Author Topic: ColpittsOscillator - realistic limit with through-hole components  (Read 850 times)

0 Members and 1 Guest are viewing this topic.

Offline bmdalyTopic starter

  • Contributor
  • Posts: 18
  • Country: ie
Hi,

I've built a Colpitts oscillator based on the attached circuit diagram.
I'm looking to get it oscillating at 200MHz+, if possible.

I'm using ceramic capacitors with nominal values of 10pF, and an inductor with nominal value 100nH.
The JFET is a J112 (NChannel JFET with Vgs off = -3.6V) - I cannot find its cutoff frequency in the datasheet, so would be interested if anyone can provide it.

All components are through hole (with leads trimmed), and soldered directly together, with no board (essentially, hanging from the power leads). Power is from a bench power supply, with 30cm leads.

Predicted theoretical frequency is approx. 225MHz (with zero allowance for parasitic reactances), but I'm measuring approx. 80MHz on my oscilloscope.

I realise that leads etc. will add inductance, the probe has input capacitance in the order of 20pF, and I’m sure there are many other parasitic capacitances and inductances at play.

I’m wondering what a realistic upper frequency is achievable with this simple direct soldered setup, or what I would need to change in the circuit, or its construction, to get the frequency higher.

Thanks in advance for any input.

Note that this is purely for educational purposes – I’m not interested in buying a purpose made signal generator in that frequency range!
 

Offline daqq

  • Super Contributor
  • ***
  • Posts: 2321
  • Country: sk
    • My site
Re: ColpittsOscillator - realistic limit with through-hole components
« Reply #1 on: July 19, 2019, 08:10:04 pm »
Well, throughhole resistors are generally a ceramic cylinder with a spiral pattern metal film on them, so that's another inductor. The transistor doesn't seem all that fast, see http://www.htmldatasheet.com/micro-electronics/j112.htm and look at the times mentioned.

Also, if you wan't to limit the effect of your scope, you can couple high frequencies indirectly - just place a probe close to the measured system and set the gain to max. Or make a 1 turn loop with your scope ground.
Believe it or not, pointy haired people do exist!
+++Divide By Cucumber Error. Please Reinstall Universe And Reboot +++
 
The following users thanked this post: bmdaly

Offline Wolfgang

  • Super Contributor
  • ***
  • Posts: 1827
  • Country: de
  • Its great if it finally works !
    • Electronic Projects for Fun
Re: ColpittsOscillator - realistic limit with through-hole components
« Reply #2 on: July 19, 2019, 08:40:00 pm »
... things like this can be made to oscillate well in the low GHz range. Just keep wires short and use a high-ft transistor.
 
The following users thanked this post: bmdaly

Offline bmdalyTopic starter

  • Contributor
  • Posts: 18
  • Country: ie
Re: ColpittsOscillator - realistic limit with through-hole components
« Reply #3 on: July 19, 2019, 08:45:07 pm »
Well, throughhole resistors are generally a ceramic cylinder with a spiral pattern metal film on them, so that's another inductor. The transistor doesn't seem all that fast, see http://www.htmldatasheet.com/micro-electronics/j112.htm and look at the times mentioned.

Also, if you wan't to limit the effect of your scope, you can couple high frequencies indirectly - just place a probe close to the measured system and set the gain to max. Or make a 1 turn loop with your scope ground.

Thanks for the link to the datasheet - OnSemiconductor specifies nothing about frequency on theirs.

There appears to be limited choice nowadays for JFETs - on ie.farnell.com a search for RF JFET gives me very few options.
One through hole option - BF256B - with the OnSemiconductor datasheet giving no frequency figures, only a non-committal "This device is designed for VHF / UHF amplifiers".
Eight further options in SOT23 package - MMBFJ309 might fit the bill with a somewhat more reassuring "As a common gate amplifier, 16 dB at 100 MHz and 12 dB at 450 MHz can be realized"
 

Offline bmdalyTopic starter

  • Contributor
  • Posts: 18
  • Country: ie
Re: ColpittsOscillator - realistic limit with through-hole components
« Reply #4 on: July 19, 2019, 08:51:12 pm »
... things like this can be made to oscillate well in the low GHz range. Just keep wires short and use a high-ft transistor.

Thanks for the response - it occurred to me that I might be seeing aliasing - 580MHz sampled at 1GSa/s - but increasing the capacitance/inductance brought the observed frequency back closer in line with expectations.
I'll keep trying with a faster JFET.
Will also try adding an emitter follower or source follower as a buffer to isolate the probe from the oscillator transistor.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf