O-Scope - Capturing the correct/full signal

[logictom]

I've been probing a xbox 360 pad to see what sort of signals are being used for the button presses and I came across a "problem".
Looking at one pulse of the signal on the oscilloscope I get a repeating signal of 2.32V pic A
I zoom out a little further and I get 3 pulses of 2.32V pic B
Zoom out again and I get these 3 pulses repeating at 2.32V pic C
Zoom out again and I get a very different picture - these 3 pulses repeating but each set at a higher voltage repeating 8 times pic D
Pic E shows one of the repetitions and Pic F and G show the end of one and beginning of another the last pulse measuring 2.3V and the first 1.06V.

I guess this is down to a lack of understanding of setting up the oscilloscope or something similar but why when I look at a single pulse does it show a stable voltage at 2.3V when it is clearly changing? Is there a way around this of setting up the scope correctly or is it just a case of making sure you check out the entire signal from a large to small range on the time scale?

[Simon]

you need to post the pictures or we don't know what you are talking about

[logictom]

I have posted them and I see them in the post above...
I've taken a screenshot - I'll attach it to the post and post a link to photobucket in case it's a forum issue.
http://i116.photobucket.com/albums/o33/Timmymna/screenshot.jpg

[Simon]

oh dear we need to sort this out, neither opera, IE or firefox show the images. can you still see the images your end ?

[logictom]

Yeah still see them, do you see the attached pic in the second post or only the photobucket one?

[alm]

They all seem to work fine here in both Opera and Mozilla Firefox, either it fixed itself in the past hour or it may be something on Simon's end? The appear to be BMP images, maybe that has something to do with it?

[ElektroQuark]

No problem here with images, IE 8.

[allanw]

It's aliasing. When you sample a signal with frequency f, you need to sample it at at least 2 times the frequency or else you'll get a false image of the signal

[logictom]

It's aliasing. When you sample a signal with frequency f, you need to sample it at at least 2 times the frequency or else you'll get a false image of the signal

D'oh! Of course, but how does that relate to the oscope, it doesn't sample at 100MHz all the time? I'm a little lost here because when you hack the Rigol it goes from 50MHz to 100MHz meaning you can change the time to 2ns rather than 5ns but with my problem you have to increase the time from 100µs to 20ms? Can you point out where I'm going wrong?

[allanw]

It only samples at 100Msamples/sec when you're zoomed all the way in.

The scope only has a limited amount of fast memory to write the waveforms into. When you're zoomed out, it'll cut its acquisition rate down. It should be possible somewhere to view the sample rate or it'll be specified in the manual.

[logictom]

You've lost me there.
The sample rate goes up when you decrease the time scale? So I get 100M samples at 2ns but fewer at 100ms? Hence why it takes longer to show the signal at 500ms? Surely I should see the effect more so at a 100µs than at 20ms then?

[TopherTheME]

It's aliasing. When you sample a signal with frequency f, you need to sample it at at least 2 times the frequency or else you'll get a false image of the signal

Given the time base and frequency of the signal he is sampling, I would never have guessed this. After you mentioned it I noticed that the sampling rate isn't displayed on the screen. Is this just a Rigol thing? Most modern DSOs I've used tell you the sample rate in the corner of the screen so you know if you have to worry about aliasing.

[logictom]

It's aliasing. When you sample a signal with frequency f, you need to sample it at at least 2 times the frequency or else you'll get a false image of the signal

Given the time base and frequency of the signal he is sampling, I would never have guessed this. After you mentioned it I noticed that the sampling rate isn't displayed on the screen. Is this just a Rigol thing? Most modern DSOs I've used tell you the sample rate in the corner of the screen so you know if you have to worry about aliasing.

It is shown on one menu, at 100µs it samples at 2.5M whilst at 20ms it samples at 10k.
Edit - Dual mode, doubled for single.
Just in case anyone is interested:

[PeterG]

All images are visible here on Safari, Firefox and IE8.

Regards

[Simon]

yea we have established that it is a moderator permissions issue that maybe chris can fix as he has more powers that myself and geoff