The Well Tempered Master Clock - Building a low phase noise/jitter crystal oscillator

[esgigt]

- andrea_mori : 2 x 5.6448MHz + 2 x 11.2896MHz + 4 x PCB
- Eldam : 1 x 11.2896 Mhz + 4 x PCB
- esgigt : 1 x 11.2896 Mhz + 1 x PCB
- fralippo : 1 x 22.5792 MHz and 1 x 24.576 MHz + 2 x PCB
- mravinsky : 2 x 11.2896 Mhz + 1 x PCB
- TNT 1x22.5792MHz 1x24.5760MHz + 2 x PCB, soldered if possible
-1audio : 1x22.5792MHz 1x24.5760MHz + 2 x PCB

I'd like mine soldered if possible too.

 

[HpW]

Just one suggestion in case you do ever a redesign:

There are newer logic on the sky since the old decade HC & AC series! Also there are now chip's with single inverter as from NXP - 74HC1GU04GW and also LVC series (just an example) . The benefit are shorter power connection and no cross power noise from other loads.

Important is also the mechanical shocks & vibration isolation for freq. stability... just google to get some information...

just my 2 cents

Hp

 

[Eldam]

Do you remeber those old time when diyers soldered the can body to the Gnd with a wire and puted Blue Tack on the body to avoid vibrations....😀

 

[esgigt]

I'd like mine soldered if possible too.

- andrea_mori : 2 x 5.6448MHz + 2 x 11.2896MHz + 4 x PCB
- Eldam : 1 x 11.2896 Mhz + 4 x PCB
- esgigt : 1 x 11.2896 Mhz + 1 x PCB, soldered if possible
- fralippo : 1 x 22.5792 MHz and 1 x 24.576 MHz + 2 x PCB
- mravinsky : 2 x 11.2896 Mhz + 1 x PCB
- TNT 1x22.5792MHz 1x24.5760MHz + 2 x PCB, soldered if possible
-1audio : 1x22.5792MHz 1x24.5760MHz + 2 x PCB

 

[andrea_mori]

Unfortunately going to 24.xx or 48.xx MHz crystals will require overtone operation with a more complex circuit. If there is a source for suitable crystals it is better to use one at the target frequency. The phase noise will be higher on a higher frequency crystal but less than a multiplier. Multipliers are simple but need a transformer and low noise Shottky diodes usually. A Minicircuits multiplier is cost effective: Mini-Circuits but not real cheap and then you still need to convert to LVTTL.

If you want low close in phase noise you can phase lock a higher frequency oscillator (49.152 MHz) to a lower noise low frequency oscillator (e.g. 6.144 MHz). You can get the low close in phase noise of the lower frequency with the higher actual frequency. Still more complexity but that's the price of performance.

The oscillator circuit for overtone crystals would need a suppressor (usually just an inductor) to ensure the crystal oscillates at the right frequency. The circuit would need to be tweaked to make sure the power in the crystal is not excessive. To make the circuit adjustable I think you could replace C8 with a varactor diode so you can pull the frequency a little.

Hi Demian,

thanks for the suggestions.
You are right, I abandon the idea of the multiplier. If there will be a daughter board, it will provide the dividers only and the switch between two XO.

I'm not sure that a single inductor will be enough to make this oscillator compliant with 3rd overtone crystal, anyway I will give it a try and eventually I'll update the PCB to accommodate the inductor. Moreover 3rd overtone crystals have usually higher Q and lower motional capacitance, so would be a good thing using them also at lower frequency.

Do you know if anyone has designed a PCB for the regulator you posted the link?

Andrea

 

[andrea_mori]

Just one suggestion in case you do ever a redesign:

There are newer logic on the sky since the old decade HC & AC series! Also there are now chip's with single inverter as from NXP - 74HC1GU04GW and also LVC series (just an example) . The benefit are shorter power connection and no cross power noise from other loads.

Important is also the mechanical shocks & vibration isolation for freq. stability... just google to get some information...

just my 2 cents

Hp

I would also add the movement of air, so the best way would be to put the oscillator in a box (iron better than aluminum).
But I would leave everyone free to choose his own way.

 

[esgigt]

I would also add the movement of air, so the best way would be to put the oscillator in a box (iron better than aluminum).
But I would leave everyone free to choose his own way.

