EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: ballsystemlord on May 28, 2024, 03:44:56 am
-
I was looking at oscilloscope specs (it's a bad habit I should learn to break ;) ), and I was trying to figure out how long it would take to fill up the memory of the scope. I've tried doing the math myself, but I'm pretty sure I didn't get math correct, so I'm turning to the community.
Please note, I don't really *need* the answer. I'm just really curious.
If I have a scope that samples at 5GS/s, does 64K wfm/s, and has 80Mpts of memory, how long would it take to fill the memory up?
Thanks
-
Hello,
If the memory is 80 Mpts for one channel, it takes (80 Mpts) / (5000 Mpts/s) = 0.016 s.
The device only makes the specified 64K wfm/s under certain conditions and this is certainly only the case if the memory used is <= (5000 Mpts/s) / (64000 1/s) = 78125 pts
Best regards
egonotto
-
Hello,
If the memory is 80 Mpts for one channel, it takes (80 Mpts) / (5000 Mpts/s) = 0.016 s.
The device only makes the specified 64K wfm/s under certain conditions and this is certainly only the case if the memory used is <= (5000 Mpts/s) / (64000 1/s) = 78125 pts
Best regards
egonotto
Because there's dead time, doesn't that mean that you can't just divide 80Mpts by 5GS/s? Like, you have to put the 64K-wfm/s into the equation somehow?
-
Hello,
there is no dead time in an acquisition. This only exists between the acquisition.
Best regards
egonotto
-
You can really only make assumptions for a single shot capture.
"64K wfm/s" doesn't really mean anything, it could easily be only with some obscure combination of settings and 10-100x worse otherwise even if selected time base would let you think otherwise. I actually have an oscilloscope that at fastest time bases dramatically reduces its capture rate to the level of dog shit (even with lowest memory depth) despite being pretty good in the middle :palm:.
-
Hello,
here is an example of the Micsig MHO3. The acquisitions are made in burst. The dead time between the bursts is much longer than in the bursts.
The images show the trigger impulses
Best regards
egonotto
-
I was looking at oscilloscope specs (it's a bad habit I should learn to break ;) ), and I was trying to figure out how long it would take to fill up the memory of the scope. I've tried doing the math myself, but I'm pretty sure I didn't get math correct, so I'm turning to the community.
Please note, I don't really *need* the answer. I'm just really curious.
If I have a scope that samples at 5GS/s, does 64K wfm/s, and has 80Mpts of memory, how long would it take to fill the memory up?
Thanks
You are missing one important parameter: at what time base?
Timebase defines how long (in time) will be every individual capture.
If we have 10div horizontally, and 100ns/div, every trigger event will initiate a capture of 1µs in length.
At 5GS/s that is 5000 points captured. Per every trigger.
If you have a 1MPTs scope that means you can have 200 such captures in a memory total.
ON a 10MPTs scope you can have 2000 such segments.
If your timebase is longer, there will be more memory used per trigger. So less segments available total.
Up to the point where you use up whole memory for single capture..
But this is one of those questions that are discussing a detail without explaining whole picture.
What is a problem you are trying to solve?
-
But this is one of those questions that are discussing a detail without explaining whole picture.
What is a problem you are trying to solve?
Recall from my initial post: "Please note, I don't really *need* the answer. I'm just really curious."
This is out of interest in learning about scopes. It has nothing to do with a real world problem.