Maker Pro
Maker Pro

oscillator doesn't oscillator as I expect

long

Mar 31, 2014
19
Joined
Mar 31, 2014
Messages
19
oscillator doesn't oscillate as I expect

Hello:

I am trying to test a simple oscillator to see if I can get it to work. It is based on a LC tank and uses one transistor. It uses a 10pf capacitor to turn on or off the transistor.

I was expecting it to oscillate at the 1/(2*pi*sqrt(L * C), that is about 160K HZ. However, I saw from the oscilloscope it is oscillating at about 93 MHz. Actually its oscillating frequency is not that stable. I think it might be because my scope probe loads the LC tank. The scope probe has about 90 pf.

My major question is why it doesn't work at 160K, but instead working on a much higher frequency?

Also, I found this circuit is sensitive to the base transistor, which is right now 100K. If I use a small value like 1K, the oscillation will disappear. The emitter resistor is 330, if I make it 100, the oscillation amplitude would be much smaller. So overall my feeling is that I cannot drive the transistor too hard, which may overwhelm the LC tank or heavily curb it.

Any help/discussion/point is very welcome :) I come here to learn!

have a nice day!
 

Attachments

  • oscillator_schema.jpg
    oscillator_schema.jpg
    18.7 KB · Views: 191
  • oscillator_circuit.jpg
    oscillator_circuit.jpg
    37.2 KB · Views: 174
Last edited:

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
You must have a very good scope to see 93MHz.

The high frequency oscillation is due to the coupling loops of your innovative construction.

Circuits such as this usually have a capacitor across the emitter resistor to give a low impedance drive to the transistor. Try 100pF to 1000pF
 

Arouse1973

Adam
Dec 18, 2013
5,178
Joined
Dec 18, 2013
Messages
5,178
Are you sure that's a 4.7mH inductor. Looks like a small choke to me. The circuit configuration shown is normally used for high frequency and not 160Khz.
Adam
 

long

Mar 31, 2014
19
Joined
Mar 31, 2014
Messages
19
sure, will try, and will report back what i see.



You must have a very good scope to see 93MHz.

The high frequency oscillation is due to the coupling loops of your innovative construction.

Circuits such as this usually have a capacitor across the emitter resistor to give a low impedance drive to the transistor. Try 100pF to 1000pF
 

long

Mar 31, 2014
19
Joined
Mar 31, 2014
Messages
19
I know little about the difference between choke and inductor. I bought the parts from amazon which is listed as "inductor". I have cut open one to see, I see the wire is reaaly thin.

I think you are correct, they are "choke" which is not supposed to be used in LC circuits, but good for dissipating AC energy.

I will wind a coil to test this circuit again.

thank you!


Are you sure that's a 4.7mH inductor. Looks like a small choke to me. The circuit configuration shown is normally used for high frequency and not 160Khz.
Adam
 

long

Mar 31, 2014
19
Joined
Mar 31, 2014
Messages
19
They are indden 4.7mh, I have measured it multiple times.

Question, is this kind of choke/inductor really easy to get saturated? Or, what is the ball part estimation of when they get saturated? The current in the circuit pass it is about several milli ampers.

Yes, most likely 4.7uH

Bob
 

long

Mar 31, 2014
19
Joined
Mar 31, 2014
Messages
19
right, i did not invent this circuit, but learned it from a FM radio circuit. I want to see how the LC tank is supposed to work here. I have a feeling that its frequency is more decided by the 10p feedback capacitor, other than the L c tank.

I also tried 100pf and 1000pf feedback capacitors, 100pf gives me smaller oscillation amplitude, and 1000pf completely kills the oscilation.


Are you sure that's a 4.7mH inductor. Looks like a small choke to me. The circuit configuration shown is normally used for high frequency and not 160Khz.
Adam
 

Arouse1973

Adam
Dec 18, 2013
5,178
Joined
Dec 18, 2013
Messages
5,178
Nope, the LC combination is what sets the frequency and the other capacitor is what controls the common base configured transistor drive. You wouldn't normally use a choke for this, the name gives it away a bit CHOKE. I think the type you have used are designed to have a low Q which is normally the opposite of what you want for a tuned circuit. Try winding an air core coil to start with . Use insulated wire, either plastic coated Kynar type or enamelled.
 

long

Mar 31, 2014
19
Joined
Mar 31, 2014
Messages
19
LC oscillator does not give me the expected frequency

I wound a coil, which is 70 turns and I measure it is about 120uh. So I expect it is to be resonance freq of capacitor of 220pf and inductor of 120uh, which is 980k hz.

