Discussion:
A silly question ...
(too old to reply)
Dave B via time-nuts
2018-09-27 16:55:37 UTC
Permalink
... Because I'm sure I should be able to figure this out for myself!

I have (as many of you do also) one of the venerable Trimble Thunderbolt
devices.  No problem with that.  All works fine, and is run 24/7, UPS
backup power and all...

I also have (again, as many of you do...) a free running OCXO for 10MHz
used for "other" stuff etc.   Also left running for long periods, but
only when I want to experiment with other stuff and don't want to
disturb what the TB is keeping sane..

Triggering a dual beam 'scope (Tek 465) from the TB on Ch1, and having
the output of the OCXO on Ch2, the resulting display on Ch2 of course
drifts in relation to the static waveform on Ch1.  (Both nice sinusoids.)

If I time how long it takes for the OCXO to drift through one full cycle
(co-incidence to co-incidence) relative to the TB on Ch1, how exactly do
I turn that time, and knowing the base frequency of the TB at 10MHz,
into a ppm discrepancy?

"I think", that if for example, it takes 1 second to drift one cycle,
that works out at 0.1 ppm.   If it takes 2 seconds, it's 0.05 ppm, if it
takes 5 seconds, it's 0.02 ppm etc.   Is that correct?

If not, please feel free to educate me!

As I said, a silly question that I'm sure I would have answered myself a
few decades ago, but age and medication etc...

Interestingly, after "a lot" of googling, I see that anything like this
using "analogue" or "CRT" scopes, has fallen off the radar and the
interweb, and some of the practices using digital oscilloscopes seem to
rely on the instrument itself to make the measurement, rather than from
"observation" and common sense.  (That I seem to lack at times too!)

But I did get diverted into reading up on some of the early history of
CRO's.

Regards to All.

Dave B.
--
Created on and sent from a Unix like PC running and using free and open source software.
::


_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
David G. McGaw
2018-09-27 17:38:00 UTC
Permalink
Correct.

David N1HAC
Post by Dave B via time-nuts
... Because I'm sure I should be able to figure this out for myself!
I have (as many of you do also) one of the venerable Trimble Thunderbolt
devices.  No problem with that.  All works fine, and is run 24/7, UPS
backup power and all...
I also have (again, as many of you do...) a free running OCXO for 10MHz
used for "other" stuff etc.   Also left running for long periods, but
only when I want to experiment with other stuff and don't want to
disturb what the TB is keeping sane..
Triggering a dual beam 'scope (Tek 465) from the TB on Ch1, and having
the output of the OCXO on Ch2, the resulting display on Ch2 of course
drifts in relation to the static waveform on Ch1.  (Both nice sinusoids.)
If I time how long it takes for the OCXO to drift through one full cycle
(co-incidence to co-incidence) relative to the TB on Ch1, how exactly do
I turn that time, and knowing the base frequency of the TB at 10MHz,
into a ppm discrepancy?
"I think", that if for example, it takes 1 second to drift one cycle,
that works out at 0.1 ppm.   If it takes 2 seconds, it's 0.05 ppm, if it
takes 5 seconds, it's 0.02 ppm etc.   Is that correct?
If not, please feel free to educate me!
As I said, a silly question that I'm sure I would have answered myself a
few decades ago, but age and medication etc...
Interestingly, after "a lot" of googling, I see that anything like this
using "analogue" or "CRT" scopes, has fallen off the radar and the
interweb, and some of the practices using digital oscilloscopes seem to
rely on the instrument itself to make the measurement, rather than from
"observation" and common sense.  (That I seem to lack at times too!)
But I did get diverted into reading up on some of the early history of
CRO's.
Regards to All.
Dave B.
_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
and follow the i
David Van Horn
2018-09-27 17:39:39 UTC
Permalink
Even with a DSO, if you sync to the Thunderbolt output, you can watch the free running osc drift relative to the Thunderbolt.

