10.455 Mhz Oscillator Required

[Ow@i$]

Hi,
I require a local oscillator (very close to pure sine )of about 10.455 MHz freq.
I hope someone has the circuit where i can change values and get req. freq.!!
or any other helpful link....
Thanks

 

[kripacharya]

try this ultra simple circuit which gives very good purity and works for large frequency range
i have already put in LC values for close to 10.455MHz. For this freq, a larger value of C2 is better. Say 27-33pF.
In case it does not oscillate simply adjusting R1 will change the gain, and it will oscillate.

you can take output from either LC tank itself - with proper buffer - or from top of R1. But sometimes top of R1 is not so pure.

Q of L1 should be good for better performance, but it will work with low-Q also.

Of course make sure that your supply is properly filtered and stable.

i have tested this circuit and made it work by only changing L and C upto 65MHz and down to 0.6Mhz. For R1 i used a 5K preset - reduce resistance for higher frequencies

 

[Ow@i$]

ok thanx i will try this..
but how exactly you can set frequency i mean hit and trial or is there a formula for designing at required frequency??

 

[kripacharya]

frequency formula is standard LC formula -- one by two pi root L C

where L = L1, and C = C1
however there is a very small amount of stray capacitance/ inductance in any circuit always. some is due to the other components, and some is due to the final layout. usually for this circuit it comes to about 10pF, maybe even upto 20pF

so best thing is to use a fixed cap for C1 which is lower than required, and trim it with a trimcap of around 10 - 25 pF. leave space on your board for connecting additional caps for same purpose.

frequency stability is dependant on usual parameters like temperature & power supply stability. so use low ppm drift coils, C0G/ NPO capacitors, shielding etc etc. the usual design procedures for getting high stability oscillators from LC tank based circuits.

I noticed you said about 10.455MHz, hence i suggested this circuit. otherwise you could get approx freq by using crystals.
for example, you can get a 10.4167Mhz crystal and trim (with caps) in standard colpitts / clapp/ seiler circuit for best freq.
or a combination of 18.432mhz (trim to ~ 18.45mhz) with 8mhz oscillator mixed, then band filtered at 10.45mhz

something like that.
of course if you can get a crystal cut for 10.455mhz, then that is the best !!

may i ask why you need this particular frequency ? what is the purpose ? it sounds like a combination of 10Mhz and 455khz....

You should post this in the RF forum also. Maybe you will get better ideas from experts there

 

[snail888]

Hi,as kripacharya mentioned,for LC oscillator is a good way for your application.
Could you give more exact parameter requirements?such as temperature drift,initial frequency accurate,output impedance,
operating voltage.
For 10.455M,i guest it maybe use in AM/FM application.
Do you want to change frequency autonomous by circuit?If you are,a available method is to use DDS,which could change frequency by digital signal or analog voltage.Good Luck.

 

[kripacharya]

yes of course DDS is always an option. but cost and complexity seems too high to generate just single frequency. you will need DDS ic (~$10), also crystal, uC circuitry for management along with all other circuitry and equipment for programming

i think a well designed discrete oscillator would be better here, yes ?

even using 2 colpitts oscillators at 10MHz and 455KHz (using crystals & resonators), combined using mixer ( maybe '602/ 612 ic ? or SBL-1 ? ) plus a double-tuned bpf would be easier

total cost ~ $3

 

[snail888]

yes,for a crystal oscillator would be better for cost.

 

[Ow@i$]

Thanx guys for reply!!!
I am actually designing a receiver to receive 10 MHz.I have worked on TRF confiuration but i want to design a superhet one... so gonnna need a local oscillator of 10.455 MHz and then mix it with incoming 10 MHz and get 455 KHz output!!!
have searched crystals but not close to 10.455....!!!
moreover kripacharya your oscliator doesn't oscillate i have ries changing values of r and c but no use!!!

 

[kripacharya]

i have tested this circuit and made it work by only changing L and C upto 65MHz and down to 0.6Mhz. For R1 i used a 5K preset - reduce resistance for higher frequencies

moreover kripacharya your oscliator doesn't oscillate i have ries changing values of r and c but no use!!!

i think you mean that you were not able to get it working, despite it being such a simple design.
Which fet's did you use ? try J310 or equivalent if you have any.
did you put a proper 1mH choke coil ?
did you use a multiturn pot for R1 and vary the gain while monitoring the output ?

I am actually designing a receiver to receive 10 MHz.I have worked on TRF confiuration but i want to design a superhet one... so gonnna need a local oscillator of 10.455 MHz and then mix it with incoming 10 MHz and get 455 KHz output!!!
have searched crystals but not close to 10.455....!!!

for a receiver you will need to be able to tune the oscillator, so using a crystal is not a good idea.
why don't you try using NE602/ NE612/SA612 or their variants ? You will get an oscillator and a mixer in a single package. There are many published and tried circuits using this chip.

 

[Ow@i$]

I am using 2N4416.. doing simulations in proteus .. yes choke is there!!!

 

[kripacharya]

I am using 2N4416.. doing simulations in proteus .. yes choke is there!!!

LoL !! a simulation !!

then sometimes you need to provide a very very small 'kick' to the tank to start it going. Even a 1mV pulse coupled in with a 2pF cap should work.
In a real implementation this kick would be automatically available due to noise/ imbalances in the circuit

 

[Ow@i$]

ok then!!! one more thing why not to use crystal??? since i am designing for single frequency 10 MHz.. need not to tune it !!!

 

[kripacharya]

ok then!!! one more thing why not to use crystal??? since i am designing for single frequency 10 MHz.. need not to tune it !!!

then sure you can use a crystal. so you found a 10.455mhz crystal ?

 

[Ow@i$]

no. thinking of making first 10 MHz one and one 455 Khz adding and then filtering!!!
i guess better oscillation can be achieved that way .. what say you??
as i need to be precise in oscillation !!!

 

[kripacharya]

sounds messy. you will be generating the exact freq you want to receive. somehow my gut feeling is this will not work well.