Ultra high frequency oscillators

Could some electronics guru please shed some light on this ?

How is the carrier wave for high frequency wireless communication transmitters generated ? Specifically, any oscillator would need negative impedance device, e.g., a bipolar transistor to sustain the oscillations. So, what transistors would be used in e.g., a GSM transmitter with a carrier frequency of 1.8 GHz

Thanks in advance for your help.

Reply to
dakupoto
Loading thread data ...

1.8GHz is not ultra high frequency

Google is your frind, see this link for plenty of examples:

formatting link

Cheers

Klaus

Reply to
Klaus Kragelund

An oscillator is simply a noise amplifier with frequency selective positive feedback. The amplifier gain needs to be higher than the losses in the feedback network.

Google for Barkhausen.

Reply to
upsidedown

According to ITU

UHF = Ultra High Frequency = 300 .. 3000 MHz,

so 1.8 GHz is definitively UHF.

Reply to
upsidedown

negative impedance device, e.g., a bipolar

So, what transistors would be used in e.g.,

Transistors have been much faster than that for a long time--I've built cir cuits that (accidentally) oscillated around 12 GHz, using 40-GHz SiGe bipol ar transistors. You can get discretes with f_T values of 80 GHz or more (B FP840 iirc).

Integrated transistors are getting near 1 THz, though not in silicon. A fri end of mine from grad school, Mark Rodwell, has been doing great things in that line for some years now out at UCSD. Before transistors were that good, there were klystrons, magnetrons, and tr avelling-wave tubes (TWTs), all of which still have their uses. For powers of milliwatts to watts, there are negative-resistance diodes such as Gunn a nd IMPATT (impact-avalance transit time), and for low power, high-order fre quency multipliers such as step-recovery diodes running off lower frequency oscillators.

Now that frequency-locked femtosecond lasers are so good, you can make esse ntially any frequency you like by picking two teeth of the comb and beating them together, e.g. with a nonlinear crystal or a fast photodetector such as an antenna-coupled tunnel junction.

Cheers

Phil Hobbs

Reply to
Phil Hobbs

To the original poster, one should remember that the f_T only define the frequency at which _current_ gain drops to unity.

However, you might still have voltage gain and hence _power_ gain in common base configuration above f_T, so do not be surprised, if some transistor stage oscillates above f_T.

Reply to
upsidedown

There are special components for UHF oscillators, e.g. the magnetrons (2.4 GHz) in microwave ovens.

A free-running oscillator does not usually have good enough stability for transmitting use.

The practical transmitters start with a crystal oscillator at a lower frequency (some MHz to 100 MHz), and multiplying the frequency to the destination frequency range.

For complicated modulation (like cell-phones), the multiplied frequency in mixed in an image-rejection mixer with the modulation baseband information to reach the final modulated signal.

--

-TV
Reply to
Tauno Voipio

Theoretically, but I've never seen one do that by accident. It's generally hard to manage even f_T/2 in a purpose-built oscillator unless you really breathe on it.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

You can get etched silica crystals with a fundamental mode up to at least 315MHz

formatting link

Connor-Winfield offer oscillators that go a bit higher

formatting link

When I was looking at packaged stable, low-jitter non-multiplied 500MHz oscillator a few years ago, it wasn't cheap (at around $100) but it did make life easier.

--
Bill Sloman, Sydney
Reply to
bill.sloman

The general process is to start with a low frequency crystal oscillator (which is very stable) and control an oscillator of some sort with a phase-locked loop, PLL.

The final oscillator is generally a fast transistor, a bipolar or PHEMT, or for low power like a cell phone, lately just part of a CMOS chip.

There are tons of variations. One can also amplify the harmonics of a lower-frequency oscillator, in multiple stages of 2:1 or maybe 3:1.

See wikipedia for details.

--
John Larkin         Highland Technology, Inc 

jlarkin att highlandtechnology dott com 
 Click to see the full signature
Reply to
John Larkin

The catch with this approach is that the jitter on the clock edges is essentially the jitter on the lower-frequency oscillator.

