Let's try to bring this to an end because it's going round in circles.
At a normal setup its possible to see the single event graduation, or you can exaggerate it if you want. In everyday use with the intensity set for 30% (as I typically leave it at) the intensity of the single events are clearly visible, I have to put in really odd settings to make it harder to see.
That might be true for the type of signals you work with, but some of the signal I deal with suffer from very narrow runts, so small that it's unlikely you'll see them with your eye even on a persistence display because they don't stand out enough. And that is if you even know they are there. If not then you'd probably never find out if all you do is look at persistence mode.
It should be obvious that having to rely on the human eye to recognize deviations is a flawed idea. Not saying it never works, and if you see a glitch it's really there, but if you see nothing it doesn't mean there isn't anything.
Also, how do you intend to find glitches in non-repetitive signals with persistence mode?
If you can capture the glitch in realtime mode then you will see it on the screen, just for you I tired turning the intensity control to 0% and the traces all became a similar beige but still clearly definable against the background and graticule
I said non-repetitive. Especially mult-state signals, if you put them on persistence mode then all you'll get is a wall of traces. It's worthless. for that.
and single events were not any harder to see than the rest of the overlapping traces.
Where? What setup? What events? My glasball is in repair so it would be helpful if you'd provide a bit more details what you're actually doing if you try to refer to it. Attaching screenshots are allowed, too
How are the frequencies of the features important to what is seen?
Simply because, like all signal forms, even a glitch is just a combination of sine waves at certain frequencies, and very small glitches usually consist of very high frequencies. Even if the scope has sufficient bandwidth and sample rate, if your wanted signal is a lot lower then looking at the edges at persistence mode might not be visible enough for the human eye, even if the glitch appears very often.
That may not be understandable to you but I've seen quite a few engineers get caught out by this on multiple occasions during my career.
You're trying to argue this both ways that segmented mode with deep memory is great because you can capture all the frequent glitches, but realtime mode is bad because the glitches are infrequent.
No. My argument is that high real-time update rates are great as long as it doesn't come at the cost of memory and other capabilities. My argument is that the WR8K offers a better price/performance than Keysight's current competitor which is the DSOX6kA.
Realtime mode is suited to looking for unknown transient characteristics and unusual events in high speed signals, while segmented mode is suited to known events you can define a trigger for.
Yes, on the DSOX6k. On a somewhat modern LeCroy (as well as on a modern Keysight Infiniium), you can find known and unknown events without the need for persistence display. It doesn't mean you can't use it, you can if you want to spend time starring at a screen.
On the DSOX6k you pretty much have to.
Since you have this fixation on the Keysight 6000 series
I compare against the DSOX6k because itit's in the same price and market segment as the scope on topic. Actually, price-wise I could as well add the DSOX4k to the list, but that compares even worse than the DSOX6k.
I know that you already suggested I should compare against a much more expensive Keysight scope as this will be better
perhaps you'd like to refer back to the chart with accurate numbers I posted:
| Parameter | Keysight DSOX6004 1GHz | LeCroy WaveRunner 8104 | Keysight DSOS104A | Keysight DSO9104A |
| Segmented Trigger Rearm | 1us | 1us | 4.5us | 4us |
Where you can see all those scopes have similarly fast segmented capture rates, and the Keysight 6000 offers faster triggering in segmented mode than in realtime so you're simply wrong again on that too.
That's the re-arm time, which is not the only factor determining the update rate. I'm pretty sure the InfiniiVision scopes don't reach higher waveform rates in segmented mode than in real-time mode due to their architecture, which for the DSOX6k would be ~450k wfms/s.
What I'll do is if I have some time next week or so then I can search around the labs if I can get my hold on a DSOX6004 (I think I've seen one or two in one of the racks) and do some quick waveform update rate testing. Provided we still have them, as they weren't exatcly the most favored scopes (and I do know that they don't sell particularly well).
However, as you might recall, the waveform update rate was just one thing I mentioned (but for some reason that's the only issue you're fixated on), and frankly the waveform update rate is probably the least relevant of all of them.
