| Electronics > Beginners |
| Load cell question |
| << < (3/4) > >> |
| wraper:
--- Quote from: german77 on October 10, 2018, 11:54:44 pm ---Here I'm back after testing. Just to say that it works without letting the magic smoke escape. My theory was right from the beginning, I don't need to connect E+ and E-. The only problem is that the rice lake 720i it's not calibrated any more as the HX711 introduces an offset error. --- End quote --- That's not how it's done and very likely you'll eventually blow up either HX711 or your scales. Currently current is sunk through input protection diodes inside of HX711. --- Quote ---Nothing to worry about. I could simply add resistors to increase the HX711 impedance. And this error usually is eliminated on the tare. --- End quote --- :palm:, nothing to worry about except you'll have incorrect readings on both scales and HX711 |O. And eventually something will fail. |
| german77:
There is something that you aren't taking in account. The HX711 doesn't have input protection of any kind. There is also no over voltage in the HX711 and it will not fail. For one simple thing It's not ground referenced. For example A+ pin will have 5.0003V, A- pin will have 4.9998. The only voltage that the HX711 sees is the 0.5mV. Its like touching a 5kV wire with one hand and another wire 5.001kV with the other hand while you are standing in a non-conductive floor/air. Are you going to be electrocuted to death. The answer is no. Because the voltage difference it's 1V. If the power supply weren't isolated and they shared the same ground then the HX711 would fail in a blink of an eye. The resistors solution din't work because there is not enough current on the input of the opamp as expected. I can turn down the HX711 via software and it will remove the offset on the scale. If I only take a reading really fast once in a while the scale wouldn't even notice. And I get correct readings on both devices. |
| wraper:
--- Quote from: german77 on October 11, 2018, 01:33:48 am ---There is something that you aren't taking in account. The HX711 doesn't have input protection of any kind. There is also no over voltage in the HX711 and it will not fail. For one simple thing It's not ground referenced. For example A+ pin will have 5.0003V, A- pin will have 4.9998. The only voltage that the HX711 sees is the 0.5mV. Its like touching a 5kV wire with one hand and another wire 5.001kV with the other hand while you are standing in a non-conductive floor/air. Are you going to be electrocuted to death. The answer is no. Because the voltage difference it's 1V. If the power supply weren't isolated and they shared the same ground then the HX711 would fail in a blink of an eye. The resistors solution din't work because there is not enough current on the input of the opamp as expected. I can turn down the HX711 via software and it will remove the offset on the scale. If I only take a reading really fast once in a while the scale wouldn't even notice. And I get correct readings on both devices. --- End quote --- Yeah, you completely don't understand what you are doing and shit you are talking about :palm:. The reason it still somewhat works is exactly due to ESD protection diodes on inputs of HX711 as they sink the current to either Vcc or GND. Otherwise there easily could be over 100V AC between GND and inputs (Y cap in PSU). Do you effin understand that inputs of HX711 are not isolated? And the fact that HX711 does or does not take readings does not matter. In negatively affects scales regardless of that. --- Quote ---he resistors solution din't work because there is not enough current on the input of the opamp as expected --- End quote --- Yeah, and why do you think significant current flows there to begin with? |O |
| David Hess:
The 5 volt common mode voltage from the load cell needs to be level shifted to the roughly 2.5 volt common mode voltage that the HX711 expects. Assuming that they do not share a common, this can be done by connecting one of the excitation lines with a 2.5 volt offset however the HX711 will no longer be running ratiometricly and there is a good chance they will interfere. The HX711 also might have troubles with no excitation load. While not ideal, and instrumentation amplifier could also be used to make the 2.5 volt level shift while connecting the negative excitation lines together but the same issues still apply. |
| NivagSwerdna:
--- Quote from: german77 on October 10, 2018, 04:24:55 pm ---I need to measure the weight of some trucks. I got a proper measuring device "rice lake 720i" that allows me to print receipts to a serial printer. --- End quote --- That's a very capable controller. Might be easier to suck out the measurements over serial. "11.12 Alibi Tracking" in the manual suggests they are logged. And then compare with what is declared. Or just put a CCTV camera over the station. |
| Navigation |
| Message Index |
| Next page |
| Previous page |