Bigdaddyale
Well-Known Member
C1 is optional, I think it was used to try and help with screen scramble
Native ESP8266 BrewPi Firmware - WiFi BrewPi, no Arduino needed!
- Thread starter Thorrak
- Start date
Help Support Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum:
If you haven't been disconnecting the rest of your build, do that (flash the ESP8266 with nothing else connected)
The capacitor is optional, and only lightly recommended. Most of my builds don't have it - as @Bigdaddyale noted, it was an attempt to fix screen scrambling. It didn't fix 100% of the instances, and was replaced with a software fix instead.
To my eyes, it looks like a sensor is dropping out (which would make sense with the length of wire you’re using). What size pull-up resistor are you currently using?
It looks like ALL of the sensors are experiencing drop-outs.
40 feet of 18 gauge is a lot of distance - and a lot of copper for signals that could travel on 30 gauge. Is it at least twisted pair?
Cheers!
40 feet of 18 gauge is a lot of distance - and a lot of copper for signals that could travel on 30 gauge. Is it at least twisted pair?
Cheers!
CadiBrewer
Well-Known Member
With the ESP8266 build, why do you need to have your sensors travel that far over copper?
Unfortunately I did not heed @Bigdaddyale 's earlier post of not soldering in the ESP8266 so instead of unsoldering I built another. I want four eventually I tried it again on the RJ45 board with just usb to RPi no 5v same no go. But this one I could uplug and so I did and flashed success!
Well success was short lived. Plugged the chip into the RJ45 pcb unplugged the usb wired up the 5v and......nothing. No little blue lights nothing. 5.5v to all the right places. Pretty much 10k across all the right points but only 0.11 v where I think 3.3v should be. Bad mosfets? Suggestions anyone?
I don't have an ESP8266 build. Not sure why this would matter? The Sensors are One-Wire address, (VCC, DATA, Ground), and go back to the Arduino.
I'm an HVAC Tech., so I have tons of solid copper thermostat wire around. Its Solid, with a beefy #18 gauge, so it should not have much resistance...? for such a small currant transmission.... I don't know from any experience, what drops over distance that One may see, looking at temperature probe communication lines, such as my set up.
I know 24 volts DC is not anywhere near what we see on these wires, so that's why I'm not sure WTH, I'm talking about.
My Resistor is a 4.7 kohm 1/2 watt carbon-film. Which I also do not know WTH this does...just know it made my Pi's work for the past 3 years, so I never questioned it, until you much-more-knowledgeable guy's asked me!
OK, so "NO ARDUINO NEEDED" how this Thread starts off at Post #1, but I never look at directions....just jump in AZZ first !
Reading Lee Bussy's post #4, it seems to me now that my Arduino and the copper wires, are NOT needed... Maybe I should just take the copper to the Scrap Yard, smoke the 2 Bucks, and flash a ESP8266 board when I get home ? !
Thanks for taking the time to chime in on my Quest for an Upgrade in the Membrane!
Last edited:
I don't have an ESP8266 build. Not sure why this would matter? The Sensors are One-Wire address, (VCC, DATA, Ground), and go back to the Arduino.
I'm an HVAC Tech., so I have tons of solid copper thermostat wire around. Its Solid, with a beefy #18 gauge, so it should not have much resistance...? for such a small currant transmission.... I don't know from any experience, what drops over distance that One may see, looking at temperature probe communication lines, such as my set up.
I know 24 volts DC is not anywhere near what we see on these wires, so that's why I'm not sure WTH, I'm talking about.
My Resistor is a 4.7 kohm 1/2 watt carbon-film. Which I also do not know WTH this does...just know it made my Pi's work for the past 3 years, so I never questioned it, until you much-more-knowledgeable guy's asked me!
OK, so "NO ARDUINO NEEDED" how this Thread starts off at Post #1, but I never look at directions....just jump in AZZ first !
Reading Lee Bussy's post #4, it seems to me now that my Arduino and the copper wires, are NOT needed... Maybe I should just take the copper to the Scrap Yard, smoke the 2 Bucks, and flash a ESP8266 board when I get home ? !
Thanks for taking the time to chime in on my Quest for an Upgrade in the Membrane!
I think the point around the ESP8266 question was that if you had an ESP8266, presumably your controllers were connected via WiFi and could therefore be placed near your fermenters. For most setups, you could then get away with ~6ft cables on the DS18b20s which should reduce the potential that the run length would interfere.
