Distributing reference clocks and triggers in a lab

[tnt]

Hi,

Let's say I have several instruments and most of them can take a 10MHz reference clock and all of them have both trigger input and trigger output.

Now I'd like to be able to distribute the 10 MHz from a single clock reference to all of them and also be able to configure the trigger in/out dispatch without having to plug/unplug cables.

Does any one have recommendations for one or the other ? (somehow I doubts BNC 'T' will do ... it'll half the power each time and it's already only 0.5 Vpp in 50ohm at the clock ref output)

[sacherjj]

This still may be fine.  If you connect them in parallel, they will all still be 0.5V.  The question then is if you have enough power to provide the needed current to all of them.  Generally the input requirement is much less than the outputs, so you might just be able to T everything up.

[mikeselectricstuff]

 video distribution amp may be an off-the-shelf solution - composite one may ebe marginal but an analogue VGA amp ought to do it easily

[DavidDLC]

What we do is to have a common signal distributor (10 MHz reference) and then small amplifier/conditioner/filter units right before our instrumentation. But our production line is noisy.

We don't have the trigger connection that you mention.

[tnt]

This still may be fine.  If you connect them in parallel, they will all still be 0.5V.  The question then is if you have enough power to provide the needed current to all of them.  Generally the input requirement is much less than the outputs, so you might just be able to T everything up.

Huh ... it won't be be 0.5V if the source impedance is 50R. It'll end up be 0.2V after 4 devices.

[tnt]

For the triggers, maybe I didn't make myself clear : I'd like to be able to select (with a switch or something) which instrument is the source and distribute that to all the others.

[Rufus]

There is a Chinese guy on ebay who sells a board with a 10MHz OCXO reference and 2 sine wave and 4 square wave outputs, it also has a 10MHz input. Item 300417981247 about $80.

As for triggers in and out you are going to need a lot of coax and BNCs what ever you do. I would probably bring them all to a patch panel and make connections with short BNC patch leads - is that so much harder than flicking switches? If you need to drive multiple inputs you may have the same loading issues as with clock distribution.

[sacherjj]

For the triggers, maybe I didn't make myself clear : I'd like to be able to select (with a switch or something) which instrument is the source and distribute that to all the others.

This is going to most easily done with patch cables.  Put the source manually on and go through whatever distribution amp you need and then out to all.  Otherwise you need to switch to select the source and also drop that source out of the outputs.  If it is something that must be that easy, then a relay for each source would probably work with a NC/NO pair to select for signal and drop out for output a given cable, assuming you reinforce that only one relay is selected at a time.

I have no idea how well relays work at 10 MHz. 

[tekfan]

I have no idea how well relays work at 10 MHz. 

You can get coaxial relays quite cheaply on ebay. Even ones that work from DC well into the 10GHz range. A bit of searching required but you'll eventually find the right one.

[bilko]

While browsing the internet, I came across
http://www.google.co.uk/url?sa=t&source=web&cd=39&ved=0CFUQFjAIOB4&url=http%3A%2F%2Fg4hup.com%2FDA%2FDA1-4%2520Manual%2520V1_01.pdf&rct=j&q=gps%20amplifier%20circuit&ei=FdLBTdDTJZKr8APtjNnFBQ&usg=AFQjCNEXqkI0cDYUOwHJsYUHlN0WPomK5g&cad=rja

Hope its of some use

[tnt]

To all: Thanks, lots of interesting stuff in the answers. I'll look all that up.