173.3MHz oscillator, how to?

[neazoi]

Hello, for a spectrum analyzer project I need a 173.3MHz oscillator to be made.
The author of this project used a Murata SAR173.3MB40X SAW resonator to drive a NE605.
But this resonator is obsolete. Any ideas of where to find a similar resonator or how to build such an oscillator (maybe harmonic crystal oscillator?)

 

[BigBoss]

Do it your resonator yourself on a substrate with a high dielectric.
Resonator is essentially a L-C parallel/series circuit, if you realize this circuit by using distributed elements on a substrate with high dielectric and low loss, you would be obtained a good resonator.
Even FR4 can be used with some expenses of dimensions.

 

[neazoi]

Do it your resonator yourself on a substrate with a high dielectric.
Resonator is essentially a L-C parallel/series circuit, if you realize this circuit by using distributed elements on a substrate with high dielectric and low loss, you would be obtained a good resonator.
Even FR4 can be used with some expenses of dimensions.

Hello I have found other SAW filters at 173.9MHz. Can I use them in my oscillator as the resonator element?
I am asking because I see from the datasheet that these filters are specified for their center frequency but have several MHz band pass.
**broken link removed**
So could I bring them to 173.3MHz by adjusting relevant capacitors in series or in parallel with the resonator?

 

[D.A.(Tony)Stewart]

No one will likely have stock. FOX has embedded PLL in one series to synthesize any F. Expect $60 for one.

Here are some examples but no stock https://www.digikey.ca/product-sear...t=0&page=1&quantity=0&ptm=0&fid=0&pageSize=25

If this is a reference Osc, then the phase noise must be very low and you need to learn how to control that by design or buy an oscillator.

High Q is one characteristic (>>10k) that reduces noise bandwidth.

 

[neazoi]

No one will likely have stock. FOX has embedded PLL in one series to synthesize any F. Expect $60 for one.

Here are some examples but no stock https://www.digikey.ca/product-sear...t=0&page=1&quantity=0&ptm=0&fid=0&pageSize=25

If this is a reference Osc, then the phase noise must be very low and you need to learn how to control that by design or buy an oscillator.

High Q is one characteristic (>>10k) that reduces noise bandwidth.

I agree these are rare.
But If I find one in close frequency, will I be able to bring it to the desired one, if the desired one is inside it's pass band?

 

[D.A.(Tony)Stewart]

yes, but one should be aware of phase noise specs and requirements. Just having any oscillator wont work.

 

[neazoi]

yes, but one should be aware of phase noise specs and requirements. Just having any oscillator wont work.

I realized that in the circuit of the author, the SAW center freq is 173.3M, the input is 183M and the IF filters are 10.7M. This does not make sense, since the difference of the RF and the LO does not give IF of 10.7M. I think that the author used a SAW he had available (173.3M) and brought it to frequency using the variable coils in the oscillator of the schematic.

 

[D.A.(Tony)Stewart]

1st LO is 184~384 ( but described as 180 to 380 ) if you want 0 to 200 RF input
1st IF is 184.0 {helix} [MHz] ( but shown as 183 )
2nd LO is 173.3 {SAW fixed}
2nd IF is 10.7 {SAW}

Most SA's use a 1st IF above the 1st LO range, but this often leads to 3 IF's to get best results.
This one will have a lot of leakage of 1st LO into 1st IF so the "0" to 200 is misleading.

 

[neazoi]

1st LO is 184~384 ( but described as 180 to 380 ) if you want 0 to 200 RF input
1st IF is 184.0 {helix} [MHz] ( but shown as 183 )
2nd LO is 173.3 {SAW fixed}
2nd IF is 10.7 {SAW}

Most SA's use a 1st IF above the 1st LO range, but this often leads to 3 IF's to get best results.
This one will have a lot of leakage of 1st LO into 1st IF so the "0" to 200 is misleading.

> 1st LO is 184~384 ( but described as 180 to 380 ) if you want 0 to 200 RF input
the pos400 is specified for 200 and up, but if you look at the graph in the datasheet you will find that it can go as low as about 180MHz.

> 1st IF is 184.0 {helix} [MHz] ( but shown as 183 )
As said previously the vco should be able to start from 180MHz. for 183 IF (helical) and 0-200MHz RF, I would expect the range of the LO to start at 183MHz.

