Thermoworks RFX First Impressions

[bwana_don]

Don, what are you doing to limit the reading to 3 minutes? All of mine consistently repot once a minute.

The only thing I did was set the temperature interval to 3 degrees for all of them. I do that before I take any of them out of the charger. I suspect at the moment, the termperature interval has reverted back to time, they way the first ones are shown in the videos over the summer. I also am not runing the ambient temp probe during the test right now. It also has a setting, maybe it would walk over them has well, I think it defaults to 1 minute. I don't want to mess up the current test I have going, otherwise I would see what happens changing them during the session. Once this test ends, I will be play around to see what caused them to be at 3 minutes this time, if there is a magic sequence, etc.

 

[bwana_don]

Just checked my test, at 29 hours, 1 is at 80%, the other 3 are 60%. I will be lucky if completes by on Tuesday. At this rate, my poorest performing probes will make 60 hours and my best performing probes will go 70 - 80 hours, maybe more. If they got it right, I would think they woud exceed advertised battery life, or they would have a lot of returns.

 

[Fraser]

I was curious about this so I just tested it and you are correct.

I blocked the RFX Gateway from my Wi-Fi through my Unifi UDM. Once I did, the temp updates stopped, even though I connected to the Gateway with Bluetooth to my phone via the Gateway settings prior to disconnecting the Wi-Fi.

Once I unblocked and reconnected the RFX Gateway to the Wi-Fi the temp updates resumed.

Based on that it appears the data path goes RFX Meat - RFX Gateway - Cloud - Device. There doesn't seem to be a path to the phone or app without going to the cloud.

In my commercial environment having Wi-Fi isn't an issue, but for someone who wants to use it as a stand-alone temperature probe in a remote environment without Wi-Fi it would be a problem.

For comp guys this is a big deal, which is why I think they’ll make the Gateway ‘dual freq’ soon. The only other option (which I haven’t tried) is to run the Gateway thru a local hotspot. Even so, that would decimate the range, all those other reasons to hate bluetooth.

 

[schlotz]

Just checked my test, at 29 hours, 1 is at 80%, the other 3 are 60%. I will be lucky if completes by on Tuesday. At this rate, my poorest performing probes will make 60 hours and my best performing probes will go 70 - 80 hours, maybe more. If they got it right, I would think they woud exceed advertised battery life, or they would have a lot of returns.

I can verify that by leaving the probes in the chargers then in the app set them to 3º does in fact change the reporting cycle to 3 minutes vs 1 minute. If you go to the graph selection you can see that each report is 3 minutes apart. I sure hope their claim of 50-60 hours was not based upon 3 min intervals. While 3 min might be ok for a long smoke, it certainly would be an issue for high temp searing, deep frying etc.

 

[SmoqueBBQ]

Just checked my test, at 29 hours, 1 is at 80%, the other 3 are 60%. I will be lucky if completes by on Tuesday. At this rate, my poorest performing probes will make 60 hours and my best performing probes will go 70 - 80 hours, maybe more. If they got it right, I would think they woud exceed advertised battery life, or they would have a lot of returns.

Are your tests being done under heat or just at room temp? I ask as mine were always done during actual smoker runs. Yesterday I decided to do one with the probes just sitting out on my desk with the Gateway left where it normally is.

The probe I've had the worst times with 9-10 hours when new down to 3-4 hours on my last cook is at 20% after 19.5 hours at room temp. The three other probes are at 20%, 40% and 20% after 19.5 hours when they've usually been dead or close to it after 12 hours.

It's very frustrating. I'm going to run them again tonight in the smoker (SPK-500) and in briskets and see if the times go back down again. It seems it's possible that the battery is severely limited under heat and having to go through stainless steel walls of a smoker. Or I could just have defective probes from the early batch.

 

[jimboc22]

Are your tests being done under heat or just at room temp? I ask as mine were always done during actual smoker runs. Yesterday I decided to do one with the probes just sitting out on my desk with the Gateway left where it normally is.

