Why this oscillator does not work?

[bluetooth]

This is a Veroboard optimized oscillator. It is insensitive to parasitic inductances as the RF path around Q1 consists of only four elements. C1 and C2 can each be mounted next to the transistor. The parasitic inductance of C3 10nF doesn´t matter as the coil is in series with it. The oscillator has very high margin: It starts with 2,5V and for the Q-factor of the coil a minimum of 10 is sufficient.


---------- Post added at 10:21 ---------- Previous post was at 10:20 ----------

https://obrazki.elektroda.pl/76_1289033392.jpg

 

[FvM]

This is a Veroboard optimized oscillator.

Yes, it's the right way to do it. The placement of C3 and C2 does the trick, the circuit is considering the critical branches.

I remembered, that for high frequency oscillators, the C1/C2 voltage divider is usually replaced by a capacitor in parallel to L1 and a smaller C1 value. The variant gives correct feedback phase in combination with the inductive input impedance of the common base circuit.

 

[G0HZU]

This is a Veroboard optimized oscillator. It is insensitive to parasitic inductances as the RF path around Q1 consists of only four elements. C1 and C2 can each be mounted next to the transistor. The parasitic inductance of C3 10nF doesn´t matter as the coil is in series with it. The oscillator has very high margin: It starts with 2,5V and for the Q-factor of the coil a minimum of 10 is sufficient.


---------- Post added at 10:21 ---------- Previous post was at 10:20 ----------

https://obrazki.elektroda.pl/76_1289033392.jpg

I agree with this too. That way of arranging the circuit should make it very tolerant of PCB layout.

 

[efron]

Looking sharp at the new circuit bottomside, there's no feedback connected to the emitter node. It can't oscillate.

The RF feedback to emitter seems correct (look at the picture below).

Zeller, certainly, your connections should be shorter but even though it is still possible to have high frequency oscillations on a veroboard. Once I made a simple FM transmitter and it worked fine at 100Mhz.

Strait capacitances and inductances will decrease the final oscillation frequency. I assume that your board is just a prototype.

Now you have 500 ohms at the emitter and 10Kohms at collector. By using 500 ohms at the emitter, your emitter current in DC should be around 8mA (I assume you are still using a resistor divider at the base using the same values for both resistors).

With 8mA emitter current, the dynamic resistor of the BE junction is estimated to r(be) = 25/8 = 3 Ohms.

This BE junction resistor is practically in parallel with the 331pF capacitor (C3) - rf equivalent transistor circuit in small signal - At 27MHz, the 331pF represents about 18 ohms and the 3ohm BE junction resistor is acting like a shortcut to ground. I guess this is the main reason why it is not oscillating.

You should ALWAYS have r(be) bigger than the impendance of C3, for example something like 100 should be nice. But a good compromise it to have 1mA at emitter in DC, so I would suggest to replace the 500 Ohm at the emitter by one of 4K (once again).

This SHOULD WORK FINE!! because now you have a 10K at the collector.

KEEP TRYING, we are all with you

250mA.

 

[G0HZU]

The RF feedback to emitter seems correct (look at the picture below).

Zeller, certainly, your connections should be shorter but even though it is still possible to have high frequency oscillations on a veroboard. Once I made a simple FM transmitter and it worked fine at 100Mhz.

Strait capacitances and inductances will decrease the final oscillation frequency. I assume that your board is just a prototype.

Now you have 500 ohms at the emitter and 10Kohms at collector. By using 500 ohms at the emitter, your emitter current in DC should be around 8mA (I assume you are still using a resistor divider at the base using the same values for both resistors).

If Zeller really does want to make an idiotproof 27MHz oscillator on veroboard (it might not be Zellerproof) then build the circuit by bluetooth in post #41.



With 8mA emitter current, the dynamic resistance of the BE junction is estimated to r(be) = 25/8 = 3 Ohms.

This BE junction resistor is practically in parallel with the 331pF capacitor (C3) - rf equivalent transistor circuit in small signal - At 27MHz, the 331pF represents about 18 ohms and the 3ohm BE junction resistor is acting like a shortcut to ground. I guess this is the main reason why it is not oscillating.

You should ALWAYS have r(be) bigger than the impendance of C3, for example something like 100 should be nice. But a good compromise it to have 1mA at emitter in DC, so I would suggest to replace the 500 Ohm at the emitter by one of 4K (once again).

