开云体育

ctrl + shift + ? for shortcuts
© 2025 开云体育
  1. Tinysa
  2. Topics

Measurement problems, strange step noise floor behavior between 600-800kHz

 

I just got a new TinySA Ultra (SeeSii store, Aliexpress) and get doubtful measurement results.
- I selected the cal output of 30MHz.
- connected the cal output with the SA input with an oscilloscope in between (50 ohm termination assumed to be provided by the input).
Problem 1: The scope shows about 40mV RMS which translates to -15dBm, but the TinySA shows the main peak at -35.5dBm on the fundamental 30MHz tone. Why is this?
Problem 2: The signal generator inside the TinySA claims to output -18dBm, but when measured with a scope on a 50 ohm load I measure around 285mV which corresponds to +2dBm. Why?
Problem 3: Using Nano VNA sine generator @140MHz with a 10dBm attenuator (because the NanoVNA outputs about +10dBm when set to 0dB!), measured with scope around 220mV RMS showing the tone in the TinySA of around -20dBm when it should show 0dBm. What is the problem here?
?
I also did the linearity measurement which reveals no significant deviations (a straight line) across the attenuator steps at 120MHz and also at 200KHz and at 2GHz (at the latter frequency the peak showed up at -23dBm and "reduced linearity" warning showed up). All in all, seems no problem with the attenuator.
?
Now to the strange noise floor behavior. Input was terminated and the noise floor started at 600KHz with around -85dBm and transitioned in weird steps towards -120dBm at around 1MHz. Does this indicate a problem?
?
I do have historic background in RF engineering but I am relatively new to these new small devices and appreciate any feedback on what I identified as problems.
?
Thank you in advance
Lutz
?

NOISEFLOOR.bmp

 

Have you set your o'scope for 50-ohms or using a 50-ohm feedthrough adaptor?

Dave - W?LEV


On Tue, Jan 21, 2025 at 9:21?PM lschafer via <lschafer=[email protected]> wrote:
I just got a new TinySA Ultra (SeeSii store, Aliexpress) and get doubtful measurement results.
- I selected the cal output of 30MHz.
- connected the cal output with the SA input with an oscilloscope in between (50 ohm termination assumed to be provided by the input).
Problem 1: The scope shows about 40mV RMS which translates to -15dBm, but the TinySA shows the main peak at -35.5dBm on the fundamental 30MHz tone. Why is this?
Problem 2: The signal generator inside the TinySA claims to output -18dBm, but when measured with a scope on a 50 ohm load I measure around 285mV which corresponds to +2dBm. Why?
Problem 3: Using Nano VNA sine generator @140MHz with a 10dBm attenuator (because the NanoVNA outputs about +10dBm when set to 0dB!), measured with scope around 220mV RMS showing the tone in the TinySA of around -20dBm when it should show 0dBm. What is the problem here?
?
I also did the linearity measurement which reveals no significant deviations (a straight line) across the attenuator steps at 120MHz and also at 200KHz and at 2GHz (at the latter frequency the peak showed up at -23dBm and "reduced linearity" warning showed up). All in all, seems no problem with the attenuator.
?
Now to the strange noise floor behavior. Input was terminated and the noise floor started at 600KHz with around -85dBm and transitioned in weird steps towards -120dBm at around 1MHz. Does this indicate a problem?
?
I do have historic background in RF engineering but I am relatively new to these new small devices and appreciate any feedback on what I identified as problems.
?
Thank you in advance
Lutz
?



--
Dave - W?LEV
 

On open ended generator to output to scope -> yes, otherwise no because the TinySA input provides 50 ohm impedance, right?
?
BTW, I just found out that the level difference can be accounted for in the setup. Meaning that the difference between the measured -15dBm from the calibration output that was causing a -35dBm peak can be set using the "Level Offset" in the Config|More menu. After that everything seems correct. I would have expected that the calibration procedure would take care of this rather simple task, unless the absolute level of the calibration oscillator is not known. What do you think? To be honest, I got a little scared with the initial result...
 

Remember to set your scope probe to 1 not 10.
vy73 OZ4HZ
?
 

Not using a probe going in directly with a BNC T and a 50ohm terminator when open ended. Scope input is 1M|15pF. Again, initially I did not expect that after the requisite calibration this TinySA instrument is off by -20dBm which is a lot.
 

开云体育

I concur with Lutz’s findings in Problems 1&2 and the noise floor profile. (I think Erik may have described the reason for that.) Following…

On Jan 21, 2025, at 1:21?PM, lschafer@... wrote:

