iSpindle - DIY Electronic Hydrometer
- Thread starter LBussy
- Start date
Help Support Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum:
FowlBall
Member
@tonari, been reading about your woes here and on github. I don’t have many suggestions other than it appears the WeMos connection between RST and D0 isn’t good...maybe a soldering issue. If you could post some more screenshots of your board from some other angles, especially showing the diode solder joints / traces and the connections on RST and D0, maybe I could spot something.
It looked like from the log you posted that you had the sleep time set at 5 secs, if I read correctly. Maybe try changing that to 30-60 secs as a test once you’ve charged your battery more. Let’s see if the results are the same.
It looked like from the log you posted that you had the sleep time set at 5 secs, if I read correctly. Maybe try changing that to 30-60 secs as a test once you’ve charged your battery more. Let’s see if the results are the same.
tonari
Member
- Joined
- Sep 19, 2018
- Messages
- 10
- Reaction score
- 1
Hi FowBall thanks, i indeed think that i have an issue between those two points. I checked with a multimeter and i don't have anything between RST and D0..., everything is ok at the BAT43 place, the problem is after that on the board.
I orderer new composant and will start from scratch as soon as i will have everything.
I tried to change the sleep value but without any effects.
My issue is my soldering stick, i think it's too hot, i can change the temp but still, when i try to desolder something it finished with a lifted padle...
Thanks for your help guys and i'm sorry if this is my fault from the beginning ^^
I orderer new composant and will start from scratch as soon as i will have everything.
I tried to change the sleep value but without any effects.
My issue is my soldering stick, i think it's too hot, i can change the temp but still, when i try to desolder something it finished with a lifted padle...
Thanks for your help guys and i'm sorry if this is my fault from the beginning ^^
FowlBall
Member
Ah, that would have been my first guess @tonari, sorry about your luck on the desoldering. Those pads can be a little delicate and you do have to take care when soldering / desoldering. Since you’re so close and the trace is bad anyway, you might be able to salvage it by taking a couple of small pieces of around 26AWG wire and soldering to the original pad locations to bypass the traces. Just a thought, and you don’t have much to lose at this point
FowlBall
Member
You don’t have to be too picky about the wire, just any thin-gauge wire (for flexibility) typically used in electronics projects should work. Something like a DuPont jumper wire used for breadboarding, just cut off the connectors and strip...
Last edited:
Lalo_uy
Well-Known Member
My issue is my soldering stick, i think it's too hot, i can change the temp but still, when i try to desolder something it finished with a lifted padle...
Thanks for your help guys and i'm sorry if this is my fault from the beginning ^^
You can try to plug it on a dimer or insert a 1N4007 diode in the AC line, that will drop the power a good notch.
yeehaw
Member
tonari, you can most likely still salvage the board despite the lifted pads, but you would really need some flux to help the solder flow. Scrape away the coating on the track leading to the lifted pad to expose the track, spread on copious amounts of flux and tin the track with some solder. If you get even a tiny length of the track tinned, you should be able to solder a thin wire on it. Also, tin the soldering iron and the end of the wire before trying to solder the wire on the track.
If you have a nicely tinned soldering iron, track and wire end, you should only need to touch the wire end and track for a second or so with the soldering iron, and end up with a nicely soldered joint. And it's always better to have your iron too hot than too cold - otherwise the solder might not flow well and you will use too much force when trying to free the component from the pad and end up with a lifted pad (speaking from experience).
If you have a nicely tinned soldering iron, track and wire end, you should only need to touch the wire end and track for a second or so with the soldering iron, and end up with a nicely soldered joint. And it's always better to have your iron too hot than too cold - otherwise the solder might not flow well and you will use too much force when trying to free the component from the pad and end up with a lifted pad (speaking from experience).
tonari
Member
- Joined
- Sep 19, 2018
- Messages
- 10
- Reaction score
- 1
Hi thanks a lot for the advice and the "how to"
The part is really tiny that will be difficult, but i will try to do it and see if i can save this gy521.
It's not really a price issue because it's really not expensive, but i don't like to throw away things like this.
The part is really tiny that will be difficult, but i will try to do it and see if i can save this gy521.
It's not really a price issue because it's really not expensive, but i don't like to throw away things like this.
skyblaster
Member
I did the same, and I have to ask: Is it really electrically necessary to remove the Wemos diode?
I mean if you're using pin headers and can separate the Wemos from the charging shield, then just avoid connecting a USB cable to the Wemos unless separated. That would simplify future USB flashing if over the air updates were to fail.