_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
and follow the instructions there.
Tom Van Baak
2018-09-27 18:18:16 UTC
Permalink
Post by Dave B via time-nuts
"I think", that if for example, it takes 1 second to drift one cycle,
that works out at 0.1 ppm. If it takes 2 seconds, it's 0.05 ppm, if it
takes 5 seconds, it's 0.02 ppm etc. Is that correct?
Yes. At 10 MHz one full cycle is 100 ns. So if the cycles are drifting by 100 ns per second that's 100e-9 s / 1 s = 1e-7 = 0.1 ppm.

At these levels of frequency accuracy, using a 'scope is plenty good enough. In fact, it's more educational and somehow more enjoyable to watch analog sinewaves drift past each other than it is to see the digital display of boring frequency counter.

Where the 'scope method starts to break down is when the frequency error gets down to the ppb level. At 1e-9 it will take 100 s for the waveforms to drift by one cycle. And at 1e-12 you would have to wait an entire day (100 ns / 86400 s = 1.157e-12).

On the other hand, with frequency offsets this low you don't have to sit there the whole time. One trick would be to take a photo of the 'scope once an hour, or, say, once every 1000 s. If you played that back at 1 fps you'd have a 1000x "time magnifier".

/tvb

_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
and follow the instructions there.
b***@verizon.net
2018-09-28 11:33:29 UTC
Permalink
This works for me:

1. 10 Mhz to be measured input to vertical channel
2. Standard input to ext trig.
3. Set scope for 10 ns per div. (1 cycle of 10 Mhz will fill up whole
screen)
4 Time the time it takes for trace to move 1 div. (not the whole cycle -
just one of its sides.)
5. Divide these seconds into 1x10-8 using your calculator. ( I set 1x10-8
into one of the memories)
6. The answer will give you your offset.
7. Example: If it takes 20 seconds for one of the sides of the cycle to
move, your offset is 5x10-10.
It does not take all day to get a reading,

73

Bill, WA2DVU
Cape May

-----Original Message-----
From: time-nuts <time-nuts-***@lists.febo.com> On Behalf Of Tom Van Baak
Sent: Thursday, September 27, 2018 2:18 PM
To: Discussion of precise time and frequency measurement
<time-***@lists.febo.com>
Subject: Re: [time-nuts] A silly question ...
Post by Dave B via time-nuts
"I think", that if for example, it takes 1 second to drift one cycle,
that works out at 0.1 ppm. If it takes 2 seconds, it's 0.05 ppm, if it
takes 5 seconds, it's 0.02 ppm etc. Is that correct?
Yes. At 10 MHz one full cycle is 100 ns. So if the cycles are drifting by
100 ns per second that's 100e-9 s / 1 s = 1e-7 = 0.1 ppm.

At these levels of frequency accuracy, using a 'scope is plenty good enough.
In fact, it's more educational and somehow more enjoyable to watch analog
sinewaves drift past each other than it is to see the digital display of
boring frequency counter.

Where the 'scope method starts to break down is when the frequency error
gets down to the ppb level. At 1e-9 it will take 100 s for the waveforms to
drift by one cycle. And at 1e-12 you would have to wait an entire day (100
ns / 86400 s = 1.157e-12).

On the other hand, with frequency offsets this low you don't have to sit
there the whole time. One trick would be to take a photo of the 'scope once
an hour, or, say, once every 1000 s. If you played that back at 1 fps you'd
have a 1000x "time magnifier".

/tvb

_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com To unsubscribe, go to
http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
and follow the instructions there.


_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
and follow the instructions there.
Tom Curlee
2018-09-27 21:30:27 UTC
Permalink
This is assuming that your oscilloscope is set at 100 nS/DIV.  My Tek 2465A will sweep at 5 nS/DIV normally, and 500 pS/DIV when the sweep is set to X10.  If I'm figuring correctly, this will allow 1e-12 in 432 seconds.  I use a stop watch to time the zero crossings of the sine wave - something like:

Assuming the wave moved 5 divisions in 185 seconds:   (500 pS * 5 DIV)/185 seconds = 13.5e-12 ppm.    (I think this is correct.  tvb?)

Since a 10 MHz sine wave tends to look like a flat line at 500 pS/DIV, I often set the vertical V/DIV to 2 mV/DIV.

