Hi David,

1. This is right, but should not have too much of an impact on the 
actual power consumption or the results. I am considering an active 
circuit to compensate this. Also, a smaller resistor (e.g. 1 Ohm) would 
reduce this effect, but you'd need a better instrument to measure the 
resulting lower voltage values.

2. Currently, I do not use anything special as what I see already 
corresponds pretty much to what I expected (at least the form of the 
curve, not necessarily the measured values). LEDs are in my opinion not 
good for keeping track of the state of the application as they consume a 
considerable amount of energy themselves. You could use digital I/Os and 
record the evolution with additional channels on the oscilloscope.

3. I happen to have a Tektronix DPO4000 series oscilloscope (not sure 
about the exact model) in the lab where I work, so I use this. It even 
has an Ethernet interface with which I can control the measurements from 
a computer. I cannot recommend any particular oscilloscope as I simply 
have not enough experience. If you evaluate different oscilloscopes, pay 
attention to the inherent acquisition noise (when I short circuit the 
input terminals on my scope I still see noise with an amplitude of up to 
10mV). The noise can usually be reduced by reducing the bandwidth of the 
acquisition. The bandwidth is another thing you want to look at. 
Depending on what you want to measure, this should be from a couple of 
10s of kHz to a couple of MHz. Typically, a scope will acquire a wave 
form, which you then need to save before you can acquire the next wave 
form (or you can overwrite the old one without saving it). The scope 
should have enough memory to acquire a waveform over a meaningful time 
window with enough temporal resolution.

Cheers,
Urs


David Li wrote:
> Hi Urs,
> 
> Thanks for the clarification. I think now I understand your setup. Only a
> few more questions:
> 
> 1. Using your example, then the voltage on the mote itself is the remaining
> from 3v - 0.2v = 2.8v. Is this right?
> 
> 2. How would you correlate the measurements on the Oscope to the events on
> the mote? Did you have to instrument any code?
> 
> 3. I am curious about the model of the Oscope in your test. I know Janos
> used HP34401A. Is there any model you would recommend for this task?
> 
> Thanks.
> 
> David
_______________________________________________
Tinyos-help mailing list
[email protected]
https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help

Reply via email to