Siglent SDS800X HD Review & Demonstration Thread

[gf]

Why do you use a lower signal level with the DHO1074?

It looks the same to me:
0 dBm ~= -13.5 dBV (for a 50  Z0}

Sorry, I missed dBm vs. dBV scale.

[markone]

Why do you use a lower signal level with the DHO1074?

It looks the same to me:
0 dBm ~= -13.5 dBV (for a 50  Z0}

That's correct, dear friend 

[markone]

-snip

Let's reverse the picture: if you create dual tone, let say 1MHz 0dBm combined with 1,1 MHz -80 dBm and feed that to Rigol what you will see?
Will you see -80dBm peak?
Following your logic, from your FFT plots this -80dBm should stick out nicely from noise floor.

Here we go with test signal below the  theoretical resolution limit 12x6 = 72 dB, I adjusted the vertical scale to fullfill the ADC range with 0dBm signal with 81mV/div :







Then I applied a 40dB attenuator at the SDG2042x output :



and I started to step down the signal level obtaining the following result :















as expected the DSO is going well below the theoretical limit of 72dB and I would say that I like what I see ;-)

For some reason @ -56dBm  (-96dBm after attenuator) the peak is there with correct level but it does not appear in peak list, my guess is that the Peack list function is not working on the averaged signal but instead on the "clear" trace, I would call it a bug.

Of course I expect that with a 2 components combined signal we will get some spurious here and there, but of course this will NOT cancel signal below 72dB full scale.

HDO's measure coming.

[markone]

Here below DHO measures, same signal approach with DSO input set at 81mV, first spectrum image with 0dBm signal, then the addition of 40dB attenuator and subsequent  signal level stepping down.

Being the DHO math system an utter shit compared to the SDS800x, we here are missing the average function on FFT (AFAIK), dBm scaling, so I stopped at -50dBm (last image, -90dBm at DSO input), where the signal peak is still clearly visible at 90dB under the full scale theoretical level, so the famous -80dB FS signal "stick out" with some margin (last but one image) :













Anyway, shame on Rigol for the missing FFT average feature, this would change a lot the whole experience.

[2N3055]

For some reason @ -56dBm  (-96dBm after attenuator) the peak is there with correct level but it does not appear in peak list, my guess is that the Peack list function is not working on the averaged signal but instead on the "clear" trace, I would call it a bug.

For peak to be detected, both detection level and excursion needs to be set.

[joeqsmith]

Vintage 8-bit DSO input -110dBm 1MHz applied.

[mawyatt]

Interesting we see some signal "artifacts" on the Rigol #0 but not on the Siglent #52. Signal is from AWG (SDG2042X) at 2mvpp @ 50 ohms (-50dBm). Both have Flattop Windows, Signal has 4 average (not available on Rigol). BTW these artifacts originate within the Rigol and not from the AWG.

Used a cheap 50 ohm termination and connected to both DSOs, and no the artifact is not from the Siglent contaminating the Rigol or vice versa as both were also checked by themselves!!

Both read the correct level, altho the Rigol is in dBV as it has no dBm reference ability and both produce a nice FFT however we prefer the Siglent display and FFT setup.

Edit: Added another plot with Rigol at a lower sample rate #1, and with Blackman Window #2 and Siglent with Blackman Window #53.

As always YMMV.

Best,

[joeqsmith]

-120dBm 1MHz applied.  Wrong tool for the job but gives some comparison of vintage tech. 

***
Should add that the display update rate, with the scope set to maximize computations is about 7 seconds!  It's too bad that used CPU upgrade did not work out as I am sure it would help in this case. 

[2N3055]

SDS2000xHD 1MHz -125dBm applied.

[2N3055]

SDS3104xHD 1MHz -125dBm 1kHz 100% AM modulation applied.

[mawyatt]

Impressive results from the LeCroy and Siglent

@ 2N3055 the two sidebands are down 6dB as they should be 

Best,

[joeqsmith]

Very impressive to see that AM modulation.  How long does it take to display? 

[2N3055]

Very impressive to see that AM modulation.  How long does it take to display?

Long.
Each acquisition run is 1 sec (100ms/div) , some time for FFT and then 16x averages. Didn't actually measure time. Maybe later.

[2N3055]

Impressive results from the LeCroy and Siglent

@ 2N3055 the two sidebands are down 6dB as they should be 

Best,

Mike,
that is what I wanted to show. 

Best,
Siniša

[Electro Fan]

SDS2000xHD 1MHz -125dBm applied.

2N3055, thanks for doing and posting the measurements.  A couple questions and maybe a request, please, thanks.

1. Why is the amplitude measurement different with the 2kHD vs 3kHD?
2. What happens when you run the modulation test on the 2k?  Similar results but just takes longer to complete?

[Performa01]

Since this thread is about the SDS800X HD, you shall not get the impression that it cannot do something similar. It can do it, despite its FFT producing spurious responses at lower RBW (error report has already been filed) – it is just a little less accurate and I had to look for a sweet spot in the spectrum where there were no spurious signals getting into the way…


SDS824X HD_FFT_1.77MHz_-125dBm_AM100_10kHz

The signal was -125 dBm at 1.77 MHz, 100% amplitude modulated with 10 kHz.