Etch-thinned silica crystals with a fundamental frequency of a couple of hundred MHz tend to offer 1psec or better edge jitter.

Surface acoustic wave resonators offer similar high Q values at even higher frequencies, but getting one made for a particular frequency can be expensive, though they are cheap in volume.

formatting link

Gollege does offer a fairly wide range of "standard" frequencies

formatting link

but I've no idea how much they cost.

--
Bill Sloman, Sydney
Reply to
bill.sloman

Official definition is "UHF". "Ultra high" as in "super challenging and bleeding edge" is way higher than 1.8GHz. Just as "Modern art" means artwork by someone whose been dead for decades, "Ultra"-whatever may not be so ultra any more.

--
Tim Wescott 
Wescott Design Services 
 Click to see the full signature
Reply to
Tim Wescott

Yes. The ones that I am aware of use a cross-coupled pair of MOSFETs, with a differential tank inductor of a few nanohenries or less, and switched banks of capacitors for the coarse tuning, and MOS varactors for the fine tuning. You can see the general circuit topology from the pictures in the patent US7038552, although things may have moved on a bit since then.

Reply to
Chris Jones

Exactly, that's why I wrote as I did

In the year 3000 UHF is probably very very very very high..., if the civilization is still present.

Cheers

Klaus

Reply to
klaus.kragelund

UHF is an acronym for Ultra High Frequency.

Reply to
John S

If the inductor is differential, how does the switching work? Are there two identical sides switched to earth? Any mismatch in the capacitors would affect the symmetry, but perhaps that's not important.

Clifford Heath

Reply to
Clifford Heath

The tank inductor is centre-tapped, and the tap is connected to the positive supply (if NMOS devices are used to make it oscillate). The two ends of the inductor are connected to the drains of the core devices that make it oscillate. The switching circuit for the banks of capacitors are shown in that patent I mentioned. One NMOS switch shorts together the bottom plates of two capacitors when they are wanted on. The capacitors have excellent matching, being on a chip. The asymmetry in the inductor due to cross-unders where the turns cross would be bigger.

Reply to
Chris Jones

Yes, to be able to use crystals in the lowest cost range. In the old days they'd use really expensive ones close to 100MHz and press them into the x-th overtone mode. Or use class-C stages and fish out harmonics but that's antique design.

Nowadays UHF frequencies are mostly generated by circuitry inside an IC which also contains the PLL and all sorts of other things, and costs less than 1/10th of a cup of Java at the coffee shop. It feels like cheating, like cracking open a TV dinner package in the evening.

That's how we dunnit 40 years ago but no more. When I still had hair.

Better yet the ARRL Handbook. That also tells readers the important details on how not to mess up the mechanical build or layout. I haven't read one in over 10 years but this book used to have the absolute best bang for the buck when it comes to RF design.

--
Regards, Joerg 

http://www.analogconsultants.com/
Reply to
Joerg

Multiplying is still done at microwave frequencies, where outrageous-Q resonators are available to allow high-N multiplication from SRDs or something. At low frequencies, with LC filters, multiplying by a large factor is uneconomical.

I was never interested in ham radio (all they seemed to do is talk to one another) but I did read the Radio Amateurs Handbook a lot.

I have the historical 1946 edition right here. Good year, 1946.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

Sure, but not at UHF anymore where we still had to do that in the days before PLLs and synthesizers became affordable.

That's kind of the idea :-)

In essence it was "the" technical exchange medium way before Usenet and CompuServe came about. I learned more useful stuff in ham radio than at the university. It was also "the" channel to find parts before EBay. "Yeah, I've got a roller inductor that I don't need anymore" and the method of payment often was a crate of beer. Good beer, none of the cheap stuff.

The other aspect I like is the minimalist nature of ham communication and that it largely does not rely on any infrastructure being intact. Morse code is unbeatable when it comes to getting a message across under very RF-hostile conditions. The newer kinds of more IP-based protocols never interested me.

I gave a lot of historical books away. Still got many in German, some way old, like 100 years or so. I am wondering what to do with those some day.

--
Regards, Joerg 

http://www.analogconsultants.com/
Reply to
Joerg

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.