Two Tone Test with Scope and SA

[Performa01]

I love challenges.

Quite obviously, we need a trustworthy low distortion dual tone signal in order to perform meaningful IMD tests. To generate this seems to be a challenge in some cases already – all the more so to confirm it without a high-dynamic range SA like the R&S FSEA30.

In my case it was a matter of finding the sweet spot on the existing average SA. By this I mean a combination of input signal level, input attenuation and IF-gain that adds as little distortion as possible.

My tests didn’t reveal any hints where the distortion actually comes from. Changing the input level, by internal and/or external attenuation didn’t show the expected behavior. If 1 dB lower drive level doesn’t make the 3rd order IMD products drop by more than 1 dB (theoretically 3 dB), then the analyzer frontend shouldn’t be the source of the distortion. If on the other hand, additional attenuation at the generator outputs, i.e. additional isolation, doesn’t change anything, then the generator shouldn’t be the culprit either.

That leaves the power combiner/splitter, which might introduce non-linearities even when it’s purely resistive – at least when we’re approaching distortion levels in the realm of -100 dBc. But my earlier tests got -92 dBc IMD, with a generator output of only +6 dBm and the isolation pads to the power combiner were 10 dB each, with a 20 dB attenuator at its output for a total of 36 dB of attenuation.

The actual test setup is the dual channel AWG with +10 dBm output levels for each tone, i.e. 4 dB higher than before. These signals are connected to the power combiner via 20 dB inline attenuators, but the output of the resistive wideband combiner is now terminated by only a 10 dB inline attenuator. Consequently, there is 36 dB total attenuation again. The resultant -26 dBm signal is then fed into the SA.

The attached screenshot shows what I have achieved when taking advantage on the sweet spot of the SA.

Ref_10MHz_O1kHz_-26dBm_Iso40dB

Finally I see distortion figures better than -100 dBc 😉

[Performa01]

The question about the internal combiner in the Siglent AWGs has been answered in reply #238 here:

https://www.eevblog.com/forum/testgear/how-much-noise-floor-and-other-things-matter-in-oscilloscope-usability/msg3906974/#msg3906974

The integrated digital combiner does indeed generate some IMD products on its own, so it is not suitable for reliable 3rd order dynamic tests. IMD can be as bad as 45 dBc at high output levels.

Of course, the -45 dBc are a worst-case figure and at lower frequencies and/or levels it performs a lot better than this, yet it cannot even come close to a proper external power combiner.

[rf-messkopf]

Okay, I've re-run my measurement with the ZFSC-2-6-n+ splitter, but now with 20 dB pads at each input, plus a 6 dB pad at the sum port. That gives me 70 dB of isolation between the two signal generators. Using the same generators (R&S SML and SMU200A at 10 dBm each) I get exactly the same results as before, i.e., 3rd order products at about -75 dBc.

I've also increased the reference level so that the fundamentals are now approx. 20 dB below in case there is intermodulation at an IF stage in the analyzer.

Next I tested the same setup with two function generators (Keysight 33521B and an old HP 3325A). Again, exactly the same result. Except that the function generators spit out some spurs (I think it's mainly the 3325A) -- see attachment.

I'm stating to believe that it is the darned splitter itself that is causing the IMD.

[mawyatt]

The question about the internal combiner in the Siglent AWGs has been answered in reply #238 here:

https://www.eevblog.com/forum/testgear/how-much-noise-floor-and-other-things-matter-in-oscilloscope-usability/msg3906974/#msg3906974

The integrated digital combiner does indeed generate some IMD products on its own, so it is not suitable for reliable 3rd order dynamic tests. IMD can be as bad as 45 dBc at high output levels.

Of course, the -45 dBc are a worst-case figure and at lower frequencies and/or levels it performs a lot better than this, yet it cannot even come close to a proper external power combiner.

Yeah, forgot about that thread since it's 1/2 year old...thanks for pointing there!!

My short-mid term memory is getting worse with age, but long term seems to holding up well.

Need to get a much of those long term experiences documented before they disappear!!

Not surprised the analog combine outperforms the digital combine in some cases. Siglent likely chose the easier digital path since the source of all these non-linearities are quite involved as has been shown.

Anyway, very interesting discussions.

Best,

[G0HZU]

To explore the limits of a decent modern swept spectrum analyser (swept LO with a wideband digital IF) it's best to use wide tone spacing and then select a very narrow span to zoom in and just look at the IMD tone itself. This assumes the analyser has a suite of analogue pre-filters ahead of the ADC and this is generally the case for a high end spectrum analyser. These pre-BPFs will typically be 2 or 3 times wider than the digital RBW filter and the analyser should select the optimum pre-BPF automatically. The BPF protects the wideband ADC from having to cope with the main test tones if you set the span much narrower than the tone spacing.

It should be possible to achieve the SFDR on the datasheet for the analyser using this method assuming the signal generators and the combiner are not limiting the performance. With a RBW of 1Hz some analysers should manage 115dB IMD3 SFDR. Older classic analysers like the HP 8568B use analogue RBW filters and this analyser can typically achieve just over 100dB IMD3 SFDR using the smallest RBW of 10Hz. This classic old analyser typically has a mixer IP3 of +13dBm and a DANL of -140dBm in a 10Hz RBW. So the IMD3 SFDR will typically be about 102dB.

