Thanks for the this comment, after seeing this I did similar tests on my Siglent SAA 3021X and what I am seeing is the Step Attenuator starts to act "goofy" at -60dB
You forget about the insertion loss of your cables and connections. You should run it with 0 dB ( nothing pressed) to figure out what is your cables and connectors loss. Also, you need to increase your input power to measure 90dB insertion loss, because your tinySA floor of measurement is about 80~90 dBm, so your source should be about 20dBm, in order for you to able to measure it with tinySA. Warning, do not burn your TinySA it can only take 0dBm, so you first have to measure your source signal with say 40dB Attenuation to be about say -20dBm then switch to 90dB attenuation then you have to measure it at about -50dBm. Also you may turn off the the LNA and AGC in your TinySA, when you just measuring attenuation. Also Power-meter might be a better choice for measuring attenuation than Spectrum Analyzer, unless you want to do flatness vs frequency. Also you have to do temperature testing. That is you have to put your Attenuator in temperature control chamber and do that in several temperatures.
Since it is an attenuator, and you have the insertion loss at different frequencies, just facto that in to your switch selection for the ultimate -db loss intended.
I recently purchased a similar attenuator from ebay with a model number KT3.0-90/1S-2N(2W). That's a hefty model number. Any way the unit was tested on an HP8711b SA and it checked okay at around 2.5% worst case. Insertion loss was also okay with a high around 280MHz and 440MHz of 1.45dB. That's not bad for Ham radio guys. I did normalize with the test cables so only the DUT and the adapters were in play. My test power was 10dBm.
Sounds reasonable. It tested the attenuator with my nanovna (NanoVNA V2 PLUS4) and at the 90db setting of the attenuator the display showed a bit of noise. Nothing like yours but noisy never the less. My nanovna has a specified dynamic range of 90dB so the measurement was pushing the nanovnas' limit. My nano has an output of 0 dBm. Well there is nothing like a "real" spectrum analyzer. My HP8711b was given to me at a ham club meeting about a year ago. The local communications company had no need of it and passed it on to one our club members who had no need for it. He brought it to our meeting and asked if anyone - well he didn't finished his question. I grabbed it snarling and growling and took it home. It's wonderful. It has been upgraded as per Imsai Guy to a vna. I enjoy your channel and watch it often. Ted Ivester, N7NSL.
Hey thanks for good postmortem. Which attenuator can you recommend with range upto 100 db. And one more, does the attenuation capacity changes over frequency - like more towards VHF and less in UHF range, or it remains same? (sorry not very literate in this subject)
I haven't found an adjustable that is; usable and affordable yet. I'll do a video when I find one. Depending on how the attenuator was designed you will see varied performance at different frequencies.
I'm curious what that sweep looks like if you sweep less than the entire range. I'd like to see it at HF, VHF, UHF bands. Smaller range with same amount of data points.
H'mm. Are you sure that the unit you're using to measure this isn't the problem? I don't expect a device that is made up of resistive antenuators to be noisy unless the resistors or switches are open circuit.
@@TheSmokinApe Yeah, I suspect you are right. I have a rotary attenuator that goes all the way down to 135dB. I measured it with a sig gen and a good HP spectrum analyser. I didn't have a VNA back then so haven't tried this experiment. I do have Nano's but I also have a HP VNA that I could try with too. One more for the experiments TODO stack.
Thanks for posting, interesting video. I don't think I'll be getting one. I'll keep using my inline attenuators. I built a 50 db RF sampler and could not measure it with the NanoVNA H 4. I have no idea why it would not work. I recalibrated the NanoVNA and still would not work. I then tested it with my spectrum analyzer and the sampler was spot on down 50 db. I like my NanoVNA and have found many uses for it, but it does have its issues from time to time. Barry, KU3X
@@TheSmokinApe The noise floor on my NanoVNA was well below 50 db and on 10 mhz on down, the 50 db sampler showed 79 db. The spectrum analyzer proved it was 50 db from 1 mhz up to 50 mhz.. For me, the NanaVNA failed to provide an accurate answer. Barry, KU3X
I thin the attenuator may be just fine. You have simply attenuated the signal below the noise floor of the analyzer. The analyzer doesn't have 90db of dynamic range.
I had to lol, the Nanovna doesn't have enough dynamic range to measure the 90db. There is probably nothing wrong with the device. The best way to measure is using a test signal and a spectrum analyzer.
Another good test bench video, Ape! It’s always good to identify subpar products…thanks!
Thanks for checking out the video Frank!
Great video Ape
Thanks for checking it out Hayden!
The VNA noise floor is much greater than -90dB. You cannot use it to measure such small signals.
Thanks for the this comment, after seeing this I did similar tests on my Siglent SAA 3021X and what I am seeing is the Step Attenuator starts to act "goofy" at -60dB
great review as usual. :-D
Thanks for checking it out Thump!
You forget about the insertion loss of your cables and connections. You should run it with 0 dB ( nothing pressed) to figure out what is your cables and connectors loss. Also, you need to increase your input power to measure 90dB insertion loss, because your tinySA floor of measurement is about 80~90 dBm, so your source should be about 20dBm, in order for you to able to measure it with tinySA. Warning, do not burn your TinySA it can only take 0dBm, so you first have to measure your source signal with say 40dB Attenuation to be about say -20dBm then switch to 90dB attenuation then you have to measure it at about -50dBm. Also you may turn off the the LNA and AGC in your TinySA, when you just measuring attenuation. Also Power-meter might be a better choice for measuring attenuation than Spectrum Analyzer, unless you want to do flatness vs frequency. Also you have to do temperature testing. That is you have to put your Attenuator in temperature control chamber and do that in several temperatures.
