Discussion:
10 MHz 'failover' switch?
(too old to reply)
W7SLS
2018-07-26 19:45:47 UTC
Permalink
Hello,

Looking for recommendation for a ‘failover’ or ‘redundant’ switch for 10 MHz distribution.

Not really sure of the correct term.
Something that sensed RF on primary 10 MHz, and then switched to secondary on fail of primary.

A brief search showed several very nice $$$ items, suitable for commercial applications,
but I wonder if there are some “last year’s” (but not last century) versions that would work for a home lab.

Context:

I have a GPSDO and a Rb source of 10 MHz.
The power supply on the GPSDO failed (worked enough to light up the GPSDO, but not enough to lock).
I have a new power supply on order, but would be nice to have “insurance”.

Thanks in advance for the group bandwidth.

Scott
W7SLS
_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_list
Bob kb8tq
2018-07-26 20:13:36 UTC
Permalink
Hi

They are a pretty rare item. A more common approach is a disciplined oscillator that
will do failover on it’s inputs. That’s still a rare item, but at least a possible thing to find.
The equally big problem will be getting doc’s on one if you do find it….

Bob
Post by W7SLS
Hello,
Looking for recommendation for a ‘failover’ or ‘redundant’ switch for 10 MHz distribution.
Not really sure of the correct term.
Something that sensed RF on primary 10 MHz, and then switched to secondary on fail of primary.
A brief search showed several very nice $$$ items, suitable for commercial applications,
but I wonder if there are some “last year’s” (but not last century) versions that would work for a home lab.
I have a GPSDO and a Rb source of 10 MHz.
The power supply on the GPSDO failed (worked enough to light up the GPSDO, but not enough to lock).
I have a new power supply on order, but would be nice to have “insurance”.
Thanks in advance for the group bandwidth.
Scott
W7SLS
_______________________________________________
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
Scott McGrath
2018-07-26 21:33:32 UTC
Permalink
Build one yourself, Detector diode on primary RF input when output drops use a rf relay or PIN diode switch to fail over to backup standard. All thats needed is a crossing detector and relay / switch driver

Yes there would be a momentary hit but it would work.

Content by Scott
Typos by Siri

On Jul 26, 2018, at 4:13 PM, Bob kb8tq <***@n1k.org> wrote:

Hi

They are a pretty rare item. A more common approach is a disciplined oscillator that
will do failover on it’s inputs. That’s still a rare item, but at least a possible thing to find.
The equally big problem will be getting doc’s on one if you do find it….

Bob
Post by W7SLS
Hello,
Looking for recommendation for a ‘failover’ or ‘redundant’ switch for 10 MHz distribution.
Not really sure of the correct term.
Something that sensed RF on primary 10 MHz, and then switched to secondary on fail of primary.
A brief search showed several very nice $$$ items, suitable for commercial applications,
but I wonder if there are some “last year’s” (but not last century) versions that would work for a home lab.
I have a GPSDO and a Rb source of 10 MHz.
The power supply on the GPSDO failed (worked enough to light up the GPSDO, but not enough to lock).
I have a new power supply on order, but would be nice to have “insurance”.
Thanks in advance for the group bandwidth.
Scott
W7SLS
_______________________________________________
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.

_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-n
Bob kb8tq
2018-07-26 21:45:19 UTC
Permalink
Hi

One interesting subtlety making something like this:

What if the two inputs aren’t quite on the same frequency? Purely as an example, say they are 1 Hz off from each other.
If you have 60 db of isolation in your “switch” you get a 1 Hz offset spur that is 60 db down. Even something much further
down is plenty to mess up the ADEV of the output.

Bob
Post by Scott McGrath
Build one yourself, Detector diode on primary RF input when output drops use a rf relay or PIN diode switch to fail over to backup standard. All thats needed is a crossing detector and relay / switch driver
Yes there would be a momentary hit but it would work.
Content by Scott
Typos by Siri
Hi
They are a pretty rare item. A more common approach is a disciplined oscillator that
will do failover on it’s inputs. That’s still a rare item, but at least a possible thing to find.
The equally big problem will be getting doc’s on one if you do find it….
Bob
Post by W7SLS
Hello,
Looking for recommendation for a ‘failover’ or ‘redundant’ switch for 10 MHz distribution.
Not really sure of the correct term.
Something that sensed RF on primary 10 MHz, and then switched to secondary on fail of primary.
A brief search showed several very nice $$$ items, suitable for commercial applications,
but I wonder if there are some “last year’s” (but not last century) versions that would work for a home lab.
I have a GPSDO and a Rb source of 10 MHz.
The power supply on the GPSDO failed (worked enough to light up the GPSDO, but not enough to lock).
I have a new power supply on order, but would be nice to have “insurance”.
Thanks in advance for the group bandwidth.
Scott
W7SLS
_______________________________________________
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.
_______________________________________________
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.
W7SLS
2018-07-26 20:59:17 UTC
Permalink
Hi Bob,

Thanks for the quick reply, makes sense.

73
Scott
W7SLS
Post by Bob kb8tq
Hi
They are a pretty rare item. A more common approach is a disciplined oscillator that
will do failover on it’s inputs. That’s still a rare item, but at least a possible thing to find.
The equally big problem will be getting doc’s on one if you do find it….
Bob
Post by W7SLS
Hello,
Looking for recommendation for a ‘failover’ or ‘redundant’ switch for 10 MHz distribution.
Not really sure of the correct term.
Something that sensed RF on primary 10 MHz, and then switched to secondary on fail of primary.
A brief search showed several very nice $$$ items, suitable for commercial applications,
but I wonder if there are some “last year’s” (but not last century) versions that would work for a home lab.
I have a GPSDO and a Rb source of 10 MHz.
The power supply on the GPSDO failed (worked enough to light up the GPSDO, but not enough to lock).
I have a new power supply on order, but would be nice to have “insurance”.
Thanks in advance for the group bandwidth.
Scott
W7SLS
_______________________________________________
time-nuts mailing list -- time-***@lists.febo.com
To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.co
Scott McGrath
2018-07-27 17:38:37 UTC
Permalink
Since this is time nuts ADEV is sacred,

but the question is the use case.

Are we looking at a time nuts level frequency distribution network. Or are we trying to lock benchtop instruments with poor internal standards to a common ‘Good’ reference.

If the former it can be done but will require a fair amount of design work as well as some EMC
Modeling of the system to ensure no feed through of unwanted signals and/or modulation or beat products

If the latter simple detection and switching will be sufficient but dual switch much better.

Content by Scott
Typos by Siri

On Jul 26, 2018, at 4:59 PM, W7SLS <***@gmail.com> wrote:

Hi Bob,

Thanks for the quick reply, makes sense.

73
Scott
W7SLS
Post by Bob kb8tq
Hi
They are a pretty rare item. A more common approach is a disciplined oscillator that
will do failover on it’s inputs. That’s still a rare item, but at least a possible thing to find.
The equally big problem will be getting doc’s on one if you do find it….
Bob
Post by W7SLS
Hello,
Looking for recommendation for a ‘failover’ or ‘redundant’ switch for 10 MHz distribution.
Not really sure of the correct term.
Something that sensed RF on primary 10 MHz, and then switched to secondary on fail of primary.
A brief search showed several very nice $$$ items, suitable for commercial applications,
but I wonder if there are some “last year’s” (but not last century) versions that would work for a home lab.
I have a GPSDO and a Rb source of 10 MHz.
The power supply on the GPSDO failed (worked enough to light up the GPSDO, but not enough to lock).
I have a new power supply on order, but would be nice to have “insurance”.
Thanks in advance for the group bandwidth.
Scott
W7SLS
_______________________________________________
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
Bob kb8tq
2018-07-27 21:06:45 UTC
Permalink
Hi

If you are feeding a frequency counter with the “10 MHz + spur” signal, it can indeed mess up the modern
computing counters.

Bob
Post by Scott McGrath
Since this is time nuts ADEV is sacred,
but the question is the use case.
Are we looking at a time nuts level frequency distribution network. Or are we trying to lock benchtop instruments with poor internal standards to a common ‘Good’ reference.
If the former it can be done but will require a fair amount of design work as well as some EMC
Modeling of the system to ensure no feed through of unwanted signals and/or modulation or beat products
If the latter simple detection and switching will be sufficient but dual switch much better.
Content by Scott
Typos by Siri
Hi Bob,
Thanks for the quick reply, makes sense.
73
Scott
W7SLS
Post by Bob kb8tq
Hi
They are a pretty rare item. A more common approach is a disciplined oscillator that
will do failover on it’s inputs. That’s still a rare item, but at least a possible thing to find.
The equally big problem will be getting doc’s on one if you do find it….
Bob
Post by W7SLS
Hello,
Looking for recommendation for a ‘failover’ or ‘redundant’ switch for 10 MHz distribution.
Not really sure of the correct term.
Something that sensed RF on primary 10 MHz, and then switched to secondary on fail of primary.
A brief search showed several very nice $$$ items, suitable for commercial applications,
but I wonder if there are some “last year’s” (but not last century) versions that would work for a home lab.
I have a GPSDO and a Rb source of 10 MHz.
The power supply on the GPSDO failed (worked enough to light up the GPSDO, but not enough to lock).
I have a new power supply on order, but would be nice to have “insurance”.
Thanks in advance for the group bandwidth.
Scott
W7SLS
_______________________________________________
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 fol

Van Horn, David
2018-07-26 22:23:05 UTC
Permalink
Not quite what you are looking for, but I implemented a pair of thunderbolts with no common parts (dual antennas power etc) into a simple RF switch.
The production manager flipped the switch on Mondays, and if either system wasn't working I had a third system in a box ready to replace.
So for any failure of the two live systems, all he had to do was flip the switch and call me.


-----Original Message-----
From: time-nuts <time-nuts-***@lists.febo.com> On Behalf Of W7SLS
Sent: Thursday, July 26, 2018 1:46 PM
To: time-***@lists.febo.com
Subject: [time-nuts] 10 MHz 'failover' switch?

Hello,

Looking for recommendation for a ‘failover’ or ‘redundant’ switch for 10 MHz distribution.

Not really sure of the correct term.
Something that sensed RF on primary 10 MHz, and then switched to secondary on fail of primary.

A brief search showed several very nice $$$ items, suitable for commercial applications, but I wonder if there are some “last year’s” (but not last century) versions that would work for a home lab.

Context:

I have a GPSDO and a Rb source of 10 MHz.
The power supply on the GPSDO failed (worked enough to light up the GPSDO, but not enough to lock).
I have a new power supply on order, but would be nice to have “insurance”.

Thanks in advance for the group bandwidth.

Scott
W7SLS
_______________________________________________
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.
Scott McGrath
2018-07-26 22:40:30 UTC
Permalink
True but if you use a good switch or PIN 90 db of isolation is easily achievable. Yes the spur is still there but it’s 90 down and will not affect ADEV as badly.

You could drive it further down with two switches with the alternate standard connected to 1 port and a termination to the other. So a switchover would switch to the alternate port and the backup path would switch from termination to backup source. This would easily buy you 120-130 db isolation assuming use of good cabling and proper routing, grounds etc

After all this IS time-nuts after all

Content by Scott
Typos by Siri

On Jul 26, 2018, at 6:23 PM, Van Horn, David <***@backcountryaccess.com> wrote:

Not quite what you are looking for, but I implemented a pair of thunderbolts with no common parts (dual antennas power etc) into a simple RF switch.
The production manager flipped the switch on Mondays, and if either system wasn't working I had a third system in a box ready to replace.
So for any failure of the two live systems, all he had to do was flip the switch and call me.