The level measurement is 1.175 dB high, and the sidebands are -5.668 and -5.276 dBc.

Well, there should be a little gap between a low end SDS800 and the 2000/3000 series.

Regarding the speed, it takes about 1.28 seconds to produce a single 2 Mpts FFT result (including acquisition). Even tough I’ve used 64x averaging in this example, usable (but less nice looking) results can be expected with just 4 averages already.

[joeqsmith]

1MHz -125dBm 1kHz 99.9% AM modulation applied. 

You can see the center spur and side lobes but they are lower than the scope's internal noise.   Scope also will not readout this low. Turning off the 3D view does speed up the display but still a few seconds.  Guessing most modern scopes would blow the doors off it.   

My SA can easily measure down this low.  Note the -135dBm peak readout.   I normally keep a 10dB attenuator attached.   Using the right tool for the job can improve the measurements. 

[joeqsmith]

Very impressive to see that AM modulation.  How long does it take to display?

Long.
Each acquisition run is 1 sec (100ms/div) , some time for FFT and then 16x averages. Didn't actually measure time. Maybe later.

I assume it uses a round robin buffer and once loaded up, you're about a second update.  Much faster than mine. 

[markone]

For some reason @ -56dBm  (-96dBm after attenuator) the peak is there with correct level but it does not appear in peak list, my guess is that the Peack list function is not working on the averaged signal but instead on the "clear" trace, I would call it a bug.

For peak to be detected, both detection level and excursion needs to be set.

Of course this was the First thing that I checked, settings are ok but peak is not detected.

[markone]

The first thing I noted was the seemingly random usage of window functions. Hanning and Gaussian are not comparable and for a meaningful test we should stick with the Flattop window on both instruments.

-snip

I was aware about that but the focus was on spurious so I spent very little time on that so I chose the window type on the base of the peak shape well knowing to be in "uncal" condition, but to be honest ... i do not know how to select flattop on SA ... you sure it's possible with my model ? (SSA3021X)

The good news is that higher HD models behave in a much better way and the problem is (sadly for me) related to the SDS800x product line, but considering the price, I agree this is fair.

[2N3055]

1MHz -125dBm 1kHz 99.9% AM modulation applied. 

You can see the center spur and side lobes but they are lower than the scope's internal noise.   Scope also will not readout this low. Turning off the 3D view does speed up the display but still a few seconds.  Guessing most modern scopes would blow the doors off it.   

My SA can easily measure down this low.  Note the -135dBm peak readout.   I normally keep a 10dB attenuator attached.   Using the right tool for the job can improve the measurements.

I agree this kind of work is absolutely job for the SA. But like you, I was curious where the limits are.
That is the first think you should know about your instruments, what they can't do...

[markone]

Interesting we see some signal "artifacts" on the Rigol #0 but not on the Siglent #52. Signal is from AWG (SDG2042X) at 2mvpp @ 50 ohms (-50dBm). Both have Flattop Windows, Signal has 4 average (not available on Rigol). BTW these artifacts originate within the Rigol and not from the AWG.

-snip

For what is worth, my DHO1000, same signal :





Here below what my SSA read :

[2N3055]

SDS2000xHD 1MHz -125dBm applied.

2N3055, thanks for doing and posting the measurements.  A couple questions and maybe a request, please, thanks.

1. Why is the amplitude measurement different with the 2kHD vs 3kHD?
2. What happens when you run the modulation test on the 2k?  Similar results but just takes longer to complete?

One is without and other one with modulation. Difference is about 1,5dB. I'm guessing it's the AWG when modulation got on.. I didn't care because what was important was accurate 6db relation.

On 2k looks pretty much the same and pretty much lasts the same, because at this settings most of the time is acquisitions and averaging not FFT processing time.

[mawyatt]

SDS2000xHD 1MHz -125dBm applied.

2N3055, thanks for doing and posting the measurements.  A couple questions and maybe a request, please, thanks.

1. Why is the amplitude measurement different with the 2kHD vs 3kHD?
2. What happens when you run the modulation test on the 2k?  Similar results but just takes longer to complete?

One is without and other one with modulation. Difference is about 1dB. I'm guessing it's the AWG when modulation got on.. I didn't care because what was important was accurate 6db relation.

On 2k looks pretty much the same and pretty much lasts the same, because at this settings most of the time is acquisitions and averaging not FFT processing time.

The sideband power in each is 1/6 the total (-7.78dB), which leaves the carrier at 2/3 total (-1.76dB), so would expect it to be -6.02dB down from carrier as you've shown.

Your impressive results at this power level with the SDS2000X HD and 3000X HD and followed up by Performa01 with the SDS800X HD

Best,

[joeqsmith]

Channel 1 & 2 fed with a 1MHz 0dBm sinewave.   Math used to add both channels.  FFT then ran on the math channel.  This is about the limit of the old LeCroy.  It can display the separate peaks but the readouts will only display a single peak.  Again, wrong tool for the job but I am curious how well your scopes can resolve two peaks like this.  Feel free to use an external combiner rather than the math.