Misterious 'resistor network' of shorts in keithley 195 input multiplexer

[TheUnnamedNewbie]

I am working on repairing a Keithley 195 Multimeter. In the schematics, there is a component (R120) that is referred to as "Resistor, Thick Film Network" in the service manual. On the schematic, however, it is not really marked as resistor (there are 'pin' markers on the schematic). The package is a standard plastic DIP 16, marked 'DALE TF-146' ( and when I measure it, it is just everything shorted together except for the NC pins.

What is the purpose of this component? Is there any 'metrology' reason for this? I would think it is not some kind of thermal EMF compensation, because then it would have to be just in the paths that have no junctions, right?

[GeoffreyF]

I am guessing but it is simply a method of connecting all those points together in a way that facilitates manufacturing and keeps wiring short.  It's probably made as a custom resistor network just as a way to get it all done.   Your measurements are certainly consistent with the content of the schematic.  If you are suspecting the part is at fault for some reason, it could be due to some internal crack, fault, cold solder, or damage.  You would see this at the test points while wiggling it around or with changes in temperature.  If it were me, I would look elsewhere though.

[TheUnnamedNewbie]

If you are suspecting the part is at fault for some reason, it could be due to some internal crack, fault, cold solder, or damage.  You would see this at the test points while wiggling it around or with changes in temperature.  If it were me, I would look elsewhere though.

I'm not suspicious of this part, I'm simply trying to trace out the PCB (as there is no silk screen on the PCB) to figure out what components go where, and was surprised to find this component like that.

[Kleinstein]

This part is really odd. I have two possible explanations: one would be using is like 0 Ohms SMD resistors to allow testing before that part is fitted. This might only apply to the prototypes with a socket for R120.

Another possibility is  that they initially planed for some resistors (e.g. to control charge injection) and later found out that R=0 is working as well.

[TheUnnamedNewbie]

I am guessing but it is simply a method of connecting all those points together in a way that facilitates manufacturing and keeps wiring short.  It's probably made as a custom resistor network just as a way to get it all done. 

The thing is - looking at the PCB, they could easily have routed it all without this component, it's not like they need it as some kind of 'array of jumpers'.

This part is really odd. I have two possible explanations: one would be using is like 0 Ohms SMD resistors to allow testing before that part is fitted. This might only apply to the prototypes with a socket for R120.

Another possibility is  that they initially planed for some resistors (e.g. to control charge injection) and later found out that R=0 is working as well.

My unit has a socket for R120 - does that mean that I have a prototype unit?

Your theory of the charge-injection sounds plausible. Perhaps it was then omitted to reduce noise/improve speed?

[Kleinstein]

The socket could still be there for production testing, or to simplify fault finding. 

Going to 0 Ohms resistance could be cost cutting. There are already some 900 K of resistance at the other side of the JFETs, so that there would not be much noise from small (e.g. 100 K) extra resistors.