non-pipelined fast ADC

On 2020-08-30 19:41, jlarkin@highlandsniptechnology.com wrote:
> On Sun, 30 Aug 2020 19:28:23 -0400, Phil Hobbs
> <pcdhSpamMeSenseless@electrooptical.net> wrote:
> 
>> On 2020-08-30 19:17, Chris wrote:
>>> On 08/28/20 17:47, Phil Hobbs wrote:
>>>> On 2020-08-26 14:47, Chris wrote:
>>>>> On 08/26/20 19:40, piglet wrote:
>>>>>> On 26/08/2020 5:58 pm, Gerhard Hoffmann wrote:
>>>>>>> Am 26.08.20 um 18:29 schrieb Chris:
>>>>>>>
>>>>>>>>
>>>>>>>> The older high end HP spectrum analysers, the 8566 and 8568 did that
>>>>>>>> as well.&nbsp; The marketing blurb called it lock and roll, locking the
>>>>>>>> LO at start of sweep, then free running it open loop for the rest
>>>>>>>> of the sweep. Allowed them to have a stable 10Hz resolution bw and
>>>>>>>> 1KHz span at Ghz, in the late 1970's.
>>>>>>>>
>>>>>>>> I think what I was really saying was, was if the solution gets too
>>>>>>>> complex, it maybe the wrong approach :-)...
>>>>>>>
>>>>>>> IIRC, there was an article in the HP Journal about this.
>>>>>>> I think I have it on paper somewhere, but HP Journal
>>>>>>> is probably searchable somewhere.
>>>>>>>
>>>>>>> Cheers, Gerhard
>>>>>>
>>>>>> This one?
>>>>>> <http://hparchive.com/Journals/HPJ-1978-06.pdf>
>>>>>>
>>>>>> piglet
>>>>>>
>>>>>
>>>>> I think i've seen that.Amazing bit of kit for it's time. HP really
>>>>> were at the peak of their game back then with tech prowess few,
>>>>> if any could match. Much of it still in use today.
>>>>>
>>>>> Last catalog price for the 8566, iirc, was 78,000 usd and yet,
>>>>> they sold bucketloads of them...
>>>>
>>>> Back in 1982, when I was working in satcom, we had a reasonably new
>>>> Ailtech 757 (iirc). It was a manually-tuned gizmo that drifted around a
>>>> bit before it warmed up.
>>>>
>>>> Then we got a swoopy new HP 8566A, later upgraded to a B. I was hooked.
>>>> I had one for awhile at IBM, and about 6 years I bought a used one on
>>>> eBay when I needed more top end than my previous 8568B. Its close-in
>>>> phase noise is remarkably low on account of its YIG-tuned sweep
>>>> oscillator.
>>>>
>>>> Its main drawback is that you can have the RF or microwave band
>>>> displayed, but not both at once. Still, for 2 cents on the dollar one
>>>> really can't complain too loudly. ;)
>>>>
>>>> Cheers
>>>>
>>>> Phil Hobbs
>>>>
>>>
>>> Most of the older HP analysers drifted a lot and needed a warmup period
>>> of anything up to an hour before they were fit to use. Both the 8566 and
>>> 8568 were a step function improvement, in that they were ready for use
>>> almost right away, Both use the same display section and the 10Hz
>>> resolution bw option is makes them great for close in pjhase noise
>>> checking.
>>>
>>> Have one of each here, came from an rf / emc lab amd must have been on
>>> 24/7. Both display sections had a fuzzy tube, but got on to the John
>>> Miles site and restored them back to sharp focus. Don't know how long
>>> they will last, but only used occasionally, so not a problem. Still a
>>> lot of them in use apparently, probably because they have a spec that's
>>> difficult to match even now...
>>>
>>> Chris
>>>
>>
>> And getting harder.  The SDR-based analyzers are so cheap that they're
>> taking over the market.
>>
>> Cheers
>>
>> Phil "YIG-tuned forever" Hobbs
> 
> Isn't a YIG frequency determined by the mag field? How can that be
> controlled to PPB precision?
> 
> 
> 
It has a super-high Q and a narrow control bandwidth.  SDRs have all 
sorts of sample clock jitter that a DBM driven from a YIG-tuned 
oscillator avoids.  The difference is 30-50 dB close-in phase noise, 
nothing subtle at all, at all.

