In the power circuit there is a comparator that detects the presence of Vin and controls a FET switch in the USBVCC circuit.
I was hoping to connect this comparator to an I/O pin to allow the Arduino S/W to detect Vin. However experiments showed the output of the comparator (on two Elegoo PCBs) wasnt changing when Vin was inserted/removed. It was always 3.3V so the FET is always off (EDIT: See later posts. 3.3V is not high enough to turn the FET off. This should measure 5V).
It appears the potential divider resistor RN1B isnt correctly grounded on the PCB layout. The GND plane stops short of Pin 7 on the resistor pack RN1.
I include PCB layout and photo of the PCB marked up with arrows showing the the area I am talking about.
Is this a known bug?
When powered by USB only the Arduino still gets power but its via the body diode rather than the FET itself. The PCB fault means that (with Vin unconnected) pin 3 of the comparator is floating so sometimes the FET is ON and sometimes OFF and the Arduino power supply voltage varies by +/- the forward voltage drop of the body diode.
You mean always "on".
The P-channel fet turns off when the gate is the same as source (5volt).
There's another problem with the Elogoo R3 (and other clones). Wrong opamp.
The schematic requires an LMV358, but an LM358 or equivalent is used.
They are not rail2rail, like the "V", so they never switch the PC/USB connection off.
Leo..
Edit: Maybe the designer of the clone didn't understand the problem he was facing with the LM358, and give it more input voltage, hoping it would solve it
I suspect the designer was more interested in saving money rather than doing a decent job. Elegoo has a bad reputation (at least with me) because of poor design like this. Also because they are sold in some parts of the world as a genuine Arduino.
Ah ok so 3.3V on the gate isnt high enough to turn the FET off if source and drain are both 5V. I hadn't spotted that.
I was mainly concerned with the strange behaviour of the comparator. The unconnected pull down resistor in the Vin divider meant it sometimes gives the right output and sometimes didnt.
I've got another board ariving today from Amazon that should be correct.
Connect a 9volt battery to the DC socket of the Uno R3. Then measure the output of the opamp (gate of the fet). If you measure ~3.5volt, then the opamp is a common LM358 (wrong). A genuine Arduino measures 5volt there (switches the USB/fet off).
Leo..
Thanks. The two Elegoo Arduinos we have only produce about 3.3V on the amplifier/comparator output. We were also able to measure 5V on the USB connector (with nothing plugged into USB) so the FET is probably ON when it should be OFF.
Real Arduino arrived today and thats got a SGM 8542 amp. Thats a rail to rail device and is producing 5V on its output. Not seeing 5V on the USB so the FET appears to be OFF (eg all working ok). The PCB also has the pull-down resistor correctly connected as per the reference circuit diagram.
I brought this manufacturing mistake to their attention, with an email, several months ago.
No reply. That makes Elegoo to me another shit brand that doesn't care.
Leo..
