Siglent SDS2504X, FY6900 and Two Tone Testing

[eloso]

I was wanting to use the AWG on my SDS2000X Plus to generate a waveform to make two tone tests of an SSB HF transmitter.   This is a learning experience in a few respects.  I haven’t generated arbitrary waveforms before and I haven’t used the scope's FFT function before although I am comfortable with FFT theory to a certain extent.   I don’t normally use the scope FFT because I have a SA but it doesn’t go down to these low  frequencies, hence the use of the scope.

I am referencing the process described here https://siglentna.com/application-note/easywave-two-tone-arbitrary-waveform/   . Note that there are no particular spurious signals on the screenshot in the article, just the general noise floor at some 60 dB down from the two tones (700Hz and 1900Hz).

So I have generated an arbitrary waveform with 7 cycles of sine wave added to 19 cycles of sine wave and uploaded that  to  SDS internal AWG.  I’ll leave aside that sending it over then network doesn’t work, no error message, EasyWave seems to do its bit, but no file arrives. So I used a USB stick which was fine.

The result is as shown in the first screenshot. I can clearly see the two wanted frequencies but there are spurii some 37-39dB down.

That is not what is shown in the article where they are using the “proper” SDG. 

As a matter of interest I tried my FY6900 and got the results shown in the second screenshot. The spurious signals are better than  50dB down

Question(s)

1   Is my FY6900 really so much better than the AWG on my Siglent ? If so, is that a bit embarrassing ? 

2      Is my application of FFT on the scope correct? - I normally use an SA but it doesn't go down to these frequencies and this is my first time playing with the Scope FFT Math function.

3.      Would I get better results using both channels of the FY6900 running them as sine waves and then using electronics to sum the two waves ? 


I’m not sure that I actually need any better performance than I can get from these methods just to check the ballpark linearity of the SSB transmitter.    There is no spec I am measuring against anyway, I just want to see that it looks ok-ish and isn’t likely to create a huge amount of IMD.

Siglent Spurii

" alt="" class="bbc_img" />

FY6900 Spurii





Regards


Eloso





Edited to use proper descriptive name for the second screenshot.

[rf-loop]

Can you explain why you run these test with quite different settings. Other show 6dBm signal and scope input set for 2V/div and other below around <0dBm and 200mV/div (are you serius or just playing or just somekind of accidental mistake)
And then sidenote. (even when it do not so much matter) You use dBm units but oscilloscope input is 1Mohm and you have not explained at all if you use external load but in FFT menu you have told to scope there is 50ohm load.
Then sidenote/advice. If you measure levels using FFT then flattop window is your friend. Other FFT window may give even several dB error depending peak real frequency related to FFT bins freq.

Then you have not explained do you run FY also using single channel output using Arb (for compare apples to apples) to generate dual tone or separate channels. It is also good to note. We can not read your mind.

Here is example how it looks with SDG1000X generated signal and using SDS2000X HD.
SDG CH1 (f1) 700 and CH2 (f2) 1900Hz sinewaves, both 10dBm level. Channels internally combined to CH1 output.
Load (oscilloscope internal) is 50ohm.
(peak marker 4) f2-f1  < -62dB_f2 and (peak marker 10) f2+f1  < -62dB_f2



In this image, AXIS LEVEL VALUES ARE ACCIDENTALLY WRONG (decimal point in wrong place!!!) and 0dB is accidentallly also in center. Buuh...
Peak table values are ok.


ETA:
Other test below. SDS2000X HD internal 1 channel AWG. Dual tone Arb waveform. Here run f1 600Hz and f2 3000Hz (internal ready made Arb) (I do not have time for make 700/1900Hz ARB)




Still f1+f2 and f2-f1 are < -59dB from f1 or f2   but as can see quality is far below SDG1000X example.
f1 peak 1
f2 peak 6
f2-f1  peak 5
f2+f1 peak 7

[eloso]

Thanks for taking the time to afford such a comprehensive reply.

Your questions are most interesting and reveal  flaws in what I am doing. Hands up to it.

I hadn't appreciated that particularly when looking at IMD products it is pretty essential to have original signals from the two systems being compared to be the same level. Kinda sounds dumb, but it had obviously passed me by. I think I was just fixated on the relative nature of the db scale and excluded any other concerns from my thoughts.   I have learnt this now and in fact since posting I have found even more reading (hint: I should do the even more reading before I make a post!)

BTW I am intrinsically nervous about the internal 50 ohm load on my nice oscilloscope and after a near miss, I decided to use a T piece with external 50ohm load where I can get away with it.  With these being audio frequencies I didn't think it was worth switching in the internal load and in fact used my external one.


I did compare apples with apples in answer to one question about whether I used both channels of the FY. No I didn't.  The FY was using a single channel in exactly the same way as when using the Siglent. I used the Feeltech software to generate the two tone mix in exactly the same way. I just didn't say that - hands up again.

Good point on the window type Flat Top being preferrable. I shall take that on board (and put it on my reading list too).

I am away from the home lab today but I shall return to this and be a little more stringent. In the meantime I note your conclusion and I think we'll find that the Siglent is not worse than the FY, but the SDS is not as clean as the SDG - which I guess isn't really suprising.

I still need to try and understand how much I am measuring issues in the scope versus issues in the generator, but at least if I ensure all test levels are identical I stand a chance of making a fair comparison, That wasn't a question - I shall do some more reading first !

Thanks again - it took time to do those tests and write it up and I do appreciate it 


Eloso

[eloso]

OK, back in the lab and this time a little more care

First screen shot Siglent SDS2000X AWG Single Channel, Arbtrary wave created by summing a 700hz Sine with a 1900Hz sine.  0dbm carrier level is the reference. 50ohm external load.

Siglent worst IMD product is -48.4dBm

Second screen shot FY6900. Same technique as with the first screenshot. Same levels, same scope settings.

FY6900 worst IMD Product is -47dBm

And for fun, a third screenshot of a two tone wav file I obtained from a PC soundcard.  Again, all the same scope settings

PC Soundcard worst IMD product is -49.7dbm

Bearing in mind this is an 8 bit scope with some kind of fiddly-faffy 10bit mode switched in then these measurements are no doubt as good as it gets. It kind of begs the question, how good is that ?  But at least I know that with the kit that I have, I can cheerfully use any of these three sources and be equally certain, or perhaps uncertain, of the final results of the SSB tests as they will be shown on the SSA.

Thanks


Eloso

[mawyatt]

Here's link you might find interesting.

https://www.eevblog.com/forum/testgear/two-tone-test-with-scope-and-sa/msg3163364/#msg3163364

Best,