Cheers

Phil Hobbs

-- 
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

On Sunday, August 30, 2020 at 4:06:50 PM UTC-7, Phil Hobbs wrote:
> On 2020-08-30 17:08, whit3rd wrote:

> >... if you lock it to the XO with any variant of phase-locking,
> > it's NO LONGER phase-locked to the trigger event, and the measurement is worthless.
> > Just cancel the trigger-caused start phase against an XO-caused start phase,
> > and ignore the absolute frequency of the LC entirely (unless you think it drifts
> > enough in a few milliseconds to matter).


> You've never actually used a state-of-the-art digital delay generator, I 
> gather.  Good ones have jitter down around 10 ps over fairly long 
> periods.  Even the SRS DG535 I bought ~25 years ago had jitter less than 
> 100 ps, and newer ones are much better.
> 
> The idea that an LC oscillator could hold the necessary phase accuracy 
> for milliseconds is ridiculous.

There's no milliseconds required on phase accuracy; that's only for the LC
frequency drift from a measurement cycle to a subsequent calibration cycle.   It's not necessary
for the LC to give an accurate sine, just a repeatable waveform, and any jitter is
irrelevant except as an addition to the thermal-noise contribution.

On Monday, August 31, 2020 at 1:59:13 AM UTC+10, Phil Hobbs wrote:
> On 2020-08-26 11:00, jla...@highlandsniptechnology.com wrote: 
> > On Wed, 26 Aug 2020 14:44:47 +0100, Chris <xxx.sys...@gfsys.co.uk> 
> > wrote: 
> > 
> >> On 08/26/20 03:24, jla...@highlandsniptechnology.com wrote: 
> >>> On Tue, 25 Aug 2020 18:42:22 -0700 (PDT), "John Miles, KE5FX" 
> >>> <jmi...@gmail.com> wrote: 
> >>> 
> >>>> On Tuesday, August 25, 2020 at 11:34:31 AM UTC-7, John Larkin wrote: 

<snip>
 
> Just rereading this thread. How about sampling the XO when the trigger 
> appears, at the same time starting the LC. If you know which half-cycle 
> the XO is on (which isn't too hard to do) you can compute the I/Q 
> coefficients that make the initial phases of the LC and XO match. (For 
> extra credit, two XOs locked in quadrature would allow you to always be 
> on a good part of the slope, as well as disambiguating the quadrants.) 

Just start a ramp and stop it on the next clock edge - actually the next clock edge but one to make sure that the ramp has been ramping long enough that initial transients have gone away, then digitise where the ramp has got to. 

Digitising a sine wave and the complementary cosine wave would have worked just as well.

Two ADCs would have been more expensive than our ramp generating circuit (which was just a couple of 5GHz broad-band transistors) but it is certainly a more elegant solution.

The cosine wave wouldn't have to be exactly complementary if you kept track of the actual phase difference from the notionally in-quadrature waveforms (which wouldn't be difficult).

The start a ramp and stop it approach  what we did back in 1988. There was a lot of auto-calibration - which got repeated every few minutes - to make sure that the ramp voltage we digitised started at the bottom of the range of the ADC we were using and always stopped just before the top of the range. 

We used slow DACs to set the starting point of the ramp and and the current charging the capaictor.

Having pairs of delay board with a built-in delay generator made this pretty straight-forward (compared with some of the other stuff we were doing).

I spent a lot of time talking to the guys who wrote the software that made the system work, but I didn't have to discuss that bit - it all worked. I'm sure that there  were holes in my original specification, but the guy that implemented the system was really good (and found that aspect of the design tricky enough to be interesting), and the guy who took over after he went away was pretty good and very thorough (and - very nominally - my boss from then to the end of the project).