Tom




From: Tom Van Baak <***@LeapSecond.com>
To: Discussion of precise time and frequency measurement <time-***@lists.febo.com>
Sent: Thursday, September 27, 2018 11:19 AM
Subject: Re: [time-nuts] A silly question ...
Post by Dave B via time-nuts
"I think", that if for example, it takes 1 second to drift one cycle,
that works out at 0.1 ppm. If it takes 2 seconds, it's 0.05 ppm, if it
takes 5 seconds, it's 0.02 ppm etc. Is that correct?
Yes. At 10 MHz one full cycle is 100 ns. So if the cycles are drifting by 100 ns per second that's 100e-9 s / 1 s = 1e-7 = 0.1 ppm.

At these levels of frequency accuracy, using a 'scope is plenty good enough. In fact, it's more educational and somehow more enjoyable to watch analog sinewaves drift past each other than it is to see the digital display of boring frequency counter.

Where the 'scope method starts to break down is when the frequency error gets down to the ppb level. At 1e-9 it will take 100 s for the waveforms to drift by one cycle. And at 1e-12 you would have to wait an entire day (100 ns / 86400 s = 1.157e-12).

On the other hand, with frequency offsets this low you don't have to sit there the whole time. One trick would be to take a photo of the 'scope once an hour, or, say, once every 1000 s. If you played that back at 1 fps you'd have a 1000x "time magnifier".

/tvb

_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
and follow the instructions there.



