Battery + micro SD shield for Teensy 3.x

Status
Not open for further replies.

ERoussel

Member
Hi,
I made a battery and uSD shield for the Teensy 3 with the following features:
- 1-cell LiPo battery power for standalone use of Teensy board
- Automatic switch to battery-powered or USB-powered when connected
- On-board battery charger (from USB), with charge status LED
- Automatic switch everything off when battery voltage lower than threshold (default: 2.95 V) => battery protection
- uSD card read/write
- Low-noise 3.3V regulator for reduced ADC noise compared to Teensy board alone
- Same size as Teensy

I currently use it as a datalogger with an ADC input.
Up to now I achieved up to 64 KSamples/second, 16 bits ADC, and uSD logging up to 2 KS/s (average on 32 samples), Teensy running at 48 MHz. This could probably be improved in the future.

DSC09157-r.jpgDSC09154-r.jpg

I would like to release code and schematics so that everyone can contribute to the project. What do you think ? Would you be interested ? Do you have any suggestion ?

Thanks for any feedback ! :)

[EDIT] sources available here : https://bitbucket.org/rousselmanu/teensy-datalogger
 
Last edited:
Looks nice. Yes, interested. Would love to see the noise difference in the ADC samples being powered by USB vs your low-noise 3.3V regulator.

I'd like to see the ADC differential inputs accessible by routing them to the side.
 
@pixelk: I'd love it ! For the moment I only have components and PCB for a few prototypes. But I'll definitely think about it.
@linuxgeek: Ok, I'll do the noise comparison in coming days.
 
Last edited:
Hi linuxgeek,
Here is a noise comparison between:
1- Teensy board alone, USB-powered
2- Teensy board + uSD/battery shield, USB-powered
3- Teensy board + uSD/battery shield, battery-powered
ADC and uSD logging configuration is as specified in the first post.

ADC-comparison.png

Zoom on second figure:
USB_Pboard_std0.2992.png

Zoom on third figure:
BAT_Pboard_std0.2704.png

Standard deviations are:
1- std=1.17 mV
2- std=0.30 mV
3- std=0.27 mV

[edit] : I should mention that for the three cases, the ADC input voltage is a given by a resistor voltage divider, going through a follower op amp. The circuit has its own 3.3V regulator (even in the first case). So in the first case, noise level would probably be much higher if using internal Teensy regulator.
 
Last edited:
Just a remark from my side: Looking at the data, imagining a histogram then the zoom on the second figure looks more or less gaussian, whereas the zoom on the third figure looks bimodal.
 
I'm surprised there wasn't more difference between usb-powered and battery-powered. Does your voltage regulator supplying 3.3V when it's usb-powered?
I wonder if you can record more samples per second. When I've done something like this before I would get transient spikes, and maybe you've managed to subdue those.

Thanks for sharing.
 
Does your voltage regulator supplying 3.3V when it's usb-powered?
Yes it does. The shield has low-noise 3.3V regulator with good filtering. That's why there is not much difference between USB and battery-powered when the shield is used. The voltage source is USB if connected, battery if not.
 
What do you mean ? Are you speaking about A10 and A11 ? They can be used as ADC diff input and are accessible on the Teensy board

I read this 'ADC diff' as wanting some of the underside trapped pins #24-33 brought out.

This board is using the SPI on the flash?

I don't suppose you put on a crystal and fed the RTC?
 
Status
Not open for further replies.
Back
Top