Which tends to the addition of a temperature controlled oven???? 🙄

 

[andrea_mori]

Which tends to the addition of a temperature controlled oven???? 🙄

No, AT-cut crystal does not need an oven. The box acts as a shield against air motion and RF.
Air motion could affect the phase noise very close to the carrier, less than 10 Hz.
Personally, I prefer to place the oscillator as closest as possible to the DAC, rather than close it in a box using relatively long cable.

 

[randytsuch]

- andrea_mori : 2 x 5.6448MHz + 2 x 11.2896MHz + 4 x PCB
- Eldam : 1 x 11.2896 Mhz + 4 x PCB
- esgigt : 1 x 11.2896 Mhz + 1 x PCB, soldered if possible
- fralippo : 1 x 22.5792 MHz and 1 x 24.576 MHz + 2 x PCB
- mravinsky : 2 x 11.2896 Mhz + 1 x PCB
- TNT 1x22.5792MHz 1x24.5760MHz + 2 x PCB, soldered if possible
-1audio : 1x22.5792MHz 1x24.5760MHz + 2 x PCB
-randytsuch 2x25.0000MHz + 4 x PCB

 

[TNT]

No, AT-cut crystal does not need an oven.

Then you need to change the name 🙂

//

 

[1audio]

Hi Demian,

thanks for the suggestions.
You are right, I abandon the idea of the multiplier. If there will be a daughter board, it will provide the dividers only and the switch between two XO.

I'm not sure that a single inductor will be enough to make this oscillator compliant with 3rd overtone crystal, anyway I will give it a try and eventually I'll update the PCB to accommodate the inductor. Moreover 3rd overtone crystals have usually higher Q and lower motional capacitance, so would be a good thing using them also at lower frequency.

Do you know if anyone has designed a PCB for the regulator you posted the link?

Andrea

Most of the overtone oscillators I have looked at seem to get away with a single coil. You just need to suppress the undesired modes enough to have it only oscillate on the desired mode. The links I posted have some examples.

I have used the regulator circuit in a few applications and have provided the piece of layout with both the regulator and the associated oscillators if that helps. I would do a new design using SMT, the old ones were through hole and used a lot of real estate unnecessarily. I have attached the relevant portion of the circuits. This would be used with either DIP oscillators or the Crystek surface mount oscillators, or SMA cables from external oscillators (where the regulators are unnecessary. . .) The outputs are transformer isolated so ground noise etc. won't get back into the circuits. Its in ExpressPCB format. It all works quite well.

 

[TNT]

- andrea_mori : 2 x 5.6448MHz + 2 x 11.2896MHz + 4 x PCB
- Eldam : 1 x 11.2896 Mhz + 4 x PCB
- esgigt : 1 x 11.2896 Mhz + 1 x PCB, soldered if possible
- fralippo : 1 x 22.5792 MHz and 1 x 24.576 MHz + 2 x PCB
- mravinsky : 2 x 11.2896 Mhz + 1 x PCB
- TNT 1x22.5792MHz 1x24.5760MHz + 2 x PCB, soldered if possible
- 1audio : 1x22.5792MHz 1x24.5760MHz + 2 x PCB
- randytsuch 2x25.0000MHz + 4 x PCB
- TNT or maybe 1 x 11.2896 Mhz + 1 x PCB, soldered if possible

 

[Myint67]

- andrea_mori : 2 x 5.6448MHz + 2 x 11.2896MHz + 4 x PCB
- Eldam : 1 x 11.2896 Mhz + 4 x PCB
- esgigt : 1 x 11.2896 Mhz + 1 x PCB, soldered if possible
- fralippo : 1 x 22.5792 MHz and 1 x 24.576 MHz + 2 x PCB
- mravinsky : 2 x 11.2896 Mhz + 1 x PCB
- TNT 1x22.5792MHz 1x24.5760MHz + 2 x PCB, soldered if possible
- 1audio : 1x22.5792MHz 1x24.5760MHz + 2 x PCB
- randytsuch 2x25.0000MHz + 4 x PCB
- TNT or maybe 1 x 11.2896 Mhz + 1 x PCB, soldered if possible
-myint67 1x11.2896mhz + 1pcb also soldered if possible

 

[kf_tam]

