You really need to mix down to a low frequency beat note as in the
HP5390 system.  There is a huge increase in sensitivity by doing
that.  It is not difficult to make a mixer circuit.

Rick Karlquist N6RK

John Green wrote:

2009/7/31 John Green wpxs472@gmail.com:

What's the difference between a HP 5370A and a HP 5372A for this type
of measurement?

73,
Steve

--
Steve Rooke - ZL3TUV & G8KVD
A man with one clock knows what time it is;
A man with two clocks is never quite sure.

Hi John,

The hp 5370A or SR620 or other sub-ns resolution time interval
counters allow you to easily compare frequency standards at
mid- to long-term. In fact, if measuring the daily drift rate of Rb
or OCXO against your GPSDO is all you need then almost any
nanosecond counter will do the job. Collect data for minutes or
perhaps days -- and you will end up with some useful phase or
frequency deviation plots as well as frequency drift calculations.
It's quite simple.

But for short-term measurements, the main limitation that 5370
and equivalent counters have is their single shot resolution of
around 20 ps. This sounds impressive, but it clearly limits your
phase measurements to 2e-11 at tau 1 second, or 2e-12 at 10
seconds, or 2e-13 at 100 seconds, etc. This means you cannot
adequately measure devices with short-term stabilities better
than that using a 5370.

Since good OCXO have stabilities down in low -12's at tau 0.1 s
to 10 to 100 seconds, this means your counter is the limiting
factor when you compare your DUT against your REF. And by
"limiting" here I sort of mean "useless".

Now it depends on what your needs really are. If you are mostly
interested in frequency accuracy or drift rate over time, then I
think a 5370 is all you need. On the other hand, if you are
interested in short-term stability then you do need something
better -- at least 10x to 50x better.

Dual-mixer methods are the solution for this. Although there are
some high-end products out there, it seems to me a fairly basic
setup is all you need. I mean, you don't need 1e-13 or 1e-14
level of performance -- none of the oscillators you mentioned
are that good to begin with. Can it be done in less than $50?

Perhaps someone on the list can suggest the minimum effort
required to obtain, say, 1e-12 or 5e-13 level of dual-mixer
stability measurement?

I'm not talking about an extreme engineering solution that gets
down in the -14's or -15's -- just a simple, cheap, home-brew
solution that improves on a 5370 short-term resolution by a
modest factor of 10 or 100. I think that's all John needs.

/tvb

----- Original Message -----
From: "John Green" wpxs472@gmail.com
To: time-nuts@febo.com
Sent: Thursday, July 30, 2009 9:01 AM
Subject: [time-nuts] Methods for comparing oscillators

Tom

Without elaborate temperature control of the mixer and ZCD it isn't
possible to maintain picosecond stability (at the mixer input).
The greatest problems are crosstalk and isolation between the 3
oscillators.
Designing a ZCD to achieve 1E-12/tau or slightly better resolution is
relatively simple especially if one uses the Collin's technique for
which spreadsheets have been created to simplify the process.
Choosing a low dc offset and offset drift mixer is critical.
Something like an Minicircuits MPD-1 is a good start as the dc offset is
low and specified as is the isolation between RF1 and RF2  (> 70dB for
frequencies > 15MHz).
It also allows low frequency isolation of IF and RF grounds to reduce
low frequency crosstalk.
If one were to use simple isolation amplifiers with 40 dB or reverse
isolation together with a splitter with output to output isolation of
30dB to drive a pair of  offset frequency isolation amps that then:

1. DUT to offset source isolation of 110dB is achievable.

2. DUT to DUT isolation of ~ 140dB is achievable.

Which should be adequate for these purposes.

If one used a 1Hz offset frequency then a ZCD output jitter of 10us or
less would be required.
This is easily achieved using a 3 stage limiter with a slope gain of 10
or more.
With a few inexpensive modifications (optimised mixer matching and
optimal IF port termination ) considerably better performance is
possible (limited primarily by drift.

Bruce

Tom Van Baak wrote:

A 1Hz offset may be convenient as its easy to adjust an 10811A or
similar to achieve this.
A 10Hz offset is harder to achieve.

Bruce Griffiths wrote:

Can easily add an isolation amplifier between the offset oscillator and
the source to boost this by another 40dB.
Such an amplifier may be required to boost the offset oscillator output
to a suitable level to drive the splitter.

That should be 210dB (inadequate shielding may limit this).
The isolation from one DUT to the LO input of the other phase detector
would be about 140dB.

Bruce

Would either of these be 10 Hz solutions?

1. open the 10811 and change the xtal load capacitance, or
2. change the oven temperature set point until the resonator
   is 10 Hz off. You'd loose a little in tempco but...

/tvb

Tom

The specified mechanical adjustment range for the 10811A is greater than
(-10Hz, +10Hz) so in principle one should be able to achieve an offset
of up to 10Hz or so.
I'd try doing this first, then of the 2 methods suggested changing the
crystal load capacitance if its feasible to do so would be the preferred
option as the oscillator stability will be less impaired than if the
oven temperature were varied, at worst only a relatively small change in
capacitance will be required.
If the motional parameters of the crystal were known then the required
capacitance change can be estimated before dissassembly.

Looking at the oscillator circuit the 1-30pF mechanical trimmer cap is
connected in parallel with a 100pF varicap which is in series with 16pF
so adjusting the frequency down if the trimmer is at its lowest
capacitance will be difficult, adjusting the frequency up by connecting
a small cap high quality across the trimmer should be easy to do.

Bruce

Tom Van Baak wrote:

The spec for the 10811A's trimmer range is +/- 10 Hz, actually.  Seems
doable.

There was a good thread on the list back in late 2006 where some of the
pitfalls of the DMTD technique in general were discussed.  It seems that
achieving optimal performance is a lot less trivial than anyone would like.

If someone wanted to build a "minimum effort" system with a lower short-term
floor than the 5370, my understanding is that it would require:

• a 10811-class OXCO tuned to generate beat notes of several Hz
• a mixer of some sort with good IF response down to DC
• four good isolation amps (which thanks to Bruce's recent work is not a
  difficult or expensive issue)
• some sort of delay compensation on one of the inputs may be needed to
  avoid decorrelation of the transfer oscillator's effect between the two
  channels
• a sound card with good differential-phase stability

Of course a 5370 or other two-channel TIC could be used instead of an FFT
back end, but I'd be interested in seeing what level of performance could be
expected with a sound card, too.  Bruce, you were looking at differential
jitter and delay on stereo sound card channels at one point, weren't you?

If differential phase stability is a problem, then perhaps one of the
channel IFs could be heterodyned up to a different carrier frequency that
could share a single physical ADC channel.  Isolation and phase-delay
compensation would again be a big deal in any such scheme.  Probably the
heterodyne LO would need to be derived directly from the transfer
oscillator.

-- john, KE5FX