to solve it, but we're going to need a much bigger data set of cooks to do it well. To get to that, we need to finish adding cloud connectivity to our apps, so that we can make it easy for our customers to "donate" their data anonymously to help train the algorithm. I'm hopeful that we'll get there before summertime. But our team is pretty small and there is a lot to do. A few folks have commented on the lack of formal reviews. I think that this is mostly because we have not set out any samples to reviewers yet (yes, we're getting asked a lot). Reviews have a pretty big impact on a new product like this, and I made the decision that I would prefer to have a few months after scrambling to ship in December to work out the inevitable bugs. So, right now, what we've been focused on is reading as much feedback as our customers give us, using this to prioritize what we're working on improving, and generally stabilizing production and our supply chain. Feel free to ask questions or make some comments in this thread. I'll try to keep an eye on it. You can also always email hello@combustion.inc. I read nearly all of those emails myself and make sure that feedback or concerns gets responded to. --Chris Young

Yes, the minimum 2 inch insertion depth is designed to protect the battery and sensitive electronics that are all stuffed up in the tip. They can safely operate to 100C/212F, so you can use it without inserting to that minimum as long as you keep the first 2 inches of the probe below 100C/212F. One thing I'm excited about it using it as a quick instant read thermometer that automatically finds the lowest temperature neat the tip. In my experience it's pretty easy to not find the exact core with a conventional thermometer, and my digital signal processing and applied math colleagues have been doing a wonderful job building very good instant read algorithms that also interpolate between the first three sensors to find the the lowest temperature so that you won't miss the true core temperature as long as you're kind of close.

Supply chains are a mess. Shipping is a mess. And new products are always the lowest priority for everyone. It only takes one component being delayed to up-end everything. Moreover, everything is a lot harder and slower because travel is restricted or impossible. Normally we'd have our engineers on the factory floor working through issues with our manufacturing partners; that isn't possible right now, so you have to FedEx things around and use Zoom calls to figure things out. Things that could be solved in a day can easily take a week because you can't all be together on the spot.

Launch will be when I'm confident that manufacturing won't become a disaster. Price will be when I'm confident that all of the costs are nailed down. I would like both of these things to be as soon as possible.

Darch: Yes, bimetallics, thermocouples, and other temperature sensing technology generate a small amount of voltage differential from the thermoelectric effect, and if you combine a large number of these junctions you build a Peltier device that could, in theory, harvest some heat to charge the battery—or even eliminate the battery altogether. But in practice it's a finicky and relatively expensive technology that isn't particular good at being miniaturized. It's what I would call a research project. Super capacitors and ultra capacitors are quite interesting, and are something we looked at closely. They weren't a good fit: the temperature limits are not adequate, to get to 3.3V you need to chain several together, which makes them too bulky, and they are quite expensive. This is likely to change over the next 5 years, but this is still a niche technology. In terms of charging quickly, you only get the benefit if you can supply enough current at a high enough voltage from the charging source. Since our charger uses common AAA alkaline batteries, we would have to kill the alkaline battery life to charge a capacitor at a very high rate. As it stands, our battery takes about 10 minutes to get an 80% charge and ~20 minutes for a full charge. In testing, we're seeing around 30 hours of battery life from the probe. Your concerns about deep discharging are very real for many battery chemistries like lithium polymer, but not a concern for the chemistry that we're working with. And the long-term capacity loss from charging/discharging cycles is minimal. TL;DR there are a lot of ways you could power a probe, we've looked at this closely and selected an approach that balances cost, maturity of technology, safety, and overall consumer experience.

Adey73: Each probe logs all of its data internally once per second, and it will be possible to use the app to save or offload those temperature logs. I have a bunch of ideas for cool things we can let people do involving log files, but they probably won't make the cut for our first release. The timer will probably support 8 probes. Maybe a couple more, maybe a couple less, technically we could support thousands, but the trade-off is memory and battery life for the timer. This is something we're actively optimizing right now, and 8 looks like the right balance of trade-offs. Speaking of battery life, the probe is quite a challenge. We use a somewhat exotic rechargeable chemistry that has the very nice properties of handling very high temperatures and not doing anything unpleasant if it does get too hot. The downside is the energy density it really low—like terrible. To get the probe a lot thinner, we had to go with an absurdly *tiny* battery. This gives us a few mAh of power in the probe, so to get 24+ hours out of the battery we just have to be somewhat clever with how we power up the sensors, process the measurements, write it to memory, and transmit it over Bluetooth. The probe has no WiFi simply because the power requirements are much too high. One happy accident was that while designing our antenna, our RF engineer had the very clever idea to use the food as part of the antenna to get more transmission efficiency at lower power. As for the concern about nasty things leaking out of the probe... Everything we used is ROHS compliant (so nothing really nasty in our electronics) and a leak would be exceedingly unlikely. Once we bond the ceramic handle to the stainless steel tube, it's not coming apart again.

Hi folks, Chris Young here. It took me a little while to get validated. There are a lot of good questions here, and I'll try to answer them as best I can. First, how does the predictive algorithm work? What about bones, fat cap, fiber orientation, salt, etc. I think the easiest way to answer this is that what it's doing is predicting how long it will take the coldest location along the thermometer to reach the target temperature that you set on the timer or in the app. So if you poke the thermometer in the wrong place and miss the geometric center of the food by a lot, it won't do a great job. That said, it's not really difficult to stick a thermometer in the thickest/deepest spot in the food. Heat comes from all sides and very well may reach the center faster/slower from different directions. But in practice, we find that the algorithm having even a rough idea of what the temperature is at the surface and immediately around the food gives it a huge advantage at predicting what will happen in the center later as the heat propagates. Sure, the surface is hotter under a fat cap or around a bone, but it's not wildly different. And so the algorithm can constantly check it's predictions and adjust as it goes, so that if the center is heating faster than it initially expects, it will adjust it's internal model and update the timer accordingly. And the closer it gets to done, the more accurate it becomes. And by having some sense of how much heat energy has accumulated at the surface and beneath the surface, we can do a good job of estimating the amount of carry over cooking that will occur. Second question, how do the battery and other electronics survive inside an oven? These components are in the front ~50mm or so of the probe, and are buried beneath the surface of your food. Because food tends to be wet, once they're even a few mm below the surface, the temperature they experience is below 100 °C / 212 °F and this keeps them from exceeding their operating limits. The other end of the probe only contains components that can survive at 300 °C / 570 °F. Third question, doesn't heat travel down the meal probe and mess up the measurements. A tiny bit, but not much. The probe is stainless steel, which is a pretty terrible conductor as metals go. And it's very thin walled, so heat doesn't propagate far before it's absorbed into the surrounding food. So the sensors do a good job at only measuring the temperature immediately next to their location within the probe. WiFi/Blueooth questions. The probe uses Bluetooth. Right now it support the BLE 4.0 and BLE5 (benefits of BLE5 are increased range and better battery life). The timer has both Bluetooth and WiFi and can repeat the Bluetooth signal to your phone if you're further away, or it can route the data over the internet with WiFi so you can get it via the app when you're out and about. To be clear, you do not need to use your phone or connect any of this stuff to the internet with that isn't your thing. If you just want to stick the probe into some meat to measure the temperature, the timer will show you the temperature. And, yes, you can also just use the timer as a kitchen timer.