Re: Distorted Sine Wave

On Sun, 2 Jun 2024 20:58:45 -0000 (UTC), Cursitor Doom
<cd999666@notformail.com> wrote:

On Sun, 02 Jun 2024 14:08:48 -0400, Joe Gwinn wrote:
>

On Sun, 2 Jun 2024 16:55:28 -0000 (UTC), Cursitor Doom
<cd999666@notformail.com> wrote:
 

On Sun, 02 Jun 2024 12:19:05 -0400, Joe Gwinn wrote:
>

On Sun, 2 Jun 2024 11:31:33 -0000 (UTC), Cursitor Doom
<cd999666@notformail.com> wrote:
 

On Sun, 2 Jun 2024 11:17:58 -0000 (UTC), piglet wrote:
>

Cursitor Doom <cd999666@notformail.com> wrote:

On Sat, 1 Jun 2024 22:00:58 -0000 (UTC), piglet wrote:
 

Cursitor Doom <cd999666@notformail.com> wrote:

On Sat, 1 Jun 2024 15:44:17 +0200, Jeroen Belleman wrote:
 

On 6/1/24 14:07, Cursitor Doom wrote:

 

I've taken a shot of the waveform into the 50 ohm input. It's
around 850mV peak-peak. Hopefully the slight distortion I spoke
about is visible; the slightly more leisurely negative-going
excursions WRT their positive-going counterparts. So it's not a
pure sine wave as one would expect. Does it matter? I don't
know!
 
<https://disk.yandex.com/i/7cuuBimDbOIBZw>

 

And <https://disk.yandex.com/i/z6fYbeVfPRK7aA>

The shape looks perfectly acceptable to me. This is +3dBm into
50 Ohms.
Is that what it's supposed to be? Canned reference oscillators
most often deliver +13dBm, sometimes +10dBm.

 
Is it? I only make it about half your figure: +1.65dBm.
I admit I'm frequently prone to careless errors, so stand to be
corrected,
but here's my method:
850mV peak to peak is 425mV peak voltage. Average of that is
0.425x0.636 =
0.27V. Average power is average volts squared divided by the load
impedance of 50 ohms = 1.46mW = +1.65dBm.
 
I shall consult the manual to see what it ought to be - if I can
find it, that is, as PDF manuals are a nightmare to navigate IME.
 
 
 

Use 0.71 for RMS instead of 0.636 ! I make that about 1.8mW or
+2.6dBm ?

 
Thanks, Erich. But there's no such thing as "RMS power" strictly
speaking IIRC, so that's why I took the average figure; not that it
makes much difference in practice. it does seem a bit on the low
side, but despite reading through the most likely sources (the
service manual and the trouble-shooting/repair manual) I can find
nothing stated for what that signal level should be! This may be
due to the user-unfriendliness of very large PDF manuals; I just
don't know. Anyway, not very satisfactory! Later today I plan to do
a direct power meter measurement of the ref osc (since none of us
here seem to agree on what 850mV vs 50 ohms equates to!!)
 
 
 
 
 

Since you have a power meter, a signal source, and an oscilloscope
why not measure the peak to peak voltage on the scope and power on
the power meter and see which calculation 0.636 vs 0.707 gives the
closest agreement?

>
It wouldn't prove anything one way or ther other, though, since that
power meter hasn't been calibrated for "quite a while" so to speak. :)
It'll give a 'good enough' reading for my purposes, but won't be
accurate enough to meaningfully test your otherwise fine suggestion.

 
 
The 0 to +10 dBm range I mentioned came from the service manual.
 
Looking at your scope picture, it looks like a 3 Vpp signal, which is
+13 dBm, a very common distribution level, but one that exceeds the
analyzer's allowed range.  All that's needed to fix this is a 3dB
inline attenuator.  Here is one for SMA connectors:
 
.<https://www.amazon.com/MWRF-Source-Male-Female-Attenuator/dp/

B07MP9D9GC?th=1>

 
Just buying a few of these and doing some experiments will be far
cheaper and faster than the various alternatives discussed.t
 
Joe Gwinn

>
I think you're looking at the first picture with the signal into the
scope's 1 Meg input. The 50 ohm trace is only 850mV peak-to-peak or
thereabouts and when I measured it with an actual power meter, came out
at about +2.5dBm so within the range you stated; no attenuation needed
(thanks for the range, by the way; I needed to know that).

 
What we don't know is exactly how you made the various measurements. If
you are observing the signal from the 10 MHz reference where it enters
the analyzer, I would expect that there is a T-connector with the scope
(set to 1 Mohm) listening in to passing signals.

>
You did ask me this before and did post an answer. See Message-ID:
<v3fsbp$2u0a6$1@dont-email.me>
>
You also still appear to think that the 10Mhz signal is going into the
analyzer. It isn't. It's coming out. Again, see Message-ID:
<v3fsbp$2u0a6$1@dont-email.me>

I did read that, but didn't know what to make of it.  I think an
annotated drawing is required.

On this drawing, where do the various scope traces mentioned up-thread
come from?

Joe Gwinn

In this case, the load seen by the incoming reference is that provided
by the input on the analyzer.  Which input is +10 dBm max.  If you set
the observing scope input to 50 ohm, the reference will see a 25 ohm
load, cutting the signal seen by the analyzer by 3 dB.  Which will take
+13 dBm down to +10 dBm, which is in range.
 
A 3dB attenuator in line will drop the signal to 10 dBm as well.
 
I've built lots of systems like that.  The 10 MHz reference is delivered
to everybody at +13 dBm, and it is the receivers' responsibility to
attenuate it to whatever they need.
 
 

I've now measured the 100Mhz oscillator and that seems fine, although I
only saw 0.61V p-p into 50 ohms, so somewhat less than the 10Mhz
oscillator's output.
So far, I've not measured anything which screams "the fault's here!" as
all the expected signals are present - although admittedly I have many
more to test. But certainly all the *major* signals within this complex
beast are present. It's looking like it could be an issue with one of
the phase detectors or LPFs. Sigh....

 
To my eye, it does scream.
 
Joe Gwinn