EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: blkptr on February 20, 2015, 02:18:57 am
-
Hello everyone, my first post!
I have a Digilent Analog Discovery from fleabay and was wondering if anyone else is experiencing this.
When I first connect the unit and open the voltmeter instrument, the DC and True RMS readings seem
high with nothing connected. The Channel 1 is higher than Channel 2, typ. on the order of 3.5 to 4.0 mV (0.0035V).
The other channel is smaller, but still in the mV range. I have another Discovery that does not exhibit this
behavior. On the other unit, voltages are usually in the 100 - 300 uV range, typ. hovering around 300 uV (0.0003V) and eventually seem to settle around zero (0.0000V).
What's interesting, is that on the unit that has the higher voltages, over time, say about 20 min. or so,
the voltages go down (as it heats up?). When they do go down, they seem to hover around 800 uV Channel 1 and 300 uV
on Channel 2.
Does anyone know if this is normal variation between units?
I've done some other minimal testing, and everything else seems ok so far...
Edit: ranges correction
-
Honestly, I've never used the voltmeter on mine - just the scope, network analyzer and AWG. But fwiw I just fired mine up. The picture below shows the values soon after it was opened. They did drift down into the microvolt range after a few minutes.
I really don't think the open circuit readings matter much. The voltmeter will have a high input impedance and like many other meters will read values in the mV range when open circuit. Of course what maters is what it reads when measuring.
-
Thanks for that mtdoc. I'll keep running tests... hopefully everything else is ok.
-
I wouldn't take much notice of readings with unconnected inputs. Connect the two inputs to a common voltage and see what the readings are.
-
In case you need more accurate measurements you may want to recalibrate the device.
See the following post: https://forum.digilentinc.com/index.php?/topic/275-offset-on-oscillosope/ (https://forum.digilentinc.com/index.php?/topic/275-offset-on-oscillosope/)
-
In case you need more accurate measurements you may want to recalibrate the device.
See the following post: https://forum.digilentinc.com/index.php?/topic/275-offset-on-oscillosope/ (https://forum.digilentinc.com/index.php?/topic/275-offset-on-oscillosope/)
Ha! Thanks attila, I didn't know it was possible! I checked earlier against my new Brymen 869s and one channel was slightly down, the other slightly up on the brymen, then I thought it would be great if I could calibrate. Nice one :-+
-
Thanks attila, I will take a look at that thread. Last night I was tired and running from memory.
Today I plugged the unit back in for a second look, and noticed that actually the voltages are
even higher than what I mentioned in my first post. I guess the Voltmeter instrument automatically
changes the number of digits. The column for Channel 1 was showing 4 digits, while Channel 2
was showing 5. What threw me off even more (which I only noticed today) was that Channel 1 range
changed pretty drastically from 4 digits to 5, and back again. For example, Channel 1 was showing
0.026V, and then all of a sudden 0.0026V, then back to 0.026V again a few seconds later. This went
on for some time... then seemed to settle at 5 digits. I did not notice such sudden changes on
Channel 2.
Attached is a screenshot of the Voltmeter instrument a few seconds after being plugged in.
-
Thanks attila, I will take a look at that thread. Last night I was tired and running from memory.
Today I plugged the unit back in for a second look, and noticed that actually the voltages are
even higher than what I mentioned in my first post. I guess the Voltmeter instrument automatically
changes the number of digits. The column for Channel 1 was showing 4 digits, while Channel 2
was showing 5. What threw me off even more (which I only noticed today) was that Channel 1 range
changed pretty drastically from 4 digits to 5, and back again. For example, Channel 1 was showing
0.026V, and then all of a sudden 0.0026V, then back to 0.026V again a few seconds later. This went
on for some time... then seemed to settle at 5 digits. I did not notice such sudden changes on
Channel 2.
You need to understand the concept of quantisation of a digitised signal, and consider that there will be a little noise on the input. Then all will become clear.
In the meantime, I suggest you connect the inputs to any convenient voltage in the range 0.5V to 10V (e.g. the 5V PSU output) and repeat your experiment. I'll bet you see 5V+-2mV or so.
-
Thanks tggzzz, I do actually need to understand much better the concept of quantization, digitization, etc...
At the moment I'm merely interested in knowing if such seemingly large variation between units and channels of a single unit is something
that represents a defect, or something that can be adjusted. The fact that the unit starts out at tens of mV on Channel 1, and only on Channel 1,
and then slowly goes down seems interesting to me. I understand that DVMs in general will always flirt around the low mV and uV ranges
open circuit...
I'm going to grab hold of an accurate meter and go through the calibration process pointed to by attila. Meanwhile I thought I would post more
about my findings in case this thread should become useful to someone else in the future.
-
Thanks tggzzz, I do actually need to understand much better the concept of quantization, digitization, etc...
At the moment I'm merely interested in knowing if such seemingly large variation between units and channels of a single unit is something
that represents a defect, or something that can be adjusted. The fact that the unit starts out at tens of mV on Channel 1, and only on Channel 1,
and then slowly goes down seems interesting to me. I understand that DVMs in general will always flirt around the low mV and uV ranges
open circuit...
I'm going to grab hold of an accurate meter and go through the calibration process pointed to by attila. Meanwhile I thought I would post more
about my findings in case this thread should become useful to someone else in the future.
The figures you quoted most definitely aren't a "large variation". To make it more concrete, consider what you would see if you had a voltmeter accurate to 1uV, and were measuring a signal specified to the 0V +-1mV.
For quantisation, what's the resolution (in bits) and the full-scale reading of the ADC - from that it is trivial to work out the nominal size of each step.
You might as well find out about "offset errors" and "gain errors", and why the two are independent.
As for open circuit - well you are measuring whatever's in the air!
-
I went through the calibration steps with the most accurate meter I have, and it did make a difference
for the better. It's still not as good as my other (newer) Discovery, but at least now I know that
it would respond to adjustment.
Until I get an even more accurate meter, I don't know if I could get the two very close to each
other... but it'll do for now.
Much thanks to all who have been helpful.