_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
and follow the instruction
Bill Byrom
2018-09-28 12:04:45 UTC
Permalink
Triggering a dual beam 'scope (Tek 465) from the TB on Ch1, and having> the output of the OCXO on Ch2, the resulting display on Ch2 of course> drifts in relation to the static waveform on Ch1. (Both nice
sinusoids.)
The Tek 465 analog cathode ray oscilloscope was/is a very flexible
instrument. But this flexibility allows you to set up the instrument in
ways which will not allow this commonly used oscillator comparison
technique to work correctly. Since you are interested in these
instruments, here are some details about setting up the instrument for
such comparisons.
(1) The Tek 465 is not a dual beam oscilloscope. Dual beam oscilloscopes
(such as the Tektronix 556 and 7844) use a special CRT which
incorporates two independent electron guns. Each electron gun
assembly has a set of vertical and horizontal deflection plates.
There are two vertical amplifiers (one for each electron gun) and
two horizontal sweep systems (one for each electron gun). If you had
a dual beam oscilloscope you could compare oscillator#1 to
oscillator#2 while simultaneously comparing oscillator#3 with
oscillator#4. It's like having two independent oscilloscopes sharing
the same CRT display.
(2) The Tek 465 single beam oscilloscope can display two traces on the
display using one of two methods:(a) Chopped trace display: This mode works well at low sweep rates (such
as 1 ms/div) but causes trouble at fast sweep rates (such as 1
us/div). The displayed trace is switched between Channel 1 and
Channel 2 at a fixed rate of about 500 kHz.(b) Alternate trace display: This mode works well at high sweep rates
but is hard to see at low sweep rates. The scope alternates between
displaying one sweep of Channel 1 and one sweep of Channel 2.
(3) The trigger source setting is crucial to using this technique to
compare oscillators. The technique does not require you to display
two channels. What is important is that you display one oscillator
while triggering on the other oscillator. The trigger source can
be set to:(a) CH 1: The Channel 2 display will drift if the two signals have a
varying phase relationship.(b) CH 2: The Channel 1 display will drift if the two signals have a
varying phase relationship.(c) NORM (normal): The trigger system gets input from the channel being
displayed at that moment. So in chopped trace display mode the
trigger is rapidly switched between CH1 and CH2, and in alternate
trace display mode the trigger alternates between CH1 and CH2 on
alternate sweeps. In all cases, you should not use NORM trigger
source with both channels displayed when comparing oscillators!(d) EXT: You apply the trigger signal to the external trigger input
connector. This works well well when comparing oscillators. If you
use alternate trace display mode and an external trigger, you can
compare oscillator#1 (on CH 1) to oscillator#0 (on the external
trigger input) while you are also comparing oscillator#2 (on CH2)
oscillator#0. So you could compare two oscillators (one on CH1 and
the other on CH2) to a GPSDO (on the external trigger input).
(4) When comparing oscillators, the fractional frequency difference
(such as ppm Parts Per Million or ppb Parts Per Billion) you can
measure depends on the oscilloscope sweep rate. What you are really
measuring is the drift of the time delay between the edge (or zero
crossing of a sine wave) of one signal relative to an edge or zero
crossing of another signal. The relationship is:
Fractional difference = (observed timing change) / (measurement
interval)Here are some examples:
Fractional difference in ppm = (time delay drift in us) per second of
observation timeFractional difference in ppb = (time delay drift in ns) per second of
observation time
(5) As you can see in my previous section, you need a very fast sweep
rate (small time/div) to measure small fractional frequency
differences. This means that for a small fractional frequency
difference with a moderately low measured oscillator frequency (such
as 1 MHz), you may not see any edges for a long time when you use a
small time/div. The Tek 465 has a delayed timebase, and you can use
this feature to move the signal edge (or zero crossing) onto the
screen. You can then watch the signal for a few seconds to determine
the timing drift rate. If the edge is drifting at 10 ns per 10
seconds, the fractional difference is 1 ppb (1 part in 10^9). If the
displayed oscillator edge is drifting to the left (earlier in time),
the displayed oscillator frequency is higher than the reference
oscillator you are using for the trigger. If the displayed
oscillator edge is drifting to the right (later in time), the
displayed oscillator frequency is lower than the reference
oscillator you are using for the trigger.
(6) If the edge rate is not very fast (such as when you are measuring
sinewave signals), the waveform edge you see at a fast sweep rate
will appear to be nearly horizontal (spread out across many
divisions). You normally want to measure the displayed signal at the
midpoint of the peak to peak voltage swing. For a sinewave this will
be the zero crossing, and for a square wave this will be the 50%
point on the edges. You can get better resolution on determining the
edge timing by increasing the vertical gain (reducing the volts/div)
setting on the oscilloscope. But you probably only want to increase
the gain so the signal is off the screen by a factor of 2 to 5,
because too much gain may result in overdrive recovery problems in
the vertical amplifier. The trigger signal (on a display channel or
external trigger input) gain should also be increased to get lower
jitter triggering.
(7) The Tek 465 input impedance (of CH1, CH2, and the external trigger
input) is 1 M ohm in parallel with about 20 pF. If you are using 50
ohm cables, it's best to use 50 ohm feedthrough terminators on the
two connectors to which the oscillators are connected. With low
frequency (no higher than around 10 MHz) sinewave sources a lack of
proper termination doesn't cause many problems, but if a signal has
fast edges (small values of risetime/falltime) an improper or
missing termination can result in reflections. This can cause
distortions in the waveform near the rising and falling edges which
add jitter and cause unstable triggering of the scope. So it's good
engineering practice to properly terminate the cables at the
oscilloscope BNC connectors.--
Bill Byrom N5BB
Tektronix Application Engineer for past 31 years.
First used the Tek 465 about 42 year ago.

_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
and follow the instructions there.
Dana Whitlow
2018-09-28 15:30:26 UTC
Permalink
Hi,

There is one other issue that can bite you if you fail to properly
terminate the output of a source:

Depending on the source's design, an essentially unloaded output
can have a substantially higher voltage swing than expected (by
2X if the source impedance is actually 50 ohms), possibly leading
to the output stage's going into clipping, which can in turn distort
the timing, possibly even in an unstable manner.
So if you want to play the "unterminated game", at least take a
look at the waveform to be sure it's still a clean sinewave. I've
noticed such distortion on my PRS-10, for example, although I've
seen no evidence of unstable timing results. But in this arena,
it generally pays to be fussy.

