The only working oscillator with Sc-cut 3rd overtone 45/49 MHz crystals is the TWTMC-D.
It supports 3V3 CMOS output with 50% duty cycle, as specified in the assembly manual.
Attachments
The oven is an option but not mandatory.
You can find the price of the TWTMC-D bare board in the order form at post #323 of the GB thread
The Well Tempered Master Clock - Group buy
Only bare board is available, no finished oscillators, you have to build it yourself.
The files seem to be spread out thru this LONG thread, with no clear index. This is why I suggested the TS re-edit the org. thread, and include the essential PDFs.Damn Hell... would we need to include a finger in the GB and someone to push it on the mouss for KHashmi
I think there are about a dozen PDFs, but I have no idea which page (post #) on which the latest, e.g. TWTMC-C.pdf , version is located.
Now we are testing a pair of frequency doubler, the new GB will start within 15-20 days.
That's interesting Andrea if we can have the good behavior of low speed Laptechs with more upsampling liberty.
Do all things added, solder points, parts, are worsting the fragile signal ?
For instance, how the quality of our soldering skills matter for mounting a clock boards, please ? I'm sure experts already measured such things here seing the high discussion level. Have I to be worried about that ? Maybe, smd soldering paste with 200 C oven will be nicer than hand iron soldering devices for clean results ? But maybe the fragile crystal soldered at the end fastly with hands ?!
Do all things added, solder points, parts, are worsting the fragile signal ?
For instance, how the quality of our soldering skills matter for mounting a clock boards, please ? I'm sure experts already measured such things here seing the high discussion level. Have I to be worried about that ? Maybe, smd soldering paste with 200 C oven will be nicer than hand iron soldering devices for clean results ? But maybe the fragile crystal soldered at the end fastly with hands ?!
We expect that the frequency doubler adds much less phase noise to the lower frequency oscillator than using another one with the crystal at doubled frequency.
No worry about soldering, there are no difficult parts to be soldered. The crystals are cold welded parts, so no problem with soldering.
Anyway we are working also to provide finished boards.
Will you consider making a 4x multiplier for the 5/6Mhz SC cuts? These are superb.
poting
2 boards or use 4 times one board did you mean ?lol...
Andrea, thank you, well I am ok with smd, I am just not able to make the soldering as clean and britle as through holes. If quality result is not affected with ears I'm ok with that.
also for the experts of clocks and layout, what about vibration damping and cooling philosophy if good enough not to use ovens for the 5 and 6 Mhz sc cut pcb please ? Lead box 1 cm thick in a wood one feeled of sand, or does this sealed metal box will get above 80 °C ?
Too much ? I remember the early 90 when some putted the cd player mechanism hanged and inside a lead block... where to stop to be good enough with this good clock please. We less care cause the fhz is low, not 100 Mhz ?
Andrea, thank you, well I am ok with smd, I am just not able to make the soldering as clean and britle as through holes. If quality result is not affected with ears I'm ok with that.
also for the experts of clocks and layout, what about vibration damping and cooling philosophy if good enough not to use ovens for the 5 and 6 Mhz sc cut pcb please ? Lead box 1 cm thick in a wood one feeled of sand, or does this sealed metal box will get above 80 °C ?
Too much ? I remember the early 90 when some putted the cd player mechanism hanged and inside a lead block... where to stop to be good enough with this good clock please. We less care cause the fhz is low, not 100 Mhz ?
Last edited:
i meant using two multiplier (2x) in series... to get a 4x multiplication in the end
The new oscillators don't need the oven, we have tested the Driscoll and the differential oscillators with and without the oven and we have measured exactly the same phase noise, no appreciable difference. So both oscillators will run at room temperature avoiding to manage high temperature.
Vibrations are a great issue because they affect the phase noise performance close to the carrier.
We have reached the best performance suspending the oscillator box with elastic bands.
But obviously everyone is free to follow his own way isolating the oscillator from vibrations.
I have measured the SC cut crystal I took from the defective Morion MV89A crystal oven.
Pics are some pages earlier. I have checked the crystal parameters at room temperature
and at it's intended temperature.
nominal at room temp.
F = 4999868.18548896
Q = 2058123.87937723
L = 5.31149294399851
C = 1.90768447144417E-16
R = 81.074358418583
C0= 2.8136087840004E-12
B-Mode at room temperature / 27°C
F = 5456268.74855202
Q = 2070228.75819907
L = 5.60406600257822
C = 1.51825777903696E-16
R = 92.8026816958027
C0= 2.50200142407676E-12
fundamental at room temp.
