Need 45 MHZ xtal oscillator circuit

[GizmoWizard]

I've been looking for (with no luck) an xtal oscillator circuit for a 45 MHz crystal: AT-cut, fundamental.

I can only find third-overtone circuits for this frequency. Can someone help please?

 

[JimB]

Why do you want such a circuit?

Most crystals intended for 45MHz are third overtone types because a fundamental mode crystal operating at that frequency would be rather small and fragile.

JimB

 

[GizmoWizard]

Hi, thanks for your response.

I'm building an up-conversion shortwave receiver and need a 44.545 MHZ oscillator for the second mixer.

All of the available crystals from Mouser or Digi-Key for that frequency are fundamental mode.

 

[Diver300]

Well the big difference between a 3rd overtone circuit and a fundamental circuit is that the 3rd overtone circuit contains a high-pass filter to stop the fundamental mode happening. So if you have a 3rd overtone 44.545 MHz circuit, it will be filtered to stop the crystal oscillating at around 15 MHz, the fundamental frequency.

As the 44.545 MHz fundamental crystals won't have any resonance lower than 44.545 MHz, it isn't necessary to prevent lower frequency oscillation, but it doesn't hurt to prevent it and the circuit should just work.

Just remember that a 44.545 MHz crystal is around 37 μm thick, and is made out of quartz, which is about as brittle as glass. The crystal can only be supported at its edges, so you are not talking about a robust item.

The Mouser / Digikey part is made by CTS, and has a load capacitance of 13 pF, and a maximum drive level of 300 μW. The ESR is anything up to 30 Ω. To keep below 300 μW at 30 Ω, the current must be less than around 3 mA. If the load is 13 pF, that means that the crystal impedance is around 300 Ω, and you should keep the voltage across the crystal to 0.8 V rms (2.33 V peak-peak).

It would be better to keep the power down to 100 μW, and that calculation depends on the load capacitance being correct. If your circuit's load capacitance is different, the frequency will be slightly off, and there will be a different impedance.

 

[MikeMl]

Crystal oscillators for

Well the big difference between a 3rd overtone circuit and a fundamental circuit is that the 3rd overtone circuit contains a high-pass filter to stop the fundamental mode happening. So if you have a 3rd overtone 44.545 MHz circuit, it will be filtered to stop the crystal oscillating at around 15 MHz, the fundamental frequency.....

I have always been under the impression that the filter required to prevent an overtone crystal from oscillating on its fundamental must be added to the oscillator circuit (is not in the crystal can). My evidence for this is that if I plug a third or fifth overtone crystal into the "crystal tester" built-in to my frequency counter, it shows the fundamental frequency, not what is stamped on the can. e.g., a 45MHz crystal would show about 15MHz when tested...

If you look at the detailed data sheets for the ~45MHz crystals on DigiKey, you will find that the data sheet implies that they are third overtone, not fundamental.

 

[Diver300]

Crystal oscillators for


I have always been under the impression that the filter required to prevent an overtone crystal from oscillating on its fundamental must be added to the oscillator circuit (is not in the crystal can). My evidence for this is that if I plug a third or fifth overtone crystal into the "crystal tester" built-in to my frequency counter, it shows the fundamental frequency, not what is stamped on the can. e.g., a 45MHz crystal would show about 15MHz when tested...

If you look at the detailed data sheets for the ~45MHz crystals on DigiKey, you will find that the data sheet implies that they are third overtone, not fundamental.

I would agree with all of that, except that several very small surface mount crystals up to 50 MHz or so, and the only 44.545 MHz crystal at DigiKey, are fundamentals:-

https://www.digikey.com/product-detail/en/cts-frequency-controls/406C35B44M54500/CTX678TR-ND/604571

At 48 MHz, Digikey have 417 Fundamental, and 17 3rd overtone crystal in stock, although many of those may be repeated due to different packaging etc. and I haven't looked in detail, I've just applied the filters on the DigiKey website for the oscillation modes.

It seems that in the very small packages, the upper frequency limit for fundamental crystals has been raised quite a lot in recent years.

 

[GizmoWizard]

Gentlemen, thank you for the information.

