[time-nuts] Thunderbolt Harmonics

Tom Holmes tholmes at woh.rr.com
Wed Jan 18 21:25:56 EST 2017

Jim Lux...

Note that the Keysight 33622 is basically an ARB, and not really intended to 
be an RF signal generator. It's output is DAC based, not oscillator based. 
Hence it will tend to have higher distortion products than a good RF 
generator. As you point out, a suitable filter for the intended frequency 
certainly can clean that up, of course.

Tom Holmes, N8ZM

-----Original Message-----
From: time-nuts [mailto:time-nuts-bounces at febo.com] On Behalf Of jimlux
Sent: Wednesday, January 18, 2017 9:45 AM
To: time-nuts at febo.com
Subject: Re: [time-nuts] Thunderbolt Harmonics

On 1/18/17 4:33 AM, Artek Manuals wrote:
> R
> Is what your seeing a harmonic (2nd? 3rd?) or a spur i.e what frequency
> is the "harmonic" ?
> How are you measuring this ? (Spectrum analyzer ? make/model?)
> More importantly and at the risk of displaying my naivety, what is the
> application that you are using the 10MHz source to feed and why if this
> harmonic is 60db down (or even only 40db down, the quoted spec) why
> would one care? What is the predicted error you will get in your
> application as a result?

One application that needs low harmonic content is where you are
measuring the harmonic generating (or lack thereof) of a downstream

I have an application where we're measuring the performance of a RF
chain followed by a digitizer.  An easy test is to feed in a nice sine
wave (at a frequency that is NOT a submultiple of the samplerate) and
look for harmonics in the power spectrum of the sampled data stream.

the first time we ran the test (using a Keysight 33622 signal generator)
we saw significant 2nd and 3rd harmonics (50-60 dB down, but easily
detectable).  A quick review of the data sheet.. Oh, the signal
generator spec is only -43 dBc for frequencies above 10 MHz.

Another case where low harmonic content is when doing two tone IMD tests
- if the sources have significant harmonic content, you might be seeing
intermod between the harmonics of the source, rather than intermods
between the fundamental of the source.

For 10 MHz, you can get minicircuits filters for 10.7 MHz that are
fairly wideband and work pretty well... about 20-30 dB of harmonic
suppression per filter I'd use the low pass flavor

loss at 20MHz is 26.84
loss at 40MHz is 41.22
loss at 50MHz is 46
est loss at 30 is 35?

loss at 10 is 0.65 spec
loss at 20 is 31.35 spec	measured -33
loss at 30 measured -60
loss at 34 is 47.26 spec
loss at 40 measured -77
loss at 67.5 is 69.85 spec

Yeah, they might have a significant tempco, but you're running all this
stuff in an underground lair with small temperature variations, right?
It only looks like a small volcano from the outside.

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.

More information about the time-nuts mailing list