Discussion:
Spectracom 8140T Line Tap Schematic
(too old to reply)
gandalfg8--- via time-nuts
2018-06-12 22:05:09 UTC
Permalink
This is probably now well past its sell by date but over the years there have been a number of requests here for schematics of the Spectracom 8140T Line Taps as used with the 8140 Frequency Distribution Amp.


The standard reply has always been that no schematics were available from Spectracom, which is true, and that tracing one out wasn't a viable option either as the 8140T modules were potted in some kind of foam, but a few years ago I made an interesting discovery about the 8140T that enabled me to quickly produce a sketched schematic followed by the intention of posting it here just as soon as I'd turned it into something a bit more legible.


Unfortunately those good intentions got left behind as other projects came along and it was only the recent posts regarding removal of the DC feed from the outputs of the 8140 that reminded me this was still a work in progress, or more to the point, remarkably little progress:-(


BTW, with reference to removing the DC supply on the 8140 outputs, although it is possible to effect this by removing internal components by far the most straightforward arrangement, and most easily reversible, would be to use an inline DC blocking capacitor of between say 100 and 470 nF, just as is already very commonly used with remotely powered HF active receiving antennas.



Anyways back to the 8140T schematic, I have now posted on mediafire the schematic of an 8140T10, the basic 10MHz in 10MHz out version, complete with far too many assorted internal photos plus details of said interesting discovery, with extra brownie points to anyone who might already be thinking foam and cardboard sandwiches:-)



http://www.mediafire.com/file/rx9np48l8nwbe3g/8140TLineTap.zip



Nigel, GM8PZR







_______________________________________________
time-nuts mailing list -- time-***@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Charles Steinmetz
2018-06-14 03:28:48 UTC
Permalink
Post by gandalfg8--- via time-nuts
with far too many assorted internal photos plus details of said interesting discovery, with extra brownie points to anyone who might already be thinking foam and cardboard sandwiches:-)
Oh, my, that brings back memories! That method of "encapsulation" was
used by a number of Rochester electronics firms of the era. It was
developed by a good friend of mine with whom I worked at another firm
about a decade before Spectracom was founded, and was introduced at
Spectracom when he moved there.

It's hilarious now, looking at the crude PC layouts, the PC cards that
look like they were separated with hatchets (in reality, they used
heavy-duty, office-type guillotine paper cutters), and the "definitely
not near mil-spec" hand soldering job using at least 5x the optimum
amount of solder. Spectracom didn't even have a dip-soldering
operation, much less a wave-soldering system.

It was a much, much simpler time, one rung up the ladder from hippies in
a basement....

Charles


_______________________________________________
time-nuts mailing list -- time-***@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Bob kb8tq
2018-06-14 12:59:48 UTC
Permalink
Hi

I think you would find a *lot* of smaller outfits “back in the day” running hand soldering lines. Small scale wave solder for through hole did not
catch on the way reflow has for SMT. Dip solder was a rare item ….

Bob
Post by gandalfg8--- via time-nuts
with far too many assorted internal photos plus details of said interesting discovery, with extra brownie points to anyone who might already be thinking foam and cardboard sandwiches:-)
Oh, my, that brings back memories! That method of "encapsulation" was used by a number of Rochester electronics firms of the era. It was developed by a good friend of mine with whom I worked at another firm about a decade before Spectracom was founded, and was introduced at Spectracom when he moved there.
It's hilarious now, looking at the crude PC layouts, the PC cards that look like they were separated with hatchets (in reality, they used heavy-duty, office-type guillotine paper cutters), and the "definitely not near mil-spec" hand soldering job using at least 5x the optimum amount of solder. Spectracom didn't even have a dip-soldering operation, much less a wave-soldering system.
It was a much, much simpler time, one rung up the ladder from hippies in a basement....
Charles
_______________________________________________
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
_______________________________________________
time-nuts mailing list -- time-***@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
gandalfg8--- via time-nuts
2018-06-14 13:31:32 UTC
Permalink
Hi Julien,



Earlier replies that discussed removing the DC supply from the 8140 outputs gave me the impression your requirement was to connect directly the outputs of the 8140 without line tap modules, which is what I was addressing with my recent suggestion of a series capacitor, but can see now that's not the case.



If you were to build your own termination the same principle would still apply of course but perhaps another option for a reasonably compact solution, aside from a coaxial DC block followed by a 50 ohm termination, might be a DC blocking attenuator followed by a BNC shorting dust cap, as the latter are generally quite small.


For example, current Ebay item 332461304064 is a DC blocked 20dB attenuator, they are also available in pairs at a slight discount, and that fitted with a shorting cap would give you a return loss of 40dB, which is probably as good as you'd get from most 50 ohm terminations anyway and would save having to build your own.


The VersaTaps phase lock an internal crystal, any frequency between 4 and 20MHz, to the 10MHz input and then provide an output at the crystal frequency or the crystal frequency divided by a fixed integer between 1 and 8,192, so in that sense yes they're a synthesiser but the few I've seen were still supplied preset for a specific frequency.


