TiltBridge - Tilt-to-WiFi Device for Tilt Hydrometer

PM sent.

 

Hello, I have manually installed tiltbridge onto a Lolin D32. I can access it through my web browser and see correct readings. I have followed the cloud service setup guide to configure the fermentrack URL. I can add it as a tiltbridge in Fermentrack (it sees the mDNS, I get a message saying the tiltebridge url is configured) and can use that to add a tilt (connection type = tiltbridge). Once I have added it Fermentrack does not show any values. If I go to manage sensor the connection type is Bluetooth. If I try and add another tiltbridge it shows no unbound tiltbridges are available until I delete the current sensor.

Please let me know what other infomation I can provide.

Thanks

 

Are you on master or dev?

 

I'm on Dev.

I left it running overnight and it started logging eventually.

 

In my case, it appears to have flashed correctly but while the display is lit, there's no text. I did see it broadcasting on WiFi and started the setup to connect to my router, but seemed to lose the connection before I finished. Now it's no longer broadcasting it's SSID. Any help is appreciated!

 

I recently bought a Tilt and stumbled upon Thorrak's excellent Tiltbridge documentation. I happened to have a Lolin D32 laying around and so I had a flashed and running Tiltbridge of my own in very short order. I can open its web server and see it connected to my Tilt and reading out coherent values. What is not happening is no data is being sent to my Google Sheet. I went through the initial setup instructions twice but the data never makes it to the Google sheet. I know that there are so many potential places that could be the problem but I cannot figure out a way to even test it. Any suggestion on how to debug? If there was a simple python script where I could paste in the Google Sheet URL and get back error messages (or perhaps it works), that would be very helpful. Also, a tab on the tiltbridge web server with an error log would be useful so you could get some feedback on what is or is not happening.

I am really amazed that you could get a single ESP32 to be a bluetooth-to-WiFi bridge.

 

Of course, my hardware shows up 12 hours before I leave town for a week! I'm be attempting this soon though!

 

Plenty of time to get it running!

 

Got mine flashed and up and running. Initial setup was pretty easy. But I have an issue accessing the web page of the device once it is on my wireless network. The d32 has an IP that it shows on startup. It sees the TILT no problem. I can see the device from my router device list, and I can ping the IP so it is on my network. But when I connect to the tiltbridge IP with http:// there is just a blank screen in my browser.

 

What browser are you using?

 

Chrome, Firefox, and IE

All the same result

 

That actually sounds like the SPIFFS didn’t flash properly.

 

Forgot to mention that I did try flashing again. After that I cleared the IP using the touchscreen and I could get into the the web interface to connect back to the wireless network. But once on the network again -- nothing.

 

Seems like this *might* be that bug I keep hitting (and none of the lib devs believe it) where the AP doesn't seem to completely go into STA mode. Try to erase the flash, then re-flash the firmware and reconnect.

 

Oh how I wish that this was advanced to plug and play status! I'm mechanically savvy, but not electrically or systems-wise. I see great potential applications like this one and Raspberry Pi tools, but which have taken those who are very competent in these areas countless months and many hundreds of posts to try and get the bugs worked out, and they are still trying. Good on you, I appreciate these are still in development mode. Sadly, it is not practical for me to educate myself to the point of participating in these issues, so I must watch from the sidelines. I guess my question out of ignorance is (and it is NOT a criticism): since the general technology and tools are available, why does it take so long to sort through these matters and develop the applications? I know it's not through a lack of capability, effort and good collaboration.

 

Lee- For those of us that are still challenged to get this working, can you tell us how to erase the flash? Sorry if this sounds like a fundamental question but this is very new to me! I've tried re-flashing, believing it would overwrite everything but maybe it's not that simple?

 

When people volunteer their time and knowledge with no expectation of financial gain, they tend to do it at their own pace... I'm sure @Thorrak and @LBussy have dedicated more of their own time (without compensation) than we can imagine.

When I first learned of all things BrewPi (and brewing in general) I had to take the time to teach myself a few things. And I've benefited from all of them.

 

You can download a tool from espressif at:
https://www.espressif.com/en/products/hardware/esp32/resources expand the tools section and look for "Flash Download Tools (ESP8266 & ESP32)" There's an erase button at the bottom. This is for Windows only.

 

Some configuration information, including WiFi credentials and access point information, is stored in an area of memory that is not overwritten when you re-flash. This causes confusion as you can imagine but also that area can get corrupt and will stay that way until a specific action is taken.

How to do it depends on your OS and the tool you are using to do the flashing.

 

We won't talk about the time it takes or the money we spend. My wife's divorce lawyer would find it sometime in the future and it would come back to haunt me.

