SparkFun Forums 

Where electronics enthusiasts find answers.

Discussions on how to get your MSP JTAG programmer up and running.
By ibelieve7
Hi, I've been working with the ez430-RF2500 boards and so far everything is OK with the programming. I was trying to get the ADC input to work with an analog accelerometer. My code was tested to be fine with a variable voltage source to fake the input. But when I connect the accelerometer output (0 - 3.0V), it only reads in 0.0V. I think there is something to do with the input impedance of MSP430 chip. But I can't find it any where except this document:

says the Ri is about 2kohm. The accelerometer output is hooked up to a 1kohm and 0.1 uf filter. I am not sure if I need to add a high impedance in parallel with the output.

I hope someone can help, thanks.
By ibelieve7
The accelerometer output was tested with a DMM.
User avatar
By leon_heller
It won't be an impedance mismatch, you should still get something. Check your connections carefully.

By ibelieve7
Thanks Leon. But I check connections usually.

I tried a 1M ohm resistor in parallel to the input and it worked, at least got me the readings I expected. I'll do a little further test.
By OldCow
I am curious. A 1 M-Ohm in parallel with a 1 k-Ohm will only result in a 0.1% change in resistance. How could that have any significant effect?

Can you show us that 1 k-Ohm and 0x1uF filter you mentioned in your first posting and what is this 1 M-Ohm in parallel with?
Code: Select all
      1k    |
           === 0.1uF
            V Gnd
By ibelieve7
I didn't put the 1MOhm in parrallel with the 1kOhm. My last post may be a little confusion. But the circuit is as above.

The reason I added it this way is that I found when I use a DC power supply as the ADC input, the ADC channel sinks about 40 mA of current which is pretty high for this application. I tried several ADC inputs to make sure it's not because of one broken channel.

With that 1M resistor, the source impedance is much higher than the ADC Ri (2k according to that TI link, might be wrong). And it works. One draw back is that it has a RC delay of ~ 40 us (assuming 40 pf charging capacitor in the ADC channel, again according to the TI link)..

If you have other solutions, please bring it up.

Code: Select all
>----^^^----+--------------+-----------------> ADC input
       1k   |              |
            |              |
           === 0.1uF       >
            |              > 1 MOhm
            |              > 
            V Gnd          | 
                           V Gnd
By dries
Hi ibelieve7,

i think i have a similar problem then you had before implementing
the 1 M resistor.

My problem goes like this: when i put a DC-voltage with a voltage source onchannel0(A0),the converting goes very nice. Putting the same voltage
on a different channels(A1,A2,...) always gives different values for the same voltage-level.

So, is it possible that different channels have different impedances? And that this is the reason why the sampling always give different results/samplingNumbers?

Any help would be highly appreciated

Best Regards,