Also, I finally connected my first iSpindel to Ubidots
If I'm understanding the Education account correctly, I should be able to retain all 5000 credits if I only ever register one device. I built a few additional iSpindels for a couple guys in my homebrew club and I'm wondering if they might each need multiple accounts so that they don't run down their credits because of the 2nd device?
If your using the mikmonken shield, all I know for sure is that if you don't remove it, the Wemos will crash and burn after a few seconds. According to the GitHub it provides a short circuit to the LiPo charger. I found out the hard way, as I forgot to remove the diode before attaching a new wemos to the shield in order to flash it.
Last edited:
flemming
Active Member
Hrm, I have not had the same experience. I didn't remove the diode or the led and didn't have any trouble using the mikmonken shield. Looking at https://github.com/universam1/iSpindel/issues/119 perhaps that's because I've always made sure the switch was off.
Robert
I have had 1 built for a while now and just hadn't done a brew for ages. Now I have a brew on, I am getting motivated to calibrate it and get it working properly. I am hoping to integrate it into BPL and I am getting the final parts I need to finish that project too.
I guess my STC1000+ has been doing a good job and I have really not made BPL a priority. But with iSpindel together I am really looking forward to getting this going before my brewing season starts.
I am also developing the iSpindel into a STEM project for Year 10 students. I'm hoping to teach them problem solving. There will be build instructions to detail the building and flashing. But assembling the unit will be their task. They will have a variety of sizes of petling and sled options which they will be able to modify if needed and 3D print. Also getting the tilt angle correct will be up to them. They can test, modify and re-check. Then they will perform a calibration using sugar solutions and do mathematical modelling. Hopefully they can verify the SG of a solution provided to them, which I know the SG and see whose build is the best. (hopefully they all are good). So far the assembly instructions are finished and have been student tested. I just need to accumulate different petling and methods of balast for testing. I might have to upload the build instructions to the cloud to see if people find them useful.
I guess my STC1000+ has been doing a good job and I have really not made BPL a priority. But with iSpindel together I am really looking forward to getting this going before my brewing season starts.
I am also developing the iSpindel into a STEM project for Year 10 students. I'm hoping to teach them problem solving. There will be build instructions to detail the building and flashing. But assembling the unit will be their task. They will have a variety of sizes of petling and sled options which they will be able to modify if needed and 3D print. Also getting the tilt angle correct will be up to them. They can test, modify and re-check. Then they will perform a calibration using sugar solutions and do mathematical modelling. Hopefully they can verify the SG of a solution provided to them, which I know the SG and see whose build is the best. (hopefully they all are good). So far the assembly instructions are finished and have been student tested. I just need to accumulate different petling and methods of balast for testing. I might have to upload the build instructions to the cloud to see if people find them useful.
tonari
Member
- Joined
- Sep 19, 2018
- Messages
- 10
- Reaction score
- 1
Hi, i'm working on a dashboard to display the ispindel data and i'm stuck on the wifi RSSI data.
Maybe i missed something...
I use the last ispindel firmware and a php file to send the data from the ispindel to my online mysql databse.
And then display the data on my dashboard.
All informations are stored on my database except the RSSI data and i don't know why.
I use the HTTP protocol on the ispindel settings.
Someone have a clue ?
Maybe i missed something...
I use the last ispindel firmware and a php file to send the data from the ispindel to my online mysql databse.
And then display the data on my dashboard.
All informations are stored on my database except the RSSI data and i don't know why.
I use the HTTP protocol on the ispindel settings.
Someone have a clue ?
Hi guys,
I've just build and calibrated this device and it works pretty good, however I've noticed that data are not transmitted to the ubidots all the time.
If I lay ispindel freely on the table maybe 5m from router with very thin wall, it shows -60 RSSI. Seems it sends data reliably. But if I put it into the same place in a big glass with water it's not transmitting maybe 8 from 10 records. When record is transmitted the signal strength shown is -77 RSSI.
Do you know what is a good value for signal strength and what is too low ? Is -77 bad value ? Thank you.
I've just build and calibrated this device and it works pretty good, however I've noticed that data are not transmitted to the ubidots all the time.
If I lay ispindel freely on the table maybe 5m from router with very thin wall, it shows -60 RSSI. Seems it sends data reliably. But if I put it into the same place in a big glass with water it's not transmitting maybe 8 from 10 records. When record is transmitted the signal strength shown is -77 RSSI.
Do you know what is a good value for signal strength and what is too low ? Is -77 bad value ? Thank you.
