Path: ...!eternal-september.org!feeder3.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail
From: Cursitor Doom <cd999666@notformail.com>
Newsgroups: sci.electronics.design
Subject: Re: Distorted Sine Wave
Date: Sat, 1 Jun 2024 19:21:29 -0000 (UTC)
Organization: A noiseless patient Spider
Lines: 158
Message-ID: <v3fsbp$2u0a6$1@dont-email.me>
References: <v38973$1atf2$1@dont-email.me>
	<3lcf5jd7li0a3c0fgddt7o8lnfocvls2pr@4ax.com>
	<48bd78e1-7da8-3bba-2879-d22962203fa3@electrooptical.net>
	<9olh5j9al34fhrebr4grqq8h6c8javjpp1@4ax.com>
	<seth5jlmgu2gv6lr61m31jk2q94073rvtk@4ax.com>
	<1n0i5jh257hiinlj2dhaatlo11s33m5n0e@4ax.com>
	<9k2i5jpfhu3ncfpm28ukusrok4hugal80s@4ax.com>
	<c6rj5j1l1gfoskul3nnvudf3nc57017k84@4ax.com> <v3d1eb$2anif$1@dont-email.me>
	<go6k5jd7754174pb4k7ko1suh89acqervm@4ax.com> <v3dia1$2anif$4@dont-email.me>
	<rcom5jpoi28rc5hucm2p2dcsiu6es8mt6m@4ax.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Injection-Date: Sat, 01 Jun 2024 21:21:30 +0200 (CEST)
Injection-Info: dont-email.me; posting-host="722a1db93b34a23f770c1bceecb5bb8e";
	logging-data="3080518"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX19WgPXjvGN7b8V9yl3i9C9shAvWnGUEWHA="
User-Agent: Pan/0.149 (Bellevue; 4c157ba)
Cancel-Lock: sha1:70BIoq4I4PwSrX7nEdkf3kGFHUA=
Bytes: 8111

On Sat, 01 Jun 2024 14:28:35 -0400, Joe Gwinn wrote:

