开云体育

Sometimes you need a swept sine generator and you don't have one available.
How about building one? Sounds complex but it is actually fairly simple.
If you search ebay for "12V RF controlled oscillator" there are many suppliers for a version that sweeps from about 500MHz to 1000Mhz for a price of 15$ per piece.
You need two of these modules, one 10dB attenuator and one mixer such as an ADE-1, either on a ready module for 10$ or as bare mixer for substantially less.
Connect one VCO with tuning input connected to ground to the mixer LO input? and one VCO with the tuning input used for frequency control with the 10dB attenuator to the RF input of the mixer.
Finally you add a 4 fold op-amp, two variable resistors (one for center frequency and one for span) and some more small components to generate the frequency control voltage and you have a wide range swept sinus generator.


The IF output of the mixer will deliver from 0 till 500MHz at about -10dBm and 2nd, 3rd and 4th harmonic are at least 40dB below the fundamental.
If you can not go down to zero Hz you have to interchange the VCO modules as the module with the lowest output frequency with tuning input connected to ground should be used as the variable VCO.
Have fun!
--------------------------------------------------------------------
HBTE Files section:?/g/HBTE/files
Erik, PD0EK

Do you have a prefered version of the op-amp circuit for a sweeper like this?

The design from Hans Summers works very well and is easy to build



You could use a 10 turn potmeter for the center frequency
Make sure you select a opamp that can drive to ground.
--
HBTE Files section:?/g/HBTE/files
Erik, PD0EK

You should be able to construct a ramp ot a triangle wave generator with most op amps or a programmable waveform generator like the AD9833. It depends on how fast/slow this oscillator can be swept. At this point I don't know the answer having never used it.

Jerry, AC9NM

The VCO can sweep very fast. Full range within 10mS.
No PLL to settle
--
HBTE Files section:?/g/HBTE/files
Erik, PD0EK

Hi all,

I received my oscillator modules today and hooked them up:

And connected the IF output to the spectrum analyzer:

So here we can see that the generated output has harmonics around 40dBc but there are also mixing products as well and they are higher.
Peak 3 which I think is 2*LO-RF is the worst and its around 27dBc. It can be reduced by putting a suitable filter on the mixer LO input, I only had a 800MHz LPF laying around which does not suppress 2*LO (1019MHz) that much but it reduced peak 3 a bit.

So how would one go about to reduce the mixing products in order to build a cleaner signal gen?
A LPF on the LO to get rid of its first harmonic?
And a LPF on the IF output to remove the LO leakage, however this would sacrifice the max frequency or increase power rolloff unless the filter contains a trap for the LO.

Any other ideas?

Best regards
Marcus

On Wed, Apr 29, 2020 at 01:10 PM, Marcus Gustafsson wrote:

So how would one go about to reduce the mixing products in order to build a cleaner signal gen?
A LPF on the LO to get rid of its first harmonic?
And a LPF on the IF output to remove the LO leakage, however this would sacrifice the max frequency or increase power rolloff unless the filter contains a trap for the LO.

Any other ideas?

I can newbie chime in on this one. I still haven't gotten past my filter designs, though did receive the 5.5x2.5mm barrel plug and opened up yesterday for the 64.5W laptop power supply to power the SH72 and see how that works interim as a most cost effective portable T12 tip soldering iron.?

A band pass filter that is tunable comes to mind.? I'm thinking that would be best.? Maybe you can gang a few ranges that can be switched.? ?

Like another way to think of is a high pass filter that is lower than the selected frequency range as well as a low pass filter that is higher than the selected frequency range.?

I'd like to read about other ideas... since I can't think of anything else off hand.? I'm guessing there might be solid state digital ASIC's now days I've not read about.?

On Wed, Apr 29, 2020 at 01:10 PM, Marcus Gustafsson wrote:

So how would one go about to reduce the mixing products in order to build a cleaner signal gen?

Hi Marcus.
Filtering the LO is not a good idea as the LO harmonics will also be generated in the mixer as it is overdriven to have good switching.
So a low pass filter at the output is the best option and yes, it will reduce the max output frequency.

?
--
HBTE Files section:?/g/HBTE/files
Erik, PD0EK

HI all,

Erik, I realize that LO harmonics is also generated in the mixer but would it not be beneficial to remove some of them beforehand? I'm aware of the fact that most mixers works better with square waves which is a fundamental frequency with an infinite number of harmonics.

Anyway I put together a LPF, it turned out to become a 400MHz one.


A quick sweep of the filter. Good performance up to 1.8GHz.

And the same spectrum picture as last time, now with the LPF on the output.

Looks much better. However when increasing frequency mixing products enters the spectrum again.

Ideally I think one would need switchable filter banks to clean this up further or just be happily aware of this simple signal generator's spurs.
I'll see if I build another LPF or a BPF for the LO just to see what happens. If I do, I'll make another post here.

Best regards
Marcus

Hi all again,

I did another thing instead. I replaced the LO with another signal generator. One with lower harmonics, 42dBc compared to 24dBc.
The result is very similar to before, so now I've verified what Erik said. There is no need to filter the LO.

After I took the screenshots I realized that the LO power in this case was 7dBm but with the module around 4.8dBm which might explain some differences.
All in all, I've learned something today.
Cheers
Marcus

Marcus

You could have a look at the required drive level for you mixer.
If? it is too low harmonics will increase with higher frequencies.
--
HBTE Files section:?/g/HBTE/files
Erik, PD0EK