FowlBall
Member
The short answer is, yes, -77 is a marginal RSSI value. For reliable packet delivery, an RSSI of >-70 is recommended. Anything <-80 is recommended just to maintain basic connectivity to the router. RSSI is a relative term and how it’s measured varies by manufacurer, but it does give at least some measure of how well the device can “hear” the transmitter (router).
If your router isn’t close to where you’ll be running the iSpindel, you may want to consider adding a repeater close by. Another alternative is to build it using a WeMos D1 Mini Pro with an external antenna attached. I’ve built one of each version, and the external antenna provides a much stronger connection.
Thanks for response. I've bought a repeater already
I want to ask also about battery life. It shows 4.25V when it is fully charged and then it goes down slowly. I have 600 seconds interval. What's the lowest voltage when the device stops working ? Thx
Lalo_uy
Well-Known Member
The protection IC cuts the power at 2.90~2.99 V
Thank you!
FowlBall
Member
Also note that the current code defines a low battery state at 3.3 volts. When it reaches this threshold, it drops the WiFi transmit power to minimum, and lengthens the sleep time to at least 1 hour (or 3 times your specified sleep time, whichever is longest). It’s purpose is to try to survive a bit longer.
FWIW, I ran mine for almost a month reporting every 10 mins. recently. It was only charged to about 4.2V when it started, and it was still at about 3.8V when my fermentation was done.
That's pretty nice result. What is your RSSI ? I've noticed if I put it next to the router it's -40, but 2m away is already -60 - is this normal? In plastic fermetor I have usually -75 RSSI. For me it looks like this chip has a very poor wifi range. I also have 10 min interval and battery goes down by 0.01V after appx. 10 - 12 transmissions which is appx. 2 hours - 0.12V per day which seems to me pretty much. What could be the problem ?
FowlBall
Member
I’d actually not looked too closely at the RSSI since mine always reported fine. I went back and took a look at what had been logged and it averaged between -80 to -85! But, as I said earlier, this value is a relative term so comparing across devices / environments isn’t necessarily helpful. So long as you are getting good connectivity and reliable reporting, I wouldn’t worry about it much. And, yes, your reported RSSI values are very typical at those distances (and -60 is an excellent signal quality).
What battery are you using? That of course will make a difference in how long it will last. (I’m using the Samsung 30Q recommended by @cherryphilip74, since I’m using his board). Also, the battery voltage will likely not decline linearly, so extrapolating like you’re doing won’t necessarily give you the max runtime.
As far as range goes, I’ve been using these ESP boards (the WeMos / Lolin design) for several years, and have always been impressed by the range they provide for connectivity, especially given the small size of the antenna. That said, there are now MANY knockoffs made now and I’m sure quality control is about the last thing with which the makers of these are concerned. So, I’m sure YMMV depending on the source of your board. Given that my board was in a Petling, in a fermenter full of wort, inside a chest freezer, I was darned impressed it even was able to get a connection!
What battery are you using? That of course will make a difference in how long it will last. (I’m using the Samsung 30Q recommended by @cherryphilip74, since I’m using his board). Also, the battery voltage will likely not decline linearly, so extrapolating like you’re doing won’t necessarily give you the max runtime.
As far as range goes, I’ve been using these ESP boards (the WeMos / Lolin design) for several years, and have always been impressed by the range they provide for connectivity, especially given the small size of the antenna. That said, there are now MANY knockoffs made now and I’m sure quality control is about the last thing with which the makers of these are concerned. So, I’m sure YMMV depending on the source of your board. Given that my board was in a Petling, in a fermenter full of wort, inside a chest freezer, I was darned impressed it even was able to get a connection!
Thank you. Yes, it's strange with that RSSI. Recently I've installed a repeater (in roaming mode), pretty close to the router an it shows RSSI usually about -60 to -70 but despite of this I have sometimes a blackout. It's not reporting maybe for an hour, then it starts normally. So I'm not sure it's the wifi strenght. What could be the problem not all the readings are transmitted ? It's iSpindle issue or network issue ? I don't know.
I have the same board from cherryphilip v3 and use the same battery as you. But yes, maybe what it is different is the chip. I've ordered wemos board but I think I've received a clone because I cannot see wemos logo anywhere. What I see is only ESP-12F - ESP8266MOD. I don't know if this can be a culprit the reliability is not that great. Have you got one with a wemos logo on it? Where did you buy yours?
As for battery readings I had 4.06V two days ago, now it reads 3.88V. With 600s report rate I still think this is too much, isn't it ?
FowlBall
Member
It’s hard to say what the periodic reporting failure could be, as there are so many variables. I would probably lean towards some sort of network / WiFi issue if I were troubleshooting it. Perhaps it could be some sort of interference with the WiFi signal; maybe try changing the channel on your router and observe for a while to see if you have any better luck.