- andrea_mori : 2 x 5.6448MHz + 2 x 11.2896MHz + 4 x PCB
- Eldam : 1 x 11.2896 Mhz + 4 x PCB
- esgigt : 1 x 11.2896 Mhz + 1 x PCB, soldered if possible
- fralippo : 1 x 22.5792 MHz and 1 x 24.576 MHz + 2 x PCB
- mravinsky : 2 x 11.2896 Mhz + 1 x PCB
- TNT 1x22.5792MHz 1x24.5760MHz + 2 x PCB, soldered if possible
- 1audio : 1x22.5792MHz 1x24.5760MHz + 2 x PCB
- randytsuch 2x25.0000MHz + 4 x PCB
- TNT or maybe 1 x 11.2896 Mhz + 1 x PCB, soldered if possible
-myint67 1x11.2896mhz + 1pcb also soldered if possible
-EdwardTam: 2x 16.9344MHz + 1x PCB

 

[BDL]

- andrea_mori : 2 x 5.6448MHz + 2 x 11.2896MHz + 4 x PCB
- Eldam : 1 x 11.2896 Mhz + 4 x PCB
- esgigt : 1 x 11.2896 Mhz + 1 x PCB, soldered if possible
- fralippo : 1 x 22.5792 MHz and 1 x 24.576 MHz + 2 x PCB
- mravinsky : 2 x 11.2896 Mhz + 1 x PCB
- TNT 1x22.5792MHz 1x24.5760MHz + 2 x PCB, soldered if possible
- 1audio : 1x22.5792MHz 1x24.5760MHz + 2 x PCB
- randytsuch 2x25.0000MHz + 4 x PCB
- TNT or maybe 1 x 11.2896 Mhz + 1 x PCB, soldered if possible
-myint67 1x11.2896mhz + 1pcb also soldered if possible
-EdwardTam: 2x 16.9344MHz + 1x PCB
- BDL : 1 x 11.2896 MHz + 2 x PCB

 

[thorstenlarsen]

- andrea_mori : 2 x 5.6448MHz + 2 x 11.2896MHz + 4 x PCB
- Eldam : 1 x 11.2896 Mhz + 4 x PCB
- esgigt : 1 x 11.2896 Mhz + 1 x PCB, soldered if possible
- fralippo : 1 x 22.5792 MHz and 1 x 24.576 MHz + 2 x PCB
- mravinsky : 2 x 11.2896 Mhz + 1 x PCB
- TNT : 1x22.5792MHz 1x24.5760MHz + 2 x PCB, soldered if possible
- 1audio : 1x22.5792MHz 1x24.5760MHz + 2 x PCB
- randytsuch : 2x25.0000MHz + 4 x PCB
- TNT or maybe : 1 x 11.2896 Mhz + 1 x PCB, soldered if possible
- myint67 : 1x11.2896mhz + 1pcb also soldered if possible
- EdwardTam : 2x 16.9344MHz + 1x PCB
- BDL : 1 x 11.2896 MHz + 2 x PCB
- thorstenlarsen : 2 x 11.2896 MHz + 2 x PCB

 

[Zoran]

- andrea_mori : 2 x 5.6448MHz + 2 x 11.2896MHz + 4 x PCB
- Eldam : 1 x 11.2896 Mhz + 4 x PCB
- esgigt : 1 x 11.2896 Mhz + 1 x PCB, soldered if possible
- fralippo : 1 x 22.5792 MHz and 1 x 24.576 MHz + 2 x PCB
- mravinsky : 2 x 11.2896 Mhz + 1 x PCB
- TNT : 1x22.5792MHz 1x24.5760MHz + 2 x PCB, soldered if possible
- 1audio : 1x22.5792MHz 1x24.5760MHz + 2 x PCB
- randytsuch : 2x25.0000MHz + 4 x PCB
- TNT or maybe : 1 x 11.2896 Mhz + 1 x PCB, soldered if possible
- myint67 : 1x11.2896mhz + 1pcb also soldered if possible
- EdwardTam : 2x 16.9344MHz + 1x PCB
- BDL : 1 x 11.2896 MHz + 2 x PCB
- thorstenlarsen : 2 x 11.2896 MHz + 2 x PCB
- Zoran : 1x22.5792MHz 1x24.5760MHz + 2 x PCB, (soldered or not...)