> 2nd LO is 173.3 {SAW fixed}
Now I believe the author used this filter because this should be available to him. with the RF out of the filter at 183MHz and the IF being 10.7MHz, the OSC should be at 172,3MHz actually and not 173.3. Since these SAW filters have several MHz bandwidth, I believe that is the reason he uses an additional transistor oscillator and bring the oscillator frequency at 172,3MHz when tuned, despite the mentioned center saw frequency is 1MHz up. How does it sound to you?

> 2nd IF is 10.7 {SAW}
Yes they are ceramic filters of more than 100KHz BW. I think I will order some crystal filters with 6KHz BW as well, to switch them in order to achieve maximum resolution at low sweep rates.

 

[D.A.(Tony)Stewart]

I stand corrected..

if 2nd IF is 10.7 and
2nd LO is 172.3 then
1st IF (BPF)= 183 (the sum) then if
1st LO sweeps 183 to 383 then
RF input is 0 to 200, but since LO starts at IF (183), the marker for 0 Hz will be the LO saturating the 1st IF detector from leakage in the mixer. (e.g. -30 dB imbalance in the balanced modulator )

This is inexpensive way but poor design for low RF performance so minimum RF 0 to 200 stated, is likely to be >11 to 200MHz..

which was my point depending numbers...

 

[neazoi]

I stand corrected..

if 2nd IF is 10.7 and
2nd LO is 172.3 then
1st IF (BPF)= 183 (the sum) then if
1st LO sweeps 183 to 383 then
RF input is 0 to 200, but since LO starts at IF (183), the marker for 0 Hz will be the LO saturating the 1st IF detector from leakage in the mixer. (e.g. -30 dB imbalance in the balanced modulator )

This is inexpensive way but poor design for low RF performance so minimum RF 0 to 200 stated, is likely to be >11 to 200MHz..

which was my point depending numbers...

>RF input is 0 to 200, but since LO starts at IF (183), the marker for 0 Hz will be the LO saturating the 1st IF detector from leakage in the mixer. (e.g. -30 dB imbalance in the balanced modulator )
This is inexpensive way but poor design for low RF performance so minimum RF 0 to 200 stated, is likely to be >11 to 200MHz..

I do not believe the coverage would be that much limited because of the overloading and the mixer leakage, so that the SA starts from 11MHz and up. I may be wring though..
Yes of course you would not expect 0Hz, in fact it should start from a few KHz or 10s of KHz. The reason I am telling that is because the SBL-1 (or the SRA-1 I use) are double balanced mixers. This means that the local oscillator "ideally" would not appear at the mixer output. The local oscillator would appear at the mixer output only by the amount of the internal imballance of the mixer components.
I also think a narrower 10.7MHz filter is a must for low KHz operation.
To be honest below 20KHz no one needs a SA because the sound card is more than enough.
However what would be considered as a better design for the low KHz range?

 

[BigBoss]

This circuit can essentially not handle 0Hz because of coupling capacitor at the input ( 10nF).
It can measure-perhaps-starting from few MHz.

 

[neazoi]

This circuit can essentially not handle 0Hz because of coupling capacitor at the input ( 10nF).
It can measure-perhaps-starting from few MHz.

You have a point here. The 10nF cap acts as a crude HPF.
I may increase that to 100nF, perhaps even 220nF, if that is the problem, so that I can measure down to the LW band at least?

 

[D.A.(Tony)Stewart]

The LO to IF leakage is 40dB typ with 5dB conversion loss, thus if LO level is +5dBm,

it will appear as if there is RF input at -30dBm (+5-40 +5) over the width of the IF filter which is significantly above the normal noise floor.

 

[BigBoss]

You have a point here. The 10nF cap acts as a crude HPF.
I may increase that to 100nF, perhaps even 220nF, if that is the problem, so that I can measure down to the LW band at least?

Maybe..
Because SLB-1 has a lower frequency as 1MHz.At lower frequencies, VSWR is worsening.

If we consider Low Side Injection..
Min. Freq=183-180MHz=3MHz
Max. Freq=183-380MHz=197MHz
Meanwhile the VCO crosses passing 183Mhz and 0Hz theoretically is possible but 10nF will limit the lowest frequency.
Try and see by augmenting this cap. to 220nF.

 

[neazoi]

Yes thank you, that is interesting!
I will probably start the sweep from 183MHz, there is no need to fold the spectrum.