Time Mark generator

Craig Sawyers

Hi list

One of the problems with living in the UK is that we are in the wilderness
for Tek gear.  You only have to look at eBay UK and US and compare the two
to get the drift!

Faced with not being able to get a hold of a time mark generator on my side
of the pond (and since shipping costs for bulky gear from the US are
prohibitive), I tried to think of a way of generating accurate time marks so
I could calibrate my 545A and 7603/7B53A.

Now I have a counter/timer (HP 5328A - cost me all of ???22UK), whose
calibration accuracy is unknown, and I also have a signal generator whose
calibration is unknown.  However, if the 5328A's calibration was on the
button, I would have a traceable method of generating time marks up to at
least 100MHz (<10ns spacing).

So what I have done is build a simple tuned RF receiver, made from a high-Q
honeycomb coil (built according to
).  This measured 350uH on my
bridge.  I then resonated it at 60kHz with two 10nF capacitors in parallel
and connected it in differential mode to my 7A22.

Why 60kHz?  Well there is a time code transmission in the UK on 60kHz,
traceable to the NPL (equivalent to the NBS in the US) to an accuracy of 1
part in 10^11 (or 10 cycles at 60kHz per year).  Much to my delight, the
signal from such a rudimentary arrangement was 500uV, just using the coil as
the antenna (and with the hf and lf filters on the 7A22 set correctly for
the frequency to reduce noise).

An alternative, not yet tried, is to use the Radio 4 long-wave transmission
at 180kHz (UK of course).  This is Caesium clock referenced at 180kHz, and
is again set to an accuracy of better than 1 part in 10^11.  Quite something
for a broadcast radio station!

I now need an easy op-amp based high-Q filter tuned to 60kHz (over the
holiday break) to reduce noise and boost the signal by a factor of 100 or
so.  Then I have an absolute method of checking the calibration of my timer.
Now by tuning my signal generator to a particular frequency as registered by
the now-calibrated timer, I can generate time marks of any period for
calibrating my 'scopes!

I'll let you know how I get on....

Craig

Don Black

#1138

Hello Craig,
                    If you can measure the colour oscillator frequency of a
colour TV while receiving a transmission it will be 4433618.75 Hz., use that to
check the frequency meter. The standard used to be +/- 5 Hz. but if you check
with the BBC I think they use an atomic standatd now and is extremely accurate
(even 5 Hz. is only 1 part per million).
In the US the frequency is 3.579545 MHz.
A very Merry Christmas and Happy New Year to all. I hope there's (another)
Tektronix in your future.
                                            Cheers, Don Black.

Craig Sawyers wrote:

Show quoted text

Hi list

One of the problems with living in the UK is that we are in the wilderness
for Tek gear.  You only have to look at eBay UK and US and compare the two
to get the drift!

Faced with not being able to get a hold of a time mark generator on my side
of the pond (and since shipping costs for bulky gear from the US are
prohibitive), I tried to think of a way of generating accurate time marks so
I could calibrate my 545A and 7603/7B53A.

Now I have a counter/timer (HP 5328A - cost me all of ???22UK), whose
calibration accuracy is unknown, and I also have a signal generator whose
calibration is unknown.  However, if the 5328A's calibration was on the
button, I would have a traceable method of generating time marks up to at
least 100MHz (<10ns spacing).

So what I have done is build a simple tuned RF receiver, made from a high-Q
honeycomb coil (built according to
).  This measured 350uH on my
bridge.  I then resonated it at 60kHz with two 10nF capacitors in parallel
and connected it in differential mode to my 7A22.

Why 60kHz?  Well there is a time code transmission in the UK on 60kHz,
traceable to the NPL (equivalent to the NBS in the US) to an accuracy of 1
part in 10^11 (or 10 cycles at 60kHz per year).  Much to my delight, the
signal from such a rudimentary arrangement was 500uV, just using the coil as
the antenna (and with the hf and lf filters on the 7A22 set correctly for
the frequency to reduce noise).

An alternative, not yet tried, is to use the Radio 4 long-wave transmission
at 180kHz (UK of course).  This is Caesium clock referenced at 180kHz, and
is again set to an accuracy of better than 1 part in 10^11.  Quite something
for a broadcast radio station!

I now need an easy op-amp based high-Q filter tuned to 60kHz (over the
holiday break) to reduce noise and boost the signal by a factor of 100 or
so.  Then I have an absolute method of checking the calibration of my timer.
Now by tuning my signal generator to a particular frequency as registered by
the now-calibrated timer, I can generate time marks of any period for
calibrating my 'scopes!

I'll let you know how I get on....

Craig


To unsubscribe from this group, send an email to:
TekScopes-unsubscribe@...



Your use of Yahoo! Groups is subject to 

Michael Dunn

#1140

Why such concern about accuracy?  Even the crappiest crystal osc or frequency counter will be 100 times more accurate than the best possible with a scope!

Craig Sawyers

#1141

    Why such concern about accuracy?  Even the crappiest crystal osc
or frequency counter will be 100 times more accurate than the best
possible with a scope!Yeah, I know - you can only pull a crystal oscillator a tiny amount using
the parallel tuning capacitor.  I just have this thing about getting things
adjusted accurately, and at least this gets me a bang-on accurate counter to
boot *very* easily.

Craig

开云体育

Time Mark generator