I have a project for a water tank level meter built on a throw in pressure sensor, an esp8266 (nodemcu), a 24v battery (maybe) all running through home assistant and esphome.
I am keen to have a more reliable source of power than a battery with solar and was thinking that I could run a poe cable from a shed (60m max) to provide power both for the to be installed poe camera and to run this project.
That being the case my poe switch I will be runnning from provides about 53 volts. So I would need to use a dc-dc step down to bring this to 24v for my project. I have another dc-dc converter to provide 5v power for the esp and the current to voltage converter required.
Has anyone successfully grabbed power from a poe connection?
I have some ubiquiti equipment and have just ordered a poe pass through switch. This should provide me with an output to be able to grab a consistent power source from. Does this sound kosher?
Hi Patrick,
Sorry we missed your post.
Your plan sounds solid to me, the only catch I can think of is that there are several POE standards for active POE which require end devices to negotiate what power they require from the supply and there is passive POE in which no negotation takes place.
It sounds like your system will be passive POE so as long as you double-check your power injectors, splitters and regulators before powering it up it should work.
Trent
thanks for that. I have a splitter around here so will give it a go. If I could ask another question. I have an esp32 project linked to a gravity throw in water level sensor (bought from you guys). I have programmed my esp32 to take in a deep sleep cycle and will only require power for the sensor for a period of 2 minutes at a time. I want to power the whole shebang on solar and battery. I was wondering if I could take the low voltage output of the battery (likely a parallel 18650 setup) through a dc/dc step up to provide a constant 18v for that two minutes. I believe that the sensor only uses 4-20mA? I am planning that it cycles only about 10 times per day.
Pat
Hi Pat,
A suitable step up converter should be able to power that sensor without too much trouble.
The sensor itself outputs a 4-20mA current loop signal to a current to voltage converter board which is included. It doesn’t refer to the power consumption of the sensor itself. The product wiki has the full information about how to read from the sensor.
You can read a bit more about current loop measurements in this short article.
Trent
I did have a look at the links but no data on power usage. I have tested the complete circuit and see that it uses about 200mA when collecting data and about 160mA when not. I believe that i am using power when in deep sleep by having the 5v relay being the switching mechanism for the sensor. I am assuming that if I change this to a suitable mosfet then this part of the system should use practically nothing until the mosfet receives the on signal from my esp32?
I am also curious about optimum voltage for the throw in water sensor. It states that it operates on 12v to 24v. My rule of thumb was choose the middle ground but ensure that the 18v is constant as I have previously checked that if the power fluctuates then the end reading also does.
Given that this only uses 200mA total for a period of 2 minutes I should be able to run the entire thing from a battery made of 18650?
Patrick
Hi Patrick
I haven’t had a close look at what you are doing here but this does not sound right.
If the spec says it will operate from 12 to 24V then it should be accurate to within its specs in that voltage range.
Where are you measuring. Are you sure you understand what a 4 to 20mA loop is. If you don’t arrange to keep the signal current within that range anything could happen.
Cheers Bob
Bob
I maybe need to double check. I have added an ADS1115 between the current to voltage converter that came with the sensor and my esp32. This was because it is apparent that the adc on the esp32 is pretty shocking.
However when the adc only was connected I tried with my benchtop power a range of voltage from 12v to 23v. The esp32 was measuring slightly different values on the adc pin depending on the voltage (all going through the current to voltage converter of course).
This is all bench tested and has not been put in position at my water supply tank. i have followed the system setup as described here…
I have inserted in this circuit the following
- a 5v relay that is switched from the esp32 to open the relay after waking from deep sleep for a set time.
- a dc/dc step down that delivers 18v to the sensor through the relay
- a dc/dc step down to 5v dc that powers both the esp32 and the current to voltage converter
- the ADS1115 is powered by a 3.3v pin from the esp32.
Patrick
Hi Patrick
You show it as powered from the 5V converter output. Why would you power it from the esp32 3.3v anyway. 5V will give you a bit more head room.
That configuration should work. Looks like the SEN0262 is a 4-20mA receiver. What sort of variation are you getting? What are you measuring with? It is feasible that a slight variation is within spec. I don’t know.
Basically the sensor (the 4-20mA loop transmitter) is a variable constant current supply which varies from 4-20mA. The supply has to be high enough to maintain 20mA through the TOTAL resistance of the loop. If for instance the loop R was 1000Ω and the supply 18V the max current would be 18mA and the system would not work properly. I have seen loops so long the supply voltage has been 100V or more. Think about this and you will get the idea. The loop transmitter will supply the 20mA while the loop R is low enough or the supply is high enough. That is the TOTAL including the sense resistor inside the SEN0262 which would be pretty low.