A 4.7 kohm resistor is what the DS18b20 spec sheet calls for - but if you happened to follow my instructions for the ESP8266 build, you might have used a 10k resistor instead. That was because I was lazy when sourcing resistors and wanted all the resistors to be the same - 4.7k is actually the better (correct) one to use. The higher the value of the pullup resistor the longer it will take the voltage to reach "high" after being pulled low by the device which is a problem for longer cable runs. I don't have any experience with (or knowledge of, or recommend...) smaller resistors - its possible that swapping the 4.7k for a 4k or 3k could help in your situation -- but it's also equally possible that it would break your sensors or prevent them from working entirely.
If you're looking to go as wireless as possible I do recommend going down the ESP8266 route -- though if you go down that route, I highly recommend buying D1 minis from the source, rather than knock offs.
No idea. That certainly sounds plausible.
I've had issues with that style of D1 mini before - though generally with the flash, not the power chips. Does it still work (light up, etc.) when powered via USB?
Is there any way I can verify that the ESP8266 that I got from AliExpress is actually D1 that accepts one-wire inputs. It eventually flashed as esp8266EX but only when disconnected from your RJ45 pcb but won't power up on 5v w/o usb when back on the pcb. It doesn't look at all like the one you referenced from AliExpress.
ESP8266EX is the same chip that Lolin (Wemos) uses on the boards that I recommend (and have had no issues with). All of the ESP8266EX variants should accept one-wire inputs without issue. From what you're describing, it sounds like there's some kind of either wiring issue, soldering issue, or issue with a component other than the ESP8266 board so long as the ESP8266 is (still) responding normally when connected via USB.
That said, try disconnecting the white/black wires from the "power in" screw terminals (completely disconnect your PSU) and see if it responds that way when connected via USB. Then, go through and readd the components individually to see what combinations it works/doesnt work in. You might find it works with USB + RJ-45 breakout board/temp sensors, but doesn't work with the power supply.
OK unplugged 5v psu and plugged in USB, 3 sensors are in RJ45 ESP 8266 is flashing blue light but when I try and detect the sensors it says it "Unable to reach brewpi-script"
I checked that the mosfets are switching/opening without the ESP8266 plugged into the pcb. Is it possible that the voltage to the gate on the mosfet is too low at on the D1 & D2 pin hence only the 0.1v instead of the 3.3v? I'm now suspicious of the ESP8266s I got so I've ordered more this time from @Thorrak 's recommended source. Meanwhile I'll keep playing with the 2 I've built.
Bigdaddyale
Well-Known Member
I wonder if you fried the diode on the ESP8266 when you had power from both the usb port and the PCB?
Possible. I didn't know you could do such a thing. Eventually the intention is to have local 5v power when running on WiFi no USB but maybe because it wasn't flashed first maybe. I've got two more so I'm going to try and flash another.
I have been able to flash 3 of the ESP8266EX using the Fermentrack utility without them connected to the RJ-45 PCB or sensors. When they were flashed I then went into the "add a new controller" guided again and indicated they were flashed so it went into setup and gave them each a different name. It worked on two of the 3. The 3rd one flashed but I don't know why it won't take what I submit as a new name in setup. It won't show up as a choice under "choose a device".
When the chips are flashed is the SSID and password taken from the Fermentrack setup and already added or does it have to be added with each setup as it is on Wifi i.e not connected by usb?
Bigdaddyale
Well-Known Member
OT, but cool!
Geekcreit® ESP32-CAM WiFi + Bluetooth Camera Module Development Board ESP32 With Camera Module OV2640
Geekcreit® ESP32-CAM WiFi + Bluetooth Camera Module Development Board ESP32 With Camera Module OV2640
OK using the advaanced route I was able to add the 3rd controller. But like the others I added when I try and configure the pins I get "Unable to reach brewpi-script. Try restarting brewpi-script." I reboot and samething. I'm probably missing something obvious.
CadiBrewer
Well-Known Member
Try erasing the EEPROM. There's a button under the advanced settings to do it.
When I hit that button I got "Unable to reach brewpi-script for device FV3_TempCntrl". If I uninstall and and try to reinstall it as a serial with it connected to the RPi by USB do I have to reflash it as a serial device first?
CadiBrewer
Well-Known Member
My understanding is yes, if it was flashed with the wifi firmware but you want to use it as serial via USB, you'd need to reflash it with the serial firmware.
I guess my worry with doing this is that at some point it needs to be wired to 5v too so it can drive a relay. @Bigdaddyale theorized that wired to 5v and usb could have fried the chip on initial attempt
at this.
CadiBrewer
Well-Known Member
That shouldn't be an issue, but you can use the USB to connect to your Pi, and then use the 5v power supply to power the relay. You would then have just the wire for the data between the ESP and the relay.