I had been leaving them out on a counter where the temp in my house would only fluctuate 10 degrees or so, which I knew would not be very indicative of real world use. I appreciated your comments with results from use while actually cooking. Even with them just sitting out on the counter I was only getting 8-18 hours until this last firmware update, so something was obviously wrong. If they were going to last 50+ hours during a cook, they should definitely last 50+ hours sitting out at room temp. I'm anxious to see your results during a smoker run with the new ones you receive soon.

 

[schlotz]

While conducting the forth test with probes set to 3º which actually is yielding 3 minute reporting vs 1 min, I
also had a long talk with Tech support. Bottom line, an RMA has been issued and I will be sending both probes and gateway back to them. Expect they will agree and send replacements, especially since the evidence is in their cloud that proves both probes are not performing as advertised. Going to let the forth test complete before sending it all back to them.

BTW: he was surprised that by setting the probes while in the charger to 3º actually caused the reporting to go in 3 minute intervals vs 1.

 

[bwana_don]

I can verify that by leaving the probes in the chargers then in the app set them to 3º does in fact change the reporting cycle to 3 minutes vs 1 minute. If you go to the graph selection you can see that each report is 3 minutes apart. I sure hope their claim of 50-60 hours was not based upon 3 min intervals. While 3 min might be ok for a long smoke, it certainly would be an issue for high temp searing, deep frying etc.

Just a reminder, this is what what was covered up on the probe documentation that I received.

Update Rate , 10 min. charge (80%) , 20 min. charge (100%)
Sear Mode, 4 hrs 30 min , 5 hrs 45 min
10 sec , 5 hrs 36 min , 7 hrs
15 sec , 8 hrs , 10 hrs
30 sec , 20 hrs , 25 hrs
60 sec , 52 hrs , 65 hrs

At least before they covered it up, it was based on 60 seconds.

 

[bwana_don]

Are your tests being done under heat or just at room temp? I ask as mine were always done during actual smoker runs. Yesterday I decided to do one with the probes just sitting out on my desk with the Gateway left where it normally is.

The probe I've had the worst times with 9-10 hours when new down to 3-4 hours on my last cook is at 20% after 19.5 hours at room temp. The three other probes are at 20%, 40% and 20% after 19.5 hours when they've usually been dead or close to it after 12 hours.

It's very frustrating. I'm going to run them again tonight in the smoker (SPK-500) and in briskets and see if the times go back down again. It seems it's possible that the battery is severely limited under heat and having to go through stainless steel walls of a smoker. Or I could just have defective probes from the early batch.

My tests have all been just on the table. Asked them early on when they were not performing well if that was a valid test. They said yes. When I mentioned that you were seeing battery degradation, they started throwing out disclaimers - mostly related to how people may not put them all they into to the minimum/recommend insert points.

I needed a starting point to verify they are going to do what I want, if they can't perform on the table, they will not perform in my smoker or pig roaster. My smoker would be relatively friendly from a temperature standpoint. Made it out of a Henny Penny 900 and it does not go above 220 degrees. I will be starting my penetration tests and distance after they succeed in simply duration tests. The smoker is double steel walls, always outside or in a garage. The pig roaster is a 250-gallon fuel tank that is quite a distance from my house. Those are the biggest reasons I bought them.

The only time I tried to use one probe for cooking was for 11 minutes at the end of a meat loaf cook. Turned off the gateway too soon, that is how I ended up with a zombie gateway.

Don

 

[bwana_don]

While conducting the forth test with probes set to 3º which actually is yielding 3 minute reporting vs 1 min, I
also had a long talk with Tech support. Bottom line, an RMA has been issued and I will be sending both probes and gateway back to them. Expect they will agree and send replacements, especially since the evidence is in their cloud that proves both probes are not performing as advertised. Going to let the forth test complete before sending it all back to them.

BTW: he was surprised that by setting the probes while in the charger to 3º actually caused the reporting to go in 3 minute intervals vs 1.

My tests are still going with all on the 3 setting. 56 hours, 1 at 40%, 1 at 20% and 2 at 10%. Speculation again. They have 2 development teams. One for the application, one for the web interface. One or both are handling the database side. Suspect they are colliding with each other. Could be they are divided by device, that might explain why the firmware updates do not have the same release numbering scheme. In any event seems like they have some software issues that need to be resolved so settings, etc. are doing as expected. Definitely need some consistency. Then it can be determined if the hardware is meeting expectations or not. We had a choice in one of the companies I worked at when it came to releasing software. You could either revert back to the old version a known commodity if there were issues or you could try to fix the issues in the latest release. The choice was always to fix the current, the mantra became - FAIL FORWARD.

