Oh and can someone link to a 0-1Ghz LNA worth buying. And also a link for a backup 50ohm calibration kit? Or how do I test if i buy one to see if its 50ohms??

Links to attenuator worth buying?

And can someone explain the VLM-33W-2W-S+ part to me? That ones going way over my head.

Hi,

Our of the above mentioned items, I bought a set of EMC SMA probes for a few €.

They are very useful for servicing radio equipments without disturbing their circuitry.

You can measure frequencies, tune,,,

Best regards,

Georges F6DFZ?

Erik.... are you really the guy that made these? That's awesome to be a beginner and get an answer from the creator. So thank you for that.

What accessories would you recommend?

Is a VLM-33W-2W-S+ something I should look into getting? Here's a link to what was mentioned.

https://www.minicircuits.com/WebStore/dashboard.html?model=VLM-33W-2W-S%2B&srsltid=AfmBOorOt7cyUDyF3pdAGc_SH-yOIsH96Bhd95iHfDJP0W3xAr0t1ozV

?

If anyone has any other recommendations can you post some links to them so I don't mess up ordering? I think I already ordered some wrong parts lol

Dear all,

?

Finally the parts arrived, and I started the IC changes.

?

First, I removed the U22 0 ohm resistor, working as jumper, and place the RF Switch, and do recall. During recall it fault again, so I decided to change the LNA, U16.

After changed it and recall, the calibration from 100kHz to 5.327GHz goes succesfuly, and I passed to calibrate abovde 5.327GHz.

I've no idea that you need to generate signal from other device, so I used an ADALM-Pluto to generate the signal and power enough for calibration, and succesfuly it worked again.

?

So now, my device is perfectly working and calibrated and I can play again with it.

?

Many thanks to everyone for the help and we're in touch.?

?

Sincerely,

?

Jorge EA1FWI.

No need for self righteousness. There are RF environments where strong signals can flood an antenna. I can see why such devices would be useful. But I am nowhere near such a RF environment.?

Trying to replace

Input RF switch U22:? AS179-92LF
LNA U14: BGA2817

might need to be replaced
Attenuator U28: PE4312C-Z

Good luck.

Impossible to tell

--

Designer of the tinySA
For more info go to

Thanks. I inaugurated my cheap hot air station and the RF switch gently came off. Now i could jumper it to check the attenuator, but i prefer waiting for the spare part and soldering paste, and keep soldering on the board a minimum. Btw the accident occured with a unfinished homebrew LISN (TVS diodes didn't fit...), so it taught me that a transient limiter does make sense :)

Is this one a clone? because if not, I'd be keen to try and replace the attenuator, or go through the trouble shooting more closely and try and fix it.?

On Sat, Mar 15, 2025 at 10:50 PM, a4notebook wrote:

I bought tinysa ultra from facebook. It fails attentuator and signal level. Can it be fixed? can it be used as a signal generator?

Maybe, no

?

--

Designer of the tinySA
For more info go to

What you are observing is indeed possible as the switch can fail in?different ways

?

--

Designer of the tinySA
For more info go to

Hi?

?

I bought tinysa ultra from facebook. It fails attentuator and signal level. Can it be fixed? can it be used as a signal generator?

Hi Erik,

yes i performed the manual routine and think it is useful. I have a low signal level on my Ultra 407, down by 40 dB, i.e. at -75 dBm. Now your manual says that even with a bad RF switch, if the signal stays within 6 dB of that level when the LNA is enabled, then the LNA is fine. However, on mine, when i enable the LNA, it even goes up to where it should be, -35 dBm. Have you ever had this before ?

The manual testing of the attenuator behaves as it should, the noise floor goes up by as much as the attenuation, maintaining the level @30 MHz.

I ordered new switches.

My conclusion is that my switch makes a low impedance only when it is in normal (non-LNA) mode, and this is rather unusual because typically you expect it makes a low impedance in either position. I hope it is not the attenuator that causes the low impedance but else behaves normally in producing the correct relative levels ?

