[Elecraft] AT1 Attenuator

[Joe Subich, W4TV]

Looking at the AT1 manual, the switches and traces are certainly
not 50 Ohm transmission lines.  As such, one gets all kinds of
stray behavior.  While you do not say at what frequency or range
of frequencies you made your test, I would suspect you will see
higher SWR as one goes up in frequency and that the departure
from 50 + j0 will be higher on 144 MHz than on 1.8 MHz.

As a general matter, one may be able to "flatten" the SWR somewhat
by placing (and "tuning") shunt capacitors across the input/output
BNC connectors to compensate for the inductance of the traces and
switches.

73,

    ... Joe, W4TV


On 9/22/2023 10:14 AM, Steve Gray wrote:
> I have purchased a second user AT1 recently, largely due to the ruinous TransAtlantic logistics costs, but also the ongoing absence of any support here in The UK.
> 
> Although purely from observation of the circuit topology it seems reasonable to assume that it is not a constant impedance array of resistors, I was quite surprised to see the high SWR readings recorded after connecting a good dummy load and an antenna analyser- encroaching as it does well into double figure readings. My received wisdom being that these devices need to look into a 50 Ohms load to preserve accuracy.
> 
> I have checked all the resistors and they are extraordinarily accurate leading me to believe that it has not been cooked by the/a previous owner, and confirming the quality of Elecraft dedigns. I also noted that the greater the level of attenuation, the lower the SWR is with the lowest attenuation displaying the worst result, and the highest displaying a near perfect match, but I guess that is not surprising because it is attenuating the reflected RF too. What I would like is some reassurance that such elevated SWR readings are to be expected for a nominally 50 Ohms stepped attenuator of this nature set to the lower levels of attenuation.
> 
> Steve, G4VRR.
> e&oe ...