I buy all of my boards from AliExpress, but from various vendors. Most all are clones but have worked satisfactorily for me. You generally won’t see a WeMos or LoLin logo on the clones, but they use the same main chip from Espressif. I have ordered components directly from the LoLin store on AliExpress before (an option if you want to try an original), but they are a few bucks more expensive.
And regarding battery usage, I’d try fully charging the battery, then letting it run until it dies to see what your runtime is. You’ll probably not have much luck in just extrapolating based on a shorter runtime like you’re doing. You can always back off the reporting frequency to extend it if necessary. So long as it covers your fermentation time, I wouldn’t sweat it. And your SG isn’t likely to change much in 10-15 mins anyway.
Lalo_uy
Well-Known Member
The WiFi range of ESP-12 modules is wonderful. I have in other application one broadcasting data and 3 display, them start to show some missing packets at 80m and -93 RSSI.
Can someone help me out?
I ordered the parts to make an iSpindel a while back and finally got around to putting it together today. It was simple to put together using the connection board provided by mikmoken in this thread and I didn't have any issues until I went to turn it on after I put it together. The temperature sensor is getting extremely hot and even started to smoke when I tried to turn it on. I already double checked and I put all 3 of the resistors in the correct place, so I can't figure out what's going on. Even when I have the 2 sections separated and it's just the battery, connection board and battery charger without the wemos or accelerometer and I try to turn it on the temp sensor gets hot and smokes. Any ideas?
I ordered the parts to make an iSpindel a while back and finally got around to putting it together today. It was simple to put together using the connection board provided by mikmoken in this thread and I didn't have any issues until I went to turn it on after I put it together. The temperature sensor is getting extremely hot and even started to smoke when I tried to turn it on. I already double checked and I put all 3 of the resistors in the correct place, so I can't figure out what's going on. Even when I have the 2 sections separated and it's just the battery, connection board and battery charger without the wemos or accelerometer and I try to turn it on the temp sensor gets hot and smokes. Any ideas?
Last edited:
Can you post a photo oif the temp sensor? If I had to guess, you have it mounted backwards.
Yeah I can post a pic. I have it mounted and oriented the same way it shows in the instructions on github.
I agree w/ @Thorrak. If it's getting hot, it's probably in backwards. I've built at least 10 of these without any of the temp sensors overheating. A picture would most certainly help.
Also, most recommend that if the sensor has heated up in this manner, you replace the sensor. I would especially recommend it if you've seen smoke.
Alright I'm starting to get really discouraged now. I tried flipping the temp sensor around and it is still heating up and I realized that it's not just the temp sensor it's the entire assembly that is getting hot. I built a whole new unit and it's still heating up and the led won't turn on on the Wemos when I switch it on and hit the reset button. See pics for reference and let me know if you guys have any ideas, I'm about to give up on it.
See my post above, let me know if you have any suggestions.
I don't see any major flaws and the temp sensor is correct in those pictures. Does it heat up on both battery power and when the charging module is plugged in? Switch on/off? What about if you plug in the wemos to USB? If the whole thing is heating up, my only guess would be a short in the charging module.
If you're willing, you could send it to me and I can take a look and hopefully fix it. PM me if interested.
If you're willing, you could send it to me and I can take a look and hopefully fix it. PM me if interested.
I don't get any led lights when the Wemos is plugged in directly to usb and it doesn't heat up either. I also don't get any lights when there is a usb plugged into the charger and it doesn't heat up, whether the switch is on or off. It seems like it only heats up on battery power. The only thing I can think is when I had the temp sensor in backwards initially and tried to turn it on a few times and hadn't noticed it was getting hot it may have fried something in the unit before I noticed and turned it off and now it's just shorting out or something.
I only ordered one charger, accelerometer and wemos, so I took it all apart last night and put it back together using a new connection board and it did the same thing. What I may do is go ahead and order a new accelerometer, wemos and charger and put a whole new unit together when the parts get here. If it still does the same thing at that point then I may try to send it to you and see if you can figure out what's going on.
I recommend ordering extras, if you can swing it. Let me know about putting one together for you.
Yeah I'm kicking myself for not ordering extras the first time since shipping from China takes so long. I'm definitely ordering at least 5 of each part this time around. I'll let you know if I still run into issues after getting the new parts in.
fwiw, whenever I start a new project I always buy enough parts to build two. With luck I end up with a backup/clone, which can be handy in many ways...
Cheers!
Cheers!
