Author Topic: Unable to build stable oscillator  (Read 4411 times)

0 Members and 1 Guest are viewing this topic.

Offline j.a.mcguire

  • Contributor
  • Posts: 29
Unable to build stable oscillator
« on: May 31, 2015, 11:20:58 am »
Ive been trying to build a simple oscillator circuit using schematics found online. The scope im using to monitot thr signal is new and I wanted to test it against a reliable signal so I have some confidence in the scope.

I've tried two schematics so far but the frequency generated jumps around all over the place.

The last one I attempted was a collpits oscillator based on the schematic here:

http://www.electronics-tutorials.ws/oscillator/osc31.gif?81223b

From the following guide:
http://www.electronics-tutorials.ws/oscillator/crystal.html

The result was the Rigol 1054z showing the frequency jumping from a few kHz up to 100mhz and just bouncing all over the place:

http://s22.postimg.org/gyo62tka8/IMG_20150531_115622.jpg

I used a 1.8mhz and a 5mhz crystal identified as:

AQ1.8432F
AQ5.0000F

If I remove the crystals from the live circuit it seems to make jack all difference to the output signal.

The transistor is a PNP 2N3906.

All other parts are per the schematic.

Pic of setup:
http://s10.postimg.org/7krvu2mw7/IMG_20150531_115750.jpg

http://s14.postimg.org/g501c6m9r/IMG_20150531_115758.jpg

Could anyone with more electronics insight advise if the schematic im following looks correct and should work.

Is this a common problem is there something fundamental I'm missing?

Thanks

James
 

Online PA0PBZ

  • Super Contributor
  • ***
  • Posts: 3991
  • Country: nl
Re: Unable to build stable oscillator
« Reply #1 on: May 31, 2015, 11:33:22 am »
you are missing the connection from the 2 220pF capacitors to the emitter.
And clip your scope directly to the output, throw away the blue and yellow wires.
Keyboard error: Press F1 to continue.
 

Offline j.a.mcguire

  • Contributor
  • Posts: 29
Re: Unable to build stable oscillator
« Reply #2 on: May 31, 2015, 11:59:39 am »
OK thanks you're totally correct I spotted that and added it in but no change. I'll try without the croc leads
 

Online PA0PBZ

  • Super Contributor
  • ***
  • Posts: 3991
  • Country: nl
Re: Unable to build stable oscillator
« Reply #3 on: May 31, 2015, 12:04:42 pm »
Oh wait...  :palm: You can't just replace a NPN transistor with a PNP version  :--
Keyboard error: Press F1 to continue.
 

Online Andy Watson

  • Super Contributor
  • ***
  • Posts: 1696
Re: Unable to build stable oscillator
« Reply #4 on: May 31, 2015, 12:06:45 pm »
Check the power supply - I notice that you haven't connected the sense terminals on the TTi supply.

Edit: Ignore that. I've just checked the manual and the sense terminals are connected (loosely) inside.
« Last Edit: May 31, 2015, 12:17:33 pm by Andy Watson »
 

Offline j.a.mcguire

  • Contributor
  • Posts: 29
Re: Unable to build stable oscillator
« Reply #5 on: May 31, 2015, 12:13:45 pm »
Nice spot I just checked the data sheet again and I'm using the wrong transistor.  Total doh moment mixing up my pnps with my npns

P.s. not sure what sense terminals are for I suspected probing output of supply for correct voltages etc? seems to work regardless
 

Offline j.a.mcguire

  • Contributor
  • Posts: 29
Re: Unable to build stable oscillator
« Reply #6 on: May 31, 2015, 12:41:16 pm »
No further forward in afraid now using NPN 2N3904

There's some sort of signal but the scope can't identify the frequency

http://s12.postimg.org/jw53pyv8b/IMG_20150531_133706.jpg

http://s22.postimg.org/4nnuwrf8v/IMG_20150531_133716.jpg

Totally confused
 

Online PA0PBZ

  • Super Contributor
  • ***
  • Posts: 3991
  • Country: nl
Re: Unable to build stable oscillator
« Reply #7 on: May 31, 2015, 12:47:09 pm »
You should build it more compact, without the parts going all over the board (example: 470K & 220K).
(you can fold one lead over the resistor and so effectively end up with a very small footprint)

I doubt if the blue capacitors are 220pF, they look more like something in the nF range. (picture?)

Measure the voltages on C, B and E relative to ground.


« Last Edit: May 31, 2015, 12:48:49 pm by PA0PBZ »
Keyboard error: Press F1 to continue.
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 12740
  • Country: gb
  • Hero999
Re: Unable to build stable oscillator
« Reply #8 on: May 31, 2015, 12:59:42 pm »
Breadboard is also notoriously unreliable, especially for circuit like this.