But as my pictures show, the freq is instead of 80.65M hz. The peak-to-peak voltage of oscillation is 3.68v, and max is 7.36v, the power supply is 4 AA batts of about 5.2V total.

Any thoughts? thank you.

Nope, the LC combination is what sets the frequency and the other capacitor is what controls the common base configured transistor drive. You wouldn't normally use a choke for this, the name gives it away a bit CHOKE. I think the type you have used are designed to have a low Q which is normally the opposite of what you want for a tuned circuit. Try winding an air core coil to start with . Use insulated wire, either plastic coated Kynar type or enamelled.
 

Attachments

  • osc_1.jpg
    osc_1.jpg
    44.3 KB · Views: 147
  • osc_2.jpg
    osc_2.jpg
    67.8 KB · Views: 150
Last edited:

Arouse1973

Adam
Dec 18, 2013
5,178
Joined
Dec 18, 2013
Messages
5,178
I don't really know what this issue is but I think you might need a lower base resistor for extra drive. Here is one I just designed at it oscillates at about 956KHz.
Adam
 

Attachments

  • OSC.PNG
    OSC.PNG
    4.7 KB · Views: 184

long

Mar 31, 2014
19
Joined
Mar 31, 2014
Messages
19
Cool.

Actually I thought about my circuit, I realized that the 10p feedback capacitor will present a pretty large reactance to the frequency I wanted it to oscillate.

The first frequency I want is : for 160K HZ, a 10p capacitor has about 100K reactance; for 1M Hz, it has about 16K.

Looking at your working circuit, I increasingly believe that my problem is that I expected it to run at a frequency too low while no enough feedback. Your employs a 220p feedback capacitor!

I will try again!

And also extremely thank you for sharing your working circuit! :D

I don't really know what this issue is but I think you might need a lower base resistor for extra drive. Here is one I just designed at it oscillates at about 956KHz.
Adam
 

long

Mar 31, 2014
19
Joined
Mar 31, 2014
Messages
19
Here is what I am now:

First modification: I did not change my circuit, but I cut short the leads of the 220pf C in the LC tank, I soldered the air coil as close to the C as possible. I note that the oscillation frequency now becomes 50M Hz. This gave me an impression that the lead length of the C does have some effect. But there is just so much can be do done about cutting leads short.

Second modifitcation: I almost literally copied your circuit, replace the 10pf feedback C with a 220pf, and arrange the base transistors to be a voltage divider, the top one is 16k, the bottom one is 6.8k. Now the oscillation disappears.

There are still some difference between yours and mine:

1. you uses a 2N3904, but mine is 2N2222. While it is hard to me to think 2N2222 cannot handle 1M oscillation, while it actually had worked on 80MHz and 50MHz.

2. your supply voltage is 12V, mine is 5V. I have tried 12V too, but still no luck.

Seems this thing has gone into a dead end :(

I also have some 2N3906 PNP, I might give that another try.

I don't really know what this issue is but I think you might need a lower base resistor for extra drive. Here is one I just designed at it oscillates at about 956KHz.
Adam
 

Arouse1973

Adam
Dec 18, 2013
5,178
Joined
Dec 18, 2013
Messages
5,178
Hey Long don't panic.

I tried your circuit and got the same results as you. It was oscillating at high frequency. This circuit as I mentioned earlier is used for small FM transmitters. I remember years ago trying to get one of those to work and I did but with only a few turns of wire an a couple of pF of capacitance it oscillated at 100MHz.