This SHOULD WORK FINE!! because now you have a 10K at the collector.

KEEP TRYING, we are all with you

Your post doesn't help. The board you are showing is for the awful circuit posted in post #29.

This oscillator circuit in post #29 should be forgotten because it is basically a bad design. Putting 10k in the collector will simply turn this circuit into a DC switch.

I won't comment on the rest of your post because it refers to the post #29 crcuit and this is a daft design for reasons I have given in previous posts.

If Zeller really does want to build a bulletproof 27MHz oscillator on veroboard then build the circuit in post #41 by bluetooth. However, I suspect the circuit won't prove to be Zellerproof...

 

[lakshminarayanakg]

Hi all,,,
I have one more Colpitts oscillator circuit, i have attached it you can change tank circuit C1/C2/L values to get desired frequency and also check the MAXIMUM frequency readable by your oscilloscope (Many CRO's read upto 1MHz only).

 

[efron]

Your post doesn't help. The board you are showing is for the awful circuit posted in post #29.

Hi Gohzu,

As you can expect, I do not agree with your statement. My point of view is just different from yours. You are suggesting a new-like design for a Colpitts oscillator while I am trying to understand why the last Zeller's circuit is still not working without suggesting a new design.

The board I was showing in my last post was referring to a previous post where it was said that there was no feedback connection to emitter. I only wanted to confirm that the feedback connections were OK.

On the other hand, I fully agree with you (and with other members as well) on the fact that Zeller's board design could be strongly improved for RF use. Nevertheless, I'm still not sure that this is the reason (the only reason) why he is not yet getting oscillations at the end.

Be certain that I also get profit of all the "good ideas" from all the posts of this thread, including yours, to improve my own knowledge.

Efron

 

[G0HZU]

The board I was showing in my last post was referring to a previous post where it was said that there was no feedback connection to emitter. I only wanted to confirm that the feedback connections were OK.

But you weren't looking at the same board. The comments about missing feedback weren't aimed at the original vero PCB. Like I said your comment was of no use.

As you can expect, I do not agree with your statement. My point of view is just different from yours. You are suggesting a new-like design for a Colpitts oscillator while I am trying to understand why the last Zeller's circuit is still not working without suggesting a new design.

Please, build the post #29 circuit yourself and put a 10k ohm resistor in the collector and see what happens to the DC voltage at the collector. Forget about making an oscillator, this circuit becomes a saturated DC switch.

You did some maths to calculate the emitter current so what do you think will happen to the DC voltage at the collector if you try and pull about 8mA through 10k ohm at the collector?
(You can't ever get 8mA emitter current if you put 10k in the collector anyway)

The circuit in post #29 is junk. It might be possible to get it to oscillate if you put the 1k ohm resistor back in the collector but it is such a poor design it should be buried forever in favour of a more conventional design.

As you can expect, I do not agree with your statement.

Do you agree with me now? Your (I say not useful) 10k + 500R circuit suggestions made an already poor design into a DC switch instead of an oscillator

 

[efron]

You did some maths to calculate the emitter current so what do you think will happen to the DC voltage at the collector if you try and pull about 8mA through 10k ohm at the collector?
(You can't ever get 8mA emitter current if you put 10k in the collector anyway)

Et voilà ! You're right.

I was too much focused thinking about oscillator's theory that I forgot to apply the basis of a good amplification phase. Sorry for that.

 

[bluetooth]

...and now I would like to see whether my circuit (post #41) is zellerprooved. Its minimum partcount, simulated with ansoft and microcap and contains 30 years oscillator building experience...
Zeller, please try.

 

[zeller]

...and now I would like to see whether my circuit (post #41) is zellerprooved. Its minimum partcount, simulated with ansoft and microcap and contains 30 years oscillator building experience...
Zeller, please try.

And what would be the frequency of oscilltion of that circuit?

---------- Post added at 20:47 ---------- Previous post was at 20:30 ----------

Please note I'm here to learn (and many have learnt a lot from this discourse i'm sure!!!) and this was my first circuit!!!