Dana Whitlow
Triggering a dual beam 'scope (Tek 465) from the TB on Ch1, and having>
the output of the OCXO on Ch2, the resulting display on Ch2 of course>
drifts in relation to the static waveform on Ch1. (Both nice
sinusoids.)
The Tek 465 analog cathode ray oscilloscope was/is a very flexible
instrument. But this flexibility allows you to set up the instrument in
ways which will not allow this commonly used oscillator comparison
technique to work correctly. Since you are interested in these
instruments, here are some details about setting up the instrument for
such comparisons.
(1) The Tek 465 is not a dual beam oscilloscope. Dual beam oscilloscopes
(such as the Tektronix 556 and 7844) use a special CRT which
incorporates two independent electron guns. Each electron gun
assembly has a set of vertical and horizontal deflection plates.
There are two vertical amplifiers (one for each electron gun) and
two horizontal sweep systems (one for each electron gun). If you had
a dual beam oscilloscope you could compare oscillator#1 to
oscillator#2 while simultaneously comparing oscillator#3 with
oscillator#4. It's like having two independent oscilloscopes sharing
the same CRT display.
(2) The Tek 465 single beam oscilloscope can display two traces on the
display using one of two methods:(a) Chopped trace display: This mode
works well at low sweep rates (such
as 1 ms/div) but causes trouble at fast sweep rates (such as 1
us/div). The displayed trace is switched between Channel 1 and
Channel 2 at a fixed rate of about 500 kHz.(b) Alternate trace
display: This mode works well at high sweep rates
but is hard to see at low sweep rates. The scope alternates between
displaying one sweep of Channel 1 and one sweep of Channel 2.
(3) The trigger source setting is crucial to using this technique to
compare oscillators. The technique does not require you to display
two channels. What is important is that you display one oscillator
while triggering on the other oscillator. The trigger source can
be set to:(a) CH 1: The Channel 2 display will drift if the two signals have a
varying phase relationship.(b) CH 2: The Channel 1 display will drift
if the two signals have a
varying phase relationship.(c) NORM (normal): The trigger system gets
input from the channel being
displayed at that moment. So in chopped trace display mode the
trigger is rapidly switched between CH1 and CH2, and in alternate
trace display mode the trigger alternates between CH1 and CH2 on
alternate sweeps. In all cases, you should not use NORM trigger
source with both channels displayed when comparing oscillators!(d)
EXT: You apply the trigger signal to the external trigger input
connector. This works well well when comparing oscillators. If you
use alternate trace display mode and an external trigger, you can
compare oscillator#1 (on CH 1) to oscillator#0 (on the external
trigger input) while you are also comparing oscillator#2 (on CH2)
oscillator#0. So you could compare two oscillators (one on CH1 and
the other on CH2) to a GPSDO (on the external trigger input).
(4) When comparing oscillators, the fractional frequency difference
(such as ppm Parts Per Million or ppb Parts Per Billion) you can
measure depends on the oscilloscope sweep rate. What you are really
measuring is the drift of the time delay between the edge (or zero
crossing of a sine wave) of one signal relative to an edge or zero
Fractional difference = (observed timing change) / (measurement
Fractional difference in ppm = (time delay drift in us) per second of
observation timeFractional difference in ppb = (time delay drift in ns) per second of
observation time
(5) As you can see in my previous section, you need a very fast sweep
rate (small time/div) to measure small fractional frequency
differences. This means that for a small fractional frequency
difference with a moderately low measured oscillator frequency (such
as 1 MHz), you may not see any edges for a long time when you use a
small time/div. The Tek 465 has a delayed timebase, and you can use
this feature to move the signal edge (or zero crossing) onto the
screen. You can then watch the signal for a few seconds to determine
the timing drift rate. If the edge is drifting at 10 ns per 10
seconds, the fractional difference is 1 ppb (1 part in 10^9). If the
displayed oscillator edge is drifting to the left (earlier in time),
the displayed oscillator frequency is higher than the reference
oscillator you are using for the trigger. If the displayed
oscillator edge is drifting to the right (later in time), the
displayed oscillator frequency is lower than the reference
oscillator you are using for the trigger.
(6) If the edge rate is not very fast (such as when you are measuring
sinewave signals), the waveform edge you see at a fast sweep rate
