Microwave 1.7GHz VCO Oscillator

In this video, we are going to take a look at a microwave VCO oscillator that can be tuned from 700MHz to 1.7GHz.
The design uses an active device, in common emitter configuration, and a variable tuned phasing network.
The phasing network is tuned using three varactos diode, spreading the capacitance variation across the network.
Oscillation is maintained by the S21 transfer of the active device. Unlike a negative impedance oscillator, this topology has greater phase performance, because the contribution of phase shift from the active device is smaller.
00:16 - Introduction
00:46 - Negative Impedance Oscillators
01:26 - Oscillators using two port devices
03:11 - Circuit description
05:42 - Usage for signal generators
06:57 - Final considerations
Matjaz also use this topology with an interdigital microstrip resonator, providing greater phase noise and higher maximum frequency.
An interdigital phase network can provide higher phase shift than 360degrees, making the contribution of the low Q phase shift of the active device low.
The topology shown was first published by Matjaž Vidmar.

[@AllElectronicsChannel]

Support the channel becoming a Patron patreon.com/allelectronics

[@paolomonai9511]

The principle of negative resistance for designing microwave VCO is widely used. I have been designing this type of voltage controlled oscillators for more than 20 years and PLO, DRO, CRO and synthesizers as well. The phase noise performance is only partly due to the noise in bias circuit of the active element. Q of resonant circuit is of paramount importance when low phase noise is requested. To guarantee wide band tuning, one can choose specially designed varactor (the so called "hyperabrupt" junction). In fact, if you add varactors, the package parasitcs increase...reducing the advantage of more varactors. The necessity of an "isolation/buffer" amplifier is a good trick to improve other performances (i.e. frequency pulling, for example) or to increase output power. At higher microwave frequencies (say >2GHz) only the negative resistance approach is commonly used. Anyway, nice video. Regards from Italy.

[@AllElectronicsChannel]

Thank you! I'm thinking to construct a 50meg to 1Ghz discrete synthesizer.. Probably an oscillator tuned from 500MHz to 1GHz mixed with other constant frequencies or dividers. What topology do you recommend me? Regards from Brazil!

[@paolomonai9511]

@@AllElectronicsChannel There is no unique and magical recipe as far as topology is concerned. It depends on the application. Is it a fine step synth, or a fast switching type? What are the phase noise specs? As a general rule of thumb, try to minimize frequency excursion range in the main loop (say one octave) using mixing with secondary fixed tones (tipycally, low noise obtained thru multiplication and filtering of reference xtal) and also switching the downconverted signal to a different divider chain. Frequency dividing/prescaling can be used but it is better to have low divider ratio (small numbers) so you do not ruin phase noise.

[@tr_2sc1970]

Great video, I got a bunch of RF transistors and I wanted to test their liability, so I'm going to make a GHz VCO for the matter. Said to myself, I look for some suggestions and I came to your video!

[@AllElectronicsChannel]

Sounds great! Good luck man!

[@MauricioPagano]

Thank you so much! excellent explanation. I know Matjaz's work there on his website we can find many interesting things.

[@gabrielfurtado4932]

Very nice!!

[@amirb715]

excellent design and results! please also show the process of assembly and fabrication.

[@estebanbriceno1625]

Yes, it will be fantastic de process, maybe in a accelerated video, an you describing the process

[@camilosoveral7294]

Excelente, já dei like. Abraços

[@AllElectronicsChannel]

Thanks!

[@omsingharjit]

can you make cheaparduino spectrum analyser using this vco or crt tv tunner ?

[@zetadoop8910]

Great video as usual BUT it often lacks values of capacitors, transistor models.

[@AllElectronicsChannel]

Sorry! This video is only an overview of the topic

[@gilmarribas]

Top

[@AllElectronicsChannel]

Thanks!

[@colinhoof-ee9rg]

The circuit you used here uses npn and PNP devices. What is the part used as the pnp device in this circuit. Is this part of a complementary pair ?

[@AllElectronicsChannel]

Any PNP will work, it is just a current source. I probably used a BC557

[@xDR1TeK]

What about the miller multiplier in the CE layout?

[@AllElectronicsChannel]

Will reduce the potential maximum frequency!

[@xDR1TeK]

@@AllElectronicsChannel my issue is that CE isn't reliable when working for accuracy, since, you know, capacitance across the feedback path where tranconductance can play a role, the collector - base capacitor will appear at the input as 1000x times higher than the input emmiter-base capacitance. Wouldn't you rather test a CB?

[@fjs1111]

Your board design is horrible!! The leads are way too long, at those frequencies leads and traces make a huge difference..

[@AllElectronicsChannel]

Thanks!

[@andrewandrosow4797]

Hello! The LC tank load in one side on 100-300Ohms input impedance.The same situation on the output. So, the LC circuit has low quality factor, and has bad stability (but the PLL with a quartz is in charge for it). Am I right?

[@MeEncantaKiley]

Dude you can affect GPS around 1.5 GHz! I’ll get this too !