The constant current characteristic is the useful point here and means the loop R can vary considerably (due to length possibly) without any degradation as long as the maximum R is considered and sufficient V is maintained.
When you get to use this set up the long cable run should be the 4-20mA loop. The SEN0262, ADS1115 and MCU should be co-sited. If you start having voltage drops or variations between the SEN0262 and the ADS1115 you are losing all the advantages of the current loop system. The I2C bit should be as short as possible anyway.
Cheers Bob
PS Just had a look at this thing on Core site and noted the following
- This sensor requires a pre-heating period of about 30 minutes with the power supply connected before accurate readings can be made
So maybe your 2 minutes every so often with the device unpowered in between might not be that accurate
Bob
thanks for that. The loop that I have is the cable that came with the device. The box containing all equipment will be mounted at head height outside the tank and the sensor will go in through a hole and rest o n the bottom of the tank. Loop will be the length of the cable supplied - 5m. All other bits will be inside the box and the shortest cable runs possible.
I have seen a diagram like this one attached. it shows connecting the vdd pin on the ads1115 to gnd on the esp32. Is this wrong?
Patrick
Maybe I have just answered my own question.
If my program is asking the esp32 to compare the AO to gnd then the vdd should be connected to ground?
Patrick
oops - that doesn’t really make sense. The ads1115 still needs power. Maybe they have the wrong esp32picture and the pins are not aligned properly?
Pat
Bothconnections should be moved down 1 place - VDD to 3V3 and GND/ADDR to GND.
Hi Patrick
As Jeff says, that drawing is incorrect. The Gnd should connect to Gnd (all Gnds should be connected) and VDD to 5V or 3.3V. You originally show VDD connected to 5V. Which one is it?
Cheers Bob
I’ve just connected the ads1115 to 5v and am monitoring the results I get now.
Could you explain the gain component of the ads1115 in more simple terms. Mine is currently set to 4.096 (just copying other people’s programs)
Pat
Hi Patrick.
I am not familiar with this device but you will find most of the info available here
You will have to step through the various pages to find all the info you want.
I just had a quick look and it seems you connect VDD to whatever voltage your processor uses, ie; if 3.3V use that. So you could have been correct in connecting this device to 3.3V.
Cheers Bob
Hi there,
still refining this project. I am trying to minimise battery size etc. Currently running off a 24v lithium pack.
Given my low power usage and use of deep sleep I should be able to run this with a dual 18650 battery in parallel.
Can I use the SKU: DFR0264 to charge a parallel 18650 battery setup?
Pat
Hi Pat,
That solar charger is designed for Lithium pouch batteries but since it has a cutoff voltage of 4.2V it should still work with 18650 cells provided they are in parallel (a series combination would require a balance changer).
Try to keep your batteries shaded and as cool as possible as they will heat up during charging, however you’re unlikely to be charging too quickly on solar anyway.
Hi Patrick
18650s in parallel will only give you a nominal 3.7V. A bit short of the 24V you have at the moment.
Even 3 18650s in series will still only be a nominal 11.1V. Still short of the minimum 12V operating range of the sensor.
To get into the quoted operating range of the sensor you would need 4 18650s in series (nominal 14.8V) and balanced charging. In my opinion you would be better off and much easier sticking with your 24V lithium and charging that. The charging balance and other battery smarts will be within the battery assembly and you should only have to apply the required voltage and current limiting to charge it.
Your 24V battery make up is unknown but if it is made up of 18650s in a series parallel configuration it will be highly unlikely to be a nominal 24V. It could be 22.2 or 25.9 or a bit different depending on what type of cell is inside. The charging voltages and current limits should be printed on the battery case. Take notice of these.
Cheers Bob
Bob,
thanks for that handy info. I am still planning what power source to use. The equipment I am using will use very little power. I am trying to mimimise space and a 3 x 18650 battery setup would work. I only need a step up converter to drive the sensor for about 2 minutes every hour - very little power draw.
I have been reading a bit about setting up a proper power path for solar/battery projects with concerns about the batteries being constantly on a minimum trickle charge - apparently not good for the 18650?
Then again I would also apparently need an appropriate lithium charge controller? Or does the 3s module SKU: FIT0869 do that job for me?
Pat
Hi Patrick,
FIT0869 is a battery management system that should cover all the protection needs you would normally need covered; short circuit, overcharge, overdischarge and also ensure the three cells are balance charged and don’t end up with a varied state of charge between the individual cells.
It won’t handle regulating the output so you get a constant 12V output at all times, you’ll just get whatever the battery level is, but it will protect the batteries and manage the safety aspect of using them in your circuit.
