Alternative to spectrum analyzer / VNA

[okw]

I trying to adjust a 2.4GHz bluetooth signal from an ESP32. According to the hardware design guidelines, I have uploaded the "RF certification and test" firmware, then use ESP software to start a "TX tone" at 2.4GHz and inspect the signal, and tweak MCU crystal load capacitors, to offset the RF signal until it's centered at 2.402GHz (antenna matching is quite good, experience a S11 of 35-38 dBi).
Problem is, I dont have a spectrum analyzer. I have oscilloscopes with FFT (TDS5104B and DPO4054), but they are only up to 500GHz and 1GHz (problem?).
I also have an SDR (RTL2838UHIDIR max 1.7GHz) and PS200 (Z / R / X / S11 / VSWR), similar to NanoVNA. (It says up to 2.7GHz, but I'm not sure.)
How can I (with some degree of accuracy), get by with what I have?
I'm also considering a Nano VNA v2 (or any other recommendation?), but if what I have will give me reasonably good result, I'm fine with that.

[veedub565]

Might be worth a look ? not sure what frequency it goes up to.

https://learn.adafruit.com/freq-show-raspberry-pi-rtl-sdr-scanner/overview

[David Hess]

What are you trying to do?

Spectrum analyzers and vector network analyzers are not substitutes for each other.

[selcuk]

An alternative approach can be setting a hardware PWM at 40MHz and measuring the output frequency with the oscilloscope.

Or measuring directly with the method in the attached document here:
https://www.eevblog.com/forum/beginners/quickly-accurately-repeatably-measuring-xtalosc-frequency/msg5267730/#msg5267730

[u666sa]

I'm also considering a Nano VNA v2 (or any other recommendation?)

v2 no bueno

You need NanoVNA H4 and LiteVNA 64. The first one can be changed to TinyPFA with a simple firmware flash.
For narrow-band measurements of quartz filters for example, you kinda have to have NanoVNA H4, because of LiteVNA and others will have problems with step and different stuff.

Here is a good guide for you in a language you don't know -- https://mysku.club/blog/misc/99119.html
Enjoy!

As far as for what you need. I think you need a spectrum analyzer.

TinySA Ultra. That's what you need.


So you need two things. TinySA Ultra and LiteVNA 64, based on frequences you working with. You might also need NanoVNA H4, especially if you need TinyPFA.

[okw]

Might be worth a look ? not sure what frequency it goes up to.

https://learn.adafruit.com/freq-show-raspberry-pi-rtl-sdr-scanner/overview

Thanks. But it only goes up to 863MHz.

[okw]

What are you trying to do?

Spectrum analyzers and vector network analyzers are not substitutes for each other.

I want to see the amplitudes around 2.4-2.5GHz.
If I understand the ESP32 correct, the bluetooth frequency will shift with crystal load capacitance (I'm using ESP32-C3 chip, not prebuilt module). So changing the load caps until the top of the amplitude is centeret at the chosen frequency (e.g. 2.402GHz).

[ftg]

What are you trying to do?

Spectrum analyzers and vector network analyzers are not substitutes for each other.

I want to see the amplitudes around 2.4-2.5GHz.
If I understand the ESP32 correct, the bluetooth frequency will shift with crystal load capacitance (I'm using ESP32-C3 chip, not prebuilt module). So changing the load caps until the top of the amplitude is centeret at the chosen frequency (e.g. 2.402GHz).

So you care about both the carrier amplitude or in other words the output power and then also want to measure the frequency error of the transmitter.
The first you could likely do with a TinySA Ultra.
The second with a frequency counter that goes up to 2.5GHz and either has a good internal reference or can be connected to a good external reference.

I guess a dumb power meter would be enough for output power, but it cannot be used to see adjacent channel power and other EMC certification related things.

[okw]

What are you trying to do?

Spectrum analyzers and vector network analyzers are not substitutes for each other.

I want to see the amplitudes around 2.4-2.5GHz.
If I understand the ESP32 correct, the bluetooth frequency will shift with crystal load capacitance (I'm using ESP32-C3 chip, not prebuilt module). So changing the load caps until the top of the amplitude is centeret at the chosen frequency (e.g. 2.402GHz).

So you care about both the carrier amplitude or in other words the output power and then also want to measure the frequency error of the transmitter.
The first you could likely do with a TinySA Ultra.
The second with a frequency counter that goes up to 2.5GHz and either has a good internal reference or can be connected to a good external reference.

I guess a dumb power meter would be enough for output power, but it cannot be used to see adjacent channel power and other EMC certification related things.

Thanks. I don't necessarily care about the power level of the amplitude, just to see the center frequency so I can adjust the offset. Down the line, I would probably look into certification and stuff, but not at this stage.

I'd really like to get a TinySA Ultra, but it takes some time to receive it.
I do have a 6GHz signal generator (Rohde & Schwarz SME06) and a 225MHz frequency counter (Philips PM6680). Would this help me along with the 1GHz FFT scope?
I also have a Spectran V5 (handheld 9kHz to 20GHz) in unknown condition, but I'm afraid to feed an active signal into it.

[ftg]

Just use a suitable attenuator with the Spectran and it will be fine. Or measure off air and try to deal with all the other users of the spectrum.
I think that will be more straightforward than building your own poor spectrum analyzer around the generator and oscilloscope.

[okw]

Just use a suitable attenuator with the Spectran and it will be fine. Or measure off air and try to deal with all the other users of the spectrum.
I think that will be more straightforward than building your own poor spectrum analyzer around the generator and oscilloscope.

I agree
Seems the ESP software can attenuate in 15 steps of 0.25dB. Do you know how much more attenuation I need to safely feed the Spectran?

[ftg]

Use a fixed, external attenuator between the Spectran and the DUT.
I would not trust software alone to protect the spectrum analyzer.
I usually have a 10 - 30dB attenuator in line when I'm using a spectrum analyzer.

[tggzzz]

Use a fixed, external attenuator between the Spectran and the DUT.
I would not trust software alone to protect the spectrum analyzer.
I usually have a 10 - 30dB attenuator in line when I'm using a spectrum analyzer.

... and a DC block, unless specified to be unnecessary.

[tszaboo]

Use a fixed, external attenuator between the Spectran and the DUT.
I would not trust software alone to protect the spectrum analyzer.
I usually have a 10 - 30dB attenuator in line when I'm using a spectrum analyzer.

I would target 0dB for the input of the small spectrum analyzers coming from any radio, compared to the maximum output of the radio. Not only will the attenuator protect the input, it helps with the impedance matching. And I find very important is having a DC block for any conducted RF measurements.

[okw]

Use a fixed, external attenuator between the Spectran and the DUT.
I would not trust software alone to protect the spectrum analyzer.
I usually have a 10 - 30dB attenuator in line when I'm using a spectrum analyzer.

I would target 0dB for the input of the small spectrum analyzers coming from any radio, compared to the maximum output of the radio. Not only will the attenuator protect the input, it helps with the impedance matching. And I find very important is having a DC block for any conducted RF measurements.

Thanks. The ESP32-C3 has a max output of 20dBm, so a 20dBm and DC block it is