Distributing Output from a Pulse Generator

[slomotion]

I need to distribute a 1MHz to 10MHz, 50% duty cycle TTL signal coming from a pulse generator. I need to switch between 5 - 8 different outputs. The propagation delay needs to be very small less than <=5nsec through the device. I have become over whelmed with the amount of choices and became very confused with which IC chip would provide the best solution or maybe a relay? I am aware that a relay would time time to setup which is ok. But once the channel is turned on it needs to have a minimal impact on the signal.

 

[VEBill]

Reading between the lines, you're talking about a pulse generator as in Test Equipment? You're not designing a circuit that requires this function, and you're not looking for an embedded solution? Right?

How about a good old fashioned 12-position rotary switch with a knob? You would have to wire it up with some basic attention to impedance matching, but it's not that difficult if you know what you're doing (regarding characteristic impedance and what affects it).

If you need a "more professional" solution (as opposed to home brewing a rotary switch), then there are many mechanical switch matrix devices available. They're typically a metal box with a bunch of coaxial connectors lined-up on various sides. And then some low voltage terminals for the control lines. Probably many hundreds of dollars each.

 

[ScottyUK]

Does the timing requirement apply to the output signals relative to each other or is it an absolute requirement relative to the source? Could you accept a propagation delay if it applied equally to all output channels? That may open up some mainstream electronic solutions without getting into exotica.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[slomotion]

Ok, I should elaborate on what I am looking to do. We are measuring switching speed on a pin diode switch filter bank. Typical switching speed ranges from 50nsec to 100nsec. We have various models ranging from SP2T to SP8T. I am tasked to automated the measurement. Currently we have a tech manual connect each output bit. Cycling through all the bits.
00001
00010
00100
01000
10000
Since this will be part of an ATE. I want to find an IC that could be controlled through our computer. We currently use a toggle switch to flip from channel to channel but really want a complete automate measurement. Still new to the site. I will upload a test setup diagram later.

 

[waross]

Magnetic reed switches?

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[electricuwe]

The most simple choice would be to use a 74ACT138 1-to-8 channel demultiplexer.

Unfortunately propagation delays are a few ns higher that the 5 ns requirement you specified, but you may consider the most part of the delay as fixed and take it into consideration when calculating the delay time of your pin-switches.

Take into consideration that for timing in the ns-range also cable lenght and termination my have important impact.

 

[benta]

There are excellent analog switches for this task that would minimise distortion of the signal.
Try the Vishay/Siliconix wvsite for more informantion.

Cheers,

Benta.

 

[benta]

Sorry, should have read website, not wvsite.

 

[VEBill]

If you're measuring the turn-on and turn-delay of the PIN diode switches within the UUT, then you should make a *relative* measurement between your control signal input, *monitored at the terminal*, and the rising or falling edge of the RF at the output. With this technique, there's no concern about propagation delays from the test equipment to the UUT; only that the propagation delay *from* the UUT to the 2-channel 'scope should be equal (trivial to achieve and monitor).

 

[VEBill]

Correction: "turn-off delay"