The REAL Inflammatory Post of the Week... (gyrator filter fun)

analog

unread,

Nov 20, 2005, 9:25:16 PM11/20/05

to

Jim Thompson wrote:
> The Phantom wrote:
>> Jim Thompson wrote:
>>> The Phantom wrote:
>>>
>>>> If you use a GBW as high as 10MHz, then what advantage is there
>>>> to use the A3A topology (your preferred topology), rather than
>>>> the A1A topology?
>>>
>>> Some time ago it was demonstrated that my topology came closer
>>> to ideal than the others, for a given GBW.
>>
>> I found this posting by you from July 2002. Is this the
>> demonstration you are referring to?
>> -----------------------------------------------------------------
>> "Comparison of the "classic" gyrator BP topology and mine versus
>> OpAmp GBW is posted at...
>>
>> http://www.analog-innovations.com/SED/GyratorConfigCompare.gif
>>
>> The "dash" number in the Window title is the GBW of the OpAmps.
>>
>> My output is the green trace, the "classic" is the red trace.
>>
>> Note that the lower the GBW the more peaking the "classic"
>> topology has in comparison to my topology. (In all cases only
>> GBW was varied, phase-margin was always 45°.)
>>
>> Some poking around shows that the peaking comes from the
>> development of a negative real term in the synthetic inductor,
>> both filter topologies exhibit this effect, the "classic" just
>> more so.
>>
>> ...Jim Thompson "
>> -----------------------------------------------------------------
>>
>> When you said "classic" gyrator BP topology, which of the 24
>> topologies I posted at the beginning of this thread were you
>> referring to? Back in the July 2002 thread Winfield Hill gave
>> ASCII schematics of the "Riordan" topology (my designation B2B)
>> and the "Antoniou" topology (my designation A1A).
>>
>> You say: "The "dash" number in the Window title is the GBW of the
>> OpAmps."
>>
>> When I look at that .gif, I see dash numbers of -5, -10, -20 and
>> -40. Does -5 mean 5 MHz, etc.?
>>
>> This 30 KHz filter is apparently different than the 20 KHz one
>> shown in:
>> http://www.analog-innovations.com/SED/GyratorFilters-BP-LP-HP.pdf
>>
>> Could you redo the comparison of your topology to what you've
>> called the "classic" using the 20 KHz filter values, or else give
>> the complete circuit used for the 30 KHz filter comparisons?
>
> I'll dig thru that and see if I can recall what was used.

Well, Jim, what did you find??

It seems that the Phantom might be on to something, er, interesting
here, and given your obvious skill with Pspice, if needed, you
could have easily regenerated the results several times over by now.

Certainly it's not the case, but the lack of attention to this
thread in deference to other threads about "white flag[s]" and backs
with "yellow stripe[s]" down them could leave one with an ironic
impression.

Please don't drop this very fascinating (and very *on topic*) thread.

Regards -- analogspiceman

Jim Thompson

unread,

Nov 21, 2005, 9:41:36 AM11/21/05

to

On Sun, 20 Nov 2005 18:25:16 -0800, analog <ana...@ieee.org> wrote:

>
>Jim Thompson wrote:
[snip]

>> I'll dig thru that and see if I can recall what was used.
>
>Well, Jim, what did you find??
>
>It seems that the Phantom might be on to something, er, interesting
>here, and given your obvious skill with Pspice, if needed, you
>could have easily regenerated the results several times over by now.
>
>Certainly it's not the case, but the lack of attention to this
>thread in deference to other threads about "white flag[s]" and backs
>with "yellow stripe[s]" down them could leave one with an ironic
>impression.
>
>Please don't drop this very fascinating (and very *on topic*) thread.
>
>Regards -- analogspiceman

I haven't had time to search. It takes no time to throw sand in the
air, but it DOES take time to work a project that's (as usual)
stretching the capabilities of a process.

Jim Thompson

unread,

Nov 21, 2005, 12:04:25 PM11/21/05

to

On Sun, 20 Nov 2005 18:25:16 -0800, analog <ana...@ieee.org> wrote:

>
>Jim Thompson wrote:
>> The Phantom wrote:
>>> Jim Thompson wrote:
>>>> The Phantom wrote:

[snip]

>>>
>>> Could you redo the comparison of your topology to what you've
>>> called the "classic" using the 20 KHz filter values, or else give
>>> the complete circuit used for the 30 KHz filter comparisons?
>>
>> I'll dig thru that and see if I can recall what was used.
>
>Well, Jim, what did you find??
>
>It seems that the Phantom might be on to something, er, interesting
>here, and given your obvious skill with Pspice, if needed, you
>could have easily regenerated the results several times over by now.
>
>Certainly it's not the case, but the lack of attention to this
>thread in deference to other threads about "white flag[s]" and backs
>with "yellow stripe[s]" down them could leave one with an ironic
>impression.
>
>Please don't drop this very fascinating (and very *on topic*) thread.
>
>Regards -- analogspiceman

I found it in archive, and posted it at....

http://www.analog-innovations.com/SED/GyratorConfigTest-SCH.pdf

It appears we were examining insertion of a series R to add a zero at
the GBW point, and also some other possible error sources.

I don't have any other information on it. Maybe someone can come up
with the original thread so we can see the sequence of discussion.

Or you can post some specific questions/situations and I'll try to
address them.