How to generate harmonics from Low-power frequency sources without preamp stages?

[Terminator3]

For example, if we need 2.5 to 5GHz doubler. The best soulution i found is compression region of few integrated amplification ICs, but it needs two preamp stages to generate enough output! Even for 0dBm!

So, my question, is it theoretically possible to get >=-5dBm output (before filtering) from -5dBm input with using one transistor/integrated amp?

 

[goldsmith]

For example, if we need 2.5 to 5GHz doubler. The best soulution i found is compression region of few integrated amplification ICs, but it needs two preamp stages to generate enough output! Even for 0dBm!

So, my question, is it theoretically possible to get >=-5dBm output (before filtering) from -5dBm input with using one transistor/integrated amp?

Hi for a doubler , you can use an amplifier e.g Gali-5 or somethings better . and then tune it for your desired frequency as well .
Best Wishes
Goldsmith

 

[Terminator3]

Thanks, that a good one. I am wondering about "Input 1dB Compression Point" parameter from datasheets, for example for MAX2644 it is -13 dBm. Is there any devices that come to non-linear region at lower input levels, say -15dBm. The bad thing that almost all integrated IC amps have integrated harmonic supression.

 

[Terminator3]

I discovered, that generating harmonics with non-linear devices is pretty easy, and can get some high order unwanted harmonics (noted for future use), but performance becomes unsatisfactory as input power is near and below 0dBm.

Problem 1: when i put 2 amplification stages, they consume too much power. I can sense heat on pcb with my finger. I did not measure total power consumption, but it would be a problem for battery usage.

Problem 2: Ampification leads to unwanted harmonics and unwanted LO signal high. So overall design needs some enclosure/more filtering.

So idea of keeping maximum signal at 0dBm with output of -10dBm is not so easy to implement.

 

[FvM]

You should consider that even harmonics are only generated by even terms in transfer characteristic. Rather than utilizing the essential unwanted nonlinearity of amplifiers, a dedicated even characteristic, e.g. a full-wave rectifier is promising better frequency doubling efficiency.

 

[goldsmith]

I discovered, that generating harmonics with non-linear devices is pretty easy, and can get some high order unwanted harmonics

Hi again
of course every nonlinear system can create gang of harmonics and we can easily take our desired one from it ! it is a usual way . ( according to the fourier series )

Problem 1: when i put 2 amplification stages, they consume too much power. I can sense heat on pcb with my finger. I did not measure total power consumption, but it would be a problem for battery usage.

It means you've dissipation . and it will deliver low efficiency . if you use a high efficiency stage you won't have any problem ( for instance either class C or E or F stage )

Problem 2: Ampification leads to unwanted harmonics and unwanted LO signal high. So overall design needs some enclosure/more filtering.

if you need multiplying , it would be a forced behavior ! ( analog multiplier requires filter )

e.g. a full-wave rectifier is promising better frequency doubling efficiency.

I agree with FvM about it . it will deliver better doubling but there would be a problem ! it's out put isn't a sine wave . it has been rectified and it is a problem . isn't it ?

Best wishes and best regards
Goldsmith

 

[jiripolivka]

I think you have a bad idea. Generating harmonics in any non-linear element requires certain power input . If you have less than such power, you MUST use an amplifier to get it. Period.

 

[vfone]

The highest efficiency (best conversion gain) of a frequency doubler can be done using a single transistor stage having a tuned circuit at the output (on 2fo).

The initial step is to find the performances of the active device at fundamental frequency, looking to parameters as: transconductance gm, transition frequency ft.
The transconductance (gm) has a direct impact in the device’s power performance and multiplication gain, and ft and fmax determined the limits to be used as a frequency multiplier.

Was discovered that the best efficiency in this situation is when the transistor is biased for Class-B, so the conduction angle for the transistor is approximately 180°. Thus the transistor conducts only half of the time, either on positive or negative half cycle of the input signal.

 

[jiripolivka]

Any PASSIVE non-linear device like a P/N junction achieves typically 10-13 dB conversion loss for frequency doubling ad higher one for tripling, etc. The diode must be saturated, so a diode pair requires >+13 dBm input to operate.

Active devices like transistors do offer a conversion gain but the input must be not too lower than for diodes (a Class C amplifier must rectify the input signal by e-b diode, to "open" and amplify). And the primary concern of the requester was "to avoid amplifiers", so no transistors are wanted.

 

[FvM]

And the primary concern of the requester was "to avoid amplifiers", so no transistors are wanted.

Realistically, he requested not to use more than one amplifier stage. So he apparently knows that the specification can't be achieved without an amplifier.