Anyway I added some decoupling across the supply and it stopped altogether is must have been using the connections as stray components. So what I did is giggle it about a bit and designed a new one which oscillates really well. Here it is with results.
Adam
 

Attachments

  • COLOSC.PNG
    COLOSC.PNG
    8.8 KB · Views: 158
  • CIMG8423.JPG
    CIMG8423.JPG
    25.4 KB · Views: 169
  • CIMG8426.JPG
    CIMG8426.JPG
    16.7 KB · Views: 156

long

Mar 31, 2014
19
Joined
Mar 31, 2014
Messages
19
Arouse1973,

thank you again for sharing your experience. It is important for me to know that what I run into is not something so unique or my components are so out of whack! :)

From your pictures, looks like it is a Colpitts oscillator. And you use two inductors. One is used in the oscillator circuit, another one works with a capacitor to form a LC tank. While the second inductor is put close to the first inductor in the oscillator to coupling out the output signal which is what we see in the middle picture.

So I assume the two inductors are identical, and what is the value of the capacitor in the LC tank which is connected to the oscilloscope?

Thank you.
 

Arouse1973

Adam
Dec 18, 2013
5,178
Joined
Dec 18, 2013
Messages
5,178
Yes correct I only used the pickup coil as to not load the circuit with my scope. Any old coil will probably do. Even connecting the scope clip to the tip and drape the wire over the coil should pick up something.
Adam
 

long

Mar 31, 2014
19
Joined
Mar 31, 2014
Messages
19
Hi Arouse1973:

I invested a bit more time on this, just tried:

1. to eliminate the spurious high-freq oscillation. Which, I assume is from a LC tank formed by the 220pf capacitor and a wire, so for 80MHz, the wire should have about 17nH. Ths number "17nH "is about in the ball park range of a one-inch wire.

2. to see if I still can get the old circuit to work.

So I re-wire a circuit, but this time I use a 20pf capacitor instead of 220pf to form a LC tank with the air coil. And again I make the critical path from the capacitor to the 2n2222 as compact/short as possible. If this 20pf wants to interact with some wire to oscillate, that frequency will be out of the working range of 2n2222, so it will be effectively suppressed.

Also I make a simple buffer stage to make measuring the frequency easy. Directly attaching the scope probe to the LC tank or using a coil nearby to coupling out signal, looks like not working well for me yet, too much noise and distortion.

The resulting result is not exactly what I want: it is supposed to be 6.3MHz(20pf and 32hH), but I got 4.42MHz. But I am glad the high-freq is gone, and somehow relaxed the LC tank is working.

I am not sure the number of "20pf" is accurate or not, and 32uH is what I measured from an inductor meter.

So my remaining questions are:

1. if an inductor meter tells me an air coil is 32uH, is it really 32uH when working @ 4.42MHz?

2. perhaps the 20pf is not really 20pf, I don't want to remove it from the circuit to test it now... But may try to measure a capacitor from same batch to see how much off it is. But measuring 20pf might prove useless as my meter usually show some non-zero number even it is measuring nothing, it might be capacitance of my hands or leads of the probe.







Yes correct I only used the pickup coil as to not load the circuit with my scope. Any old coil will probably do. Even connecting the scope clip to the tip and drape the wire over the coil should pick up something.
Adam
 

Attachments

  • new_osc_1.jpg
    new_osc_1.jpg
    19.8 KB · Views: 140
  • new_osc_2.jpg
    new_osc_2.jpg
    41 KB · Views: 138
  • new_osc_3.jpg
    new_osc_3.jpg
    67.4 KB · Views: 122
Last edited:

Arouse1973

Adam
Dec 18, 2013
5,178
Joined
Dec 18, 2013
Messages
5,178
Glad you are making progress. An inductor will change it inductance as the current distribution changes due to skin effect. But at those frequencies and the small diameter wire you are using I doubt it will change much. It's more likely to be the capacitor. What type of capacitor are you using.
Adam
 

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
Inductors are measured at low frequencies so that the effect of the inductor capacitance is minimised. At some highish frequency the inductor will be self resonant and above this, will behave as a capacitor.

Hence the effective inductance will be less than the measured inductance.
 

Arouse1973

Adam
Dec 18, 2013
5,178
Joined
Dec 18, 2013
Messages
5,178
Surely this is only to improve the measurement side of things. So are you saying that the parasitic capacitance effect the actual inductance of the coil? For the inductance to change surely the flux distribution would need to change.

Adam
 
Top