Products > Test Equipment
MiniCircuits Attenuator : BW-40N100W+ : Test Procedure
(1/2) > >>
killingtime:
Hi,

I picked up a used MiniCircuits Attenuator 40dB (BW-40N100W+). The data sheet can be found below.

https://www.minicircuits.com/pdfs/BW-40N100W+.pdf

It shows typical performance to 4GHz of 40dB +/- 0.5dB, with guaranteed attenuation accuracy at +/- 1.6dB. So that's 38.4 - 41.6 dB.

I'm testing the attenuator on a SAA3021X+. Overall amplitude spec for this instrument is 0.7dB. The results I'm getting are attached and look very different to the spec sheet. An average of -37.5 dB with a dip to -35.75 dB and a peak to -44.1 dB. That's over 8.35 dB of variation and well outside spec.

Either the attenuator is damaged, or I'm not testing it correctly.

I've set the RBW to 10KHz to lower the noise floor. BW step is 1MHz. Trace average is 100. Tracking gen is on and I connect both N cables together via a barrel connector. Let the unit complete one full sweep, then activate 'Normalization' to null out any cable loss. I get a flat trace after this. I then insert the attenuator between both cables keeping the barrel connector in the signal path. RG142 cables an N connectors are spec'd to 3GHz.

Can anyone see anything wrong with this test procedure?

Thanks.

joeqsmith:
Looks like your SAA3021X+ was spec'ed for 9 kHz ~ 2.1 GHz but you show it sweeping to 3GHz:   

https://siglentna.com/spectrum-analyzers/ssa3000x-plus/    

Regardless if you modified the Siglent or not, how did you verify your setup prior to trying to measure the Mini-Circuits device?
killingtime:

--- Quote from: joeqsmith on May 21, 2023, 01:28:19 am ---Looks like your SAA3021X+ was spec'ed for 9 kHz ~ 2.1 GHz but you show it sweeping to 3GHz. Regardless if you modified the Siglent or not...

--- End quote ---

Yes, you can extend the frequency range of this device by entering a valid licence key. No modification required. This instrument shares the same hardware as the 3.2GHz version SAA3032X, which is why the key works.


--- Quote from: joeqsmith on May 21, 2023, 01:28:19 am ---how did you verify your setup prior to trying to measure the Mini-Circuits device?

--- End quote ---

Via the Normalization feature. The analyzer sweeps to 3GHz with the cables in the signal path and you get a flat trace back. Without Normalization on I see a gradual attenuation up to 3GHz, which is what you'd expect for RG142 and some connectors. Also put a different 20dB attenuator in using the same cables. Same settings. Got 20dB with 0.7 dB variation. Not too bad. TekBox20dB.jpg attached.

After some more trial and error I think I've figured out why I got the original attenuation plot.

1. I noticed one of the N connectors was a bit lose. It was already reasonably tight. Tightened it up by no more than 1/8 a turn and the huge dip at 2GHz disappeared.

2. Reduced the internal attenuator to 0dB (from -20) and the plot smooths out and gets closer to -40dB. 0Atten.jpg attached. The attenuator is now (just) within the 1.6dB spec given by MiniCircuits.

With the cables tightened up, I put the -20dB internal attenuator back on, and got the 20dBAtten.jpg plot. Much more variation and the whole plot shifts +1dB.

The internal attenuation is having an effect but I'm not sure at this stage why. Could be that with the TG set at -20dBm, and with -20dB of internal attenuation and -40dB of external attenuation I'm pushing the signal close to the noise floor of the analyzer (-90dBm). The internal pre-amp is off.


joeqsmith:
Member Xrunner was testing some low cost 40dB attenuators with a Siglent as well.   I posted some similar data using one of the low cost VNAs.   

https://www.eevblog.com/forum/rf-microwave/cb-and-ham-radio-techs-love-their-bird-wattmeters/msg4423153/#msg4423153
Weston:
The cable and input port of the VNA are not a perfect match to 50 ohms, which can cause a change in S22 with frequency.

By changing the internal attenuator measurements you might be changing the input match of the VNA itself.

Its pretty common to take S21 measurements with a ~10dB or so attenuator at the DUT end of the return coax cable (side connecting to the input port of the VNA). Try calibrating the through measurement with the 20dB attenuator in place and see if that makes a difference. If the noise floor is an issue you can enable averaging. 
Navigation
Message Index
Next page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod