🚀 go-pugleaf

On Fri, 11 Jan 2002 01:47:14 GMT, "Spehro Pefhany"
<speff@interlog.com> wrote:

>In sci.electronics.design Russell Shaw <rjshaw@iprimus.com.au> wrote:
>
>> I need to make small resistors (10-200ohms) operate at a
>> few GHz, and so be non-inductive/capacitive. I can do it
>> by depositing carbon paste with a stencil.
>
>You should be able to buy this stuff from the makers of inks for
>thick-film hybrids. You can silk screen it on (control of viscosity,
>pressure and squeegee angle and spacing from the substrate are all
>important in controlling the thickness). Then it is "fired" at a high
>temperature. You can trim said resistors with abrasive or lasers. 
>
>Is it really such a big deal with regular resistors? I can't imagine an
>0402 or 0201 resistor 10-200R resistor having much inductance or
>capacitance at those frequencies. 
>
>Best regards,


Regular surface-mount resistors are quite good up to a few GHz. One
important thing to worry about is the transition into the resistor.
For example, if you wanted to terminate a microstrip that was 0.1"
wide, it's better to use a bigger resistor that matches trace width,
or a couple of smaller ones in parallel, to avoid the inductance
associated with necking down the trace into one skinny resistor. That
sort of thing starts to matter as you approach maybe 10 GHz.

Some people think it's best to mount the resistors upside down, so the
resistance element is against the pcb. 

John