Comment regarding the interval time of 32.6msec that I used. ?An exact multiple of 1/60 would have been 33.3¡­..msec but to get a stable display I had to set it to 32.6msec. ?Difference is likely related to accuracy of the TinySA clock and/or timing errors/accuracy within the TinySA code.

?

just FYI,

Don wd8dsb?

?If I remember right, the HP network analyzers have a ¡®skip isolation¡¯ option in their cal routine, making me think it won¡¯t make much difference in most cases - especially in a Nano.

Hi Erik and gang,

?

I owe Erik an apology as not to long ago I said the interval triggering feature that Erik implemented for me a few years ago on the TinySA was not working properly on my TinySA Ultra.? In a recent post that Erik responded to someone on a different topic he mentioned setting the RBW to 850 KHz when using Zero Span mode and that made me revisit the interval triggering feature as I had ignored the RBW setting when using interval triggering.

?

Today I tested the interval triggering feature with the display sweep time set to 20 msec, trigger interval set at 32.6msec which is a multiple of 1/60 (60 being the power line frequency in the US), set the RBW to 850 KHz, and the Zero Span frequency set to 146 MHz which is a frequency I know I can hear RFI that's generated by an LED light dimmer in my family room that I use for testing the interval trigger feature.? I then turned on the LED lights and I can easily see the light dimmer going into conduction at a frequency of 120 Hz which would be each half cycle of the 60 hertz sinewave and it's a very stable display (apparently in sync with the 60 Hertz line frequency which is what I was hoping for).? I've attached a screen shot showing the display generated by my LED light dimmer driving LED can lights (antenna used was a homebrew Moxon beam designed for 146 MHz).

?

I originally wanted this interval trigger feature to help track down RFI like power line noise that's in sync with the electrical grid frequency, and will now continue my work on this subject now that I better appreciate the setting of RBW when in zero span mode.? Having the ability to trigger in sync with the electrical grid is beneficial when there are other signals present which are not necessarily in sync while you are attempting to see signals in sync with the grid.

?

Bottom line is that the interval trigger feature appears to working great and having the RBW set to 850 KHz really helps.

?

Big Thanks to Erik,

Don (wd8dsb)

WinSpectrum III v3.0.1.1 ¨C Update & Fixes
Some users reported that nothing was displayed on the graph when using tinySA (basic). The issue was caused by WinSpectrum's default scan setting of 450 points, which the tinySA (basic) does not support (it has a 290-point limit).

Fixed in version 3.0.1.1: WinSpectrum III now detects tinySA (basic) failure when the number of points exceeds 290, adjusts the scan to 290 points, and correctly displays the graph¡ªno user intervention required.

Additionally, this update includes several improvements and fixes:

? Added LNA on/off command.
? Start/Stop Frequencies are now displayed at the bottom of the graph.
? Fixed logfile reading issue.
? Various marker-related fixes and improvements.
? Added Delta calculation for markers.

Tested successfully on tinySA (basic) version 1.4-40, tinySA Ultra (non plus) version 1.4-156 and tinySA Ultra+ versions 1.4-196 and 1.4.197.

About the marker's 'Tolerance' field
Some users may wonder why WinSpectrum III includes a Tolerance (%) setting for peak markers when this parameter is not explicitly found in most VNAs, Spectrum Analyzers, or tinySA devices.

This feature is not mandatory and does not modify the raw measurement data. Instead, it is an optional tool that helps locate peaks within a user-defined range rather than just a single frequency point.

For example, if a peak frequency is set to 10 MHz with a 5% tolerance, WinSpectrum will search for the highest peak within the 9.5 MHz to 10.5 MHz range, rather than just at 10 MHz. This can be useful in cases where signals fluctuate slightly due to environmental factors or device tolerances.

The goal of this feature is to make peak identification more flexible and user-friendly, especially when dealing with noisy signals or spectral variations. Of course, if users prefer a fixed peak frequency search, they can simply ignore this setting.

73?

So help me understand... This is basically an attenuator, but only for signals above a certain threshold, letting weaker ones through without attenuating them??

?

Matthew?

KD6KVH?