[RoGeorge]

I'm stating to believe that it is the darned splitter itself that is causing the IMD.

Might be, especially if that splitter has some magnetics core, but I don't have enough practice to know what IMD levels to expect.

From the datasheet it's not clear if that model is a resistive splitter.  Since the specs for it specify a minimum frequency, that might be a hint that the splitter has some ferrite inside.  Ferrites have hysteresis and saturation curves, they are not linear devices.  Therefore, ferrites materials are expected to introduce distortions (and thus to produce IMD).

Would be interesting to measure the intermodulations using a pure resistive splitter, or any other kind of splitter that doesn't have magnetics.  I suspect there is close to no intermodulation produced in the output stage of the generators (because those stages are usually an amplifier with a very low output impedance in series with a 50 ohms resistor, therefore any backfed signal will appear much smaller to the output amplifier).

[G0HZU]

I think what normally happens within an ALC based sig gen is that the tone from the other sig gen will cause a very small low frequency error term at the ALC detector diode. The ALC error amplifier within the sig gen will then amplitude modulate the sig gen to try to cancel the error term assuming it all happens within the loop bandwidth of the ALC system. The net result is the sig gen modulates itself to keep the ALC happy and this usually means you see what look to be classic IMD3 terms on the signal generator output.

This is why it can be useful to use a sig gen that can turn the ALC bandwidth right down. If the tone spacing is much wider than the ALC loop bandwidth then the ALC system will be too slow to correct the error at the ALC detector diode. So it doesn't add any undesirable 'corrective' AM modulation to the output.

[rf-messkopf]

So I've been playing around a bit. I've also tested a Mini-Circuits ZFSC-2-1 (5 MHz to 500 MHz) and an Anaren 10014-3 quadrature hybrid (500 MHz to 1 GHz) at various frequencies and tone spacing. It seems that at very narrow tone spacing like 1 kHz I'm indeed hitting a limitation of the analyzer. I currently have the ZFSC-2-1 hooked up and I'm getting 3rd order products lower than -97 dBc with 3 kHz spacing using the method described by G0HZU, i.e., zooming in on the 3rd order product. This is a bit challenging because sweeps tend to get slow, and one has to determine the right amount of input attenuation to keep the noise floor at reasonable levels without introducing IMD in the first mixer of the analyzer.

Not entirely sure, but maybe with some more effort one could make better measurements also at 1 kHz spacing. I think this contribution summarizes it very well:

To explore the limits of a decent modern swept spectrum analyser (swept LO with a wideband digital IF) it's best to use wide tone spacing and then select a very narrow span to zoom in and just look at the IMD tone itself. This assumes the analyser has a suite of analogue pre-filters ahead of the ADC and this is generally the case for a high end spectrum analyser. These pre-BPFs will typically be 2 or 3 times wider than the digital RBW filter and the analyser should select the optimum pre-BPF automatically. The BPF protects the wideband ADC from having to cope with the main test tones if you set the span much narrower than the tone spacing.

It should be possible to achieve the SFDR on the datasheet for the analyser using this method assuming the signal generators and the combiner are not limiting the performance. With a RBW of 1Hz some analysers should manage 115dB IMD3 SFDR. Older classic analysers like the HP 8568B use analogue RBW filters and this analyser can typically achieve just over 100dB IMD3 SFDR using the smallest RBW of 10Hz. This classic old analyser typically has a mixer IP3 of +13dBm and a DANL of -140dBm in a 10Hz RBW. So the IMD3 SFDR will typically be about 102dB.

For the FSIQ the IMD performance is quoted for tones at –30 dBm with ∆f >5 x RBW or 10 kHz, whichever is greater, and 3rd order products at < -74 dBc (TOI 22 dBm typically). So it seems I'm asking for too much here when I try to measure everything in one sweep.
Edit: After thinking a bit and reading some application from the manufacturers, the excessive IMD I have seen was likely due to the ADC being used to implement the digital RBW filters. Quoting from this application note:

ADCs are non-linear components whose intermodulation distortion follows rules which are different from other common RF components. Their intermodulation distortion is not specified using the TOI, but is included in the spurious free dynamic range (SFDR) specification. SFDR covers not only the intermodulation products, but all unwanted signals. Since SFDR specifications are given in "dB below full scale" (dBFS), it is essential to scale the signal correctly before applying it to the ADC. On a spectrum analyzer, the so-called IF gain, which is often coupled to the reference level, is used to optimize the signal level in front of the ADC.

The intermodulation products of an ADC remain more or less constant, independent of its input signal level. This in return means that the SFDR is dominated by the signal level, which means that a higher ADC input level will result in the same increase in SFDR.

Recommendations:
1. The best method to get around the ADC related IMD contribution is to avoid having two tones at the ADC input simultaneously by selecting an appropriate tone spacing (e.g. > 5 MHz for the R&S FSW and R&S FSV).
2. If the tone spacing is fixed and cannot be changed, the CW tones should be close to the full scale level of the ADC. The R&S FSW in default setting automatically takes care about the signal scaling for the ADC, using as much of the ADC’s scale as possible and avoiding an ADC overload at the same time.

