ADC reference problem

[korda]

I see that RFE's A/D convector reference is wired to a VCC.
This introduces problem of measuring +VBAT if we running RFE from a battery. And there is no point to measure +VBAT if we running off VEXT.

Attenuating VBAT for ADC readings does not help - as battery goes down, VCC and HREF go down the same way, and so is the input voltage on the A/DC input!
Do you have a solution for this? We need to solve this ASAP, as we about to deploy hundred of RFEs for a test! Blind battery maintenance of 100's of nodes will be a problem.

[kbanks]

I've talked it over with some of our hardware guys, and we think we know what external components it will take to provide useful battery monitoring.

I think we can have schematics and example SNAPpy scripts by end of Friday (I have to order a part)

[korda]

Kevin, it is the end of Friday and we did not hear from you about ADC HREF problem - solution. Any luck with fix for it?

[kbanks]

Sorry I did not get back with you - the part I am going to test with did not come in on Friday.

Here is what we are thinking will work (but as I said, we have not been able to confirm this yet).

Analog Devices ADR510
Digi-Key # ADR510ARTZ-REEL7CT-ND


________
GPOUT >---/\/\/\/-----O-----| |---------O GND
10K | 1 | ADR510 | 2
| |________|
ADC IN <-------------- 3(nc)


( current draw when active is 150-250 uA )

# GPOUT may be driven high just during measurement to conserve power,
# or instead connect straight to Vcc to save an IO pin.

# Since Vref is 1.0V, the calculation is very straightforward.

Vcc = 1024 / readAdc(0)

[korda]

I think in the suggested circuit ADC will always read 1V.
I would try swapping resistor and the shunt - may be then we can read voltage changes across resistor. What do you think?

[kbanks]

We built it up and tested it using a variable bench supply.

Look over in the SNAPpy script exchange for a new entry...

[martinnossel]

FYI

The SNAP REFERANCE MANUAL, PG 8 SAYS

"
Channel 8 is an internal low voltage reference.
Channel 9 is an internal high voltage reference
"
Reading values from one rfe showed the opposite?.

Martin

[kbanks]

You are right, the docs have it backwards. I will correct them this morning, and the updated versions will be in the next posted software update.

Sorry for any confusion.