Beiträge von thinkorange

I have made progress!

Sticking with the 19v HP laptop power supply measurements of 9v battery are accurate and steady. It may vary 0.01 of a volt.

Going back to the 500 ohm resistor method

I tried physically distancing the Shelly from the power supply and shielding it in a metal box and these made no difference.

My breakthrough is in powering it with a 18v cordless drill lithium ion battery. Steady readings!! No fluctuations and behaves as it should.

So the 2 totally different power supplies 19.5 and 12v both introduce severe noise affecting measurements but only when the pressure sensor is in circuit. BTW The sensor has always been a couple of meters away inside the tank.

So I have a way forward. I can look further at using separate power supplies or somehow suppressing the noise. No access to an oscilloscope so guessing a bit!

Thanks for the help!

Thanks for the reply,

I switched ECO mode off and reduced R1 to 500 ohms then to 250 Ohms. ADC readings are still fluctuating by 10-20%.

I have also tried a current to voltage converter instead of the resistor method. Sometimes known as HW-685. Mine is: Generic Current-Voltage Transducer-Signal Converter 0/4-20mA to 0-3.3V5V10V : Amazon.com.au: Electronics

With the converter I also get a stable voltage output as measured with a multimeter but the same bad results on the Shelly UNI ADC.

I only have one pressure sensor and one shelly uni so can't swap them out.

Next logical steps as I see them are below (they may take me a while)

1. get a 24V and/or 36V supply (I don't have these and the e-bike battery voltage is too high)

2. power the sensor and the shelly separately

3. try the voltage measurement with a different ADC - I may have an ESP32 somewhere (which I will also need to power separately).

Meanwhile I will lift the lid on the tank to check its level

Still using the 19.5V PSU, I tried the filter. It made zero difference - volts read by ADC are still fluctuating wildly. Filter was a 1k ohm resistor in the line between the pressure sensor minus and a 47nF monolithic capacitor between ADC and Sensor GND.

Thanks for your reply thgoebel.

I don't have a 24V supply but I did just try a 19.5V DC supply with 2A output and the result is the same.

The sensor I have is this one ....

Water Level Sensor, 4-20mA Throw in Type Water Level Sensors, Anti Bending 0 5m Range Depth Input Liquid Level Transmitter, 6m/19.7ft, 316 Stainless Steel, with Waterproof Cable: Automotive: Amazon.com.au

Some of the description says 24V, some 36V but also "9-36V Wide Voltage Power Supply".

I may yet try it with the 36V battery from my e-bike.

I reconfirmed that at the same time that the ADC of the Shelly UNI is fluctuating the multimeter is quite stable.

I wonder if I need to add a low pass filter (resistor with cap to ground) on the input?

I will look into that further and report back if I have any luck.

Hi

I wonder if someone has some ideas to help?

I have a simple circuit with a Shelly Uni to measure depth of a water tank using a 2 wire 4-20mA pressure sensor.

I cannot get the ADC reading stable although the ADC is stable when measuring other things (VCC itself or a battery source).

The pressure sensor has stable current (to 0.1mA) when measured with a multimeter.

The pressure sensor shows a stable voltage (to 0.01V) across R1 with a multimeter.

The Problem: Voltages measured by the ADC are in the 4-5V range but are bouncing around by 0.5 to 1V. No two readings are the same.

I am using a 12V DC 2A power supply and it remains stable when measured with a multimeter.

My thoughts:

- ground loop? (I can't find it)

- damaged ADC on the Shelly Uni (it is new and has not had any other use).

Does the ADC have any special requirements such as pull-up resistor?

Thanks for your help.

pasted-from-clipboard.png