will appear to be nearly horizontal (spread out across many
divisions). You normally want to measure the displayed signal at the
midpoint of the peak to peak voltage swing. For a sinewave this will
be the zero crossing, and for a square wave this will be the 50%
point on the edges. You can get better resolution on determining the
edge timing by increasing the vertical gain (reducing the volts/div)
setting on the oscilloscope. But you probably only want to increase
the gain so the signal is off the screen by a factor of 2 to 5,
because too much gain may result in overdrive recovery problems in
the vertical amplifier. The trigger signal (on a display channel or
external trigger input) gain should also be increased to get lower
jitter triggering.
(7) The Tek 465 input impedance (of CH1, CH2, and the external trigger
input) is 1 M ohm in parallel with about 20 pF. If you are using 50
ohm cables, it's best to use 50 ohm feedthrough terminators on the
two connectors to which the oscillators are connected. With low
frequency (no higher than around 10 MHz) sinewave sources a lack of
proper termination doesn't cause many problems, but if a signal has
fast edges (small values of risetime/falltime) an improper or
missing termination can result in reflections. This can cause
distortions in the waveform near the rising and falling edges which
add jitter and cause unstable triggering of the scope. So it's good
engineering practice to properly terminate the cables at the
oscilloscope BNC connectors.--
Bill Byrom N5BB
Tektronix Application Engineer for past 31 years.
First used the Tek 465 about 42 year ago.
_______________________________________________
To unsubscribe, go to
http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
and follow the instructions there.
_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
and follow the instructions there.
Bill Byrom
2018-09-28 23:22:55 UTC
Permalink
In addition to nonlinear issues with output amplifiers, filters have
poor performance when improperly terminated. This can lead to harmonic
distortion and that can be a problem. You want the duty cycle to be
exactly 50%. See:https://tf.boulder.nist.gov/general/pdf/1437.pdf
--
Bill Byrom N5BB
Post by Dana Whitlow
Hi,
There is one other issue that can bite you if you fail to properly
Depending on the source's design, an essentially unloaded output
can have a substantially higher voltage swing than expected (by
2X if the source impedance is actually 50 ohms), possibly leading
to the output stage's going into clipping, which can in turn distort
the timing, possibly even in an unstable manner.
So if you want to play the "unterminated game", at least take a
look at the waveform to be sure it's still a clean sinewave. I've
noticed such distortion on my PRS-10, for example, although I've
seen no evidence of unstable timing results. But in this arena,
it generally pays to be fussy.
Dana Whitlow
On Fri, Sep 28, 2018 at 7:06 AM Bill Byrom
Post by Dave B via time-nuts
Triggering a dual beam 'scope (Tek 465) from the TB on Ch1, and
having>>> the output of the OCXO on Ch2, the resulting display on Ch2 of
course>>> drifts in relation to the static waveform on Ch1. (Both nice
Post by Dave B via time-nuts
sinusoids.)
The Tek 465 analog cathode ray oscilloscope was/is a very flexible
instrument. But this flexibility allows you to set up the
instrument in>> ways which will not allow this commonly used oscillator comparison
technique to work correctly. Since you are interested in these
instruments, here are some details about setting up the
instrument for>> such comparisons.
(1) The Tek 465 is not a dual beam oscilloscope. Dual beam
oscilloscopes>> (such as the Tektronix 556 and 7844) use a special CRT which
incorporates two independent electron guns. Each electron gun
assembly has a set of vertical and horizontal deflection plates.
There are two vertical amplifiers (one for each electron gun) and
two horizontal sweep systems (one for each electron gun). If
you had>> a dual beam oscilloscope you could compare oscillator#1 to
oscillator#2 while simultaneously comparing oscillator#3 with
oscillator#4. It's like having two independent oscilloscopes
sharing>> the same CRT display.
(2) The Tek 465 single beam oscilloscope can display two traces on
This mode>> works well at low sweep rates (such
as 1 ms/div) but causes trouble at fast sweep rates (such as 1
us/div). The displayed trace is switched between Channel 1 and
Channel 2 at a fixed rate of about 500 kHz.(b) Alternate trace
display: This mode works well at high sweep rates
but is hard to see at low sweep rates. The scope alternates between>> displaying one sweep of Channel 1 and one sweep of Channel 2.
(3) The trigger source setting is crucial to using this technique to>> compare oscillators. The technique does not require you to display>> two channels. What is important is that you display one oscillator>> while triggering on the other oscillator. The trigger source can
