Frequency measurement with EasyESP

[Olbi]

A very very low cost soil moisture sensor with NE555 Timer IC and EasyESP

It's very simple to get soil moisture dependent frequency by using some isolated wires, two resistors and a NE555 Timer IC (see Attachement-Picture). The great problem is, to measure this frequency with a NodeMCU or a simple ESP8266!

kapazitiver Sensor.jpg (157.49 KiB) Viewed 11604 times

Yes, it's possible to take a low pass filter, a voltage devider and so on for using the ADC Pin. But in reality without spetial calibrating the usable Range of the 10bit resolution was very very small, in my case:

ideal
Air: 32 kHz (19 pF) 1,34 V ADC 404
Water: 8 kHz (60 pF) 1,17 V ADC 347

real
dry soil 10,7 kHz (45,1 pF) 1,21 V ...
wet soil 9,4 kHz (51,2 pF) 1,19 V ...

... that's not fine and it's not my target to use the ADC


I would like to (direct) measure the frequncy with EasyESP (see Attachement Circuit Diagram). It does not have to be 100% accurate!

If I use the "Pulse Counter" Device, it's counting ..., but it's counting without end! Is there a possibility to count the pulses within a special time? ... with "Rules"...? Or is there a other possibility?

555 and NodeMCU.png (19.02 KiB) Viewed 11604 times

[toffel969]

A very very low cost soil moisture sensor with NE555 Timer IC and EasyESP

It's very simple to get soil moisture dependent frequency by using some isolated wires, two resistors and a NE555 Timer IC (see Attachement-Picture). The great problem is, to measure this frequency with a NodeMCU or a simple ESP8266!
kapazitiver Sensor.jpg


Yes, it's possible to take a low pass filter, a voltage devider and so on for using the ADC Pin. But in reality without spetial calibrating the usable Range of the 10bit resolution was very very small, in my case:

ideal
Air: 32 kHz (19 pF) 1,34 V ADC 404
Water: 8 kHz (60 pF) 1,17 V ADC 347

real
dry soil 10,7 kHz (45,1 pF) 1,21 V ...
wet soil 9,4 kHz (51,2 pF) 1,19 V ...

... that's not fine and it's not my target to use the ADC


I would like to (direct) measure the frequncy with EasyESP (see Attachement Circuit Diagram). It does not have to be 100% accurate!

If I use the "Pulse Counter" Device, it's counting ..., but it's counting without end! Is there a possibility to count the pulses within a special time? ... with "Rules"...? Or is there a other possibility?
555 and NodeMCU.png

May be an arduino via i2c/serial?

[rayE]

Dont you just set up a pulse counter device and set it to be delta/total/time? iv not tried it but im presuming the delta value will be pulses per second?

[rayE]

I gave the pulse counter a try today. Set up as follows.

My setup:
v2.0.0-dev12

Device = Pulse Counter
I/O = CH5
Type = delta/total/time
Mode = Rising
Delay = 10s

I injected a 3.3V p-p 1Hz signal into the I/O pin. For the first minute or so it seemed to work, the time variable gave an accurate representation of the frequency. Im not quite sure what delta does, it just seemed to count up and reset at some random point. It looks like it can only measure very low frequency's as the resolution of the time variable is 1ms.

Then after running for a short while the ESP started rebooting frequently. I disabled the Pulse Counter device and no more reboots so it looks to me like there is a problem in the device code?

Here is a dump of a reeboot.