-- 
Bill Sloman, Sydney
 

On 2020-08-30 20:58, whit3rd wrote:
> On Sunday, August 30, 2020 at 4:06:50 PM UTC-7, Phil Hobbs wrote:
>> On 2020-08-30 17:08, whit3rd wrote:
> 
>>> ... if you lock it to the XO with any variant of phase-locking, 
>>> it's NO LONGER phase-locked to the trigger event, and the
>>> measurement is worthless. Just cancel the trigger-caused start
>>> phase against an XO-caused start phase, and ignore the absolute
>>> frequency of the LC entirely (unless you think it drifts enough
>>> in a few milliseconds to matter).
> 
> 
>> You've never actually used a state-of-the-art digital delay
>> generator, I gather.  Good ones have jitter down around 10 ps over
>> fairly long periods.  Even the SRS DG535 I bought ~25 years ago had
>> jitter less than 100 ps, and newer ones are much better.
>> 
>> The idea that an LC oscillator could hold the necessary phase
>> accuracy for milliseconds is ridiculous.
> 
> There's no milliseconds required on phase accuracy;

As the wise man said, "When you find yourself in a hole, stop digging." ;)

Reiterating point 1: you've obviously never used a state-of-the-art DDG,
at least not at a level where you needed to knew what it did.

> that's only for the LC frequency drift from a measurement cycle to a
> subsequent calibration cycle.   It's not necessary for the LC to give
> an accurate sine, just a repeatable waveform, and any jitter is 
> irrelevant

An accurate, repeatable, low-jitter delay on an asynchronous waveform is
what a DDG is all about.  See point 1.

 > except as an addition to the thermal-noise contribution.

How would you estimate the thermal noise contribution exactly?

Cheers

Phil Hobbs

-- 
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

On Sunday, August 30, 2020 at 6:48:58 PM UTC-7, Phil Hobbs wrote:
> On 2020-08-30 20:58, whit3rd wrote:
> > On Sunday, August 30, 2020 at 4:06:50 PM UTC-7, Phil Hobbs wrote:
> >> On 2020-08-30 17:08, whit3rd wrote:
> > 
> >>> ... if you lock it to the XO with any variant of phase-locking, 
> >>> it's NO LONGER phase-locked to the trigger event, and the
> >>> measurement is worthless. Just cancel the trigger-caused start
> >>> phase against an XO-caused start phase, and ignore the absolute
> >>> frequency of the LC entirely (unless you think it drifts enough
> >>> in a few milliseconds to matter).
> > 
> > 
> >> You've never actually used a state-of-the-art digital delay
> >> generator, I gather.  Good ones have jitter down around 10 ps over
> >> fairly long periods.  Even the SRS DG535 I bought ~25 years ago had
> >> jitter less than 100 ps, and newer ones are much better.
> >> 
> >> The idea that an LC oscillator could hold the necessary phase
> >> accuracy for milliseconds is ridiculous.
> > 
> > There's no milliseconds required on phase accuracy;
> 
> As the wise man said, "When you find yourself in a hole, stop digging." ;)
> 
> Reiterating point 1: you've obviously never used a state-of-the-art DDG,
> at least not at a level where you needed to knew what it did.

But this is about a triggered LC generator, not about a delay.
Why would a digital delay generator be involved?

> > that's only for the LC frequency drift from a measurement cycle to a
> > subsequent calibration cycle.   It's not necessary for the LC to give
> > an accurate sine, just a repeatable waveform, and any jitter is 
> > irrelevant

> An accurate, repeatable, low-jitter delay on an asynchronous waveform is
> what a DDG is all about.  See point 1.

>  > except as an addition to the thermal-noise contribution.
> 
> How would you estimate the thermal noise contribution exactly?

Minimize, not estimate; you have to buffer the LC waveform into an ADC,
and that means the buffer amp has some high input impedance, and presumably
introduces its own noise (or if you lower the input Z, pulls the LC amplitude or
frequency).