Great points, thanks for the detail Aduedc!
Yup, good ol' ground one normalization.
Hi, are you referring to the Tiny Spectrum Analyser in your comment by mistake instead of Nano VNA ?
cheers .😊
You're getting some new fancy instrument Ape!
It matches my fancy pants!
Since it is an attenuator, and you have the insertion loss at different frequencies, just facto that in to your switch selection for the ultimate -db loss intended.
Curious if you ran a test on each switch independently, it seems like you have a bad switch or resistor in there somewhere.
I don't recall doing that, not a bad idea.
I recently purchased a similar attenuator from ebay with a model number KT3.0-90/1S-2N(2W). That's a hefty model number. Any way the unit was tested on an HP8711b SA and it checked okay at around 2.5% worst case. Insertion loss was also okay with a high around 280MHz and 440MHz of 1.45dB. That's not bad for Ham radio guys. I did normalize with the test cables so only the DUT and the adapters were in play. My test power was 10dBm.
Oops! Forgot to say that the DUT was only swept to 500MHz. Us Hams don't knowingly dabble in those higher frequencies.
I think part of my "problem" with this one was too low of an input signal and not enough dynamic range on the VNA. Thanks for the comment!
Sounds reasonable. It tested the attenuator with my nanovna (NanoVNA V2 PLUS4) and at the 90db setting of the attenuator the display showed a bit of noise. Nothing like yours but noisy never the less. My nanovna has a specified dynamic range of 90dB so the measurement was pushing the nanovnas' limit. My nano has an output of 0 dBm. Well there is nothing like a "real" spectrum analyzer. My HP8711b was given to me at a ham club meeting about a year ago. The local communications company had no need of it and passed it on to one our club members who had no need for it. He brought it to our meeting and asked if anyone - well he didn't finished his question. I grabbed it snarling and growling and took it home. It's wonderful. It has been upgraded as per Imsai Guy to a vna.
I enjoy your channel and watch it often. Ted Ivester, N7NSL.
That is one heck of a score!
I plan to use a 60 dB one for radio direction finding
They supposedly work very well for that 👍
Hey thanks for good postmortem. Which attenuator can you recommend with range upto 100 db. And one more, does the attenuation capacity changes over frequency - like more towards VHF and less in UHF range, or it remains same? (sorry not very literate in this subject)
I haven't found an adjustable that is; usable and affordable yet. I'll do a video when I find one. Depending on how the attenuator was designed you will see varied performance at different frequencies.
there is also a similar 0-82db step attenuaror that uses switches instead of buttons, wonder how that performs?
I have been on the look out for a different one, thanks for watching Andrew.
Looks almost like a waveguide, with attenuation that your just putting in the way or not
I'm curious what that sweep looks like if you sweep less than the entire range. I'd like to see it at HF, VHF, UHF bands. Smaller range with same amount of data points.
James, curiosity is what killed the cat...
H'mm. Are you sure that the unit you're using to measure this isn't the problem? I don't expect a device that is made up of resistive antenuators to be noisy unless the resistors or switches are open circuit.
Nope, I’m not. In retrospect it could have been the dynamic range of the unit 👍
@@TheSmokinApe Yeah, I suspect you are right. I have a rotary attenuator that goes all the way down to 135dB. I measured it with a sig gen and a good HP spectrum analyser. I didn't have a VNA back then so haven't tried this experiment. I do have Nano's but I also have a HP VNA that I could try with too. One more for the experiments TODO stack.
that stack keeps getting bigger 👍
Thanks for posting, interesting video. I don't think I'll be getting one. I'll keep using my inline attenuators.
I built a 50 db RF sampler and could not measure it with the NanoVNA H 4. I have no idea why it would not work. I recalibrated the NanoVNA and still would not work. I then tested it with my spectrum analyzer and the sampler was spot on down 50 db.
I like my NanoVNA and have found many uses for it, but it does have its issues from time to time.
Barry, KU3X
Someone posted that the noise floor wasn't low enough on the Nano, don't know if that is true or not. Thanks for checking out the video Barry.
@@TheSmokinApe The noise floor on my NanoVNA was well below 50 db and on 10 mhz on down, the 50 db sampler showed 79 db. The spectrum analyzer proved it was 50 db from 1 mhz up to 50 mhz.. For me, the NanaVNA failed to provide an accurate answer. Barry, KU3X
Could this have been tested on a TinySA?
Yeah, it could have.
I thin the attenuator may be just fine. You have simply attenuated the signal below the noise floor of the analyzer. The analyzer doesn't have 90db of dynamic range.
I think you are right, thanks for the info Robert!
I had to lol, the Nanovna doesn't have enough dynamic range to measure the 90db. There is probably nothing wrong with the device. The best way to measure is using a test signal and a spectrum analyzer.
Fair point, thanks for watching German
It almost looks like a musical instrument.
It kind of does, thanks for checking it out TO
That be a good trick for the one team that always wins the fox hunts ☺️😂
Haha
Parallel resistance?? 1/R
Beats me!
Hi Ape,
Judging by your tests, I think you will need to get some discrete attenuators that will add up to the total you need. Stay safe. 73 WJ3U
I think you may be right, thanks for checking it out Don!