Electron schematic help

I’ve had one Electron Asset Tracker fail and am nervous about trying out a second one without feedback on how I’ve wired it up.

I’m attaching a screenshot of my schematic (drawn in Excel). I’d appreciate any feedback on the design that may cause problems for the Electron board itself. (the GPS board seems to be fine)

  • The LEDs are a NeoPixels strip. Amazon
  • The servo is a generic hobby type and requires 5-7v to work Amazon
  • The beeper is a generic piezo beeper Amazon
  • The second battery pack has 4 AA LiOn cells totaling 7.2v maximum Amazon
  • I’ve connected the 4-cell pack to the GPS board to keep the electrons LiPo charged longer.

What is the intent of the connection between the Neopixels an VBAT?
VBAT is not the terminal for the LiPo output but is connected to 3v3 via a 0Ohm resistor and is meant as backup power input for the Backup RAM page and RTC.

I think you’d be better off using Vin or Li+ (depending on your intent) to power the LEDs.

1 Like

Thanks!!!

I didn’t know that, I was thinking VBAT was a positive power source.

This is exactly why I posted the question. Something messed up one of my Electron boards. And I didn’t want my backup to get ruined as well. The ruined board works OK until I attempt to use the GPS, then it goes into an endless reset loop.

The weird thing is, the device worked just as expected until it just quit. (when I set the LED’s on full brightness maybe?)

If I use Vin or Li+ do you think everything else looks OK?

Thanks again for your help!
Don

I’m sorry… I don’t see a Li+ pin, is it the one labeled 3.3 maybe? (connected to the switches in my schematic)

This is the Li+ "pin" - it's just a solder hole without a pin attached

The max 3v3 output current is 800mA on the Electron. Each NeoPixel when set to full white can draw 60+mA

This would require further investigation what exactly triggers the misbehaviour to understand the possible causes.

2 Likes

Thank you ScruffR,

It’s not lost on me that you always seem to help everyone in the community.
Thank you so much for all you do

I’m running the LEDs with the FX_MODE_CIRCUS_COMBUSTUS effect.

There are 62 LEDs in the strip, and probably more than 40 at max power when the brightness is set at maximum. I would have definitely exceeded the capacity.

1 Like

Do you mind explaining the difference between the two?

The device is a warning beacon, unless in maintenance mode... when the cover open switch is made the LED's light up to warn people of a safety risk.

The servo swivels to 90 degrees, the beeper chirps at regular intervals, and email messages are sent with GPS coordinates.

It's highly likely the cover open switch will actually power the device up if I can figure out how to do it with the LiPo battery.

When the device is permanently powered via USB or via Vin with no more than 5V you should go with Vin. Otherwise you should go with Li+ as Vin will either supply no power or the input voltage may exceed the voltage rating for the Neopixels.

Hence, with your 7.2V that will directly go from the extra LiPos to Vin, you most likely would go with Li+.

Thank you for your help,

Here’s the new schematic

Powering that many LED's will require over 2.5A (more like 3.5A). The USB port CANNOT provide more than 2A based on the USB spec. I don't believe powering off Li+ will give you enough voltage (4.2v max) or current to power the neopixels which require 5v. You should consider using a switching buck regulator to step down the primary 7.2v to 5v at 3-5A to power the neopixels separately.

Note that at 3A continuous draw for the neopixels, those batteries won't last very long.

wow, that certainly is a sneaky "pin" - thanks ScruffR!

While going with a 5V buck would be a good addition to satisfy the Neopixel specs, I'd just feel inclined to ask @thresholdengineering did your Neopixels work as expected off the 3.3V before the board failed?
If so, the 5V may be a "soft" requirement.

But 3.5+ A (LEDs + servo) will still be a challange.

I've been running the Particle InternetButton via the SparkFun Battery Shield and didn't have any issues with the Neopixels.

1 Like

True that!

1 Like

Thanks peekay123,

So far the Neopixels have not been a problem with the Electron’s LiPo alone.

I’ll test out the pixels with the Li+ source as soon as I can to see if that’s a problem.

The servo is getting it’s power directly from the extra battery pack. Only the signal/data wire is connected to the Electron.

Yes, that’s what caught me by surprise, I had the whole prototype up and running. Everything was working great after running quite a few tests. I shipped the device to my client and it was working on his end as well until… it just quit working.

That’s why I’m being so over cautious going forward. We are down to our last kit and I don’t want to have to buy another one if I can avoid it.

1 Like

It looks like the neopixel strip is rated at 5-6v (6v max), the servo is rated at 5-7.2v.

If I used 4 standard 1.5v AA batteries everything would be fine, but for remote deployment with extended idle periods (months) I don’t think alkaline batteries would work.

The Energizer ultimate lithium batteries spec sheet says 1.5v nominal voltage, but I checked them with a voltmeter and they are actually 1.8 volts

I found a 3-cell AA battery holder, or I can use the 4 cell holder and tap it at 3 cells for the neopixels.

I’m open to any advice you may have?

I found a 3-cell battery holder, this will drop the aux power from 7.2v to 5.4v with the AA LiOn cells at 1.8v each.

I’d like to not use the Electron’s LiPo so I can power it off completely for long term remote deployment.

I thought I saw a command to allow full functionality without the electron’s LiPo connected however I can’t find that command anywhere?

This is similar to what I was remembering, what’s unclear is without the LiPo I still need cellular and GPS when the power comes on.

https://community.particle.io/t/disabling-the-electron-red-charging-led-when-not-using-the-lipo/26604?u=thresholdengineering

That could also be achieved by splicing-in a switch on the LiPo wires.

Are you looking for the still undoucmented pmic.disableBATFET()?

Yes, I think so?

The idea is when the cover switch is open there is no power to the system at all. Once the switch is made everything powers up including GPS, cellular, and cloud communication.

When deployed remotely the device is idle for weeks or months until a hazard is detected at which time an alert is dispatched to emergency responders.

My fear is even at low-low-power mode the lipo will still drain and the unit will not function. The external LiOn battery pack should remain charged since it is disconnected, but with a dead lipo my understanding is the GPS and cellular will not function?