Re: Will this 25 MHZ oscillator work?
The oscillator shows considerable gain reserve in a LTSpice simulation, using common crystal parameters. I've also seen a lot of similar crystal oscillators working fine.
Here is the PCB layout with a buffer at the right I physically built it but got no signal!
The buffer has no power supply and the oscillator won't work if R4 is actually 470K instead of 470 ohm.
Do you show the PCB from the bottom (solder) side? If it's from the top side, BC547 has E and C reversed.
In any case, a measurement of DC levels would reveal, if the circuit has a suitable bias point.
Hint... It would work if you used the values that E-design used in the circuit he showed you :) changing values is a sure path to failure.
Did you figure out what FVM meant about the power supply to the buffer transistor ?
Hint --- look at where the negative rails are going to :)
have fun experimenting
cheers
Dave
You must be doing something wrong. Although I knew it was a waste of time, I took a few minutes at my bench and actually tested the circuit I posted.
I substituted your transistor with another general purpose 2N4401 I had around. The osc started without problems using crystals from 10-27MHz
I have used this circuit and its variations many times without any problems.
Thanks for your time.The resistor is okay, true value is 470 ohm, the transistor position should be okay(see attached), I think you are right power supply to buffer is missing although i doubt if that will prevent oscillations.Im using 27mhz crystal. See the attached.
yes but with no power to the buffer transistor you are not going to measure an output ! :)
Also you really do need to do some resoldering ... the cct is full of dry/bad joints
Im not suprised it doesnt work
do you have an experienced electronics friend nearby that can give you some teaching on how to solder correctly?
Dave
Attached is some notes I put together many years ago for such an oscillator. The 500 nH coil will depend on if you are trying for series mode or parallel mode. If series mode the resistor across the crystal should be low (1k) to prevent spurious modes.
The notes are for series resonant mode. You will need to supply the crystal parameters.
You can 'kick start' a simulation to speed up startup by putting a small resistor in series with crystal and creat a short burst from a generator at expected oscillation freq. upon beginning simulation run.
Zeller,
What transistor are you using in your circuit?
If it is a 2N5769 like the original circuit it is connected the wrong way.
Im using transistor BC548B, I have also connected power to the buffer transistor bu still no luck!
Well maybe you need someone with better electronic skills and test equipment to check it out for you.
Maybe the crystal is faulty. There will be a good reason why it will not work.
show us your modified cct with power to the buffer transistor
Dave
PC board outline now attached with power to buffer, I will rework the circuit from scratch though a modification to the existing one did not work.
ok here's the story....
built up the circuit on a breadboard, using your components, zeller. I didnt add the buffer stage, I just wanted to prove the oscillator. The BC548 worked ok up to 20MHz. But wouldnt work with the 24MHz xrystal I had (didnt have a 25MHz lying around). it just locked on the 8MHz fundamental of the crystal. Changing the transistor to a MPS2222 (plastic version of the metal cased 2N2222) it was the same result. Had a RF transistor, a 2SC2644 lying around and tried it. perfect 24MHz on the freq counter.
Lesson to be learned zeller ... When some one gives you a list of components to use as e-design did in his schematic, use them Dont change the values and then complain that the cct doesnt work ;)
you can see its just thrown together on the breadboard, no real respect for keeping legs of components short at 25MHz region, the trim cap is somethere around 15pF, something that was handy in my spare parts bin
and you can see the chrome telescopic whip from the output 100nF capacitor to the freq counter
cheers
Dave
As expectable. A overtone crystal won't easily oscillate at it's intended frequency without using a resonant circuit, or at least a frequency selective means supressing the fundamental.
24 MHz fundamental mode, parallel resonant crystals are readily available. I am surprised to see a third overtone crystal at 24 MHz since it takes more quartz (a thicker blank) then a fundamental at 24 MHz would. Many GSM cellphones use a 26 MHz fundamental mode, 10 pF load crystal.
You need the inductor in series with crystal for overtone series resonance operation along with about 1k resistor across crystal to avoid it starting up on fundamental. (need series coupling cap to avoid shunting base bias with resistor across crystal). The value depends on crystal Lm and osc feedback caps values, but is probably in 1.2 to 2.0 uH range for 24 MHz. You can also put a choke in series with emitter resistor to reduce the loading of the oscillator by emitter bias resistor.
The feedback cap that you have going from emitter to ground is better connected to collector instead of ground which gives it a direct path that doesn't have to go through your rather long path collector bypass cap.
agreed, but it did work quite well with a change in transistor type to an RF one :) as can be seen on the freq counter.
I suspect the BC548 just didnt wanna work too well in that cct and as said, zeller should have stuck to the design he was given earlier in the thread :)
My main purpose was just to prove that the basic circuit as it stood works, that was a success
Dave
Absolutely. I think however, that using an overtone crystal imposes an unnecessary complication to the design. Inductorless crystal oscillators are quite common for lower frequencies, but they don't use overtone crystals as far as I'm aware of. With a fundamental crystal, the circuit works well with general purpose transistors like BC548.
I didnt specifically/purposely choose an overtone crystal, I just pulled a few crystals of various freq's out of the component drawer. I was suprised to see it oscillating happily at its fundamental 8MHz with the BC548. If I get the urge I may include the series inductor and parallel resistor and see what happens with the BC548
I guess the thing is we dont really know what kind of crystal zeller is using in his osc. that he's been trying to make work.
maybe he's also happened to have a overtone crystal and discovering like I did that it wont oscillate in that basic cct with a BC548 without the use of the series inductor etc
All indications are so far of a lack of basic understanding of soldering technique, circuit layout etc :(
There's times when I would like to be able to just spend real time with a beginner like him and give several hours on hands on basic training :) They would go on to have much more success in their electronics endevours
cheers
Dave