Code: Select all

Exception (0):<CR><LF>
epc1=0x40218600 epc2=0x00000000 epc3=0x00000000 excvaddr=0x00000000 depc=0x00000000<CR><LF>
<CR><LF>
ctx: sys <CR><LF>
sp: 3ffffc20 end: 3fffffb0 offset: 01a0<CR><LF>
<CR><LF>
>>>stack>>><CR><LF>
3ffffdc0:  00000005 c0032025 00000000 40218614  <CR><LF>
3ffffdd0:  40107334 00000005 00000000 40107382  <CR><LF>
3ffffde0:  ffffffff 00000020 00000001 40107382  <CR><LF>
3ffffdf0:  ffffffff 00000020 00000000 00000022  <CR><LF>
3ffffe00:  3fffc200 401072fc 3fffc258 4000050c  <CR><LF>
3ffffe10:  400042e4 00000030 00000016 ffffffff  <CR><LF>
3ffffe20:  60000200 00000000 3fffff00 08000000  <CR><LF>
3ffffe30:  400042db 00000000 00000000 00000000  <CR><LF>
3ffffe40:  0000ffff 3fffc6fc 3ffeee3c 3fff76fc  <CR><LF>
3ffffe50:  000002f4 003fc000 60000600 00000030  <CR><LF>
3ffffe60:  000002f4 0848ddde 60000600 00000030  <CR><LF>
3ffffe70:  40243578 00000030 0000001c ffffffff  <CR><LF>
3ffffe80:  40243633 00000008 ffffffff fffffffb  <CR><LF>
3ffffe90:  3ffea854 40243620 00000000 bfffffff  <CR><LF>
3ffffea0:  ffffffff 3fffc6fc 3ffeee3c 3ffeed74  <CR><LF>
3ffffeb0:  3ffeee60 0848ddde 60000600 00000030  <CR><LF>
3ffffec0:  40000f58 00000000 00000020 00000000  <CR><LF>
3ffffed0:  00000013 40243620 3ffeee3c 00000001  <CR><LF>
3ffffee0:  ffffffff 3fff37d8 3ffeee3c 3fffdab0  <CR><LF>
3ffffef0:  00000000 3fffdcb0 3ffeee88 00000030  <CR><LF>
3fffff00:  00000000 400042db 00000064 60000600  <CR><LF>
3fffff10:  40004b31 3fff76fc 000002f4 003fc000  <CR><LF>
3fffff20:  40104e5e 3ffeee60 3ffeecc0 40107dbc  <CR><LF>
3fffff30:  402430e5 3ffeecc0 3ffeee60 0848ddde  <CR><LF>
3fffff40:  3fff76fc 00001000 40243586 00000008  <CR><LF>
3fffff50:  4010561c 00000000 40243633 3ffeed74  <CR><LF>
3fffff60:  3ffeee60 02969f0f 3ffeee60 60000600  <CR><LF>
3fffff70:  40246331 3ffeed74 3ffeee60 0848a2a7  <CR><LF>
3fffff80:  40246376 3fffdab0 00000000 3fffdcb0  <CR><LF>
3fffff90:  3ffeee70 00000000 40000f65 3fffdab0  <CR><LF>
3fffffa0:  40000f49 00033e03 3fffdab0 40000f49  <CR><LF>
<<<stack<<<<CR><LF>
bBÊdžBrûΦjzÿÿR†ÃgrBec<DEL>¤çüINIT : Free RAM:26032<CR><LF>
INIT : SPI not enabled<CR><LF>

Regards
Ray

[Domosapiens]

How about, first dividing by 100 with a 555 circuit
http://www.talkingelectronics.com/proje ... ts.html#I2
and then a a rule that counts for 100 seconds?

[rayE]

Even /100 would give you a very poor resolution! as i said from my brief experiment when the pulse counter worked it looked to me like the minimum resolution is 1mS. So i would think divide by 1000 BUT the pulse counter looks like it still has a problem? maybe someone else with a frequency generator can try it out? I could be wrong

Ray

[Olbi]

Now, after some experiments I learned:

• "Delta" in Counter Type equals to "Count" in Values Name,

• Counter Type "Delta" counts pulses per "Delay" seconds in Values Name "Count",

• max. frequency with "Debounce Time" 0 is 1000 Hz,

• "Mode Type" LOW don't count, RISING and FALLING counts normal, CHANGE counts double

Example
Debounce Time: 0
Counter Type: Delta
Mode Type: CHANGE
Delay: 5 [seconds]
GPIO pulses with 200 Hz
Count = pulses * CHANGE * Delay = 200 * 2 * 5 = 2000

Ok, max. 1000 Hz is very low. So I have to use a / some frequency devider ...

[rayE]

Hi Olbi,
What firmware version are you using? i take it you had no problems with the esp rebooting?

[Olbi]

I'm using GPIO 10 on a NodeMCU V2 with firmware v2.0.0-dev12 without deep sleep connection and I think there are no reboots.

[Drum]

I have a few of these:
https://www.tindie.com/products/aprbrot ... or-rev-21/

They do work, but it is not a perfect solution. Putting a PCB in wet soil will eventually cause it to fail without really god sealant, and to be accurate you need good voltage control. (in air at 3.3V readings are 860 to 900, at 2.8V readings are 710 to 740). Of course coatings also affect the readings, and it is a little short for my use as well.

I need something with at least 2 different sensors at different depths and external antenna. Finding an inexpensive and reliable sensor is the biggest challenge so far.

[harada57]

A space is missing after -m! And a host name with -h would be nice....

วิธีเล่นบาคาร่า