Debugging WISP Hardware

This page is more focused on how to bring new hardware online.
Questions? - yeagerd yeagerd and - alanson alanson maintain this page.

Introduction

In the process of developing and manufacturing WISPs, the folks at Intel have come up with some general steps for verifying WISPs are doing what they should. This document is a work in progress, please update with your knowledge and insights if you go through hardware debugging!

See Also

For firmware issues, see Debugging Firmware.
For RF Front End issues, see WISP RF FE Measurement Procedures.

General Testing Procedure


1. Can the WISP rectify voltage from a reader? With a voltmeter, measure VOUT (the unregulated supply) while the MCU is in LPM4 (sleep) (see code examples page). The WISP should easily be able to rectify 4V to 5V within a few feet of the reader.

Debugging:
- power wisp VOUT with a power supply and measure current draw with a multimeter. It should be less than 10uA (probably 1-3uA).
- check the front end with a VNA (see alanson's tutorial on the HW FAQ)
- make sure your reader is on (can it read commercial tags?)
- the modulator might be burned out. remove the BJT or FET that is directly connected to the antenna

2. The sleep code can easily be modified to turn on the demodulator (I think here is some commented out code to this effect, otherwise, just ask if you don't know how). Turn ON the demodulator, and check the output of the level shifter with an oscilloscope (the data input to the MCU) to see if there is activity (bits from the reader).

Debugging:
- Measure the demodulator enable voltage to make sure its actually ON.
- Check the level shifter input to see if bits are happening, but not making it through the level shifter
- Make sure you replace the 1uF with 100nF as Michael Buettner corrected (however, this wouldn't cause you to not see ANY bits).
- Check the comparator inputs. The supply / neg ref should float around the average of the pos input.

3. Check the output of the supervisor while manually sweeping the voltage on VOUT with a power supply. DON'T GO ABOVE ~5V. In fact, 3V is sufficient to test, so play it safe! The supervisor output should transition from low to high in the 2.0V to 2.2V region.

Debugging:
- There's not a lot going on here. There's the supervisor input, output, level shifter enable, level shifter supply (regulator output). Check all the signals for sensible voltages.
- Its easy to make solder errors on the level shifter. Reflow or resolder the connections. A fine tipped multimeter probe can be used to test for signals at the pads.

3. If all is well so far, load up regular WISP code. Do you see activity on the data out (modulator) line with an oscilloscope?

Debugging:
- If you see modulator activity, but the reader is not receiving data, there's several possible problems. (1) the protocol isn't happening correctly due to timing errors, protocol mismatch (not Miller 4, ex), MSP430 oscillator error (not a problems these days), errant data sent by WISP (shouldn't be a problem with stock firmware). (2) the modulator is blown out. you can run through alanson's VNA tutorial and turn ON and OFF the modulator in the firmware to measure that the impedance is getting modulated properly (S11 changes by > 10dB).
- VOUT should hover around 2V at 1-3 meters distance. Less than 1m, VOUT should increase above 2V.
- Run the code on the debugger. Set break points to see if certain areas of the code are being reached, such as areas of the receive EPC state machine. We can chat if you get to this step.





The last modification was made by - yeagerd yeagerd on Sep 22, 2010 5:08 pm