[time-nuts] E4437B phase noise spurs. Any ideas?

Sat Jul 18 21:20:55 UTC 2015

Hi Bill,

Thank you for your detailed response.

First of all, I'm happy with the SMIQ and I included that only
to show that the spurs do not come out of the spectrum analyzer
(which is an FSU indeed).

Regarding the E4437B: You are right that the data sheet specs
allow this but I find it rather unusual for such a high profile
device, especially since it's the high spectral purity version.
I have an E4432 (3 GHz and not designed for high spectral
purity) and an SMIQ and they both do not show such strong and
so many spurs.

Also, the phase noise plots in the data sheet do (in my view)
show fairly narrow features; at least don't look like highly
smoothed plots.

The E4437B spurs appear only above 2.4 GHz. Virtually no spurs
are visible below 2.4 GHz. And they look quite exactly the same
for every frequency above 2.4 GHz. Same offset, similar amplitude.
Both when changing the frequency a couple of kHz and when changing
by GHz.

Also, I find that asymmetry relative to the carrier "interesting".

So what I would like to understand is if I caught an unlucky
device of the E443x series and can expect to see better
performance with other ones. Or if that is a pre-failure sign.
And if other people are seeing them as well.

Thanks,
Wolfgang DL1SKY

On 2015-07-17 07:47, Bill Byrom wrote:
> Hi, Wolfgang. Both of your generators appear to be well within their
> datasheet specifications based on the spectrums you posted.
>
> The datasheet for that ESG-DP (digital modulation with high spectral
> purity) model is at:
> http://cp.literature.agilent.com/litweb/pdf/5965-3096E.pdf
>
> The E4437B nonharmonic spur specification is on page 5 (use the ESG-
> DP columns):
>
> So according to that table, above 2 GHz the E4437B spurs at >3 kHz
> offset **should be below -68 dBc**. It appears to me that the largest
> spur at such offsets (at about +3.5 kHz offset from the carrier) is
> **actually about -74 dBc**. The largest spur is about -63 dBc at +1 
> kHz
> offset (which is closer than the spur specs).
>
>
> The R&S SMIQ03B specs are:
>
>
> The SMIQ03B nearby spurs in your screen capture are at about +/- 9 
> kHz
> offset from the carrier (about -85 dBc), so are too close to be 
> covered
> by the spec above. The spurs at >10 kHz offset are about -90 dBc.
>
> I'm guessing that the spurs might be affected by several adjustments
> performed during factory alignment. In most modern complex 
> instruments
> (including both of these signal generators, I believe) there are far 
> too
> many alignments to be made by a human, and a long automated process
> using a rack of equipment is used to align the instrument and store
> certain constants in nonvolatile memory in the instrument. But there 
> is
> no reason to worry based on your results, since they are well within
> instrument specifications.
>
> So ... assuming that the spurs aren't coming from your spectrum 
> analyzer
> or other sources (and I'm guessing that's a FSU which is very good),
> both generators appear to be within specifications (after warmup 
> within
> the calibration interval, of course).
>  * *The E4437B closest spur covered by the >3 kHz spec is at about 
> +3.5
>    kHz, and it's about 6 dB better than the spec.*
>  * The SMIQ03B has a 6 dB better spurious spec than the E4437B 
> between 2
>    and 3 GHz, but the SMIQ offset for their spur spec is wider than 
> the
>    ESG (10 kHz rather than 3 kHz).
>  * At other carrier frequencies the spurs may be higher or lower in
>    amplitude (and at different offsets). Try tuning the frequency in 
> 1
>    kHz steps over a wide range -- I would guess that the spur offsets
>    and levels will change in a complex fashion due to the 
> synthesizer.
>  * The phase noise specifications normally apply to a smoothed trace 
> and
>    do not include narrowband spurs such as the ones you see in your
>    screen capture.
> --
> Bill Byrom N5BB
> Tektronix RF Application Engineer
>
>
> On Thu, Jul 16, 2015, at 04:15 AM, Wolfgang DL1SKY wrote:
>> Hi,
>>
>> I just got a used/refurbished E4437B which I wanted to use as a
>> all-purpose RF generator primarily for 3-4 GHz.
>>
>> Unfortunately, I'm seeing strange spurs for frequencies above 2.4 
>> GHz,
>> see the green curve in the attached image.
>> The yellow curve is an SMIQ03 for comparison.
>>
>> Observations:
>>
>> - For frequencies below 2.4 GHz none of the spurs appear.
>>
>> - It has an OCXO and I left the device in standby (oven on) for
>>   12 hours.
>>
>> - If I leave the device ON for 1-2 hours, the spurs go down.
>>
>> Anybody else seeing this? Any ideas how to fix this? Does this look
>> like a pre-failure sign?
>>
>> Regards,
>> Wolfgang DL1SKY
>> _________________________________________________
>> time-nuts mailing list -- time-nuts at febo.com
>> To unsubscribe, go to
>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>> and follow the instructions there.
>> Email had 1 attachment:
>
>
>>  * E4437B-phase_noise_problem-1.png
>>   34k (image/png)
>
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to 
> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.

-- 
Wieserlabs UG (haftungsbeschränkt)
http://www.wieserlabs.com

Sitz der Gesellschaft: München, Oberbayern
Registergericht: Amtsgericht München, HRB 181414
Geschäftsführer: Wolfgang Wieser
USt.-ID-Nr: DE267137112