Here is the "instruction" for measuring the max. waveform update rate from siglent...
Somehow, the pdf was broken (thx to sergio for the info) - Now I´ve upload it again.
This Siglent "How to test the waveform capture rate" is, how I tell it polite and nicely, just poor.
What is purpose for it. Is it that Siglent here tell what is quality of producing products specifications or what is this.
In this example they give visual result is NOT average max wfm/s update rate what is meaningful if calculate example probability to get rare glitch detected in some time gap.
Then they just take one cycle and inside there measure two trigger event time interval equal frequency and get 135kHz. Now without more explanation then tell that wfm updata rate is over 120kwfm/s. Some may now think it can also be 135kwfm/s.
Even more, in practice with true signals in one special test setup ... do we really measure 10MHz pure sinewaves. Yes this is for test and also other brands do same. So well... borderline ok.
But in my thinking error is that they do not talk anything what is difference between this possible Peak wfm/s value and continuous average wfm/s. Are they shy for show this cap in trigged stream when start TFT update phase. Or why not explain bit more.
There is pause in capture before it start capture again, some other Siglent pauses around 25 times in second due to display process.
I have used external oscilloscope for measure trig out so that I measure whole display and other processing cycle or better if over two cycles. Yes there can be thousands of trigs in one cycle but this whole cycle period need measure for average wfm/s what is much more important than fastest peak value between two triggers what they now nearly show in instructions but without searching most short single period. Other method, without external oscilloscope, is use external counter what can set for this purpose. But there need be careful and user need really know what he is doing and know his counter principles. I have used example Agilent 53131 but it need manually set for this (trigger level and enough time for count so that this scope periodic capturing pause do not mess result).
I have not measured (yet) SDS2000XPlus but I do not claim 120kwfm/s is wrong. I think with my experience that it is perhaps enough right for tell wfm/s up to average what is usually what all advertise. I hope not so many advertise up to max peak speed.
Due to normal oscilloscope use and overall wfm/s value still this is not so important if it is highly accurate value because still it is true only with just said combination of settings.
Btw. It is fun to see that every time I have tested lot of more deeply these many models wfm speed there have been one finding. Nearly right value can get also without signal at all. Just trigger mode auto and without signal soon it start autotrig using minimum trigger interval.
Sequence mode is very different.
In this paper they just measure random speed what is perhaps near truth.
But this test performance is very extremely poor.
My opinion is that in Sequence mode we need know guaranteed speed, guaranteed minimum trigger interval and it can not at all measure like they do it in this paper.
This around same can also do without even external scope, so that take in signal and long enough sequence, then look segments time stamp from start to end and calculate average and this is just average it also is not guaranteed speed what is only what really matter. Why they did not tell it.
But....
This give some kind of value what is perhaps near right but it is not guaranteed max speed. Here with these Siglent models I have tested is some differences between models. Some model average speed and max guaranteed speed difference is more than in some other model.
Measuring guaranteed maximum is bit more complex. And there is also two methods for it.
What is guaranteed max speed.
With this constant speed of trigger events no ANY event is missed.
This do not even need trigger output, naturally.
Of course one other method is where also it can do but then need compare scope input signal and trigger output signal so that no any missing events but I do not explain it now because there is also possible traps depending method.
But one method is easy and bullet proof but bit slow to do.
This method use enough stable pulse generator what can work in single burst mode and there must not be high cycle to cycle jitter if want go to near guarateed max speed border reliable.
If we set oscilloscope for Single Sequence mode and example 50ns/div time scale.
Then we set max amount of segments. I do not know how many in special model just with this time scale, So if max is 50000 then set it.
Now also pulse generator need set for generate exactly 50000 pulse single period and set it for waiting start button or other start method and set for pulse speed what speed we want test if it pass. Say example 500kHz. (2us period)
Then scope in said mode (naturally trigger adjusted so that it trigs reliable with this signal we have set=.
Then start start scope for single and now it is waiting..
Start generator burst. Look if Sequence is ready. If it get all segments we can now say that guaranteed max speed is least 500000 segment/s after we repeat this test enough times without any fail.
Also we can now rise this pulse speed to example 501kHz and look if it still pass until we fuind limit... then step back and tests enough and if all pass this is max quarateed speed with this setup (selected used channel and used time scale and used memory length if we have tested some very different setup max speed)
Example I have tested some other models max quareteed speed have been even more than Siglent specs.
Many times this is not very important detail but in some cases it is very important to know thuis guaranteed speed so that we can trust that sequence have not dropped any events out.
This speed is very very different when we change time scale and also dependent about segment sample length. With slower time scales we can rise amount of max segments reducing segment length what also naturally then reduce samplerate.
In segment mode is perhaps much more important to really know its max guaranteed speed.
In normal visible mode we many times do things where wfm/s update rate is not very important.
If sequence mode drop out some segments it may be total end of game and whole test goes to garbage. There need be perfect sure. So user need evaluate his scope and finds also how to do it reliable way.
Home
Help
Search
Login
Register
