Discrete JFET "passive" R2R attenuator

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Just a spontaneous idea. What do you think? Is it worth?

Possible schematic (only 4 bits are shown)
ejrHlJf.png
 
If you like making things complicated yes, rather use a CD4069 = all-in :)
Otherwise what's the fun? :)

Don't forget to take account for the internal resistance of the drivers.
Mona

If you mean JFETs' resistance, it's insignificant in comparison to ladder resistance. Working currents are very small, so JFET resistance nonlinearity is insignificant too.

If you mean driver stage, it can affect input impedance a bit, because when the given bit is grounded, it steals some current from the source through the gate resistor.
 
If you like making things complicated yes, rather use a CD4069 = all-in :)
Don't forget to take account for the internal resistance of the drivers.
Mona
The 4069 will not work: the input protection diodes will conduct towards Vdd, which is the signal, when a high logic state is applied.

The 4049 would solve this problem, because of the different protection scheme, but another one remains: there is a substrate diode between Vdd and Vss, and this would limit the signal to less than 1Vpp.

The 4007 could work, but only for two pairs: the third one is hardwired to Vss and Vdd, and thus unusable.

A better option is the CD4053: it has three usable switches, there are no limitations provided it is correctly supplied, and it incorporates a logic voltage translator
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.