Yes, I know there is not trigger out but it can test also with different methods. Also segmented acquistion maximum trigger/acquistion speed is extremely easy to check. Just calculate from segment time stamps over example 1000 or 10000 segments.
That method doesn't work because you'll get the average time for triggering/re-arming and don't know whether you missed a pulse or not. The only way which works is to send a burst with 100, 1000 pulses (or even more) and adjust the time between each pulse so that the segmented recording captures each pulse (you end up with a number of segments equal to the number of pulses send). That way you know the worst case trigger/re-arm time.
Of course, who have told anything else. It give only information what kind of speed range we are. But this test even first time novice can do without aany experience, without any knowledge, with very simple equipments. Of course if people know and can do more perfect test it is better. More perfect more better.
But still it looks that no one who have hands on GDS2kE do not want do even simpliest tests for even some data. Of course if test method is explained reader can then think how to read these results.
So, why this silence about real data.
(Sidenote: It looks like "beauty contest" is more interesting. What this tell? Perhaps many peoples use oscilloscopes and other tools only for decorate they hobby corner. This is good wink also for manufacturers. They may start producing hobby corner decorating "looks like professional" boxes. Something like "credibility on the street" in camera world.
This business idea is free for all. Let's make "professional" "High-End" looks like cheap boxes for decorating hobby labs. (this same was times ago in audio world when some Japanese companies start making "professional" looks like total junk for peoles who want feel they are "audio enthusiasts". I remember well this trend in 80-90 era.) Of course if triggering/re-arming works exactly without any variations it give quite accurate result. In this case if input frequency is enough high we do not care dropped but we have pushed it to capturing segments as fast as it can. Oh, if nitpick then not exatly. In worst case if trigger event happend just before it is ready (name it time A) for next capture we need wait next trigger event in our input signal. If it is 100MHz we need wait 10ns - how much before A previous edge was. In worst case we need wait 10ns - very short time. So, in this case if scope can do exactly 1us period trig/capture we get this sad period 1010ns.
Situations change lot if oscilloscope triggering/re-arming have lot of variations. More variations and more wrong this method result is if we are looking speed what do not miss any pulse even in lot of segments. This is true.
But method is still so easy that even novice can do it without any good tools but some kind of signal generator what give enough high frequency.
But, if there is lot of variations in triggering/re-arming result is average and if variation is high lowest bulletproof segments/s speed may be lot of below average when we do not accept any pulse lost.
It is really really sad if segmented acquisition have widely random triggering/re-arming time. But all kind of shit can still name "segmented memory acquisition" in sellers advertisemenents. Good example is in Rigol where is just some kind of slow "frame recorder" together with very badly implemented mask test. Then we wander why lot of users underestimate these tools power in use. Of course they underestimate these tools because they do not have these tools or tools they have are just crap toys. After then we hear opinions that these tools are nearly for nothing. Yes, if they are shit they are just names in sellers advertisemenst features list. Of course they still may have some useful use.
Siglent SDS2000 is faster than SDS1000X. Processing power is faster and acquisition / digital side triggering engine is faster.
So, I take this slower SDS1000X and test it.
In new FW P06 they have touched to Sequence acquisition (I do not explain this change here)
There is big improvement in this function useability value but also perhaps some small disadvantages. (there is not free lounges)
Test: (test repeated also so many times that I can be enough sure (for me) and result is not just random lottery win)
This scope reach maximum segment/s speed with 50ns/div where segment lenght is 700 points and one channel in use.
With this setting maximum amount of segments is 45526.
I have previously measured max ~500k segments/s speed. Still now if use low amount of segments say example 100 it can "randomly" capture all without lost if burst 100.
Bit more hard test.
45526 segment.
50ns/div
CH1
OR CH2 in use.
Here do not matter anything if sinc on or off and if dots or lines etc. (of course, this is segmented memory acq.)
Signal 100ns pulse.
Burst lenght 45526 pulse.
Pulse frequency 486 kHz.
I run this 20 times sequentially and every time it capture every pulse. Exactly.
No any single misssed pulse. It can say that speed maximum is least 486ksegment/s
All other settings same but now signal 80MHz sinewave. (with this freq it do not reach maximum average, but near. Because cycle period is 12.5ns. Best case trig happend just when system is ready and worst case 12.5ns later if nitpick)
Read time stamps difference over 45525 intervals and result is average speed 497.25 ksegment/s
SDS2000 is bit faster than this SDS1000X
Home
Help
Search
Login
Register
