Attenuator Calibration

[PurelyNonsense]

Just a quick opinion question for everyone. I was in the Air Force PMEL for a few years before going to commercial calibration, now I work for the Army. My question is on methods of calibrating Attenuators. The Air force uses the method of connecting the Signal Generator to the TI through a couple of Attenuator pads and go straight to the Measuring Receiver. The Army connects the signal source to a power splitter which has port A to the Measuring Receiver's Sensor head and Port B to an 8485D Power Sensor with a 30 dB attenuator in line with the 8485D. If you use the Army's PIMAT program, it does all the set ups itself but the TB manual does not say why you use the 8485D. I believe it's for leveling to ensure that the 0 dB, or whatever initial value is, is the same as before.
My question is, what method is best? I can see the benefits of the Army way but is that introducing too many possible errors? Or does it reduce the drift that may occur between connections?
Thanks

[Bryan]

Just from your wording I think the USAF is more typical. Here is what I typically do:
Source/6dB pad/UUT/6dB pad/Measuring receiver, before test the UUT is not included and the 6dB pads are connected and a relative reading is taken to reference.  Insert UUT between the pads and have at it.
Army procedure sounds like an absolute amplitude is measured at both UUT input & output then calculate the difference.  More potential for error in my book, splitter may contribute as well as dealing with 2 absolute accuracy measurement vs. 1 relative msmt plus it is good practice to use the attenuators for minimizing the effects of reflection, army method appears to lack that.  However for testing multiple frequencies it does look like it might run faster with less operator interaction from removing & inserting UUT as would have to be done using relative at each frequency measurement.    My $0.02.

[CalibratorJ]

I believe it's for leveling to ensure that the 0 dB, or whatever initial value is, is the same as before.
Or does it reduce the drift that may occur between connections?
Thanks

This. And it allows for not only connection repeatability but for multiple devices in rapid succession.

You could just use your E8364C.....

[Bryan]

"You could just use your E8364C....."
ding ding!

[PurelyNonsense]

"You could just use your E8364C....."
I would if I had one. In the Army Cal lab I'm at a T level, meaning they don't give us piddley  peons that kind of equipment. That's the one problem I have with the Army, they segregate the cal techs and even if I have the experience and know how, I'm limited by my Wage Grade level.

[BrodGasko]

Hi...in my case Source/6dB pad/UUT/6dB pad/Measuring receiver, before test the UUT is not included and the 6dB pads are connected and a relative reading is taken to reference.  Insert UUT between the pads and have at it. Splitter may contribute as well as dealing with 2 absolute accuracy measurement vs. 1 relative msmt plus it is good practice to use the attenuators for minimizing the effects of reflection, army method appears to lack that.

turn-key pcb