Path: ...!eternal-september.org!feeder3.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail From: Cursitor Doom Newsgroups: sci.electronics.design Subject: Re: Distorted Sine Wave Date: Sat, 1 Jun 2024 21:46:10 -0000 (UTC) Organization: A noiseless patient Spider Lines: 169 Message-ID: References: <48bd78e1-7da8-3bba-2879-d22962203fa3@electrooptical.net> <9olh5j9al34fhrebr4grqq8h6c8javjpp1@4ax.com> <1n0i5jh257hiinlj2dhaatlo11s33m5n0e@4ax.com> <9k2i5jpfhu3ncfpm28ukusrok4hugal80s@4ax.com> <0v3n5jh1mrclpfh9jru7k4n54v9ovs03q5@4ax.com> MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Injection-Date: Sat, 01 Jun 2024 23:46:11 +0200 (CEST) Injection-Info: dont-email.me; posting-host="722a1db93b34a23f770c1bceecb5bb8e"; logging-data="3080518"; mail-complaints-to="abuse@eternal-september.org"; posting-account="U2FsdGVkX18nvsdcgxILMxHHa9cxuIQxE26ufrouCa0=" User-Agent: Pan/0.149 (Bellevue; 4c157ba) Cancel-Lock: sha1:nuyPbc48ejuejAAeuBKXczJfybs= Bytes: 8731 On Sat, 01 Jun 2024 17:18:03 -0400, Joe Gwinn wrote: > On Sat, 1 Jun 2024 19:21:29 -0000 (UTC), Cursitor Doom > wrote: > >>On Sat, 01 Jun 2024 14:28:35 -0400, Joe Gwinn wrote: >> >>> On Fri, 31 May 2024 22:17:37 -0000 (UTC), Cursitor Doom >>> 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 >>>>> wrote: >>>>> >>>>>>On Fri, 31 May 2024 11:34:46 -0400, Joe Gwinn wrote: >>>>>> >>>>>>> On Thu, 30 May 2024 16:24:25 -0700, john larkin >>>>>>> wrote: >>>>>>> >>>>>>>>On Thu, 30 May 2024 18:53:29 -0400, Joe Gwinn >>>>>>>> >>>>>>>>wrote: >>>>>>>> >>>>>>>>>On Thu, 30 May 2024 14:56:30 -0700, john larkin >>>>>>>>>wrote: >>>>>>>>> >>>>>>>>>>On Thu, 30 May 2024 16:06:47 -0400, Joe Gwinn >>>>>>>>>> >>>>>>>>>>wrote: >>>>>>>>>> >>>>>>>>>>>On Thu, 30 May 2024 15:02:44 -0400, Phil Hobbs >>>>>>>>>>> wrote: >>>>>>>>>>> >>>>>>>>>>>>On 2024-05-29 18:59, Joe Gwinn wrote: >>>>>>>>>>>>> On Wed, 29 May 2024 22:11:47 -0000 (UTC), Phil Hobbs >>>>>>>>>>>>> wrote: >>>>>>>>>>>>> >>>>>>>>>>>>>> Cursitor Doom 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 >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> 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. >>>>>>>>>>> >>>>>>>>>>>.>>>>>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!) > > Maybe. But it's simpler to just try the attenuator. > > Joe Gwinn If I can find a reverse attenuator, I'd be happy to try it.