[time-nuts] OT - Old Hatfield 2105 Step attenuator specs
Alberto di Bene
dibene at usa.net
Fri Jan 10 10:50:55 EST 2014
On 1/8/2014 11:13 PM, Alan Melia wrote:
> /Hi Alberto it is quite interesting to continue that test with no attenuator
> but the shells of the coax plugs connected together. I would guess with the
> gear you have the resultant would be at least 120dB down........but this is
> not the case for all signal generators!/
Hi Alan,
quite true. I performed the test you suggested, using as generator a Rohde&Schwarz SMDU
that has a calibrated output down to -140 dBm, so it must be well shielded...
I used 10 MHz as frequency, and, given that the settings of this forum do not allow HTML (why ?)
these are the links to the screen captures stored on my Dropbox account.
As selective voltmeter I used the ELAD FDM-S1 receiver together with, guess what... Winrad :-)
This is what I see with the Hatfield attenuator set to its maximum, i.e. 100 dB :
https://dl.dropboxusercontent.com/u/15089947/hatfield-100dB.gif
Setting it to 0 dB gives this result :
https://dl.dropboxusercontent.com/u/15089947/hatfield0dB.gif
So you can see that the difference between the two measures is just 88 dB, not the theoretical 100...
And it is almost all to be attributed to internal leakage of the attenuator, because, excluding the attenuator,
and just connecting together the two BNC shells, leaving the center pin unconnected, gives this :
https://dl.dropboxusercontent.com/u/15089947/hatfield-shells.gif
So the Hatfield attenuator IMHO can be fruitfully used only if you do not pretend from it the utmost
precision at high attenuation settings.
73 Alberto I2PHD
More information about the time-nuts
mailing list