F = 1706530.674
Q =12634
L = 2.882184522
C = 3.017798817E-15
R = 2446.1
C0= 3.400172339E-12
Nominal frequency at 87.8°C
F = 4999977.73472294
Q = 2210624.24356611
L = 5.31056260615351
C = 1.90793506338953E-16
R = 75.4698598161708
C0= 2.67067707950378E-12
B-Mode at 88°C
F = 5445966.6930594
Q = 1559143.3579169
L = 4.56636349264629
C = 1.87033681366853E-16
R = 100.216511187362
C0= 3.02668789364201E-12
Fundamental @ 88°c
F = 1706366.18003318
Q = 7829.11559388434
L = 2.67725816365808
C = 3.24941770179082E-15
R = 3666.31393448872
C0= 3.28464263904697E-12
The fundamental is so weak that it does not deserve a trap. The B-mode
resonance at +9.5% is nearly as good as the nominal frequency. Depending
on the oscillator circuit one might experience frequency hopping.
88°C was produced with a Steinelt heat gun that was sitting on the
boom of the microscope. I wonder why my low cost DG8SAQ VNWA can
measure this and my R&S ZVB cannot. The reference frequency of the
VNWA was locked to my LUCENT GPS source.
There are some spurii above the nominal frequency and above the b-mode.
I did not check them in detail. The time for a sweep is half an hour because
of the required settling time and the huge number of measurement points.
I you zoom in, they look worse. The 2 attached plots show some ringing after
excitation of the resonance, they are too fast. They create nonphysical results
because it seems that more energy comes back than is sent by the VNA.
That does not matter here because I used these plots only for locating the
resonances and not for using them for the crystal estimator.
Gerhard
Pics are some pages earlier. I have checked the crystal parameters at room temperature
and at it's intended temperature.
nominal at room temp.
F = 4999868.18548896
Q = 2058123.87937723
L = 5.31149294399851
C = 1.90768447144417E-16
R = 81.074358418583
C0= 2.8136087840004E-12
B-Mode at room temperature / 27°C
F = 5456268.74855202
Q = 2070228.75819907
L = 5.60406600257822
C = 1.51825777903696E-16
R = 92.8026816958027
C0= 2.50200142407676E-12
fundamental at room temp.
F = 1706530.674
Q =12634
L = 2.882184522
C = 3.017798817E-15
R = 2446.1
C0= 3.400172339E-12
Nominal frequency at 87.8°C
F = 4999977.73472294
Q = 2210624.24356611
L = 5.31056260615351
C = 1.90793506338953E-16
R = 75.4698598161708
C0= 2.67067707950378E-12
B-Mode at 88°C
F = 5445966.6930594
Q = 1559143.3579169
L = 4.56636349264629
C = 1.87033681366853E-16
R = 100.216511187362
C0= 3.02668789364201E-12
Fundamental @ 88°c
F = 1706366.18003318
Q = 7829.11559388434
L = 2.67725816365808
C = 3.24941770179082E-15
R = 3666.31393448872
C0= 3.28464263904697E-12
The fundamental is so weak that it does not deserve a trap. The B-mode
resonance at +9.5% is nearly as good as the nominal frequency. Depending
on the oscillator circuit one might experience frequency hopping.
88°C was produced with a Steinelt heat gun that was sitting on the
boom of the microscope. I wonder why my low cost DG8SAQ VNWA can
measure this and my R&S ZVB cannot. The reference frequency of the
VNWA was locked to my LUCENT GPS source.
There are some spurii above the nominal frequency and above the b-mode.
I did not check them in detail. The time for a sweep is half an hour because
of the required settling time and the huge number of measurement points.
I you zoom in, they look worse. The 2 attached plots show some ringing after
excitation of the resonance, they are too fast. They create nonphysical results
because it seems that more energy comes back than is sent by the VNA.
That does not matter here because I used these plots only for locating the
resonances and not for using them for the crystal estimator.
Gerhard
Attachments
Last edited:
Thanks in advance for considering this!
As I requested earlier, I'm potentially considering building one of the TWTMC projects .... but am having a hard time finding the LATEST pdf files in this long thread.
Usually, for such project threads that are popular (and in which the files are added to or updated frequently), the ORIGINAL POST (page 1, comment 1), is updated .
If the original TS is not editable, you can ask the MODs to make it so.
Thanks in advance for considering this!
Thx for starting this thread.
As I requested earlier, I'm potentially considering building one of the TWTMC projects .... but am having a hard time finding the LATEST pdf files in this long thread.
Usually, for such project threads that are popular (and in which the files are added to or updated frequently), the ORIGINAL POST (page 1, comment 1), is updated .
If the original TS is not editable, you can ask the MODs to make it so.
Thanks in advance for considering this!
- Status
- Not open for further replies.