On Sun, 30 Aug 2020 21:48:49 -0400, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

>On 2020-08-30 20:58, whit3rd wrote:
>> On Sunday, August 30, 2020 at 4:06:50 PM UTC-7, Phil Hobbs wrote:
>>> On 2020-08-30 17:08, whit3rd wrote:
>> 
>>>> ... if you lock it to the XO with any variant of phase-locking, 
>>>> it's NO LONGER phase-locked to the trigger event, and the
>>>> measurement is worthless. Just cancel the trigger-caused start
>>>> phase against an XO-caused start phase, and ignore the absolute
>>>> frequency of the LC entirely (unless you think it drifts enough
>>>> in a few milliseconds to matter).
>> 
>> 
>>> You've never actually used a state-of-the-art digital delay
>>> generator, I gather.  Good ones have jitter down around 10 ps over
>>> fairly long periods.  Even the SRS DG535 I bought ~25 years ago had
>>> jitter less than 100 ps, and newer ones are much better.
>>> 
>>> The idea that an LC oscillator could hold the necessary phase
>>> accuracy for milliseconds is ridiculous.
>> 
>> There's no milliseconds required on phase accuracy;
>
>As the wise man said, "When you find yourself in a hole, stop digging." ;)
>
>Reiterating point 1: you've obviously never used a state-of-the-art DDG,
>at least not at a level where you needed to knew what it did.
>
>> that's only for the LC frequency drift from a measurement cycle to a
>> subsequent calibration cycle.   It's not necessary for the LC to give
>> an accurate sine, just a repeatable waveform, and any jitter is 
>> irrelevant
>
>An accurate, repeatable, low-jitter delay on an asynchronous waveform is
>what a DDG is all about.  See point 1.
>
> > except as an addition to the thermal-noise contribution.
>
>How would you estimate the thermal noise contribution exactly?
>
>Cheers
>
>Phil Hobbs

A pretty good triggered LC oscillator will pile up jitter at very
roughly 1 picosecond RMS per microsecond. The junk it takes to start
and stop it compromise the Q some. So it's good to close the PLL and
lock it to a crystal within a microsecond or two. 

A cheap XO will have jitter of maybe 20 ns RMS per second, less if you
get a good one by accident. A $70 OCXO will be roughly a thousand
times better. 

The XO or OCXO will improve greatly with a cover to keep air currents
from wafting over it.



-- 

John Larkin         Highland Technology, Inc

Science teaches us to doubt.

  Claude Bernard
  

On Friday, August 28, 2020 at 9:47:41 AM UTC-7, Phil Hobbs wrote:
> Its main drawback is that you can have the RF or microwave band 
> displayed, but not both at once.  Still, for 2 cents on the dollar one 
> really can't complain too loudly. ;)

Sure it can!

http://www.ke5fx.com/sweep_0_24G.gif

In single-sweep mode at least...

-- john, KE5FX

On a sunny day (Sun, 30 Aug 2020 14:08:38 -0700 (PDT)) it happened whit3rd
<whit3rd@gmail.com> wrote in
<fcef7b0c-f98d-4e4f-8577-91824ca64d2co@googlegroups.com>:

>On Sunday, August 30, 2020 at 9:11:07 AM UTC-7, jla...@highlandsniptechnology.com wrote:
>
>> >>>> I start an LC oscillator when I get a trigger, and use it to time out
>> >>>> delays. The ADC is clocked from an OCXO and observes the waveform of
>> >>>> the triggered LC oscillator, and I close a loop to lock the LC to the
>> >>>> XO. Actually, the LC frequency is whatever it wants to be. The math
>> >>>> gets ugly.
>
>> I'm thinking along those same lines. If I can (quickly!) measure the
>> phase angle between my XO and my triggered oscillator, I can seize the
>> initial phase offset and close a loop on that. ...
>
>> My triggered LC oscillator is great for a couple of microseconds, but
>> is piling up drift and jitter. It needs to be locked to a good XO
>> long-term.
>
>This makes no sense; if you lock it to the XO with any variant of phase-locking,
>it's NO LONGER phase-locked to the trigger event, and the measurement is worthless.
>Just cancel the trigger-caused start phase against an XO-caused start phase,
>and ignore the absolute frequency of the LC entirely (unless you think it drifts
>enough in a few milliseconds to matter).