For WiFi-based controllers, the setup flow is this:
- Connect the controller via USB to your Raspberry Pi and complete the flash process
- Disconnect the controller from your Raspberry Pi, and power it up
- Using a WiFi device (laptop, phone, tablet, etc.) connect to the controller's access point and set up the Controller's WiFi connection (enter your SSID/password)
- Once the controller's is connected to your WiFi, go back into Fermentrack and use the guided setup to add the controller to Fermentrack
"Unable to reach brewpi-script" means that Fermentrack isn't able to see/communicate with your controller.
To clarify a few things - did you flash the Serial or WiFi firmware? After you flashed the firmware, did you do the steps above (including connecting to the controller and configuring the SSID/password for it to connect to your WiFi)? Were you able to see the controller's mDNS ID in the guided workflow (if you were using the WiFi firmware)?
If you successfully set a controller up in Fermentrack but are still getting the "unable to connect" message, you can try using the "debug connection" option on that device's dashboard. No guarantees it will help, but it might tell you something helpful!
To clarify this point - there's two different things at play here. Both USB and - hopefully - your hardwired power supply are generating +5V. The way that the PCBs are wired, you've got the 5V necessary to run your relays regardless of whether you are using USB to power it or your hardwired power supply.
The problem (and what @Bigdaddyale was likely thinking when he commented) is that you were trying to power your board with both USB and the hardwired power supply when you had it plugged in to the wall at the same time as it was plugged in via USB. In some cases this won't cause any problems, but it can.
I tend to recommend against powering the power supply when plugged into USB in every case, but primarily because I'm paranoid that somehow I'll end up frying my laptop's USB ports. Keeping them separate makes me worry far less.
For WiFi-based controllers, the setup flow is this:
- Connect the controller via USB to your Raspberry Pi and complete the flash process
- Disconnect the controller from your Raspberry Pi, and power it up
- Using a WiFi device (laptop, phone, tablet, etc.) connect to the controller's access point and set up the Controller's WiFi connection (enter your SSID/password)
- Once the controller's is connected to your WiFi, go back into Fermentrack and use the guided setup to add the controller to Fermentrack
"Unable to reach brewpi-script" means that Fermentrack isn't able to see/communicate with your controller.
To clarify a few things - did you flash the Serial or WiFi firmware? After you flashed the firmware, did you do the steps above (including connecting to the controller and configuring the SSID/password for it to connect to your WiFi)? Were you able to see the controller's mDNS ID in the guided workflow (if you were using the WiFi firmware)?
If you successfully set a controller up in Fermentrack but are still getting the "unable to connect" message, you can try using the "debug connection" option on that device's dashboard. No guarantees it will help, but it might tell you something helpful!
To clarify this point - there's two different things at play here. Both USB and - hopefully - your hardwired power supply are generating +5V. The way that the PCBs are wired, you've got the 5V necessary to run your relays regardless of whether you are using USB to power it or your hardwired power supply.
The problem (and what @Bigdaddyale was likely thinking when he commented) is that you were trying to power your board with both USB and the hardwired power supply when you had it plugged in to the wall at the same time as it was plugged in via USB. In some cases this won't cause any problems, but it can.
I tend to recommend against powering the power supply when plugged into USB in every case, but primarily because I'm paranoid that somehow I'll end up frying my laptop's USB ports. Keeping them separate makes me worry far less.
This was very useful.
I did flash the Wifi firmware as I was hooked up to USB only.
@Bigdaddyale was right to warn me as in my original floundering I probably smoked the first chip by having both sources hooked up. But I had three more and @Bigdaddyale 's caution to move on with.
I could not figure out how the chip would get the Wifi info. I didn't realize the chip itself became an access point with its own SSID that I could pickup to enter my Wifi credentials. Then your last post finally penetrated my thick skull.
Unfortunately when the chips were tried in a RJ-45 pcb there was no sign of life and no access point so I took them off the board again and hooked them back up to the usb, got the access point and was able to populate them with the wifi info and get the to talk with my wifi.
For some reason the esp8266's don't energize while on either of the boards I soldered up. I see no blue light by chips like when they are on the USB. So that's probably why they can't be "seen". The power source is verified as 5v to the board. I've checked resistances and continuities on the boards without the chips on them. The mosfets disallow any continuity between D & S when not powered at G and limited when powered. However when the chip is on the board the voltage where its supposed to be 3.3 are about 0.1V and something like 0.12 and .24 at D2 and D1. The places where its supposed to be 5v are 5v.
I just don't know where I screwed up. This project is really my first time soldering circuit boards. I attached some photos. The pcb with the 4.7K ohm at R6 is a little hacked up since I had to de-solder the first chip I fried. Since it was hacked up I thought I'd see if I could put a regular resistor on the tiny surface mount spots where there was a 10K.
