6
\$\begingroup\$

I am doing a personal project with a Wemos D1 mini and one thing that was of interest to me was to detect a servo-motor stall. I was very glad to find an useful answer to this question: How do I detect servo stall with an Arduino?

I re-created the schematic provided by @KryanF (which seemed to be very useful) and gave it a shot, but using a submersible water pump (since my servo is still on the way). According to the pump data, it should take 0.12 A when powered at 3.3 V or 0.18 A when powered at 5 V.

I powered my micro-pump with 3.3 V and use a digital multimeter to verify the running current and it shows, indeed 0.12 A. When reading the ADC pin, I get values around 120 (they are varying between 100 and 140 with some peaks of 160.)

Now, let's start from this statement (from the above question):

Let's say for a value of 25 read on the ADC, 25 units * 4.88 mV/unit is 122 mV on the ADC input in the real world. The required current through the sense resistor to get 122 mV across it follows Ohm's law. V = IR, so 0.122 V = I * 0.1, which is 1.22 A.

In my case, the value per unit should be 3300/1024=3.22 mV/unit. So, a read from ADC of 120 translates in 386 mV. I also used a 0.1 Ω resistor, but when measured it show 0.3 Ω, so I'll be using 0.3 as R value.

My calculations shows like this: 0.386 V = I * 0.3, therefore I=1.28 A. Wow, this is A LOT for that micro-pump.

I realized that I can get the real value if I divide that by 10: 1.28 / 10 = 0.128 A, which is a much closer value to the reality. Verified this also when powering the pump with 5 V.

Well, I am not an electronics expert, just a hobby enthusiast with software experience who wants to do some home automation project. Regarding all the electronics, I just read the web.

This is the schematic I made:

drawing of schematic

Since I cannot power my micro-pump directly from a digital pin (since it outputs only 12 mA), I am using a transistor (2N2222A) connected to 3V3 and to the control pin (D1.)

I did some more measurements with the multimeter and, apparently, the A0 analog pin actually reads the correct value (no need to divide or anything else). The problem that I have is actually multiple problems:

  1. When reading A0 pin with no load on the micro-pump (D1 pin is LOW) I get 1024 (the same way as the A0 would be connected to a 3V3 pin). Is the current directly flowing through the micro-pump when it is idling?
  2. When the micro-pump is working, with D1 control pin on full load (with digitalWrite(HIGH),) I get on the A0 about 300 mV (the same, more or less as I get with the multimeter.) With Ohm's law this means about 3 A, but in reality (measured with multimeter) I have 0.8 A.
  3. When I try to power the micro-pump with a lower voltage, by controlling the D1 pin with analogWrite, for example an 80% duty-cycle with analogWrite(205) I get on voltage of about 1.2 V on A0 (and also by multimeter.)

If I apply the Ohm's law on that circuit I get, in theory, 12 A! That is crazy, indeed, because the current measured with a multimeter when the micro-pump is running on 80% duty-cycle is actually about 0.4 A.

There must be something that I am missing here, since this circuit cannot be used to actually measure the current drawn by a DC motor.

BTW: I also tried to connect the end of the second BAT42 to the transistor emitter instead of the collector, but I get the same results.

\$\endgroup\$
4
  • 2
    \$\begingroup\$ A 0.1 ohm resistor should not be 03. ohm. That's most likely a measurement error. Short the leads of your ohmmeter together and read the resistance. Next measure the resistance of the 0.1 ohm resistor, and subtract the value from the first reading from it. You could also just trust the 0.1 ohm resistor to be closer to correct than your meter. Typical hobbyist meters (in the hands of typical hobbyists) don't do well when measuring small resistances. \$\endgroup\$
    – JRE
    Jan 20, 2022 at 9:00
  • 2
    \$\begingroup\$ The microcontroller on Wemos D1, is an ESP8266, which has an ADC range of 0~1V. because the 1V is not practical for hobbyists, board makers like wemos, use a voltage divider in the input so the user can measure values between 0~3.2 V. these not-so-precise resistors add some errors and lower the resolution of the ADC. (but not 10 times worse!) \$\endgroup\$ Jan 20, 2022 at 9:19
  • \$\begingroup\$ @TirdadSadriNejad, that is indeed the case. See this question and answer on Arduino StackExchange. \$\endgroup\$
    – StarCat
    Jan 20, 2022 at 11:20
  • \$\begingroup\$ The analogue reading on the Wemos is extremely inaccurate and is also highly-dependent on whether your WiFi stack is powered up or not. The only reliable use for it is to detect if a signal is under or over a certain threshold voltage, but even then, you must calibrate your threshold by measuring what the readout is for a given voltage with and without the WiFi stack running. They will NOT be the same. \$\endgroup\$
    – pfabri
    Jan 20, 2022 at 14:02

1 Answer 1

1
\$\begingroup\$

The 2N2222A has a saturation voltage between 0.2V and 0.3V with this load current. This is the biggest part of the voltage you measure here, the additional 12mV across R2 are the minor part. You just don't measure a voltage across R2 alone here.

If you use a P mosfet as switch in the positive path of the motor and connect the lower end of R2 (and the cap) directly to GND, things are much better. Remove the lower (second) diode.

Another question is, whether the ESP ADC can properly read voltages below 0.4V, some can't.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.