-----Original Message-----
From: time-nuts <time-nuts-***@lists.febo.com> On Behalf Of W7SLS
Sent: Thursday, July 26, 2018 1:46 PM
To: time-***@lists.febo.com
Subject: [time-nuts] 10 MHz 'failover' switch?

Hello,

Looking for recommendation for a ‘failover’ or ‘redundant’ switch for 10 MHz distribution.

Not really sure of the correct term.
Something that sensed RF on primary 10 MHz, and then switched to secondary on fail of primary.

A brief search showed several very nice $$$ items, suitable for commercial applications, but I wonder if there are some “last year’s” (but not last century) versions that would work for a home lab.

Context:

I have a GPSDO and a Rb source of 10 MHz.
The power supply on the GPSDO failed (worked enough to light up the GPSDO, but not enough to lock).
I have a new power supply on order, but would be nice to have “insurance”.

Thanks in advance for the group bandwidth.

Scott
W7SLS
_______________________________________________
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.

_______________________________________________
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 instructio
Bob kb8tq
2018-07-27 00:36:04 UTC
Permalink
Hi

90 db will not drive the ADEV nuts, but it still will be a lot worse than a normal standard will deliver, even at
120 db down, if the offset is a bit above 1 Hz you will still see it in an ADEV plot.

Bob
Post by Scott McGrath
True but if you use a good switch or PIN 90 db of isolation is easily achievable. Yes the spur is still there but it’s 90 down and will not affect ADEV as badly.
You could drive it further down with two switches with the alternate standard connected to 1 port and a termination to the other. So a switchover would switch to the alternate port and the backup path would switch from termination to backup source. This would easily buy you 120-130 db isolation assuming use of good cabling and proper routing, grounds etc
After all this IS time-nuts after all
Content by Scott
Typos by Siri
Not quite what you are looking for, but I implemented a pair of thunderbolts with no common parts (dual antennas power etc) into a simple RF switch.
The production manager flipped the switch on Mondays, and if either system wasn't working I had a third system in a box ready to replace.
So for any failure of the two live systems, all he had to do was flip the switch and call me.
-----Original Message-----
Sent: Thursday, July 26, 2018 1:46 PM
Subject: [time-nuts] 10 MHz 'failover' switch?
Hello,
Looking for recommendation for a ‘failover’ or ‘redundant’ switch for 10 MHz distribution.
Not really sure of the correct term.
Something that sensed RF on primary 10 MHz, and then switched to secondary on fail of primary.
A brief search showed several very nice $$$ items, suitable for commercial applications, but I wonder if there are some “last year’s” (but not last century) versions that would work for a home lab.
I have a GPSDO and a Rb source of 10 MHz.
The power supply on the GPSDO failed (worked enough to light up the GPSDO, but not enough to lock).
I have a new power supply on order, but would be nice to have “insurance”.
Thanks in advance for the group bandwidth.
Scott
W7SLS
_______________________________________________
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.
_______________________________________________
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-
ed breya
2018-07-27 16:29:34 UTC
Permalink
Getting great isolation at 10 MHz is the easy part, given enough
switching elements and control. One question is whether the switchover
needs to be transparent (glitchless), without adding or losing any clock
cycles, and ideally with no phase shift. This would involve a much more
sophisticated system, with two or more redundant references locked
together, at least short-term.

Ed

_______________________________________________
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.
Bob kb8tq
2018-07-27 16:44:25 UTC
Permalink
Hi

Isolation in a carefully managed test setup can be done (with enough money to spend). Isolation it the real
world with grounds and cables running here and there is likely to be a challenge. At least that’s been the case
on the few dozen of these systems I’ve designed and put into production ….

Bob
Getting great isolation at 10 MHz is the easy part, given enough switching elements and control. One question is whether the switchover needs to be transparent (glitchless), without adding or losing any clock cycles, and ideally with no phase shift. This would involve a much more sophisticated system, with two or more redundant references locked together, at least short-term.
Ed
_______________________________________________
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
Loading...