Could the reduction of 5V to the 3.3v be screwed up because of a fault in the chips i bought? I do have have more chips from the source you recommended on order but they are probably on a "slow boat from ......" so it might be awhile before I can try them.
Bigdaddyale
Well-Known Member
There should only be five resistors and there is no R6 position on the board.D6 pin on the ESP8266 goes to R1 which goes to +3v3 pin
Progress!
Ok, a few more questions --
Just to be clear, are you saying that when you have the ESP8266 powered via USB it works (connecting over WiFi) but when you have them powered by the 5v PSU connected to the PCB it doesn't work?
Have you tried disconnecting the 5v PSU from the PCB, plugging the ESP8266 into the PCB, and then powering it via USB? (In other words - is it having the ESP8266 plugged into the PCB that's the problem, or is it powering it via the +5v power supply that causes it to stop responding)?
When you refer to the mosfets, are you talking about the SMD ones you soldered onto the PCB?
Can you post pictures of your PCB without an ESP8266 mounted on it?
Bigdaddyale
Well-Known Member
He soldered the D1 mini to the board so I don't think he will be able to easily remove it.
Ah - I thought that was only one of the boards.
You're right @Bigdaddyale I meant R1. I posted the wrong photos. I knew there were only 5 resistor obviously I should slow down on the beer. I've ony soldered up two boards so far. The beat up one where I desoldered (yeah it was hard) the mini and put the normal resister on and the other two photos are of the other board top and bottom.
@Thorrak yes that's right when powered by USB I can get them on Wifi and that was how I got to their acces point too. But on the board nothing, no blue ligy not signal.
The Mosfets i'm talking about are what I soldered to position Q1 & Q2. Weren't these BSS138s supposed to go there?
OK I just tried the powering by usb and plugged into the board without the 5v and nothing, no blue light and no signal. So it looks like the mini on the board is a problem.
Now we’re getting somewhere. Nothing on the board should impact the D1 mini when no sensors/LCD/relays are connected. The mosfets are only used for the LCD - the resistors are for the LCD and the temperature sensors.
That said, my guess would be that there is a short somewhere on the board. Take a continuity tester and check all adjacent solder points to see if two happen to be touching. Also, look closely at the board to make sure you didn’t accidentally touch the solder mask at some point and singe/short a trace.
Bigdaddyale
Well-Known Member
The second picture looks like a solder bridge on the RJ45 connections. Hard to tell
@Bigdaddyale and @Thorrak As an inexperienced solderer I check for lack of continuity on all adjacent solder points. I also check along traces to points where there should be conductivity. But I'll check them again.
I have a variable DC supply and a fixed 5v psu. Could I power the 3.3v and 5 v pins directly and separately when the mini is plugged into a brand new empty rj45 pcb I haven't touched yet? Then I could incrementally solder in components. Then the RJ45 next if the direct 3.3v powers the mini and talks to wifi and lets fermentrack see it.
I have a variable DC supply and a fixed 5v psu. Could I power the 3.3v and 5 v pins directly and separately when the mini is plugged into a brand new empty rj45 pcb I haven't touched yet? Then I could incrementally solder in components. Then the RJ45 next if the direct 3.3v powers the mini and talks to wifi and lets fermentrack see it.
Actually I could start with only the direct 3.3v and save adding the 5v until later in the test. I'm still on the fence about adding displays.
Update with only the 3.3v supply wired directly to the pin on the mini plugged into a bare RJ45 board I got it to link into my wifi but Fermentrack still says"
Unable to reach brewpi-script. Try restarting brewpi-script." so I reboot Rpi and samething debug reports
"ConnectionRefusedError at /devices/6/debug_connection/"
Last edited:
ARGGH!
Thorrak I blame you for this.
I was gonna see what was new with your stuff so I grabbed an SD card, re-flashed Raspbian and then realized I grabbed the wrong one and I lost this morning's work because I didn't push changes up.
GAH! All. Your. Fault.
That's incidentally why I have so many commits. I am still developing on an SD card and I don't trust the damned things.
Thorrak I blame you for this.
I was gonna see what was new with your stuff so I grabbed an SD card, re-flashed Raspbian and then realized I grabbed the wrong one and I lost this morning's work because I didn't push changes up.
GAH! All. Your. Fault.
That's incidentally why I have so many commits. I am still developing on an SD card and I don't trust the damned things.
If I suspect my software on my SD should I reflash and start over or is there a command line option for reinstalling Fermentrack?
Bigdaddyale
Well-Known Member
I would go back to square one at this point. You are having way too many troubles than most on their first experience installing Fermentrack and setting up an ESP8266.
Similar threads