Of course they are probably doing agile development and release small packages as needed. If they have appropriate tests systems and if they have enough containers for all the different packages to develop and if and the appropriate promotion plan and if it has all the appropriate user story tests automated they can do updates/new release on demand. That’s a lot of “ifs”. Oh, that also requires all the developers and engineers to play nice together.

Did I mention they covered up a timed based interval that they marketed this with for months. Then right as they were releasing it they switched to temperature based interval. That is a pretty big change and why it looks to me they are having the issues we are seeing right now.

Begs the question, who is managing the RFX Product – Marketing or Engineering?

Don

 

[SmoqueBBQ]

My tests have all been just on the table. Asked them early on when they were not performing well if that was a valid test. They said yes. When I mentioned that you were seeing battery degradation, they started throwing out disclaimers - mostly related to how people may not put them all they into to the minimum/recommend insert points.

I needed a starting point to verify they are going to do what I want, if they can't perform on the table, they will not perform in my smoker or pig roaster. My smoker would be relatively friendly from a temperature standpoint. Made it out of a Henny Penny 900 and it does not go above 220 degrees. I will be starting my penetration tests and distance after they succeed in simply duration tests. The smoker is double steel walls, always outside or in a garage. The pig roaster is a 250-gallon fuel tank that is quite a distance from my house. Those are the biggest reasons I bought them.

The only time I tried to use one probe for cooking was for 11 minutes at the end of a meat loaf cook. Turned off the gateway too soon, that is how I ended up with a zombie gateway.

Don

Seems reasonable to test it that way first.

They did ask me right away about whether I inserted them all the way. I told them I did read the instructions and was putting them fully up to the lines in briskets and the smoker temp is set to 225 so I’m way below what they’ve marketed. I even mentioned their promotional video of one sticking out of meat on a grill in open flame and that I’m not doing anything close to that.

If they are really that sensitive to heat if not inserted all the way, they probably should not advertise them to be good to 1000 degrees and deep fryer safe. I think they conduct heat way too much to be good to 1000 degrees.

I do have an RMA to send them back for replacement, but I’m running them tonight and may wait until I get the second order before sending them out.

 

[schlotz]

Decided to end my forth test after 25 hours. Both probes were at 60%. Obviously getting them to report at 3 minute intervals significantly increases the battery life. BTW: on the cloud site in settings the reporting duration can actually be adjusted all the way up to 5 minutes but in the app the limit is 60 seconds. Major discrepancy there PLUS the undocumented ability to set the probe temp setting to 3º which actually changes the reporting duration to 3 minutes. There's got to be lack of coordination within their development teams on those two fronts. Issue still remains at 60 second reporting where the life observed was in the 15-17 hr range and far from the stated 50-60+ hours. Will box everything up today then ship back with RMA#.

 

[bwana_don]

Seems reasonable to test it that way first.

They did ask me right away about whether I inserted them all the way. I told them I did read the instructions and was putting them fully up to the lines in briskets and the smoker temp is set to 225 so I’m way below what they’ve marketed. I even mentioned their promotional video of one sticking out of meat on a grill in open flame and that I’m not doing anything close to that.

If they are really that sensitive to heat if not inserted all the way, they probably should not advertise them to be good to 1000 degrees and deep fryer safe. I think they conduct heat way too much to be good to 1000 degrees.

I do have an RMA to send them back for replacement, but I’m running them tonight and may wait until I get the second order before sending them out.

Decided to end my forth test after 25 hours. Both probes were at 60%. Obviously getting them to report at 3 minute intervals significantly increases the battery life. BTW: on the cloud site in settings the reporting duration can actually be adjusted all the way up to 5 minutes but in the app the limit is 60 seconds. Major discrepancy there PLUS the undocumented ability to set the probe temp setting to 3º which actually changes the reporting duration to 3 minutes. There's got to be lack of coordination within their development teams on those two fronts. Issue still remains at 60 second reporting where the life observed was in the 15-17 hr range and far from the stated 50-60+ hours. Will box everything up today then ship back with RMA#.

