Power Issues with Feather M4 Express Moderators: adafruit_support_bill, adafruit

[Draculen]

Good Morning and Happy 4th of July!

I recently finished one side of a massive array of LED's Powered by a separate Feather M4 (It looks amazing in my opinion!)

I am currently setting up another array of Neopixel RGBW LED's with a completely new Feather Board via Circuitpython. The way the project works is it has 2 different logic pins (A2 and A4 in this case) that run there own animations side by side. On the first array of this project it worked perfectly fine wiring them up. This time around I can only seem to get power to one side of the LED's (Pin A2) and when the other Neopixels have there power connected the feather shuts down and the USB starts to heat up . The pinouts are setup as follows

For both neopixel sides/pins - 5v (power) from Neopixel to USB Pin (Spliced together), Data wires (A2 and A4, separate ofcourse), Ground on both Neopixels are spliced together and connected to the main ground Pin of the Feather.

When I De-soldered the one side of the Neopixels (Pin A4's power) the Feather boots up and works as intended. As soon as I touch the 5v/power lead to the neopixel again the board once again shuts down and the USB Jack starts to heat up again. I am completely stumped as this is the exact same setup I had with the exact same code and no matter what I try I cannot seem to have both running at once without the Feather shutting down.

The only difference is the first project I used a butt splice for the ground and power leads where as this time I made it look more fancy by wrapping the leads and soldering them together correctly (kind of like a twist tie which is then soldered and heat shrunk.

Any information about what is possibly going wrong would be greatly appreciated. Thanks

[adafruit_support_bill]

Sounds like you are overloading your power source. If you have a lot of pixels, you need to provide a suitable power source for them. And you should run the power directly to the pixels - not via the Feather.

Pixels can draw as much as 60mA each (80mA for RGBW). 'Typical' applications will average around 20mA per pixel.

[Draculen]

I will be upgrading the power source after I finish this side off the project. As I said in my previous post using the same technique I am powering 90 neopixels without any issues or dimming

My issue is I'm using the same technique/wiring only this time I'm only powering 10 neopixels. Only one side is receiving power and when the other power lead connects the whole feather shuts down and starts to heat up. So knowing that I've already powered way more neopixels I am just at a loss at what could be causing this issue when no variables have changed other then its a different m4? Any insight would be greatly appreciated!

[adafruit_support_bill]

My issue is I'm using the same technique/wiring only this time I'm only powering 10 neopixels.

In that case, you probably have a short circuit somewhere in your 10-pixel configuration. If you post some photos showing all the soldering and connections we can take a look.

[XRAD]

1) 90 Neopixels need their own distinct power supply, even if they work fine now
2) agree w/short...or a bad neopixel in the '10' strip. Don't use it anymore.
3) rather than soldering and deosldering (leads to trace damage), make your build modular. Add a JST/other type connectors (assortment box M/F w/wires cheap) for the feather pins , power supply , neopixels, and data lines to a proto-board so that you can easily swap out components.

[Draculen]

I apologize for the delay. I think you may be right that its a short. I forgot that electricity doesnt really move in the direction of the arrows for the neopixels so it could have a short far up the chain thats shooting backwards towards the feather?

Here are some photos. the left hand side (Red lead for data) the first 2 strips work fine, the right handed strip (white lead for data) is what disconnects when the power lead is touched from the board to the 5v pad

https://drive.google.com/file/d/1JxTdRK9r3A7eaYE6_WXXX_32KuU9shYY/view?usp=sharing

https://drive.google.com/file/d/1_ISNoqlcHbGq-09UFbZVY_dU882wW4HB/view?usp=sharing

https://drive.google.com/file/d/1b8eEIwvX6qjeRnh5m_gFintbWfSEmbeg/view?usp=sharing

https://drive.google.com/file/d/1x-VQcYIg-_2tyUdLlhKeg-6dc2yqHL5B/view?usp=sharing

https://drive.google.com/file/d/1mBF2N8btmE_qZEIASX4fwsKhpXScFx0r/view?usp=sharing


Thank you in advance for the support

[XRAD]

Nice work! lot's of soldering! many places for a short......some pixels face up, and most face down..perhaps connections got out of sync?

Tip: If you use pixel 'arrays' in your code, you could just turn on segments of the whole strip as needed......and use one continuous flexible strip for each side.......

the 'arrows' indicate the data flow, not current....so that the first pixel is addressed firstly, and each one down the arrows is in turn addressed...You can power and ground from either end of strip, or in the middle. You may consider several power connections...

[adafruit_support_bill]

I don't see any obvious short circuits in the photos. You will need to use your multimeter to isolate it.

Set your multimeter to the continuity test setting if it has one. Otherwise, set it for resistance measurement.

Touch the black probe to the GND pin at the start of the strip. Then touch the red probe to the 5v pin. If it shows continuity, then you have a short.

To locate the short, disconnect the second half of your strip and test the two halves separately. When you find out which half has the short, separate that into two halves and test those individually. Continue until you have isolated the pixel or pixels with the short.

[Draculen]

This is fantastic advice! I will start troubleshooting it straight away. Thank you for the assistance!!!!!

[Draculen]

Just wanted to update and say this worked wonders! after checking for continuity (never knew it was called that!) the strip is now working great!

I am not sure if it would be more wise to just make another post for another request but I know XRAD mentioned getting another power source (and I most certainly have to! I did an actual count of pixels and its about 400 (2 different pins have 103 , each side of the project has 206 pixels. If I could have any advice on what sort of power source or item from Adafruit would aid me in powering this project (it is a portable sword) I was thinking a 12v battery with a buck converter to step down such a high voltage? I have never needed more power then what the microcontroller needed so i'm in uncharted territory despite reading the uber guide for power some things just still sound so foreign to me!

Also here is an updated photo(ish!) https://drive.google.com/file/d/1P9ASBY ... sp=sharing

Thanks in advance!

[adafruit_support_bill]

I did an actual count of pixels and its about 400 (2 different pins have 103 , each side of the project has 206 pixels. If I could have any advice on what sort of power source or item from Adafruit would aid me in powering this project (it is a portable sword)

Worst case for 400 pixels would be 24A(!!!) if you set them all to full intensity white. You should be able to get by with considerably less - especially following some of the advice in this guide: https://learn.adafruit.com/sipping-powe ... s/insights

I was thinking a 12v battery with a buck converter to step down such a high voltage?

There is no advantage to going with a higher voltage supply and a buck converter. Any voltage conversion will incur some loss of power because the process is not 100% efficient. For portable Neopixel projects with 3.3v processors, the most efficient solution is usually just to power everything directly from 3.7v LiPo cells.

To decide on what battery or batteries to use, you need to know a few things:

How many hours do you need to run on a charge?
What is the average current draw in normal operation?
What is the maximum expected current draw?

Based on the answers to those questions, you can calculate the capacity and drain-rate requirements for your battery or batteries.

I have never needed more power then what the microcontroller needed

In your system now, your pixels are already drawing much more power than the microcontroller. Even when off, the pixels will be pulling about 400mA. That is easily 10 times what the M4 needs.

[XRAD]

As long as you don't plan on having ALL the pixels on at full brightness for any length of time, for longer lasting power and portability, I would try an 11.7v 7000mah lipo (w/deans connector) with a 10A stepdown buck convertor for the neopixel power and a separate 1.5 a stepdown for the board....you can get fancy with a latching switch https://www.pololu.com/product/2813 or just a simple SPST latching push button....for the main power ON/OFF....


or get yourself a regulated bench-top power supply and the set the voltage and run your sword, and see what the required max amps are....then you know for sure....