I am working on 500mhz signals. At these frequencies the 0.05pF of capacitance in a SMD resistor adds up. Today I am fighting inductance of resistors. Most data sheets say nothing about inductance and some just say “low inductance”. One sheet said “0.5 to 5nH” depending on resistance value.

Is there information on inductance verses resistor size? The real question; is it better to use one 1W resistor, four 1/4W resistors in parallel or 30 small 0603 in parallel. (paralleling reduces inductance)

 

You can't cheat. Size matters.
https://www.mouser.com/datasheet/2/427/ch-1764162.pdf

 

Fun reading: vishay

 

0201 & 0402 parts are smaller than flyspecks in the pepper, but they have lower inductance.
Hope you have a good VNA to characterize ALL your parts

 

Hope you have a good VNA to characterize ALL your parts

My network analyzer does not go high enough and the "company" is just starting to let engineers in to use the good stuff.

0402 parts are smaller than flyspecks in the pepper, but they have lower inductance.

Question: If I have 40 of the 0402 parts in parallel the trace inductance is high. OR I can use one big 2W resistor where the resistor has too much inductance but the traces are 5mm wide. I am thinking that at the end it will not make much difference.

 

https://www.mouser.com/datasheet/2/21/C1A101050R0J4A0_DataSheet(Rev_B)-1221860.pdf

 

https://www.mouser.com/datasheet/2/21/C1A101050R0J4A0_DataSheet(Rev_B)-1221860.pdf

While you posted that (thanks) I was reading a paper on 2512 verses 1225 resistors. The resistor in the right has twice the bandwidth.

This morning I moved to a 1020 part because the inductance is lower and there is more solder. It appears the solder resistance is a problem with low value resistors.

 

I have a few rf oscillators that I simulate on my desktop. I really need to construct an oscillator that runs at 6MHz and I have found the same issue with resistors. Specs say little about the several uH that exist in their parts. How am I suppose to shop for parts?

 

We built complete tube radios with old school carbon resistors. 6MHz is easy to design for. Turn off the sim and start building stuff.

 

I have a few rf oscillators that I simulate on my desktop. I really need to construct an oscillator that runs at 6MHz and I have found the same issue with resistors. Specs say little about the several uH that exist in their parts. How am I suppose to shop for parts?

The parasitic inductance is in the low, single digit nH not uH range. In oscillators, resistors are used mostly for biasing purposes and stray inductance doesn't really matter. For a 6MHz oscillator you can buy/use any resistor, even lumped if you prefer.

 

The parasitic inductance is in the low, single digit nH not uH range. In oscillators, resistors are used mostly for biasing purposes and stray inductance doesn't really matter. For a 6MHz oscillator you can buy/use any resistor, even lumped if you prefer.

Well, I won't use wire wound resistors but just about everything else is good.️

 

Yes, biasing that's what they are for. I measure 3.5uH on everything I've bought so far. I will get some carbon resistors and try again. Thanks for the reassurance.

 

I measure 3.5uH on everything I've bought so far.

I don't see how that kind of series inductance would even be possible. How did you get 3.5uH?

 

Extech 380193 LCR meter. it's calibrated.

 

Extech 380193 LCR meter. it's calibrated.

What do you read with a short?

 

0

 

0

What's the calibration check standard at the uH level?

 

I am not sure what you mean. meter goes as low as 0.1uH. The calibration is automated via set button presses.

 

I am not sure what you mean. meter goes as low as 0.1uH. The calibration is automated via set button presses.

Unless the device has fixed internal standard values it compares values to, that's not a 'calibration', it's really an operational check.
I mean what is your external check standard (a stable inductor of known value, that been verified by a external calibration standard or source) to make sure the button press process gives accurate values to a device under test.

 

I am gonna grab carbon resistors and find a standard to go with my meter, but I am sure it's as accurate as claimed.