What weights a lot more is the fact that the sample memory on the DSOX6k is tiny. 4MB max, turning into 2MB per channel in four channel mode. If that wasn't bad enough, this is only true as long as the sample rate is below 2GSa/s (for a 1GHz to 6GHz scope!). If you dare to use a sample rate of 2GSa/s or more then the available memory drops to 1MB in half-channel and 500k in full channel mode. Just to remember, that is a scope sold in 2016 that starts at $18k! It's a bit like Tek's TDS694C (10Gsa/s and 120k max memory) but that was back in 2000. I'd have thought we left the era of low memory scopes long behind, but apparently not.
Aside from a wide range of limitations that comes with it, the tiny memory also means that the 6Ghz DSOX6k (which can properly resolve a 6GHz signal in 2ch mode only anyways as in four channel mode the max sample rate is 10Gsa/s so well below Nyquist-Shannon) quickly runs out of sample memory and has to drop the sample rate, which reduces the usable real-time bandwidth further.
I'd really like to see how you spin this into an advantage.
The other thing is that the DSOX6k, like all InfiniiVision scopes, are built on a simple Windows CE platform. That means limited processing, limited memory (RAM), which in turn means a lot simpler and more basic tool sets. It also means no access to the OS and the ability to install user software. Although I'm sure it means nothing to you, in this class of scopes being able to run MathLab or other software is actually a thing.
With the DSOX6k, Keysight pretty much tries to compete with an entry-level platform (InfiniiVision) in the high-end market, which is a bit daft, really, even more so when that's their only option in this segment.
I have many times turned scope for infinite persistence and after lounge come back and look if there exist some anomaly - glitch or what ever. What kind of "blind point" you have here now.
It works with 20GSa/s scope and it works with lower end 2GSa/s Siglent or what ever. And what is difficulty to see these, they are there as long until you reset. Or is it just too simple way?
After this unknown anomaly is visible on the display it is more easy to think how to continue analyzing.
That is fine (and let's assume for a moment you can actually see the glitch), the difference is:
- you come back after lunch, look at your scope, see some glitches on the persistence screen and then only start thinking how to continue analyzing,
- I come back from lunch, look at my scope, I get a list with time stamps of every occurrence and the parameter of every glitch that was captured, plus a screenshot for each occurrence. Which means while you think how to analyze further I'm already on my way solving the problem
Not that I'm expecting a 2GSa/s Siglent scope (or any entry-level scope) to be capable to do this. But not having to rely on the human eye to identify a glitch is one of the advantages high end scopes have enjoyed for a very long time.
Persistence is a well defined thing, perhaps Mr W is confused from how Lecroy have implemented realtime on some of their modern scopes:
ru.tek.com/dl/48W-26394-0_0.pdf
LeCroy WaveRunner Xi-A Series [negative features include] Persistence not available in WaveStreamTM Fast Viewing mode.
As someone who actually knows these scopes and doesn't have to resort to Tek marketing material for information about LeCroy scopes I'm certainly not confused about LeCroy's implementation, thanks.
Regarding WaveStream, this is a specific mode in some LeCroy scopes that emulates the behavior of an analog scope, i.e.
including phosphor persistence. For example, on the old WaveRunner Xi-A Tek mentions the non-existing persistence is controlled through the intensity control knob in the top right corner (pressing it enables and disables WaveStream):
So yes, looks like Tek got it wrong (I guess they got confused because you can't select normal persistence mode in WaveStream which already is a persistence mode
), and maybe if they spent less time talking about the competition and invest more time building scopes that aren't old shit then maybe their market share wouldn't be in a constant decline and they wouldn't need to embarrass themselves by having to compare their scope with an older scope of a competitor
WaveStream is an additional mode, it doesn't replace the normal real-time and persistence modes that also exist in these scopes. And even in persistence mode you get more tools at hand than with the Keysight DSOX6k:
That's a simple square wave with excessive jitter on the trailing edge, as easily recognizeable on what is a standard persistence display found in any modern scope.
There's also color grading (that's also in the DSOX6k) but that doesn't give much more information in this case.
But there's also 3D persistence in the LeCroy:
The 3D image clearly shows that the jitter is linear and squarish in nature, which helps identifying the cause.
And you couldn't do that with the DSOX6k.
3D persistence is quite a nice tool if you want to examine a signal visually, as often you can immediately identify certain unwanted components, and can give a better view of the behavior of the UUT.
BTW, these screenshots are from a 10yr old WRXi I quickly put together for testing. Newer LeCroy scopes like the one on topic have a much wider array of tools available.
And you still think the DSOX6k is better value/money than the WR8k?
Home
Help
Search
Login
Register
