I have a step attenuator that I built years ago (30?) and with the NanoVNA, I was finally able to test it.

Shorted, the loss is about .2 DB @ 30 MHz.
The 10, 20 and 30 DB step attenuators are surprisingly close (according to my NanoVNA SAA2N). I'm testing between 1 MHz and 30 MHz.
I have the scale set at 10 DB/Division. (Wow, is this VNA handy!!)

However, the 60 DB attenuator starts at about -56 DB at 1 MHz and at 30 MHz, it shows -32 DB. I'm guessing that some of the resistors in the attenuator that I built may be inductive. Seems to make sense.

Trouble is, how do I test to see if the resistors are inductive. I've googled it, but I can't seem to find a way to test the resistors for inductance.

--
Mike, W0IH

Measure it like any other component.
Also 60dB in a single stage is hard.

On Thu, 28 Jan 2021 at 00:45, Michael <mcfroston@gmail.com> wrote:

> I have a step attenuator that I built years ago (30?) and with the
> NanoVNA, I was finally able to test it.
>
> Shorted, the loss is about .2 DB @ 30 MHz.
> The 10, 20 and 30 DB step attenuators are surprisingly close (according to
> my NanoVNA SAA2N). I'm testing between 1 MHz and 30 MHz.
> I have the scale set at 10 DB/Division. (Wow, is this VNA handy!!)
>
> However, the 60 DB attenuator starts at about -56 DB at 1 MHz and at 30
> MHz, it shows -32 DB. I'm guessing that some of the resistors in the
> attenuator that I built may be inductive. Seems to make sense.
>
> Trouble is, how do I test to see if the resistors are inductive. I've
> googled it, but I can't seem to find a way to test the resistors for
> inductance.
>
> --
> Mike, W0IH
>
>
>
>
>
>

It’s more likely parasitic capacitive coupling. Trying to get 60dB out of a single stage is asking too much. This is why many attenuators have a shield between stages. But this won’t work for a single stage.

Clark Martin
KK6ISP
Yet another designated driver on the information super highway.

> On Jan 27, 2021, at 3:45 PM, Michael <mcfroston@gmail.com> wrote:
>
> However, the 60 DB attenuator starts at about -56 DB at 1 MHz and at 30 MHz, it shows -32 DB. I'm guessing that some of the resistors in the attenuator that I built may be inductive. Seems to make sense.
>
> Trouble is, how do I test to see if the resistors are inductive. I've googled it, but I can't seem to find a way to test the resistors for inductance.

It is nearly impossible to make a wide band 60db attenuator in a single stage that is why you can't buy them. What you can buy however are 30db attenuators install 2 of them back to back with a shield in between them and you will be fine. If you have the space you could remove the excising resistors and install a small board instead using 2 of these with a shield .https://www.minicircuits.com/WebStore/dashboard.html?model=LAT-30%2B unless you look for higher power in that case there are ample of similar fixed attenuators that will do up to hundreds of watts.

I would estimate that the problem is not using screening box and inductive coupling between the input 50 resistor and output 50 ohm resistor in case of pi structure or capacitive coupling. Mostly the 60dB is not made in one stage but more.

I do have a screen box that I'd built to isolate the different sections. I was quite surprised that the 10, 20 and 30 dB attenuators were so VERY close.
As you can see, I used parallel resistors in many cases to get the values that I needed.

The 60 dB section is at the far right. At 1 MHz, I have -56 dB, and at 30 MHz, I have -32 dB attenuation.
So, I'm back to my original question, difficulty aside, is there a way to test to see if a resistor is inductive?

--
Mike, W0IH

What stops you from taking that resistor and measuring it on your VNA?
Am I missing something?


So, I'm back to my original question, difficulty aside, is there a way

> to test to see if a resistor is inductive?
>
>
>

I owe you gentlemen an apology.
You were right! The 60 dB is difficult to attain.

I finally got around to experimenting with the 60 dB attenuator. I put in several different designs, and they all ended up looking very much the same as the original, even when I tried to add capacitors (caps didn't make much more than a few dB of difference). Once it reached 30 MHz, the attenuation changed from the -56 dB to about -32 dB.

I did use the NanoVNA to measure the attenuation of the different resistors, and they were flat across the spectrum 1 - 30 MHz.

I finally gave up and changed that section from the 60 to a 5 dB attenuator. I figured that since I have the 10, 20, 30 (which totals 60 dB, reasonably accurately), I figured that I might need a 5 dB instead, especially considering the power levels that this unit was designed for (under 1/2 watt).

Anyway, thanks for the help!

--
Mike, W0IH