 

[walangalam]

- andrea_mori : 2 x 5.6448MHz + 2 x 11.2896MHz + 4 x PCB
- Eldam : 1 x 11.2896 Mhz + 4 x PCB
- esgigt : 1 x 11.2896 Mhz + 1 x PCB, soldered if possible
- fralippo : 1 x 22.5792 MHz and 1 x 24.576 MHz + 2 x PCB
- mravinsky : 2 x 11.2896 Mhz + 1 x PCB
- TNT : 1x22.5792MHz 1x24.5760MHz + 2 x PCB, soldered if possible
- 1audio : 1x22.5792MHz 1x24.5760MHz + 2 x PCB
- randytsuch : 2x25.0000MHz + 4 x PCB
- TNT or maybe : 1 x 11.2896 Mhz + 1 x PCB, soldered if possible
- myint67 : 1x11.2896mhz + 1pcb also soldered if possible
- EdwardTam : 2x 16.9344MHz + 1x PCB
- BDL : 1 x 11.2896 MHz + 2 x PCB
- thorstenlarsen : 2 x 11.2896 MHz + 2 x PCB
- Zoran : 1x22.5792MHz 1x24.5760MHz + 2 x PCB, (soldered or not...)
- walangalam : 1 x 11.2896 Mhz + 2 x PCB

 

[AR2]

- andrea_mori : 2 x 5.6448MHz + 2 x 11.2896MHz + 4 x PCB
- Eldam : 1 x 11.2896 Mhz + 4 x PCB
- esgigt : 1 x 11.2896 Mhz + 1 x PCB, soldered if possible
- fralippo : 1 x 22.5792 MHz and 1 x 24.576 MHz + 2 x PCB
- mravinsky : 2 x 11.2896 Mhz + 1 x PCB
- TNT : 1x22.5792MHz 1x24.5760MHz + 2 x PCB, soldered if possible
- 1audio : 1x22.5792MHz 1x24.5760MHz + 2 x PCB
- randytsuch : 2x25.0000MHz + 4 x PCB
- TNT or maybe : 1 x 11.2896 Mhz + 1 x PCB, soldered if possible
- myint67 : 1x11.2896mhz + 1pcb also soldered if possible
- EdwardTam : 2x 16.9344MHz + 1x PCB
- BDL : 1 x 11.2896 MHz + 2 x PCB
- thorstenlarsen : 2 x 11.2896 MHz + 2 x PCB
- Zoran : 1x22.5792MHz 1x24.5760MHz + 2 x PCB, (soldered or not...)
- walangalam : 1 x 11.2896 Mhz + 2 x PCB
- AR2: 1 x 22.5792 MHz and 1 x 24.576 MHz + 2 x PCB (soldering not needed)

 

[mcluxun]

- andrea_mori : 2 x 5.6448MHz + 2 x 11.2896MHz + 4 x PCB
- Eldam : 1 x 11.2896 Mhz + 4 x PCB
- esgigt : 1 x 11.2896 Mhz + 1 x PCB, soldered if possible
- fralippo : 1 x 22.5792 MHz and 1 x 24.576 MHz + 2 x PCB
- mravinsky : 2 x 11.2896 Mhz + 1 x PCB
- TNT : 1x22.5792MHz 1x24.5760MHz + 2 x PCB, soldered if possible
- 1audio : 1x22.5792MHz 1x24.5760MHz + 2 x PCB
- randytsuch : 2x25.0000MHz + 4 x PCB
- TNT or maybe : 1 x 11.2896 Mhz + 1 x PCB, soldered if possible
- myint67 : 1x11.2896mhz + 1pcb also soldered if possible
- EdwardTam : 2x 16.9344MHz + 1x PCB
- BDL : 1 x 11.2896 MHz + 2 x PCB
- thorstenlarsen : 2 x 11.2896 MHz + 2 x PCB
- Zoran : 1x22.5792MHz 1x24.5760MHz + 2 x PCB, (soldered or not...)
- walangalam : 1 x 11.2896 Mhz + 2 x PCB
- AR2: 1 x 22.5792 MHz and 1 x 24.576 MHz + 2 x PCB (soldering not needed)
- mcluxun: 1 x 22.5792 MHz and 1 x 24.576 MHz + 2 x PCB (soldering not needed)