Indeed.
Perhaps if you PLL the LC to a precision reference and set its frequency via a DAC and varicap,
that  would be like tuning a tuning fork by removing or adding some weight as in the mechanical equivalent,
and then leave it (maybe do that calibration just before the measurement, calibration button?,
Trigger would be like pinging a precisely tuned crystal / tuning fork.
The tuning needs not happen _during_ the trigger, assuming drift is OK?

Anyways he also writes mF (milli Farad) if he means uF (micro Farad) I think, so ..
factor a couple of thousand, so maybe it is not so critical ;-)
 

On 2020-08-30 23:05, John Miles, KE5FX wrote:
> On Friday, August 28, 2020 at 9:47:41 AM UTC-7, Phil Hobbs wrote:
>> Its main drawback is that you can have the RF or microwave band
>> displayed, but not both at once.  Still, for 2 cents on the dollar one
>> really can't complain too loudly. ;)
> 
> Sure it can!
> 
> http://www.ke5fx.com/sweep_0_24G.gif
> 
> In single-sweep mode at least...

I'm a happy user of your GPIB Tools via a Prologix GPIB-Ethernet 
adapter.  Thanks for making them available!

Cheers

Phil Hobbs

-- 
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

On 2020-08-30 22:46, jlarkin@highlandsniptechnology.com wrote:
> On Sun, 30 Aug 2020 21:48:49 -0400, Phil Hobbs
> <pcdhSpamMeSenseless@electrooptical.net> wrote:
> 
>> On 2020-08-30 20:58, whit3rd wrote:
>>> On Sunday, August 30, 2020 at 4:06:50 PM UTC-7, Phil Hobbs wrote:
>>>> On 2020-08-30 17:08, whit3rd wrote:
>>>
>>>>> ... if you lock it to the XO with any variant of phase-locking,
>>>>> it's NO LONGER phase-locked to the trigger event, and the
>>>>> measurement is worthless. Just cancel the trigger-caused start
>>>>> phase against an XO-caused start phase, and ignore the absolute
>>>>> frequency of the LC entirely (unless you think it drifts enough
>>>>> in a few milliseconds to matter).
>>>
>>>
>>>> You've never actually used a state-of-the-art digital delay
>>>> generator, I gather.  Good ones have jitter down around 10 ps over
>>>> fairly long periods.  Even the SRS DG535 I bought ~25 years ago had
>>>> jitter less than 100 ps, and newer ones are much better.
>>>>
>>>> The idea that an LC oscillator could hold the necessary phase
>>>> accuracy for milliseconds is ridiculous.
>>>
>>> There's no milliseconds required on phase accuracy;
>>
>> As the wise man said, "When you find yourself in a hole, stop digging." ;)
>>
>> Reiterating point 1: you've obviously never used a state-of-the-art DDG,
>> at least not at a level where you needed to knew what it did.
>>
>>> that's only for the LC frequency drift from a measurement cycle to a
>>> subsequent calibration cycle.   It's not necessary for the LC to give
>>> an accurate sine, just a repeatable waveform, and any jitter is
>>> irrelevant
>>
>> An accurate, repeatable, low-jitter delay on an asynchronous waveform is
>> what a DDG is all about.  See point 1.
>>
>>> except as an addition to the thermal-noise contribution.
>>
>> How would you estimate the thermal noise contribution exactly?
>>
>> Cheers
>>
>> Phil Hobbs
> 
> A pretty good triggered LC oscillator will pile up jitter at very
> roughly 1 picosecond RMS per microsecond. The junk it takes to start
> and stop it compromise the Q some. So it's good to close the PLL and
> lock it to a crystal within a microsecond or two.
> 
> A cheap XO will have jitter of maybe 20 ns RMS per second, less if you
> get a good one by accident. A $70 OCXO will be roughly a thousand
> times better.
> 
> The XO or OCXO will improve greatly with a cover to keep air currents
> from wafting over it.

Especially at low modulation frequencies.  Gerhard and I were discussing 
that yesterday in the tantalum caps thread--thermal drifts give rise to 
very steeply rising baseband noise.

Cheers

Phil Hobbs

-- 
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com