So I have a bunch of 1602 LCD modules with the I2c add on adaptor boards which are rated at 5v and box of ESP32 S3 dev boards which I got for a handful of magic beans on Aliexpress.
The ESP dev boards are showing quite a bit variability on the 5v pin - anywhere from 3.1v to 4.2v on the ones I’ve tested, which is a bit disappointing but not too bad for something that only cost me around 3 bucks each.
If I run the a 1602 LED module on a dev board that has an output around 4v continuously is it likely to cause any problems other than a reduced display brightness?
That doesn’t sound right. How are you powering these boards and how are you measuring the voltage? Exactly which dev board is it (the schematic will be useful, even for cheap clones)?.
I’m powering it with a USB cable from my PC & I’m measuring the voltage from the ground pin on the dev board to the 5v pin. I got a box of 10 of them for just over $30 about 4 months ago.
Edit: The USB is connected through a powered USB hub.
Maybe they offloaded them so cheap because it was a bad batch?
Everything else seems to work fine.
I read a post a while back on reddit or github where people were saying the output on the 5v pin can be as low as 1v with some of the generic S3 dev boards, that’s why I checked it before I connected anything to it.
So do you mean GND and V5IN? The reason I say that it doesn’t sound right is that dev boards for 3.3V modules (such as the ESP32) usually have the external 5V pin directly connected to the 5V supply from the USB port. 5V dev boards such as UNO have two sources for the 5V supply (USB and 9V regulated down to 5V) so they have a diode in one of those feeds, and that means the external 5V could be as low as 4.5V for a cheap clone. But for a 3.3V module there is no need for anything between the USB 5V and the external 5V, so I cannot think of a reason that the external 5V would be anything other than exactly the same as the USB 5V, and certainly not 1V! Have you confirmed that the supply from the hub is 5V?
Do you have a link to the site where you found that image? Is there a description of what the second USB port is for?
Perhaps. But if you have only paid peanuts then it’s likely worthwhile trying to find out why something was priced like that and whether or not there is an easy fix. Sellers often don’t know the value of what they are selling.
Check if that is an exact match to your board. The clue is the dual USB ports and that V5In pin, which I took to be a clone-like labelling for V5Out, but it really is an In!
We would need a circuit diagram to figure it out properly, but that link might put you on the right track. I will continue looking.
Jeff’s reddit rabbit hole find is promising! I’d additionally like to see if we could glean anything from your physical setup + front and back of your boards, if you wouldn’t mind sending a few photos.
Sorry for the slow reply, I was called out earlier & I just got back.
Yeah that’s correct I measured the voltage between GND and the 5V pins
I haven’t measured the voltage from the hub but I will make a test lead from an old cable so I can see what I’m getting from it. It seems like a likely culprit for the undervolt reading on that pin, unless there’s a faulty diode or switching IC between the USB socket and the 5V pin on the dev board.
That board in the reddit post seems slightly different to the type I have, mine don’t have the solder pads for the USB-OTG on the back or the IN-OUT pads on the front. I can’t see any unpopulated pads on my boards. The Neopixels work on mine straight out of the box too.
I have to go out again, so I search Ali & try to find a link to the boards I have later.
This seems to be the post I mentioned where they mentioned a low voltage output on the 5V pin, it seems to be specific to the model with the open jumper pads like the one in the reddit link Jeff posted.
The only obvious differences without seeing seeing the schematic is that it has USB-C sockets where mine have micro USB sockets and mine have two tiny SMD components near the socket where the USB-C model has just one. I’ll try to get a closeup pic so you can see the trace from the USB socket to the voltage regulator after I’ve fed the kids.
I ended up giving up on using those boards to drive the LCD. I think the 5v pin is just meant to be an input to power the board but there must be a faulty component allowing a transient voltage to appear on that pin when powered through a usb port. If there was a reliable 4V there like I saw the first few times I measured it I would have considered taking my chances driving the LCD with it but sometimes the voltage measured as low 2.3V so it seemed like a fools errand trying to drive a 5V component from it.
I ended up using another board I had which appears to be some incarnation of the board mentioned in the reddit link Jeff posted. The only difference is the board I have doesn’t have the RGB solder pads mentioned in the reddit post that need to be bridged to put the neopixel in circuit. The neopixel works straight out of the box but I had to bridge the IN-OUT pads to get a 5V output to appear on the 5V pin. It was actually 4.7V but it’s close enough for reliable operation of the LCD.
If it is an input then the voltage reading using a high impedance meter could be anything at all - the voltage will disappear when any sort of load is applied to the pin. Your board might be a case, often seen in clones, where a change has been done without a correct understanding of why the original was done like it was. Sometimes the change ends up making no sense at all.
A possible alternative to using a different board would be to find a separate supply of 5V, for instance a breakout from the USB cable.
That sounds like a good explanation of what I was observing. It would go a long way to explain why I was getting a lower voltage reading when I put the dev board in a unpowered breakout board compared to what I saw when it wasn’t in the breakout board.
I was going take that route and power it externally but dropping a another dev board in and soldering a bridge was a quick easy fix. The one I have now has the same pin count (44 pins) as the breakout board and the printing for the pins matches the new dev board, where the first one I was using only had 42 pins and most of the printing is blurred. So I will be less likely to mess up the wiring if I decide repurpose it or modify it some time in the future.