Not specific to Tiltbridge, but these projects in general, we have these investments ($ being monetary and T being time) in the following:

• Hosting fees ($)
  
• PCB design (T)
  
• PCB prototypes ($)
• Coding in:
  • C++ (T)
  • HTML (T)
  • CSS (T)
  • Python (T)
  • Javascript (T)
  • SQL(T)
• 3D designs (T)
• 3D prints ($)
• Logo design ($)
  
• Writing documentation (T)
• End-user support (T)

On top of that, just about the time where we think we have something solid, some jackstick in China will change a design and we're back to the drawing board (as happened with Tiltbridge.) Remember that a lot of you look for the cheapest way to accomplish a given task, so you are constantly out there looking to save $0.11 on a 100-unit purchase on Aliexpress. That means we have no idea what you will buy next.

Then there are the libraries we use for a lot of these. This is Open-Source so we're not coding our own libraries and we're sure not paying for support on the libs. We consume other people's works just like you do. So when random person in India makes a mistake deep in their library in an area that is only touched by a very specific set of circumstances, we end up with a bug that takes hours or days to find and squash. All the while we're thinking "this person depended on me and now their batch of beer is at risk."

Finally, we do this after (or at the expense of) our own jobs, families, and hobbies.

That represents a small part of a day in the life of an Open-Source developer.

How can you help? Glad you asked. When you find a bug, Google it, find the solution as we will, and create a pull request with the solution. Not a coder? Help improve the documentation. Don't know what a pull request is? Learn that then come back and do one of the other things. That's not to say we expect help, we don't. We do this because we enjoy it for the most part. But if you want things to go faster, there are only two ways: Pay for it or do it yourself.

And here's a because I love you guys, really. I'm not complaining but since you asked ...

 

It’s not that simple - there are flags that have to be added to the flash command to force it to erase everything that was previously there. WiFi settings, for example, get preserved through a typical flash.

While there are a million answers to this question, there are three key factors I’ll cite:

First, none of these projects are made from whole cloth. Fermentrack, for example, incorporates the work of 32 other projects, each of which might incorporate the work of dozens of projects more. For Fermentrack to function requires that each of these projects in turn work in a very specific way - when many of those projects are actively being developed and improved. I think that a lot of the “bugs” you cite arise from changes that come from that development. Someone else changes something, and it’s up to us/users to adapt to that change. This is the unfortunate nature of software development - but it is (typically) indicative of some kind of process, so I’ll take it.

Second, all of these projects are DIY by design. Users of these projects range from highly skilled engineers to people without any electrical background whatsoever. Things that seem self evident to me as the author of a project may not be understood by anyone else - even an experienced user might miss a solder joint and end up with unexpected results. If this was a commercial project with centralized assembly and testing it might not be a problem, but lacking that infrastructure we pay in support requests.

To this point many of my projects are designed to minimize the potential for error - either in the software or in assembly. I’ve got another project I’m putting the finishing touches on now that will hopefully make things like TiltBridge even more accessible to users who don’t live on a command line.

Finally, many of these projects are ultimately hobbies for those of us who are working on them. I have a full time job outside the Tech/Brewing industries that eat up most of my week, and as a result I have fairly limited time to work on projects like this one. Even though a change may seem trivial, if the time doesn’t exist to implement it, it may sit unimplemented for awhile. Even if I have the time to implement it, those changes may happen outside the public eye - which is what has been happening with the latest revision to TiltBridge. It’s moving, I just haven’t been speaking about it.

Regardless, I understand the sentiment - and I take it as a compliment! If these projects are working as intended, they should almost appear to be magic. The fact that they sometimes don’t is what I strive to correct!

 

Have tried reflashing multiple times using the command line and also the utility that Thorrak sent me. I also completely wiped and reloaded MicroPython per the Wemos site instructions. No such luck. Same result.

The good news is I can walk past my fermenting fridge and see gravity without pulling out my phone Just can't send the data anywhere.

 

Yeahhhhh - I’m going to dig into the code later tonight. I worry that you may have found a bug

 

Like @popquizkid I get a blank browser screen from http://tiltbridge.local/, but if I use the IP address, its connected and reading my Tilt. I still can't connect to my Google sheets, but I've always had problems with that, even from the Tilt itself so I'm probably misunderstanding something in it's setup. My TFT display still doesn't work though. It's lit, but no text?

 

Klaus thank you very match. According your procedure is now working. I had same issue as how you described: "After flashing tiltbridge i had black screen without wifi AP. I flashed by Arduino IDE your code, then reflashed tiltbridge and vuala, screen is going and i can connect to AP."

ROBOL

 

Just a heads up - I'm doing some server maintenance, so tiltbridge.com may be down for a bit.

 

I have this board https://www.aliexpress.com/item/32824819112.html?spm=a2g0s.9042311.0.0.5f054c4d1CbHFL

When I first flashed the board with esptool.py I got a blank screen, no WiFi and red led permanently on. Using kshuler's suggestion I uploaded a test simple clock .ino to the board which worked OK and then re-flashed with esptool.py and I now have TiltBridge working !

Thanks kshuler for the tip and thanks to Thorrak for making TiltBridge available.