As you might know, i am running into some hardware issues with my mkr gsm 1400 on top of the connector carrier.
I am expecting more since all sensors will be receiving 5volts from the mkr board instead of using the carrier provided 5volts. Meaning that they would have to work at the 4.66VDC the board actually provides at the 5VDC pin, likely messing calibration around and making it harder to work with.
Also i do not understand why the carrier runs power to the sensors taking it from the board when there is a power buck right there. The only reason i could think of would be power stability into VIN since if a sensor takes too much current it could bother the mkr board loosing its own.
So i am trying to find out if it would make more sense to use usb to power the mkr board, remove the connection between board and carrier on pins VIN and 5VDC, then connect those two together at the carrier and feed that same signal to AREF.
From that hack i expect to make mkr power independent from sensors power, so intensive power draw from the sensors should not steal any power from the mkr board, protecting the brains from overexerting muscles. Also the tweak should provide the sensors with stable real 5 vdc at their vcc pins so they get what they expect instead of what the arduino board can provide.
What bothers me most is that after the tweak i may loose the ability to power off the sensors, if I had it in the first place that is. I do not know yet if the lowpower mode does power off the 5vdc line to turn sensors off.
Could you advice, hint or point precautions before i break the carrier, please?
Why is it a requirement to use the 5 volts from the board? I thought I could use the AREF pin to provide the reference for ADC to work properly. Not sure how that works with the carrier chips that handle the leveling up from the 3.3V the board can read, but as long as AREF and carrier side 5V do match i believe it should work.
Can you please elaborate, i am certain I am missing something (or many things).
I didn't say it is requirement.
you asked why the carrier connectors are powered from the 5 V pin of the board. I say it is necessary for the use case when the board with the carrier is powered from the USB connector on the board
Not the best wording on my part, I see it now. English is not my mother tongue and surely there are still mistakes I cannot notice.
What i meant is that I wondered why that pin from the board was connected (soldered) in the connector carrier design without the option to disconnect it. Since the purpose of the carrier is to offer voltaje level along with connectors and a power buck, i think the designer could have chosen not to force the connection of the 5V pin or let that pin be left unconnected. Just like i was planning. Knowing why the designer did what he did may help me understand what am I risking when i hack my carrier to use the carrier 5V instead of the board's.
the MKR Connector Carrier makes many assumptions about the expected use.
the analog input pins have voltage dividers and a capacitor. it makes them unusable as digital inputs and outputs and even unusable for fast changing analog input like from an AC sensor.
it has logic level shifters on all digital pin connectors which makes it unusable with 3.3 V modules.
many Grove modules can work with 3.3 V power and both 3.3 V and 5 V signal level.
there are Grove modules which require two digital pins but most of the connectors on the carrier have only one pin connected to MCU.
so maybe just don't use it if it doesn't suit what you need.
btw. do you know to what output voltage is the regulator configured? I didn't use it.
I powered the carrier and the MCU over the 5 V rail (really over one of the Grove connectors)
The output of the power buck at the carrier is 5.09 vdc while loaded with the mkr board, so I guess it is pretty stable.
So you ignored completely the power buck and used your own. I also thought about it, but felt like wasting an included buck able to deal with the 12 volts my current solar setup provides based on a 18 vdc panel and a 10 to 16 volts lead acid battery.
But in the end what I am aiming to accomplish with my hardware hack actually targets your setup but using the internal power ic of the carrier.
I wonder if they should release a second hardware version that should allow us both to do what we need without investing into a second power stage or risking integrity of the carrier.
Have you ever removed one pin of the carrier? My soldering skills are rusty, and my sucker rubber tip is already crunchy. So any piece of advice on removing the conection between the 5V pin of the board and the same of the carrier would be very much appreciated.
disconnect VIN and 5V between the board and the carrier, so they do not take precedence over usb power when the carrier is powered
feed the mkr board with the usb, since I still need it to be connected anyway for the sms to be sent (i will check if power is provided through the usb before anything)
connect 5V pin at the carrier to AREF pin at both the board and carrier, so sensors use the carrier power directly and report with that stable voltage through the leveling stages
I just noticed that would lead to having sensors and AREF powered even if my usb is disconnected. I have read that may break the board inputs, right?
Feels like the simplest way to fix my issue in the short term is connect a tiny usb cable with an old charger and leave it in the box. Not very professional, but I must have it working by the weekend for the first trial deployment even if incomplete.
I will keep this thread opened though. I think we can fix this properly with a capacitor and a resistor while Arduino people have a look at the issue at hand. I believe the usb cable and charger are just that for the mkr board when not powered. It could include a coil, but I rarely saw those in any equivalent circuit.
Not very professional, but I must have it working by the weekend for the first trial deployment even if incomplete.