?
I just got a new TinySA Ultra (SeeSii store, Aliexpress) and get doubtful measurement results.
- I selected the cal output of 30MHz.
- connected the cal output with the SA input with an oscilloscope in between (50 ohm termination assumed to be provided by the input).
Problem 1: The scope shows about 40mV RMS which translates to -15dBm, but the TinySA shows the main peak at -35.5dBm on the fundamental 30MHz tone. Why is this?
Problem 2: The signal generator inside the TinySA claims to output -18dBm, but when measured with a scope on a 50 ohm load I measure around 285mV which corresponds to +2dBm. Why?
Problem 3: Using Nano VNA sine generator @140MHz with a 10dBm attenuator (because the NanoVNA outputs about +10dBm when set to 0dB!), measured with scope around 220mV RMS showing the tone in the TinySA of around -20dBm when it should show 0dBm. What is the problem here?
?
I also did the linearity measurement which reveals no significant deviations (a straight line) across the attenuator steps at 120MHz and also at 200KHz and at 2GHz (at the latter frequency the peak showed up at -23dBm and "reduced linearity" warning showed up). All in all, seems no problem with the attenuator.
?
Now to the strange noise floor behavior. Input was terminated and the noise floor started at 600KHz with around -85dBm and transitioned in weird steps towards -120dBm at around 1MHz. Does this indicate a problem?
?
I do have historic background in RF engineering but I am relatively new to these new small devices and appreciate any feedback on what I identified as problems.
?
Thank you in advance
Lutz
?
<NOISEFLOOR.bmp>
 

On Tue, Jan 21, 2025 at 01:21 PM, <lschafer@...> wrote:
Now to the strange noise floor behavior. Input was terminated and the noise floor started at 600KHz with around -85dBm and transitioned in weird steps towards -120dBm at around 1MHz. Does this indicate a problem?
No this is not a problem.? The tinySA noise floor gets worse below 1 MHz.? due to circuit limitations.? Mine is the same as yours as shown below.?
?
?
?
If you turn on the LNA it goes about 20 dB lower...
?
?
?
 

Thank you Roger, this gives me some piece of mind.
 

On Tue, Jan 21, 2025 at 01:21 PM, <lschafer@...> wrote:
- I selected the cal output of 30MHz.
- connected the cal output with the SA input with an oscilloscope in between (50 ohm termination assumed to be provided by the input).
Not using a probe going in directly with a BNC T and a 50ohm terminator when open ended. Scope input is 1M|15pF. Again, initially I did not expect that after the requisite calibration this TinySA instrument is off by -20dBm which is a lot.

Problem 1: The scope shows about 40mV RMS which translates to -15dBm, but the TinySA shows the main peak at -35.5dBm on the fundamental 30MHz tone. Why is this?
?
You cannot connect this way and expect to get good results.? When you connect a coax cable to the T you are loading down with an impedance that will depend on the length of the coax and the scope termination impedance at the other end.? The impedance will be very low if the length of cable is close to 1/4 wavelength.? You might think that a 10X probe would do the job if placed on the T instead.? That will place 15 pF or so across the tinySA input which is a reactance of 354 ohms at 30 MHz.? This will result in a slight drop in the measure voltage and the display on the tinySA. However the signal on your scope will be fairly noisy . The reason is that -35 dBm is 4 mV RMS and a 10x probe will reduce this to 400 microvolts into the scope input.?
?
Forget the T. The proper way to measure is to use a direct connection from CAL to RF input on the tinySA.? Then use a quality coax cable from the CAL output of the tinySA to the scope input? which is terminated with 50 ohms. make sure your scope is set to 1X or you will read the wrong voltage on the scope.? Also set your scope to read RMS voltage or divide the peak-peak by 2.83.
?
Problem 2: The signal generator inside the TinySA claims to output -18dBm, but when measured with a scope on a 50 ohm load I measure around 285mV which corresponds to +2dBm. Why?
-18 dBm is 28.2 mV.? You measured 10x this so I suspect your scope is set to 10X

?
Problem 3: Using Nano VNA sine generator @140MHz with a 10dBm attenuator (because the NanoVNA outputs about +10dBm when set to 0dB!), measured with scope around 220mV RMS showing the tone in the TinySA of around -20dBm when it should show 0dBm. What is the problem here?
?
Many questions need to be answered.? Your attenuator might not be 10 dBm at 140 MHz.? Measure it with the NanoVNA.?? What is your scope bandwidth?? What type of coax and length did you use (dB loss in cable??).? Did you terminate in 50 ohms and not use a T?
?
Roger
 

Thank you so much! You are absolutely correct! Of course, the scope was still set to the 10x probe endorsing to increase sensitivity on the TinySA.
?
I am totally embarrassed now and feel sorry to have wasted your time. After re-calibration all problems regarding my questions have resolved.
Nonetheless, I was able to hand-calibrate the TinySA with a -20dBm sensitivity difference across the entire range.
?
?
Thank you again
Lutz
 

No need to feel embarrassed. ?The tinySA takes some time to get familiar with. ?We are all in this group to share and learn from each other. ?Have fun with your new device.
?
Roger
?
 

开云体育

In my rush to try to duplicate the test, I too neglected to set my scope to x1. I should have noticed the 20dB discrepancy. <red face>

On Jan 21, 2025, at 6:07?PM, lschafer@... wrote:

?
Thank you so much! You are absolutely correct! Of course, the scope was still set to the 10x probe endorsing to increase sensitivity on the TinySA.
?
I am totally embarrassed now and feel sorry to have wasted your time. After re-calibration all problems regarding my questions have resolved.
Nonetheless, I was able to hand-calibrate the TinySA with a -20dBm sensitivity difference across the entire range.
?
?
Thank you again
Lutz_._,_._,_

1 - 12 of 12