RF delay line for 1.5 GHz to 4 GHz

[p.francks]

Dear all,
I am looking for a "true time delay" line for microwave pulses from 1.5 GHz to 4 GHz. I am looking for something that is programmable via TTL with switching times < 2 us and a max delay of around 1 ns and a step size < 150 ps. Can you recommend me any module that can do that?

Thanks!

Peter

[f4eru]

that spec looks more like a phase shift.
What do you try to achieve on which signals ?

[TimFox]

Here is a survey of programmable delay lines.
https://www.everythingrf.com/search/delay-lines/filters?page=1&country=global&stype=;Programmable;

[p.francks]

I have (square) pulses at a carrier frequency typically ~2-3 GHz that are ~10 ns long that I want to delay with respect to a second signal issued from the same pulse generator. I want to control the relative timing of these two signals in steps up to 1 ns. The application is for some magnetic resonance experiment.

[Gerhard_dk4xp]

I have made a switchable delay line of 3 ranks of
( 1:6 SMA relay, variable lengths of semi rigid coax, 6:1 SMA relay).

It has 3 rotary switches and can bring everything between 4  and >100 MHz
into phase quadrature with the help of a power splitter, for measuring
phase noise. When the delay line was done, I bought a commercial
phase noise analyzer. 

I also have two motor-driven "Arra, Inc" 9228A-28 delay lines for microwaves.
The switching speed of all that is some decades slower than what
you need.

Cheers, Gerhard

[p.francks]

Thank you very much! Yes, fast switching is really necessary. After some more research, in principle something like this would be ideal
http://www.gigabaudics.com/PADL6/padl6.html

However, this company seems to be inactive by now  :'(
Therefore I am asking? Do you know any other vendor for such a device?

[Chat GPT]

One option for a "true time delay" line for microwave pulses in that frequency range could be a digitally programmable delay line (DPDL). These modules are designed to provide precise, programmable delays to microwave signals and are commonly used in radar and communications systems.

One example of a DPDL that meets your specifications is the Hittite HMC649LP6. This module operates from 1.5 GHz to 4 GHz and provides a programmable delay range of 0 ns to 7.5 ns with a step size of 150 ps. It has a switching time of 1.5 us and is programmable via TTL.

Other manufacturers, such as Mini-Circuits, Analog Devices, and Peregrine Semiconductor, also offer DPDLs that may meet your requirements. It's worth researching and comparing different options to find the best fit for your application.

[p.francks]

Thank you very much for the reply. The HMC649LP6 seems to be a phase shifter for 3-6 GHz or did I pick the wrong device?

https://www.analog.com/media/en/technical-documentation/data-sheets/hmc649.pdf

A more general question: Would a phase shifter also work? From a pure physics point of view a phase shift is a delay but I don’t know how these devices behave for pulses that have a spectral width of here 500 MHz.

Thanks for your help again!

Peter

[Marco]

One example of a DPDL that meets your specifications is the Hittite HMC649LP6. This module operates from 1.5 GHz to 4 GHz and provides a programmable delay range of 0 ns to 7.5 ns with a step size of 150 ps. It has a switching time of 1.5 us and is programmable via TTL.

This answer feels a bit GPT.

[gf]

A more general question: Would a phase shifter also work? From a pure physics point of view a phase shift is a delay but I don’t know how these devices behave for pulses that have a spectral width of here 500 MHz.

Delay corresponds to a freqency dependent phase shift, i.e. phase(f) = 2*pi*f*delay.

[p.francks]

Yes, I get that. Maybe I should have posed the question more specifically: How is a phase shifter actually working? Is it really a delay (then it would be fine) or does it adjust the phase otherwise? Because if it is constant phase regardless of the frequency, then I will run into a problem for a pulse as this consists out of multiple frequencies. Multiple spectral components will then have different delays... Next problem is that a 360° phase shifter is only 500 ns at 2 GHz. So one would need to daisy chain.

Nonetheless, I am still on the hunt for something suitable for my delay. I find a lot of research articles but not so many products.

[Marco]

This Russian page has a nice list of products.

[TimFox]

Yes, I get that. Maybe I should have posed the question more specifically: How is a phase shifter actually working? Is it really a delay (then it would be fine) or does it adjust the phase otherwise? Because if it is constant phase regardless of the frequency, then I will run into a problem for a pulse as this consists out of multiple frequencies. Multiple spectral components will then have different delays... Next problem is that a 360° phase shifter is only 500 ns at 2 GHz. So one would need to daisy chain.

Nonetheless, I am still on the hunt for something suitable for my delay. I find a lot of research articles but not so many products.

A delay line has a constant time delay from input to output, and the phase shift for a sine wave will follow the equations given above.
A simple example of a constant 1.5 ns delay is approximately 30 cm of RG-58/U coaxial cable, properly terminated.
There are other ways to get a phase shift, which will have a different relationship between frequency and phase, such as a filter of various kinds.