 

[bwana_don]

With the RMA - are you shipping them back and getting a refund?

If not, is there an expectation that the new Gateway and/or probes will perform differently?

If so why?

You mentioned the cloud sight, I have been calling that the web interface, the app and the web interface are both dependent on the cloud and use the same data.

I had pointed out the web interface interval setting inconsistency you mentioned earlier.

1. The web interface implies it is for all devices.
2. The app interface implies it for the probe attached to the RFX Gateway.
3. If you change it to 5 on the web interface, that setting becomes available on the app interface.
4. I have asked tech support 3 times about what is true and received 2 different answers 1st and 3rd being the same. One answer is it is for how often the gateway reports data to the cloud, the answer and the one that I got the 1st and 3rd time is it is for the probe attached to the gateway. My best guess is they know what it is supposed to do per documentation, but since they have made some wholesale changes, it is probably doing something else and there is not a definitive answer.

I have recorded a number of inconsistencies between the web interface and the application. Some are annoying, some are unexplainable, some are a real problem. To date the ones I have shared I am told have been documented and turned over to development. With my latest test I have noticed more issues that I will be documenting and sharing.

My current test with all set to 3 is at 66 hours, well that depends on which probe you look at, 1 probe has magically lost an hour, can’t even phantom how the data could change in the middle of a session. 1 probe still at 20% the others are at 10%.

The 3 second reading is confirming for me that there are much bigger issues related to how often the probes are reporting and they are not working as intended.

They are going to need to convince me these things are going to work as marketed or mine will be going back for a refund before that window closes.

Don

 

[geek with fire]

You guys type faster than I can read. Sorry if this has already been discussed. But I have a theory that many of the problems talked about on this thread is a product of Tworks shifting gears on reporting intervals after the initial deployment. If you recall early on, the reporting period was based on time. But now it is based on degrees of temperature change. I would wager this could lead to a compatibility issue between older/newer probes with older/newer gateways. For folks like me who ordered a 4 pack, I wasn't surprised to see a large variance in serial numbers between the probe that came with the gateway, and the other 3 individual probes. Those last three are in sequence and are in the 15xxx range. The other one is in the 12xxx range. I'm rotating them each time I cook in efforts to try to isolate any problems with that single probe. None found yet. But primarily only grilling cook have been done.

A couple other things of note, I'm seeing a pretty huge difference in ambient temperatures as compared to to my signals probe. They they are side by side:

Possibly just an interval thing. But they are consistently over 10°F apart:

Then of course there is the temperature that RecTeq is reporting, which is worthless:

At the end of the day, I end up "measuring with my heart"...which sort of defeats the purpose of using bleeding edge technology.

Lastly, I have concerns that Tworks is turning into a digital rental outfit like all of the other software companies. They are charging for their cloud service; specifically when you exceed 6 devices on a single account. I'd love the ability to have some sort of software development kit that I could use to write an interface to access the near live data on the Gateway and Signals. There is no reason why this should not be possible, other than to hold the data hostage and charge you for access their cloud. I'd even pay for the ability to access the local data. But I'm not going to pay for data on their cloud, if I don't need it outside of my local network, when they are forcing the cloud model.

This turned into a rant...not my intent.

 

[bwana_don]

You guys type faster than I can read. Sorry if this has already been discussed. But I have a theory that many of the problems talked about on this thread is a product of Tworks shifting gears on reporting intervals after the initial deployment. If you recall early on, the reporting period was based on time. But now it is based on degrees of temperature change. I would wager this could lead to a compatibility issue between older/newer probes with older/newer gateways. For folks like me who ordered a 4 pack, I wasn't surprised to see a large variance in serial numbers between the probe that came with the gateway, and the other 3 individual probes. Those last three are in sequence and are in the 15xxx range. The other one is in the 12xxx range. I'm rotating them each time I cook in efforts to try to isolate any problems with that single probe. None found yet. But primarily only grilling cook have been done.