However, "versatile" does not necessarilly mean quick or convenient to reprogram.
Aside from perhaps needing to change the crystal the division ratio needs to be set by adding or removing links across pairs of holes in the circuit board. Those holes are on a standard 0.1inch pitch so fitting headers that would take shorting links should be straightfoward enough, and I suspect what was originally intended, but the units I have were supplied preprogrammed using resistor style wire ended zero ohm links mounted above the circuit board and soldered from below. Reprogramming one of these would involve unsoldering and perhaps removing all the connectors etc in order to remove the circuit board, so I'd certainly make darn sure they had headers fitted before being returned to the box!



Nigel, GM8PZR






-----Original Message-----
From: Julien Goodwin <time-***@studio442.com.au>
To: gandalfg8 <***@aol.com>; Discussion of precise time and frequency measurement <time-***@febo.com>
Sent: Thu, 14 Jun 2018 11:36
Subject: Re: [time-nuts] Spectracom 8140T Line Tap Schematic

I'm somewhat tempted to take the schematic and see if I can fit it in
the comparatively tiny Pomona boxes, although I have far too many side
projects already, enough 10MHz taps, and I still need to do the terminators.

The VersaTaps are supposed to be an actual synthesizer IIRC.
















_______________________________________________
time-nuts mailing list -- time-***@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Bob kb8tq
2018-06-14 15:43:30 UTC
Permalink
Hi
Post by gandalfg8--- via time-nuts
Hi Julien,
Earlier replies that discussed removing the DC supply from the 8140 outputs gave me the impression your requirement was to connect directly the outputs of the 8140 without line tap modules, which is what I was addressing with my recent suggestion of a series capacitor, but can see now that's not the case.
If you were to build your own termination the same principle would still apply of course but perhaps another option for a reasonably compact solution, aside from a coaxial DC block followed by a 50 ohm termination, might be a DC blocking attenuator followed by a BNC shorting dust cap, as the latter are generally quite small.
For example, current Ebay item 332461304064 is a DC blocked 20dB attenuator, they are also available in pairs at a slight discount, and that fitted with a shorting cap would give you a return loss of 40dB, which is probably as good as you'd get from most 50 ohm terminations anyway and would save having to build your own.
I would be very surprised if the original terminations did any better than 20 to 30 db return loss. They are not fancy devices.
I’ve also seen a lot of attenuators that only make it to about 20 db of return loss …..

Bob
Post by gandalfg8--- via time-nuts
The VersaTaps phase lock an internal crystal, any frequency between 4 and 20MHz, to the 10MHz input and then provide an output at the crystal frequency or the crystal frequency divided by a fixed integer between 1 and 8,192, so in that sense yes they're a synthesiser but the few I've seen were still supplied preset for a specific frequency.
However, "versatile" does not necessarilly mean quick or convenient to reprogram.
Aside from perhaps needing to change the crystal the division ratio needs to be set by adding or removing links across pairs of holes in the circuit board. Those holes are on a standard 0.1inch pitch so fitting headers that would take shorting links should be straightfoward enough, and I suspect what was originally intended, but the units I have were supplied preprogrammed using resistor style wire ended zero ohm links mounted above the circuit board and soldered from below. Reprogramming one of these would involve unsoldering and perhaps removing all the connectors etc in order to remove the circuit board, so I'd certainly make darn sure they had headers fitted before being returned to the box!
Nigel, GM8PZR
-----Original Message-----
Sent: Thu, 14 Jun 2018 11:36
Subject: Re: [time-nuts] Spectracom 8140T Line Tap Schematic
I'm somewhat tempted to take the schematic and see if I can fit it in
the comparatively tiny Pomona boxes, although I have far too many side
projects already, enough 10MHz taps, and I still need to do the terminators.
The VersaTaps are supposed to be an actual synthesizer IIRC.
_______________________________________________
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
_______________________________________________
time-nuts mailing list -- time-***@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Dana Whitlow
2018-06-14 15:47:31 UTC
Permalink
Hi Bob & Julien,

Why bother with the short on the downstream end of that DC-blocking
attenuator? At 10 MHz I'd expect
leakage radiation to be negligible.