The CTS crystal that I have has a power rating of 10 μW typ, 100 μW max. So I'm aiming for 50 μW to ease up on the gain/noise of the post-amp. The ESR is 40 ohm and the shunt capacitance is 4 to 7 pF.

So I can just take a standard Colpitts oscillator, for example, plug in the formulas for the two capacitors, and limit the voltage across the crystal to maintain a 50 μW drive level?

 

[Diver300]

Gentlemen, thank you for the information.

The CTS crystal that I have has a power rating of 10 μW typ, 100 μW max. So I'm aiming for 50 μW to ease up on the gain/noise of the post-amp. The ESR is 40 ohm and the shunt capacitance is 4 to 7 pF.

So I can just take a standard Colpitts oscillator, for example, plug in the formulas for the two capacitors, and limit the voltage across the crystal to maintain a 50 μW drive level?

Yes.

Please share the circuit.

At 4 - 7 pF, the impedance will be quite high so the permitted voltage will be larger.

 

[GizmoWizard]

I don't have an oscillator circuit to share just yet, as that was the impetus for this thread. If you hadn't reminded me of the crystal drive level I probably would have melted the quartz

As soon as I get something working or an idea, I'll drop it in here.

 

[Diver300]

You don't melt quartz, you crack it.

Quartz crystal are mechanical resonators. Their resonant frequency is almost entirely due to their physical properties, the mass and the stiffness. They are adjusted in production by adding mass, and the electronic circuit can only make tiny changes to the frequency.

If you overdrive a crystal, it moves too much and cracks or suffers effects from the movement being too large.

 

[GizmoWizard]

I was joking about melting the quartz.
Back when I was designing amateur radio equipment for a living, whenever I blew a power transistor someone would inevitably yell "make sure you clean that silicon off the floor."

 

[BobW]

Is there any reason why you don't just buy a 45 MHz packaged crystal oscillator?

 

[GizmoWizard]

Hi,

I need a 44.545 MHz signal to drive a quad-ring mixer in a dual- conversion receiver. If I could find a crystal oscillator module at that frequency I'd snap it up.

 

[Mikebits]

Have you considered using one of the Analog Device DDS chips, programing them is very easy. I been using the AD9833 device. They take very little space and programming is just over simple SPI interface. I assume your receiver has a microcontroller onboard. Well just a thought.

https://www.analog.com/media/en/technical-documentation/data-sheets/AD9833.pdf

You can get a premade module from ebay that would just plug into what ever pcb you have for cheap.
**broken link removed**
**broken link removed**

The module has on onboard programmable PGA with a digital pot for setting gain of the output.

 

[Mikebits]

Here is how I interfaced the Module with a microcontroller, just 4 wires.

 

[MikeMl]

Or use the "value-added custom field-programmable oscillator"service that DigiKey offers. For a small additional fee, DigiKey will deliver any of **broken link removed** to the nearest frequency possible to your specification.

 

[GizmoWizard]

Wow! Many thanks for the time and effort everyone has taken to help me with this. I've got a lot of choices to consider.

I'm embarrassed that I wasn't even aware of the existence of Digi-Key's preprogrammed oscillator service. That looks very promising, as does several of the other options on this thread.

Again, thanks everyone.

 

[MikeMl]

...I'm embarrassed that I wasn't even aware of the existence of Digi-Key's preprogrammed oscillator service. That looks very promising, as does several of the other options on this thread...

Be aware that a lot of these programmable oscillators have too much phase noise to be used as a Local Oscillator, especially if you are multiplying the frequency before using the multiplied freq for LO injection...

 

[Mikebits]

Or use the "value-added custom field-programmable oscillator"service that DigiKey offers. For a small additional fee, DigiKey will deliver any of **broken link removed** to the nearest frequency possible to your specification.

I did not know of this service. Thanks Mike.

 

[GizmoWizard]

I will not be multiplying the output, 44.545 MHz will be the LO injection frequency. Low jitter and phase noise will be important.

Since phase noise is not listed in the data sheets, do you have any suggestions on narrowing down the large number of available devices?