This is the reason why the IMD products did not depend on the reference level setting. Also, it is good to keep in mind that the ADC is not included in the TOI specification. Easy to forget when you sit in front of the instrument.

The FSIQ and FSEA only use digital RBW filters for RBW settings < 1 kHz; the larger RBW filters including 1 kHz are analog (LC and crystal). When digital filters are used, the 3rd IF at 21.4 MHz is downconverted to 25 kHz and the sampled by an 18 bit ADC at a rate of 200kHz, with the analog filter section still in the signal path. Unfortunately I cannot find anywhere in the manuals which analog filter is used with which RBW or span setting. But since there is an 1 kHz crystal filter, there is a chance that it is switched in front of the ADC with narrow spans or RBWs. So it may be possible to measure 3rd order products with these instruments even when the two tones are closely spaced.
/Edit

I'm stating to believe that it is the darned splitter itself that is causing the IMD.

Might be, especially if that splitter has some magnetics core, but I don't have enough practice to know what IMD levels to expect.

From the datasheet it's not clear if that model is a resistive splitter.  Since the specs for it specify a minimum frequency, that might be a hint that the splitter has some ferrite inside.  Ferrites have hysteresis and saturation curves, they are not linear devices.  Therefore, ferrites materials are expected to introduce distortions (and thus to produce IMD).

It's a splitter with 3 dB loss in both arms, with 30 dB of isolation, so it cannot be resistive. And I just popped the lid to confirm that. 

[G0HZU]

After a quick scan of the datasheet and the glossy brochure for the FSIQ series it looks like this analyser dates back to the late 1990s. It appears to be a crossover type of analyser that still retains a classic 95dB logamp and a full suite of analogue RBW filters. However, it also has a digital IF that can function as a vector signal analyser. It may well be the case that the digital IF has a 12 bit ADC but I'm just guessing based on the date shown on the brochure. In basic spectrum analyser mode I'd expect it to outperform the classic HP 8568B analyser as the phase noise and IP3 specs are better than the 8568B. This must have been a very expensive analyser when it was new. I think 14 bit digital IFs appeared two or three years later, at least that's when I first saw analysers with a 14 bit digital IF at work. Maybe some slightly newer FSIQ analysers also have a 14 bit digital IF?

[rf-messkopf]

After a quick scan of the datasheet and the glossy brochure for the FSIQ series it looks like this analyser dates back to the late 1990s. It appears to be a crossover type of analyser that still retains a classic 95dB logamp and a full suite of analogue RBW filters. However, it also has a digital IF that can function as a vector signal analyser. It may well be the case that the digital IF has a 12 bit ADC but I'm just guessing based on the date shown on the brochure. In basic spectrum analyser mode I'd expect it to outperform the classic HP 8568B analyser as the phase noise and IP3 specs are better than the 8568B. This must have been a very expensive analyser when it was new. I think 14 bit digital IFs appeared two or three years later, at least that's when I first saw analysers with a 14 bit digital IF at work. Maybe some slightly newer FSIQ analysers also have a 14 bit digital IF?

I just edited my post above, see there: The FSEA/FSIQ do use digital RBW filters below 1 kHz RBW; the ADC is an 18 bit one, sampling at 200 kHz. And yes, they were very expensive.

[G0HZU]

Thanks! I was kind of expecting a digital IF with fewer bits but with a higher sampling frequency and wider bandwidth. If it has a narrowband digital IF with 18 bits I would have expected it to perform really well even with close tone spacing. I've not played with an R&S FSIQ analyser so I can't really be sure what to expect from it.

Not entirely sure, but maybe with some more effort one could make better measurements also at 1 kHz spacing.

I agree. It might be worth doing a few more experiments to explore its limits.

[rf-messkopf]

Thanks! I was kind of expecting a digital IF with fewer bits but with a higher sampling frequency and wider bandwidth. If it has a narrowband digital IF with 18 bits I would have expected it to perform really well even with close tone spacing. I've not played with an R&S FSIQ analyser so I can't really be sure what to expect from it.

At least that's what I gather from the service manual I have. I would have to check if the assembly part numbers match, maybe there were changes down the line. Also there may be differences between the FSEA/FSE and FSIQ units in this regard.  R&S was never generous in terms of service manuals, and full schematics were not released. If I remember correctly, the digital IQ demodulator is a completely separate unit with its own ADC and DSP, but offhand I'm not sure.

I'm also not sure about the effective analog bandwidth of the digital IF. According to the assembly block diagrams and the wiring diagram, the analog RBW filter section is ahead of the digital IF assembly and the ADC. No idea what the actual bandwidth it uses is. Unless someone can dig out some documents that describe it, one would have to do some reverse engineering.

I agree. It might be worth doing a few more experiments to explore its limits.

I'll see what I can do.

[gf]

I stumbled across this paper which I found very interesting:
https://www.researchgate.net/publication/261321090_Low-IMD_two-tone_signal_generation_for_ADC_testing