be set to:(a) CH 1: The Channel 2 display will drift if the two signals have a
varying phase relationship.(b) CH 2: The Channel 1 display will
drift>> if the two signals have a
varying phase relationship.(c) NORM (normal): The trigger
system gets>> input from the channel being
displayed at that moment. So in chopped trace display mode the
trigger is rapidly switched between CH1 and CH2, and in alternate
trace display mode the trigger alternates between CH1 and CH2 on
alternate sweeps. In all cases, you should not use NORM trigger
source with both channels displayed when comparing oscillators!(d)>> EXT: You apply the trigger signal to the external trigger input
connector. This works well well when comparing oscillators. If you>> use alternate trace display mode and an external trigger, you can
compare oscillator#1 (on CH 1) to oscillator#0 (on the external
trigger input) while you are also comparing oscillator#2 (on CH2)
oscillator#0. So you could compare two oscillators (one on CH1 and>> the other on CH2) to a GPSDO (on the external trigger input).
(4) When comparing oscillators, the fractional frequency difference
(such as ppm Parts Per Million or ppb Parts Per Billion) you can
Fractional difference = (observed timing change) / (measurement
Fractional difference in ppm = (time delay drift in us) per second of>> observation timeFractional difference in ppb = (time delay
drift in ns)>> per second of
observation time
(5) As you can see in my previous section, you need a very fast sweep>> rate (small time/div) to measure small fractional frequency
differences. This means that for a small fractional frequency
difference with a moderately low measured oscillator frequency
(such>> as 1 MHz), you may not see any edges for a long time when you use a>> small time/div. The Tek 465 has a delayed timebase, and you can use>> this feature to move the signal edge (or zero crossing) onto the
screen. You can then watch the signal for a few seconds to
determine>> the timing drift rate. If the edge is drifting at 10 ns per 10
seconds, the fractional difference is 1 ppb (1 part in 10^9).
If the>> displayed oscillator edge is drifting to the left (earlier in
time),>> the displayed oscillator frequency is higher than the reference
oscillator you are using for the trigger. If the displayed
oscillator edge is drifting to the right (later in time), the
displayed oscillator frequency is lower than the reference
oscillator you are using for the trigger.
(6) If the edge rate is not very fast (such as when you are measuring>> sinewave signals), the waveform edge you see at a fast sweep rate
will appear to be nearly horizontal (spread out across many
divisions). You normally want to measure the displayed signal
at the>> midpoint of the peak to peak voltage swing. For a sinewave
this will>> be the zero crossing, and for a square wave this will be the 50%
point on the edges. You can get better resolution on
determining the>> edge timing by increasing the vertical gain (reducing the
volts/div)>> setting on the oscilloscope. But you probably only want to increase>> the gain so the signal is off the screen by a factor of 2 to 5,
because too much gain may result in overdrive recovery problems in>> the vertical amplifier. The trigger signal (on a display channel or>> external trigger input) gain should also be increased to get lower>> jitter triggering.
(7) The Tek 465 input impedance (of CH1, CH2, and the external
trigger>> input) is 1 M ohm in parallel with about 20 pF. If you are using 50>> ohm cables, it's best to use 50 ohm feedthrough terminators on the>> two connectors to which the oscillators are connected. With low
frequency (no higher than around 10 MHz) sinewave sources a lack of>> proper termination doesn't cause many problems, but if a signal has>> fast edges (small values of risetime/falltime) an improper or
missing termination can result in reflections. This can cause
distortions in the waveform near the rising and falling edges which>> add jitter and cause unstable triggering of the scope. So it's good>> engineering practice to properly terminate the cables at the
oscilloscope BNC connectors.--
Bill Byrom N5BB
Tektronix Application Engineer for past 31 years.
First used the Tek 465 about 42 year ago.
_________________________________________________
To unsubscribe, go to
http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
and follow the instructions there.
_________________________________________________
To unsubscribe, go to
http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com> and follow the instructions there.
_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
and follow the instructions there.

Loading...