I got the formula for the original circuit (capacitors e.tc) from a book but probably the folmulas do not work at higher frequencies.I did not originally want to build anything different from my original circuit- because the method explained in the book was so thorough and nice!!!( thanks also to Efron's advice, really appreciated) but it looks like that circuit will not oscillate ! (I have in fact built the original circuit in which R2 =10k and R3= 500 ohm making sure everything was as close as possible but it did not oscillate!

Now for the benefit of all here and myself ,I will do this.

I will build the same circuit suggested by Ghonzu in the same style(i.e not on circuit board) as Gohzu suggested and also build another one as suggested by Bluetooth (will this oscillate to around 27MHZ blue tooth? ).

Whatever happens I will post here, I believe the best way to learn is by try and error, and we are not here to score points. I haven't given up guys.So watch this space!!

---------- Post added at 20:54 ---------- Previous post was at 20:47 ----------

Efron,
Yes I agree, the feedback connection was connected, and the math relating to the new resistors- which you correctly suggested will not work( in fact the circuit did not work in practice).

I will try out Ghonzu and Bluetooth Ideas now and let you all know the progress.

 

[bluetooth]

Without parasitics it will oscillate at 27,2MHz. If you decouple power via an 1pF cap at the collector you will get 27,0 MHz. With 10pf from collector to output you get 25,5MHz. If you connect a 10pf cap from the emitter to output jack you get 27,0MHz. If you dont use 50 Ohm load the frequencies will slightly differ.

 

[zeller]

I have made Bluetooth's design and from the results it looks like it works!!!HOOORAY Finally!! Thanks bluetooth for saving me from drink!!!

I will try Gohnzu's design later.I have attached the oscilloscope display itself and the circuit board. I'm using an old phillips pm 3160 35 mhz analogue scope which is difficult to calculate the exact frequency (can anyone help)- the waves are quite narrow and i could not ezpand them- but from my rough estimations it seems like a couple of MHZs.

I want to eventually attach a mixer (microchip) to transmit an AM signal. I have no idea of phase noise, stability etc and any other jargon which one has to correct before doing the above, so how do i check that?. The signal seems clear and stable on the scope to me.Will it work eg transmit sound anyway without worrying about phase noise etc?

The circuit is not perfect, there are some capacitor long legs but at least it works!! I plan to implement this on a proper circuit later.

Others, thanks again for your help.




PS: The scope picture was taken when the scope T/DIV was at 0.1 mu s(microsecond) and the Amp/DIV at 50mV.

 

[bluetooth]

Congratulations! I see on your picture a 555 timer circuit. Do you want to modulate a square wave tone? For this you don´t need a timer. You can change the values in the oscilator circuit and it generates some khz AM and the 27MHz simultaneously.

 

[zeller]

Congratulations! I see on your picture a 555 timer circuit. Do you want to modulate a square wave tone? For this you don´t need a timer. You can change the values in the oscilator circuit and it generates some khz AM and the 27MHz simultaneously.

Bluetooth,
I was not planning to use the square wave.
But I suppose I need a mixer, to mix the other signal with the 27mhz to produce AM. IS it possible to introduce another signal in the circuit by changing capicitor values on the collector without a mixer (gilbert cell)? How is this done?

 

[bluetooth]

I need to know what receiver do you want to use! Is it a CB receiver or a shortwave AM radio? Is it a SSB radio? Is it a wideband superregenerativ receicer? This is important as the frequency stability must match the receivers bandwith.

 

[zeller]

I need to know what receiver do you want to use! Is it a CB receiver or a shortwave AM radio? Is it a SSB radio? Is it a wideband superregenerativ receicer? This is important as the frequency stability must match the receivers bandwith.

I have built a basic am receiver with a couple of JFETS resistors etc with an LC design to receive 27 mhz.

 

[zeller]

I have got hold of a good digital scope but I'm quite worried about the low reading of Voltage VPP which is registering as 260mV this seems too small, is that possible?

 

[FvM]

Can clarify, which final circuit you're referring to, and how did you probe it, also what's the probe impedance?

 

[zeller]

Can clarify, which final circuit you're referring to, and how did you probe it, also what's the probe impedance?

I have used bluetooth circuit,

https://obrazki.elektroda.pl/76_1289033392.jpg

I have inserted the probe on the emmiter of transistor Q1 just before the 4700 resistor,I'm not sure about the probe impedance but it is a standard oscilloscope probe set at x1.