> On Fri, 31 May 2024 22:17:37 -0000 (UTC), Cursitor Doom
> <cd999666@notformail.com> wrote:
> 
>>On Fri, 31 May 2024 14:48:28 -0400, Joe Gwinn wrote:
>>
>>> On Fri, 31 May 2024 17:29:47 -0000 (UTC), Cursitor Doom
>>> <cd999666@notformail.com> wrote:
>>> 
>>>>On Fri, 31 May 2024 11:34:46 -0400, Joe Gwinn wrote:
>>>>
>>>>> On Thu, 30 May 2024 16:24:25 -0700, john larkin <jl@650pot.com>
>>>>> wrote:
>>>>> 
>>>>>>On Thu, 30 May 2024 18:53:29 -0400, Joe Gwinn <joegwinn@comcast.net>
>>>>>>wrote:
>>>>>>
>>>>>>>On Thu, 30 May 2024 14:56:30 -0700, john larkin <jl@650pot.com>
>>>>>>>wrote:
>>>>>>>
>>>>>>>>On Thu, 30 May 2024 16:06:47 -0400, Joe Gwinn
>>>>>>>><joegwinn@comcast.net>
>>>>>>>>wrote:
>>>>>>>>
>>>>>>>>>On Thu, 30 May 2024 15:02:44 -0400, Phil Hobbs
>>>>>>>>><pcdhSpamMeSenseless@electrooptical.net> wrote:
>>>>>>>>>
>>>>>>>>>>On 2024-05-29 18:59, Joe Gwinn wrote:
>>>>>>>>>>> On Wed, 29 May 2024 22:11:47 -0000 (UTC), Phil Hobbs
>>>>>>>>>>> <pcdhSpamMeSenseless@electrooptical.net> wrote:
>>>>>>>>>>> 
>>>>>>>>>>>> Cursitor Doom <cd999666@notformail.com> wrote:
>>>>>>>>>>>>> On Wed, 29 May 2024 13:42:13 -0700, john larkin wrote:
>>>>>>>>>>>>>
>>>>>>>>>>>>>> On Wed, 29 May 2024 21:43:54 +0200, Arie de Muijnck
>>>>>>>>>>>>>> <noreply@ademu.com>
>>>>>>>>>>>>>> wrote:
>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> On 2024-05-29 19:07, Cursitor Doom wrote:
>>>>>>>>>>>>>>>> Gentlemen,
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> Whilst fault-finding on my HP 8566B spectrum analyzer,
>>>>>>>>>>>>>>>> I've found the 10Mhz reference oscillator is generating
>>>>>>>>>>>>>>>> an 'unsatisfactory waveform'
>>>>>>>>>>>>>>>> which may be causing the device to be unable to lock it's
>>>>>>>>>>>>>>>> main PLL. I've come across this waveshape before, but
>>>>>>>>>>>>>>>> mostly with oscillators I was building and in the process
>>>>>>>>>>>>>>>> of trying to iron out the wrinkles of and certainly NOT a
>>>>>>>>>>>>>>>> critical reference oscillator from a respected
>>>>>>>>>>>>>>>> manufacturer. Can anyone tell what's most likely going on
>>>>>>>>>>>>>>>> here?
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> https://disk.yandex.com/i/z6fYbeVfPRK7aA
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> Looks like reflections in the cable. Try the 50 Ohm
>>>>>>>>>>>>>>> termination.
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> Arie
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> If the drive is a sine wave, a cable can't generate that
>>>>>>>>>>>>>> 2nd harmonic.
>>>>>>>>>>>>>
>>>>>>>>>>>>> I don't understand how a reflection can account for it
>>>>>>>>>>>>> either.
>>>>>>>>>>>>> THe cable's only 4' long! However, with the 50 ohm input
>>>>>>>>>>>>> enabled, the 2nd harmonic disappears. It's just one of those
>>>>>>>>>>>>> inexplicable mysteries that no one knows the answer to. :)
>>>>>>>>>>>>
>>>>>>>>>>>> That?s pretty diagnostic. There must be an LC filter on the
>>>>>>>>>>>> output?mis-terminating it will cause all sorts of
>>>>>>>>>>>> frequency-response whoopdedoos.
>>>>>>>>>>> 
>>>>>>>>>>> It also occurs to me that if there is a diode in series with a
>>>>>>>>>>> resistor somewhere, the impedance presented to the feed coax
>>>>>>>>>>> may be 50 ohms for positive input voltage, and say 10 Kohm for
>>>>>>>>>>> negative.  At the very least one could get an inverted
>>>>>>>>>>> reflection on negative.
>>>>>>>>>>> 
>>>>>>>>>>> Joe Gwinn
>>>>>>>>>>> 
>>>>>>>>>>Yeah, or an emitter follower.  Good point.
>>>>>>>>>
>>>>>>>>>Plus some LC filter wiggles, to distort and smooth things.
>>>>>>>>>
>>>>>>>>>I dug around and found a copy of the HP 8566B spectrum analyzer
>>>>>>>>>service manual.  The 10 MHz ref input is an amplifier driving a
>>>>>>>>>mixer,
>>>>>>>>>with not hint of for instance a TTL input.  So, the problem must
>>>>>>>>>be elsewhere.  Or, it's just busted.
>>>>>>>>>
>>>>>>>>>.<chrome-extension://efaidnbmnnnibpcajpcglclefindmkaj/https://
>>>>xdevs.com/doc/HP_Agilent_Keysight/
>>>>HP%208566B%20Troubleshooting%20&%20Repair%20Vol.%201.pdf>
>>>>>>>>>
>>>>>>>>>Joe Gwinn
>>>>>>>>
>>>>>>>>334 pages! Where is the issue?
>>>>>>>
>>>>>>>The pagination was unclear, so I didn't see a way to say.  How I
>>>>>>>found it was to look at the schematics.  Look for module A22. 
>>>>>>>Don't think search works on such images, but visual search didn't
>>>>>>>take that long.
>>>>>>>
>>>>>>>Joe Gwinn
>>>>>>
>>>>>>Too much work for free consulting.
>>>>> 
>>>>> I took another look.  There is a page reference on the right edge,
>>>>> near the bottom, which is not marked as a page X of Y, but is.
>>>>> 
>>>>> Anyway, look at pages 59 and 77.  At 77, look for A22 in the lower
>>>>> left region.  Just above, look for INT and EXT jacks.  The EXT jack
>>>>> is where an external reference enters.
>>>>> 
>>>>> Joe Gwinn
>>>>
>>>>I don't believe the 10Mhz ref osc is the problem, John. The 2nd
>>>>harmonic distortion goes away when the scope input impedance is set to
>>>>50 ohms. There is some slight distortion on it, but not enough to
>>>>cause an out-of- lock error. Furthermore, since that osc is the
>>>>pace-setter for every other module in every other loop in this
>>>>analyzer, its failure would give rise to way more error messages than
>>>>a mere "YTO unlock" as it stands at present. The manual suggests the
>>>>most likely areas where the fault is located are in one of the boards
>>>>A19, A20, A21 or A11.  If A22 were the culprit, there would be over a
>>>>dozen error messages.
>>> 
>>> What is the 10 MHz signal power level at the EXT input in these two
>>> castes, 1 Mohm and 50 Ohm?  The expected range is 0 to +10 dBm.
>>> 
>>> Joe Gwinn
>>
>>7.68dBm on 50 Ohms 2.68V P-P on 1 Meg
>>
>>No issues there AFAIC.
> 
> Not so fast there.  It's the scope that is being set to 1 Meg or 50 ohm,
> and at 1 Meg we are observing the drive into the 50 ohm input impedance
> of the EXT ref input of the spectrum analyzer.

Unless I misunderstand what you're saying, that's not the case, though. 
What we are observing is the output of the analyzer's 10Mhz reference 
oscillator taken from a BNC socket on the rear of it which HP have 
thoughtfully provided and fed directly into a scope switchable between 1M 
and 50 Ohms.

> That 2.7 Vpp over 50 ohms is about +13 dBm, which exceeds +10 dBm.  So
> the EXT input may be over-driven. Install a 5-dB inline attenuator and
> see what happens.

It's not an external input. They've put it there in order to make it easy 
to check the frequency of the reference oscillator and adjust it to 
precisely 10.000000Mhz if necessary (after a *minimum* 72 hour warm-up!)