[sdiy] Another SSM2164 question

[Tom Wiltshire]

Thanks guys, that's helpful.

What I really needed to know was whether these are some "magic  
values" that might cause all sorts of weirdness if altered. I'm not  
thinking of leaving the network out, but I needed to know if I can  
replace it with easier-to-get cap values like 470pF or 680pF. If it's  
just a HPF then I can. Especially if I move the resistor value in the  
other direction to keep the overall frequency similar (which, by  
chance, is what my 560R/470pF combination does).

I shall also experiment with these values to see if it has any effect/ 
influence on the resonance's frequency dependence.

Thanks again,
Tom


On 12 Aug 2008, at 17:09, Tom Farrand wrote:

> Tom,
>
> Am guessing this is a roll-off network to keep the chip from
> oscillating at a high frequency.  From the datasheet, check out
> figure 14.  The device still has gain beyond 1MHz.  However, note that
> a complete phase change from 180 to -180 degrees happens about 7MHz.
> That means the device tends to (will) oscillate at high frequencies,
> which is a "bad thing".
>
> Since the desired frequency span of interest is out to maybe 50KHz
> (without introducing unwanted artifacts to the sound), what happens
> much higher is of no concern, unless something "bad" can happen.  The
> combination 500 Ohms and 560pF works out to a 90 degree phase shift at
> about 568KHz ... ten times the higher limit of "interest" in the
> signals being passed (50KHz).  That keeps the device from oscillating
> at low levels at higher frequencies while still passing all the
> frequencies of interest.
>
> AD engineers have thought that through and offer a cheap and
> predictable cure.  I suggest you incorporate their network on each
> input as prescribed.  It only costs pennies and it does work.  By not
> going into a detailed explanation as to "why" this happens, one feels
> a bit left out.  I suspect the "why" has something to do with how the
> chip works without violating existing patents.  A full disclosure
> explanation might have to reveal proprietary information which is
> something AD might not want to do.
>
> All guesses ... YMMV.
>
> Peace.
> Tom Farrand
>
>
> On Tue, Aug 12, 2008 at 9:54 AM, Tom Wiltshire  
> <tom at electricdruid.net> wrote:
>>
>> Hi all,
>>
>> The subject says it all really. The VCAs in the SSM22164 require a  
>> 500R resistor and 560pF capacitor to ground on each VCA input, as  
>> well as a resistor to convert voltage to current (30K in the  
>> datasheet).
>>
>> Now, what I'd like to know is what do the 500R/560pF do and how  
>> are their values calculated? All the datasheet says about it is  
>> that "a 500R resistor in series with a 560pF capacitor must be  
>> added from each input to ground to ensure stability". Great.
>>
>> Practically, it seems you can vary the values a bit without  
>> incident (I've used 560R/470pF and nothing bad has happened) but  
>> it's a bit hard to know when you're not sure what they're supposed  
>> to prevent.
>>
>> The datasheet is at:
>> http://www.analog.com/static/imported-files/data_sheets/SSM2164.pdf
>>
>> Thanks,
>> Tom
>>
>>
>> _______________________________________________
>> Synth-diy mailing list
>> Synth-diy at dropmix.xs4all.nl
>> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>