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

Status
Not open for further replies.
Sorry to say but the 45Mhz board has very same problem. Sometime it starts the good 45Mhz sometimes 49.182 Mhz. But in tis case it is not changed when I touched the base with sold. iron, It starts 45 or 49 Mhz randomly...

Next week I'll investigate...
The BOM for 45/49 MHz SC-Cut Driscoll oscillator does not include a trap to suppress the unwanted B-mode oscillation because all the crystals I did test (previous batch) started at the right C-mode frequency.
I have not tested the last batch, maybe they are a little different from the previous batch, and so they need the trap, like all other SC-cut crystals.
I'll test the only one I own from the last batch (all other have been already shipped).
 
Thanks. I tested it at night, the heat of the soldering iron to the transistor base its jumping to the good 49 frequency, and the heat to the collector or emitter jumps to 53Mhz. (or I touch the C3 pins around). The other 45Mhz crystal just randomly jump, when I turn on-off. is the transistor good for here? The datasheet says its Hfe is just 100 around.

Is anyone who ordered the last group buy a 45 or 49 Mhz driscoll, trying the board?
Experienced anyone other this problem too?

(The AT cut clapp board is running perfect I try that too.)
 
45/49 MHz Sc-cut crystals issue

After some investigations...

the 45/49 MHz SC-cut crystals of the last batch seem to be different from these of the of the previous batch, so they start randomly at B-mode instead of C-mode (49 MHz instead of 45 MHz and 53 MHz instead of 49 MHz).

To avoid this issue the above crystal need a trap to suppress the unwanted B-mode and start at the right frequency, like all other SC-Cut crystals at different frequencies. The capacitor C3 in the schematic does the job.
BTW, in order to avoid overdriving the crystal because of the adding of the capacitor C3, a few components value have to be changed: C2, R1, R2.

Moreover the BFR182 seems to be too sensitive, so it was deleted from the BOM. I have tried several MMBT5179 and they work fine with SC-Cut crystals from 22 up to 49 MHz, so it can be used without any selection.

Follows the update BOM.
 
Last edited:
Updated schematics, PCBs, BOMs and Assembly guide

Updated schematics, PCBs, BOMs and Assembly guide

Attached the assembly guides of all the boards available. They contain the schematic of each board, the BOM to source the parts and the instructions to build each circuit.

TWTMC-D assembly guide was updated about 45/49 MHz SC-cut crystals.

- TWTMC-C: Colpitts-Clapp oscillator revised
- TWTMC-D: Driscoll oscillator revised
- TWTMC-P: Pierce oscillator
- TWTMC-AIO: Dual Driscoll oscillator with power supply
- TWTMC-D&D: daughter board (power supply, dividers, and so on) revised to supply the Driscoll oscillator and the oven board
- TWTMC-OVN: oven board for SC-Cut crystals Driscoll oscillator
- TWTMC-DIL: DIL adapter for Ian's Fifo board
 

Attachments

  • TWTMC-C.pdf
    533.2 KB · Views: 1,305
  • TWTMC-D.pdf
    818.1 KB · Views: 1,432
  • TWTMC-P.pdf
    700.9 KB · Views: 760
  • TWTMC-AIO.pdf
    660.6 KB · Views: 839
  • TWTMC-D&D.pdf
    663.3 KB · Views: 805
  • TWTMC-OVN.pdf
    752.3 KB · Views: 790
  • TWTMC-DIL.pdf
    452.7 KB · Views: 662
  • TWRPS-pp_micro.pdf
    348.9 KB · Views: 653
Hello Andrea,

Does the new BOM apply to the first historical batch of AT crystals and PCB I have, or should have I to upgrade some pcbs ?

kind regards from France,

Eldam

Hi Eldam,

if I remember right you got the TWTMC-C board, nothing was changed for this board.
In the meantime there are 2 more oscillator boards, the TWTMC-D Driscoll oscillator and the TWTMC-P Pierce oscillator.

Andrea
 
So I test the 45Mhz SC cut Driscoll board a little, I built it with potato unbuffered squarer. Very stabile, frequency different about -20-25ppm from the nominal, its good I think.
First impresions good, very nice sound, but I cant decide what different to the AT Clapp. It have to test paralell in future.
I apply it without oven for first, but I will do test with the heating, I am curious that will there any hearable different.
Andrea, the oven works well, about 80-90 C, but is it normal the darlington works between 100-120 C measured on case? Will it damage the junction for a long term working?
 
So I test the 45Mhz SC cut Driscoll board a little, I built it with potato unbuffered squarer. Very stabile, frequency different about -20-25ppm from the nominal, its good I think.
First impresions good, very nice sound, but I cant decide what different to the AT Clapp. It have to test paralell in future.
I apply it without oven for first, but I will do test with the heating, I am curious that will there any hearable different.
Andrea, the oven works well, about 80-90 C, but is it normal the darlington works between 100-120 C measured on case? Will it damage the junction for a long term working?

To heat the crystal at 80° the PCB goes at about 100°.
The MJD122 is specified -55 to +150, so I think it's not a problem.
My oscillator works along last year without any problem.
 
To heat the crystal at 80° the PCB goes at about 100°.
The MJD122 is specified -55 to +150, so I think it's not a problem.
My oscillator works along last year without any problem.
Thanks! Tomorrow I 'll test the AT Clapp and the SC Driscoll. I connect these to the two slots of Ian's FIFO. I will test same musics so I just change the output sample rate in the music player to compare the 2 clocks.
 
Hi Andrea,

I finally got all the parts and built the 24.576Mhz SC-Cut Driscoll. First, I got a problem of incorrect "start-up" frequency. When it powers up the first time, it will always be 26.7671Mhz. Restart the board, it will become normal (24.576Mhz). I soldered the 12pf C3 and there is no waveform (stop oscillating)! So, the C3 trick doesn't work for me. When I soldered the 74LVC04, here comes the second problem (perhaps they are related). It won't oscillate unless the probe or my finger touches the circuit, especially the XO case or the input trace before 74LVC04 (like 47k R9). Any ideas?

PS. Q1 is MMBT5179 as suggested from previous post. All part values are from BOM, except 100nf ceramic which is C0G. R11 is 68k instead of trim pot. C3 was measured 1/2 VCC (3V).

Poting


twtmcd2.jpg


twtmcd1.jpg
 
Last edited:
Hi Andrea,

I finally got all the parts and built the 24.576Mhz SC-Cut Driscoll. First, I got a problem of incorrect "start-up" frequency. When it powers up the first time, it will always be 26.7671Mhz. Restart the board, it will become normal (24.576Mhz). I soldered the 12pf C3 and there is no waveform (stop oscillating)! So, the C3 trick doesn't work for me. When I soldered the 74LVC04, here comes the second problem (perhaps they are related). It won't oscillate unless the probe or my finger touches the circuit, especially the XO case or the input trace before 74LVC04 (like 47k R9). Any ideas?

PS. Q1 is MMBT5179 as suggested from previous post. All part values are from BOM, except 100nf ceramic which is C0G. R11 is 68k instead of trim pot. C3 was measured 1/2 VCC (3V).

Poting


twtmcd2.jpg


twtmcd1.jpg

Hi Poting,

C3 capacitor is needed to suppres the unwanted B-mode, without C3 the oscillator starts at 26 MHz instead of 24 MHz.

Seems the MMBT has not enough gain, try replacing with BFR182 or NE85633-R25.

Andrea
 
Hi Poting,

C3 capacitor is needed to suppres the unwanted B-mode, without C3 the oscillator starts at 26 MHz instead of 24 MHz.

Seems the MMBT has not enough gain, try replacing with BFR182 or NE85633-R25.

Andrea


With C3, it won't start. As you mentioned in the previous post, MMBT should be fine without selection (?). I guess I have to try BFR182 or NE85633 this time.
 
With C3, it won't start. As you mentioned in the previous post, MMBT should be fine without selection (?). I guess I have to try BFR182 or NE85633 this time.

I know I said that MMBT should be fine without selection, I tried several parts and the oscillator did start always at the right frequency. Also, I used the MMBT for the finished boards I have shipped without any problem.

Seems the hfe of the MMBT varies a bit from a batch to another.
As soon as I find the time I will try other BJT.
 
Status
Not open for further replies.