Electronics > Projects, Designs, and Technical Stuff

3rd order Sallen Key Unity Gain LPF - stability question

<< < (4/6) > >>

Wimberleytech:

--- Quote from: Jay_Diddy_B on April 25, 2020, 02:27:11 pm ---Hi,

(Wimberleytech - congratulations on reaching 1000 posts !!!)

I will give you the answers without too much of an explanation:



--- End quote ---

Nice work, and thanks.

Here is a spice file reflecting with Jay showed us.

Jay_Diddy_B:
Hi,
Further to my last post.

Component Variation - Tolerance

Monte Carlo Analysis can be used to simulate random component variation.

In LTspice the syntax is

{mc(nominal value, tolerance)}

{mc(10K,0.05)} - this is a 10k \$\Omega\$ 5% resistor

A dummy parameter is used to run the simulation a number of times.

Each time the model runs it uses a different set of parts. It is like building the circuit many times and comparing the results.

Monte Carlo analysis



Frequency Domain



Time Domain



I don't see any hint of instability from 5% components.

Regards,
Jay_Diddy_B

Someone:

--- Quote from: JDW on April 24, 2020, 11:36:43 pm ---My schematic currently shows 1% tolerance resistors, but I am inclined to use 5% tolerance resistors for reasons of cost.  (Yes, even pennies matter.)  But when I plug in different resistance values, all within a 5% tolerance, into that Japanese LPF calculator, there are some instances where the calculator shows an oscillation frequency.
--- End quote ---
a) what tolerance do you expect on your capacitors? how does that relate to the variation/tolerance of the other components?
b) the online "calculator" is probably incorrect


--- Quote from: JDW on April 25, 2020, 12:31:20 am ---I must repeat what I said in my opening post about the reason I used large resistors and small capacitors in combination with an op-amp, as opposed to just using a cheap R-C LPF instead of an op-amp -- "capacitor discharge time" affects my ADC acquisition time.
--- End quote ---
I'm not sure what you mean by "capacitor discharge time", as an idealised linear system the input from the sensor to the ADC can be completely described by a frequency response. Notably, most ADCs are improved by adding more capacitance on their input pin (reducing the impedance driving the input).

JDW:

--- Quote from: Someone on April 25, 2020, 11:37:10 pm ---a) what tolerance do you expect on your capacitors? how does that...

--- End quote ---

Although I appreciate your comment, it's clear you didn't read previous comments.  I went into detail about tolerance.  So for yourself and all others who later find this thread, please read through all previous comments.  A lot of your questions are probably already answered there.


--- Quote from: Jay_Diddy_B on April 25, 2020, 02:54:03 pm ---I don't see any hint of instability from 5% components.
--- End quote ---

Jay, I am humbled and eternally grateful for the time you so kindly spent on those simulations.  That is the kind of hand-holding my feeble brain really needed.  Thank you very much.

It also seems clear that the calculator either was incorrect when it cited an Oscillation frequency as I varied the 330k resistors (within a 5% tolerance) OR their "Oscillation" means something other than what I am thinking.  But thank you, Jay, for running the 5% resistor tolerance simulations to show the filter is indeed perfectly stable.  Thank you also for showing that so long as I use 5% tolerance capacitors (e.g., C0G) all will be well.  BRAVO!

Jay_Diddy_B:
JDW,

The simulation is showing the filter is stable with an op-amp that is working correctly.

The LTspice simulation do not include the MCP601.

If the real circuit doesn't work like the model, try a different op-amp.

Regards,
Jay_Diddy_B

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod