SDS814X HD FFT
A few assorted screenshots, all showing a 400 MHz span representative of you might use for EMI work.
Signal from SDG6022X, dual tone with the principle @100 MHz.
These results are only applicable to the scope settings in use. FFT 2Mpts remained unchanged although it's auto managed by the scope. Memory depth is just what it was, didn't check.
Higher FFT sweep speeds are likely possible if one takes the trouble to optimize them.
At 200us/div and max 2Mpts FFT a sweep speed of ~1/s is obtained.
100us/div = ~2/s
50us/div = ~4/s
As we increase only timebase the sweep gets faster where at 10us/div its speed is indiscernible and at 500ns/div detail is starting to be lost without changing other settings.
Max hold and Average screenshots thrown in for good measure.
All other settings should be plain to see in screenshots.
Dear Tautech,
Thank you very much to be the first to broke the ice with real facts!
This is the information I was expected from my post.
I will not answer to all of your messages but will try to summarize theme up to now. Please correct me if I'm wrong or add more if I miss something!
===========
Rigol DHO800x vs Siglent SDS800x HD
- FFT speed:
Rigol is 4 times (average) faster than Siglent
- FFT accuracy:
Siglent is better but slower than Rigol
- FFT features (Peak, Avg, Max Hold, customize parameters of the FFT such RBW, manual markers, Log scale, etc.):
Siglent is better than Rigol
- FFT color heat map:
Rigolis the only one with this feature (useful too)
- Balance between speed and accuracy: NONE of the scopes have such thing; you can chose only one of this features like speed=Rigol OR accuracy (and FFT better features)=Siglent (4 times slower)
Abot smell, Rigol have "annoying" smell, not Siglent.
=============
I hope this summary will be helpful also for the future reader in need to find an answer to those questions.
I will not conclude which one is better because it would be a false answer. Best is to say NONE of theme are good for this task or "it depends".
To be honest, only scope in less than 7000-8000 € range that has something resembling realtime SA is GW-Instek MDO-2000A series.
And that is because it literally has realtime SA mode implemented. It works well but it's ADC are only 8 bit so limited dynamic range. Also scopes are in 1500€ ranges, a bit old fashioned by now etc.....
All other scopes in that price range will have only FFT as a math function, with implementation variations.
Short version for the two scopes mentioned:
- if you want just something bouncing on the screen to see if there is something happening at some frequency, DHO800 will have faster refresh at some settings.
- if you want control of what are you measuring, accurate windows, accurate level measurements, Peak Hold and Averaging, various markers and peek markers and control of them, then SDS800xHD is the one.
If you want SA display on your scope that resembles R&S MXO4 or MXO5 or Tektronix MSO 6 digital down conversion tech, you will have to pay for those...
2N3055 (I'd liked that transistor anyways
![Smiley :)](https://www.eevblog.com/forum/Smileys/default/smiley.gif)
) thank you for this answer. I have no experience with GW-Instek scopes therefore I cannot judge your statement.
On the other hand (I'm an EMC engineer) I have experience with Keysight 20GB/s scopes used with FFT for EMC troubeshooting and they are good...in their price range.
One advantage of a scope vs RTSA is the dual more (time/frequency domain) and very wide real time bandwidth on the scope side (professional scopes I mean).
Not to mention the trigger functions that can correlate aggressor with observed emission, things that cannot be achievable with a simple SA.
A lats word about Rigol DHO804. The reason I'm stick to its FFT speed performance is that I've used it for the EMC tasks and it was enough for basic investigation but yes, in terms of markers and SA management is not very useful. The reason I've stopped to SDS824x HD is its better implementation of the FFT Math function and much closer look to a SA. Unfortunately its speed is not that good and I have no idea why they failed on that. I guess is due to the processing power available on the HW. Who knows.
Thank you very much to all of your for help.
Cristian