EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: Buzz 25W on April 11, 2013, 02:52:45 pm
-
Hi,
I am a new user of the forum and would appreciate help on this issue.
I am using a Rigol DS1102E to monitor an AM modulated carrier at a frequency of 3.875 MHz. Scope settings are 1.0 ms/div horizontal and 500 mV/div vertical with all other settings at default. I ma observing an aliasing artifact at/near the carrier frequency that makes my use impossible. Attached is a screen shot with the frequency a few hundred Hz from 3.875 MHz.
Is there any setting/method I can use to eliminate this artifact?
Thanks,
James
-
Turn Long memory on.
1ms/div it have 250kHz samplerate with short memory (it is selected with default setting). Of course it produce alias with all frequency components over 125kHz. (Nyquist limit) With this setting it can use only for signals what are under 125kHz, all over this produce alias.
With Long Memory samplerate is 20MSa/s if there is 1ms/div horizontal speed selected.
With this setting it can use up to 10MHz and all over this produce aliasing.
Also sometimes acquire mode "peak" is useful!
-
Dave recently did this video , about this subject comparing 2 Rigol scopes , for some more details .
EEVblog #451 - Rigol DS1052E vs DS2072 Oscilloscope (https://www.youtube.com/watch?v=_TSr9nFN1GU#ws)
-
Hi Buzz 25W,
Sorry to be the bearer of sad tidings, but modulated RF waveforms are the bane of all digital scopes.
The problem is the wide difference in frequency from the modulation to the carrier. You can acquire the modulation and suffer alias with the RF, or capture the RF and not see the modulation. It's an inherent limitation of the digital concept. Some expensive scopes ($10k plus) somehow simulate what an analogue scope does, but since I don't own one, I don't know how they achieve it.
If asked, I always recommend getting and keeping an analogue scope, then getting a DSO later. You'll mostly use the DSO, but will go back to the analogue scope enough times to keep it. In some ways, despite what the people flogging DSOs want us to believe, DSOs should really be called high speed waveform digitisers, since they aren't really a digital equivalent of an analogue scope.
Cheers, Colin
-
There is a simple solution: turn on 'peak detect' or 'envelope' mode.
-
Sorry to be the bearer of sad tidings, but modulated RF waveforms are the bane of all digital scopes.
I have used analog oscilloscopes tens of years. Good ones are very good for some special things and also today I use (mostly) analog scope specially in service and other things. Of course digital ones have things what analogs do not have but...I have find that I want kep my analog scopes and for some things some digital(s).
Example with <0.5k$ digital and < 0.2-0.5k$ good analog I think I can do more that with one 1-2k$ digital (in most cases and with my needs) Analog scope is always realtime and never need think is it true or alias what I see and becouse in work, time is money, they are many times much more fast to see things need look. And no need think is it some processed fake what I see or is it real.
But also with cheapest digital scopes can do something.
Here examples with Owon.
There is 5s persistence on so ther can see if there is false trigs. (exept picture where is "pek detect" mode.
First there is 3.699MHZ (just only becouse in finland this is fun freq) carrier AM modulated with 1200Hz and modulation deep 30%
(5% mod and there start severe problems for trig)
These two images with 3.7MHz carrier there is used 25MSa/s samplerate (so that more comparable with Rigol DS1000E) Of course there can use 1GSa/s if use 10M memory.
Then next two images where carrier is 145MHz, AM mod 30% with 1200Hz
Then same but AM mod depth 60%.
In all images HP8644B output RF level setting was same.
Persistence on so that there can see if triger fails.
Steady as rock.
-
I thank all for comments and "rf-loop" for the solution. Simply increasing the "memeory depth" to "long memory" resulting in changing the samplimg rate has resolved the problem at 3.875 MHz. I now much better understand the issue of aliasing of digital scopes and I undertand from "rf-loop" that this fix will work up to 10 MHz with the sampling capabilities of this low-end scope. That is probably good enough for me, and if I go higher than 10MHz and the frequency is not near one of the "bad" spots, that will be fine also.
Thanks, again for the lesson and the fix
Buzz
-
Some expensive scopes ($10k plus) somehow simulate what an analogue scope does, but since I don't own one, I don't know how they achieve it.
Cheers, Colin
$10k scope like ~$800 Rigol DS2072? Fortunately data visualization algorithms have been progressed since 90's from just drawing lines and dots. Take a look at Dave's video showing a comparison between DS2000-series scope and Tektronix analog scope, I think that the DS2000-series works just fine showing AM-modulated carrier without aliasing.
Regards,
Janne
-
Gooday Buzz,
Glad the guys on this forum could sort the problem out for you.
As I think I said "in another place",I don't have a DSO,so I had pretty much run out of ideas! ;D
73,VK6ZGO
-
Rigol DS1000 vs. Tektronix TDS2000.
Rigol DS1000 vs Tektronix TDS2000 - aliasing (https://www.youtube.com/watch?v=M40GsBdTT9M#)
-
It looks fine to me? What were you expecting?
As mentioned, try Peak Detect. If that's still not good enough, trigger on different parts of the waveform, zoomed, to get a better idea of what it's doing.
Are you trying to check linearity? Carrier distortion? There are always alternative ways of looking at the waveform that inform you of these sorts of aspects.
Tim
-
It looks fine to me? What were you expecting?
As mentioned, try Peak Detect. If that's still not good enough, trigger on different parts of the waveform, zoomed, to get a better idea of what it's doing.
Are you trying to check linearity? Carrier distortion? There are always alternative ways of looking at the waveform that inform you of these sorts of aspects.
Tim
The OP's pix show a waveform that looks more like an SSB signal---note the difference in the other displays which show a full carrier AM signal.
Note: Dave's video shows what is a "big no-no" to RF people---modulation in excess of 100%
-
The OP hasn't offered any other pictures? I don't see how you can get an SSB (or DSB) looking signal from AM unless the modulation is |sin t| to begin with (the only way to disambiguate these being carrier phase, which is invisible on any display at this scale).
But I guess it was just unlucky exactly at 3.875, so what's pictured is actually what it's supposed to be (given that the carrier (which, again, is invisible anyway, so we have no way to know) is off by a few 100Hz). So, we never even saw the problem :-// :P
Tim