A couple other things of note, I'm seeing a pretty huge difference in ambient temperatures as compared to to my signals probe. They they are side by side:
View attachment 707749
Possibly just an interval thing. But they are consistently over 10°F apart:
View attachment 707750
Then of course there is the temperature that RecTeq is reporting, which is worthless:
View attachment 707751

At the end of the day, I end up "measuring with my heart"...which sort of defeats the purpose of using bleeding edge technology.

Lastly, I have concerns that Tworks is turning into a digital rental outfit like all of the other software companies. They are charging for their cloud service; specifically when you exceed 6 devices on a single account. I'd love the ability to have some sort of software development kit that I could use to write an interface to access the near live data on the Gateway and Signals. There is no reason why this should not be possible, other than to hold the di ata hostage and charge you for access their cloud. I'd even pay for the ability to access the local data. But I'm not going to pay for data on their cloud, if I don't need it outside of my local network, when they are forcing the cloud model.

This turned into a rant...not my intent.

What are you seeing that indicates near live data with the gateway? I have yet to see anything in my tests that indicates readings that are less than 1 minute apart, other than in the first couple of minutes of a session and they are all going to the cloud. Sear mode or equivalent not happening no matter what I have them set too. I do not see any readings that are temperature based.
I totally agree on the charging for the cloud service. I was going to buy a smoke x the day they announced the RFX. I m becoming so tainted with the RFX, I don't know if I would want to bother with the Smoke X.

 

[geek with fire]

What are you seeing that indicates near live data with the gateway? I have yet to see anything in my tests that indicates readings that are less than 1 minute apart, other than in the first couple of minutes of a session and they are all going to the cloud. Sear mode or equivalent not happening no matter what I have them set too. I do not see any readings that are temperature based.
I totally agree on the charging for the cloud service. I was going to buy a smoke x the day they announced the RFX. I m becoming so tainted with the RFX, I don't know if I would want to bother with the Smoke X.

Sorry, I vastly over-simplified my comment because my post was getting long. I'm certainly not seeing anything live. What I am saying is that if I had access to read the data on the Gateway from my local network, I could ensure that there are no delays being caused by the upload and processing at the cloud level. There are delays from the probe to the gateway. Certainly that will need to exist to save battery life. But that is the only physical delay that should be required. Live data from the gateway would be an easy win. But Tworks has a vested financial interest in requiring the cloud.

 

[SmoqueBBQ]

Last night's test in the smoker has some interesting results. I'm not even sure what to think anymore. lol

They must be tweaking stuff as none of my tests are consistent now. Probe 7089 which was 9-10 hours to 0% when I first got it, went down to 3-4 hours after a week of use, then 19.5 hours on my desk and now last night in the smoker it went 15 hours and was at 60%.

I do now have an issue with range that I didn't have until the last firmware update. Two of my probes that were in my smoker furthest away from the Gateway wouldn't communicate with the gateway. One did do periodic readings sometimes hours apart and one just didn't read anything at all.

The smoker they were in is a spk-500. If I draw a straight line to the gateway from the probes they would have to transmit through the smoker wall, through a c-vap, through a normal wall and through a hood on another smoker, which is about 22-25 feet. I would imagine it should be able to do this without issue, but regardless of which probes I put in that one, they either don't register now or only do so a few times per evening.

I'll keep using them for now before I send them back for replacement, but it does seem like TW is tweaking things here and there and it's making it difficult to get actionable results from testing.

 

[dr k]

I give you guys a lot of credit for the time spent on testing, patience to trouble shoot and perseverance with constantly being in contact with TW. I would have done the same for a few days if I ordered a RFX setup and been in contact with TW but would have sent everything back for a refund and let them do the R&D on their time to update firmware/software on a brand new product. My motivation to return everything is also driven by the need for TW to deal with brand new returns and their FB high praise posts of themselves several weeks old and videos that don't reflect reality. I commented on the FB RFX threads about these issues and SMF threads on these issues. No acknowledgement from customers or TW. I recommend a mass return for a refund and get packaging and instructions with the correct pics and no stickers hiding obsolete info. I think these need to go back to TW and they need to correct this on their time not yours.

 

[normanaj]

I'm in complete agreement with dr k on this one!