Try soldering it together on as smaller piece of strip board as you can.
« Last Edit: May 31, 2015, 01:11:27 pm by Hero999 »
 

Offline j.a.mcguire

  • Contributor
  • Posts: 29
Re: Unable to build stable oscillator
« Reply #9 on: May 31, 2015, 01:08:02 pm »
You sir are a flipping genius!

They were nf caps, swapped them out for pf and low and behold:

http://s28.postimg.org/73l818qln/IMG_20150531_140219.jpg

Final circuit:

http://s10.postimg.org/zfc6y4ep3/IMG_20150531_140645.jpg

Many thanks!
« Last Edit: May 31, 2015, 01:10:00 pm by j.a.mcguire »
 

Offline j.a.mcguire

  • Contributor
  • Posts: 29
Re: Unable to build stable oscillator
« Reply #10 on: May 31, 2015, 01:15:15 pm »
The 1.8MHz waveform is interesting, its all curvy....woooh! (Only changed the crystal)

http://s24.postimg.org/ebt836qwj/IMG_20150531_141220.jpg
 

Offline j.a.mcguire

  • Contributor
  • Posts: 29
Re: Unable to build stable oscillator
« Reply #11 on: May 31, 2015, 01:19:36 pm »
Is there any way I can make this a true square wave form rather than a saw tooth or a blob?  Or is that as good as I'm going to get?
 

Online PA0PBZ

  • Super Contributor
  • ***
  • Posts: 3991
  • Country: nl
Re: Unable to build stable oscillator
« Reply #12 on: May 31, 2015, 01:29:27 pm »
Is there any way I can make this a true square wave form rather than a saw tooth or a blob?  Or is that as good as I'm going to get?

You will not get a square wave from that circuit, but you can try a 1nF cap where the scope is now and at the other end of the cap 2 diodes antiparallel to ground. Measure at the point where the cap meets the diodes. But what you really need is a limiting amplifier.
Keyboard error: Press F1 to continue.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 13303
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Unable to build stable oscillator
« Reply #13 on: May 31, 2015, 05:18:06 pm »
The signal you were seeing is just noise.  Ambient noise, mostly mains hum, switching power supplies and nearby radio transmitters.  The scale is way smaller than the oscillator should be producing (~2mV, not ~V).  You can confirm the source by clipping the probe tip to its ground clip, and waving it around.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 12740
  • Country: gb
  • Hero999
Re: Unable to build stable oscillator
« Reply #14 on: May 31, 2015, 05:42:58 pm »
If you want a squarewave then why not use CMOS gates instead?

 

Offline alsetalokin4017

  • Super Contributor
  • ***
  • Posts: 1873
  • Country: us
Re: Unable to build stable oscillator
« Reply #15 on: May 31, 2015, 07:43:18 pm »
When I first got my DS1054z I too wanted to test it on some little oscillator circuits. The first one I built was a basic multivibrator using 2 NPN transistors, to make a nice square wave output. I put in 2 sets of timing capacitors so I have two frequencies to choose from.

I also built a couple of crystal oscillators using CMOS gates. One basically identical to the circuit above, and another one using a slightly different circuit. At 20 MHz I don't get a good square wave displayed though. Scope bandwidth, or oscillator?

My comments on j.a.mcguire's problems are these:

First, I'm one of those people who think that the "Auto" button is For Emergency Use Only.  ;)   If you know the approximate voltage and frequency of your circuit, go ahead and use the Vertical and Horizontal scale controls to set the scope up to give you an intelligible picture for your anticipated signal, and then modifiy the settings based on what you see. The "Auto" button will make you lazy and mislead you as to what is going on, as it seems to have done in this case.

Second, whenever possible use the 10x attenuation setting on the probe, and make sure that the channel is properly set to 10x probe atten. as well. The 1x position loads the circuit more and reduces the effective bandwidth of the measurement system and should only be used when you really need it, like with very very small signals.

Third... the breadboard and clipleads. For circuits oscillating at high frequencies these are only going to give you problems. And 5MHz is "high enough" to qualify here. I offer one suggestion as to a better construction method, that is nearly as fast and easy as the breadboard, below, and which will provide much more stability in oscillating circuits.

Fourth: One of the great features of the Rigol scope is that it is really easy to make good screen captures. Just put a USB stick in the front panel hole and when the scope tells you that it is recognized, push the "print" button on the right hand top of the panel, under the "Help" button, and in a few seconds to a minute or so you'll have a great print of your screen saved on the USB drive. It won't have the correct date and time though, and will have a generic sequence filename, but you can figure that part out easily enough when you transfer it to your computer for upload or display. (One of the "bugs" of this scope is that it sometimes takes a long time for the file to transfer to the USB stick, like up to a minute or more, but it eventually will get there. Sometimes it goes faster if the scope is Stopped before you start saving the screenshot.)

Good luck with your new scope! I'm sure you'll like it as much as I like mine. It's a great unit for the price and its few minor bugs are easy to live with.


The easiest person to fool is yourself. -- Richard Feynman
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf