Here some examples.
MHM 2010 Active Hydrogen Maser | Active Hydrogen Maser | Time & Frequency References | Timing & Synchronization Systems | Products
Passive and Active Hydrogen Maser Frequency Standards
Phase noise is not so impressive
Phase Noise 1Hz 117 dBc
Phase Noise 10Hz 133 dBc
not better than a SOTA OCXO, that's not so expensive.
I think it would be interesting with an environment specification and some specific test cases where the DUT is exposed both mechanical and sound pressure which during the exposure, the phase noise is measured.
"Environmental Sensitivity
The MHM 2010 is designed to for low sensitivity to temperature, magnetic field and power supply changes."
.. they didn't say anything about mechanical resistance...
My idea was mostly to combat vibration from outside and specifically the acoustical one.
I suppose all the phase plots we see are in very quite and stable environment.
//
"Environmental Sensitivity
The MHM 2010 is designed to for low sensitivity to temperature, magnetic field and power supply changes."
.. they didn't say anything about mechanical resistance...
My idea was mostly to combat vibration from outside and specifically the acoustical one.
I suppose all the phase plots we see are in very quite and stable environment.
//
BOM and assembly guide
All items have been shipped today to the members who have not required the assembly service.
Next days I will ship to Phil (korben69) the items for the members who required the assembly service. Seems the following members have not
yet sent the order form to Phil, please fill the order form and send it as soon as possible:
- thorstenlarsen
- zenelectro
- JethroTull
- Lindamar
Here attached BOM and assembly guide for all the boards.
All items have been shipped today to the members who have not required the assembly service.
Next days I will ship to Phil (korben69) the items for the members who required the assembly service. Seems the following members have not
yet sent the order form to Phil, please fill the order form and send it as soon as possible:
- thorstenlarsen
- zenelectro
- JethroTull
- Lindamar
Here attached BOM and assembly guide for all the boards.
Important notice: 90/98 MHz users
In the BOM are not indicated the components value for 90/98 MHz oscillators, because I'm still working around for these frequencies.
Please, wait for a while to order components.
While for frequencies from 5 to 49 MHz I have checked all the crystals starting at right frequency, I'm encountering some issue about the higher frequencies.
I'm testing the 90.3168 MHz crystal with PCB and through hole components (C1-C2-C9-L4) using socket pin, as for other frequencies, but this crystal does not run at right frequency.
It starts at fundamental (18 MHz) and third overtone (54 MHz), but it doesn't start at fifth overtone. Calculated values are 10 pF for C1-C2 and 470 nH for L4 (C9 fixed 1 nF). With the above values, removing temporarily L4 from the circuit, the oscillator starts at 54 MHz (third overtone), but with L4 in place it doesn't start.
Next days I'll build all the circuit with SMD components (as soon as they arrive), to check if there is problem of stray capacitance with TH components.
Since we are on a diy forum, I would start a technical discussion around this.
Colpitts-Clapp circuits are negative resistance oscillators.
The network seen by the crystal should present negative resistance to overcome its motional resistance. This is the starting point for the oscillator to start and sustain the oscillation.
C1-C2 in series give us that negative resistance, but there are other parameters that affect this resistance, decreasing it for some reason.
Stray capacitance of the capacitors themselves, stray capacitance of PCB traces, shunt capacitance of the crystal (being "shunt", acts as a divider), transconductance of the active device. At these frequencies all become crucial.
As the rule of thumb (but also in practical), the negative resistance should be at least 3/4 times the motional resistance of the crystal, to guarantee enough margin.
I have tried to calculate the negative resistance of the network with the above values, considering also stray capacitance of the capacitors, shunt capacitance of the crystal and the transconductance of the jfet: the theoretical result is around -116 ohm.
The motional resistance of the crystal is around 37 ohm, so the negative resistance should be around 3 times. Not a safe margin.
The low transconductance of the jfet (8 to 12 ma/V) does not help.
In practical the negative resistance could be less than the calculated value.
I'll give a try to the circuit using a bjt, such as BFR92 or BFS20, to see how it works. If it starts at the right frequency, then the problem is the low
transconductance of the jfet.
Let me experiment a little, sorry for the inconvenient.
In the BOM are not indicated the components value for 90/98 MHz oscillators, because I'm still working around for these frequencies.
Please, wait for a while to order components.
While for frequencies from 5 to 49 MHz I have checked all the crystals starting at right frequency, I'm encountering some issue about the higher frequencies.
I'm testing the 90.3168 MHz crystal with PCB and through hole components (C1-C2-C9-L4) using socket pin, as for other frequencies, but this crystal does not run at right frequency.
It starts at fundamental (18 MHz) and third overtone (54 MHz), but it doesn't start at fifth overtone. Calculated values are 10 pF for C1-C2 and 470 nH for L4 (C9 fixed 1 nF). With the above values, removing temporarily L4 from the circuit, the oscillator starts at 54 MHz (third overtone), but with L4 in place it doesn't start.
Next days I'll build all the circuit with SMD components (as soon as they arrive), to check if there is problem of stray capacitance with TH components.
Since we are on a diy forum, I would start a technical discussion around this.
Colpitts-Clapp circuits are negative resistance oscillators.
The network seen by the crystal should present negative resistance to overcome its motional resistance. This is the starting point for the oscillator to start and sustain the oscillation.
C1-C2 in series give us that negative resistance, but there are other parameters that affect this resistance, decreasing it for some reason.
Stray capacitance of the capacitors themselves, stray capacitance of PCB traces, shunt capacitance of the crystal (being "shunt", acts as a divider), transconductance of the active device. At these frequencies all become crucial.
As the rule of thumb (but also in practical), the negative resistance should be at least 3/4 times the motional resistance of the crystal, to guarantee enough margin.
I have tried to calculate the negative resistance of the network with the above values, considering also stray capacitance of the capacitors, shunt capacitance of the crystal and the transconductance of the jfet: the theoretical result is around -116 ohm.
The motional resistance of the crystal is around 37 ohm, so the negative resistance should be around 3 times. Not a safe margin.
The low transconductance of the jfet (8 to 12 ma/V) does not help.
In practical the negative resistance could be less than the calculated value.
I'll give a try to the circuit using a bjt, such as BFR92 or BFS20, to see how it works. If it starts at the right frequency, then the problem is the low
transconductance of the jfet.
Let me experiment a little, sorry for the inconvenient.
Hi All,
Link to the Assembly Service form is this one :
https://docs.google.com/forms/d/1odqObA4PFXhxqxyqrZiAPSWgGCTsZBO19erk2LmXXHk/viewform?usp=send_form
Latecomers are welcome
I'll check each request with Andrea to ensure that right parts are used for each configurations.
Regards,
Phil
Link to the Assembly Service form is this one :
https://docs.google.com/forms/d/1odqObA4PFXhxqxyqrZiAPSWgGCTsZBO19erk2LmXXHk/viewform?usp=send_form
Latecomers are welcome
I'll check each request with Andrea to ensure that right parts are used for each configurations.
Regards,
Phil
Tip of the day
If you would invest a little more to reach better thermal and mechanical stability, you could install mica capacitors in place of NP0/X7R type (C1-C2):
100 pF 598-MC12FA101J-TF
56 pF 598-MC12FD560J-F
200 pf 598-MC12FA201J-F
33 pF 598-MC08FA330J-F
22 pF 5982-08-100V22-F
18 pF 598-MC08EA180J-F
5.6448 MHz users, should install 100 nF in place of 1 uF (C4):
100 nF 581-12102C104K
(Mouser codes)
If you would invest a little more to reach better thermal and mechanical stability, you could install mica capacitors in place of NP0/X7R type (C1-C2):
100 pF 598-MC12FA101J-TF
56 pF 598-MC12FD560J-F
200 pf 598-MC12FA201J-F
33 pF 598-MC08FA330J-F
22 pF 5982-08-100V22-F
18 pF 598-MC08EA180J-F
5.6448 MHz users, should install 100 nF in place of 1 uF (C4):
100 nF 581-12102C104K
(Mouser codes)
- Status
- Not open for further replies.