Dana
Post by gandalfg8--- via time-nuts
Hi
On Jun 14, 2018, at 9:31 AM, gandalfg8--- via time-nuts <
Hi Julien,
Earlier replies that discussed removing the DC supply from the 8140
outputs gave me the impression your requirement was to connect directly the
outputs of the 8140 without line tap modules, which is what I was
addressing with my recent suggestion of a series capacitor, but can see now
that's not the case.
If you were to build your own termination the same principle would still
apply of course but perhaps another option for a reasonably compact
solution, aside from a coaxial DC block followed by a 50 ohm termination,
might be a DC blocking attenuator followed by a BNC shorting dust cap, as
the latter are generally quite small.
For example, current Ebay item 332461304064 is a DC blocked 20dB
attenuator, they are also available in pairs at a slight discount, and that
fitted with a shorting cap would give you a return loss of 40dB, which is
probably as good as you'd get from most 50 ohm terminations anyway and
would save having to build your own.
I would be very surprised if the original terminations did any better than
20 to 30 db return loss. They are not fancy devices.
I’ve also seen a lot of attenuators that only make it to about 20 db of return loss …..
Bob
The VersaTaps phase lock an internal crystal, any frequency between 4
and 20MHz, to the 10MHz input and then provide an output at the crystal
frequency or the crystal frequency divided by a fixed integer between 1 and
8,192, so in that sense yes they're a synthesiser but the few I've seen
were still supplied preset for a specific frequency.
However, "versatile" does not necessarilly mean quick or convenient to
reprogram.
Aside from perhaps needing to change the crystal the division ratio
needs to be set by adding or removing links across pairs of holes in the
circuit board. Those holes are on a standard 0.1inch pitch so fitting
headers that would take shorting links should be straightfoward enough, and
I suspect what was originally intended, but the units I have were supplied
preprogrammed using resistor style wire ended zero ohm links mounted above
the circuit board and soldered from below. Reprogramming one of these would
involve unsoldering and perhaps removing all the connectors etc in order to
remove the circuit board, so I'd certainly make darn sure they had headers
fitted before being returned to the box!
Nigel, GM8PZR
-----Original Message-----
Sent: Thu, 14 Jun 2018 11:36
Subject: Re: [time-nuts] Spectracom 8140T Line Tap Schematic
I'm somewhat tempted to take the schematic and see if I can fit it in
the comparatively tiny Pomona boxes, although I have far too many side
projects already, enough 10MHz taps, and I still need to do the
terminators.
The VersaTaps are supposed to be an actual synthesizer IIRC.
_______________________________________________
To unsubscribe, go to https://www.febo.com/cgi-bin/
mailman/listinfo/time-nuts
and follow the instructions there.
_______________________________________________
To unsubscribe, go to https://www.febo.com/cgi-bin/
mailman/listinfo/time-nuts
and follow the instructions there.
_______________________________________________
time-nuts mailing list -- time-***@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Charles Steinmetz
2018-06-15 01:41:02 UTC
Permalink
Post by gandalfg8--- via time-nuts
I'm somewhat tempted to take the schematic and see if I can fit it in
the comparatively tiny Pomona boxes
Before getting too excited about the various 8140 module designs, it may
be worth reflecting that the performance of the 8140 system is mediocre
at best from a time-nuts perspective. There are lots and lots of much
better designs if you're building your own. Even something as simple as
a modified video distribution amplifier can be orders of magnitude
better (see, e.g.,
<http://www.ko4bb.com/getsimple/index.php?id=download&file=02_GPS_Timing/Extron_ADA_6_modifications_for_use_as_10MHz_distribution_amp_STEINMETZ.pdf>).

The point of the 8140 system is that it is a moderate performance, "do
everything" solution for large installations (hundreds of potentially
multifrequency taps spaced around acres of facility), not a
time-nuts-quality distribution system for a facility the size of a
residential home.

Best regards,

Charles


_______________________________________________
time-nuts mailing list -- time-***@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Bob kb8tq
2018-06-15 12:59:53 UTC
Permalink
Hi

The typical 8140 tap has an ADEV of around 2x10-11 at a tau of 1 second and is in the 3x10^-12 range at 100 seconds. That’s not super duper,
but it’s plenty good enough for the intended use of feeding racks full of test gear. If you are feeding it with a typical Cs or telecom Rb, it’s about
as good as the source you are feeding it with.

On the plus side, if you have an inventory of gear needs 100 KHz as a reference, just grab that tap. This other device needs 1 MHz, grab the tap.
That one needs 5 MHz …. It’s tough to do that with other systems.

Bob
Post by gandalfg8--- via time-nuts
I'm somewhat tempted to take the schematic and see if I can fit it in
the comparatively tiny Pomona boxes
Before getting too excited about the various 8140 module designs, it may be worth reflecting that the performance of the 8140 system is mediocre at best from a time-nuts perspective. There are lots and lots of much better designs if you're building your own. Even something as simple as a modified video distribution amplifier can be orders of magnitude better (see, e.g., <http://www.ko4bb.com/getsimple/index.php?id=download&file=02_GPS_Timing/Extron_ADA_6_modifications_for_use_as_10MHz_distribution_amp_STEINMETZ.pdf>).
The point of the 8140 system is that it is a moderate performance, "do everything" solution for large installations (hundreds of potentially multifrequency taps spaced around acres of facility), not a time-nuts-quality distribution system for a facility the size of a residential home.
Best regards,
Charles
_______________________________________________
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to https://lists.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Loading...