EDN Published Active BPF in error

[mawyatt]

Here's a good example why folks shouldn't always rely on well known and respected technical resources such as EDN for unquestionable technology references.

https://www.edn.com/simple-low-pass-filters-tunable-with-a-single-potentiometer/


This is stated as an Active Tunable LPF which it's not, as pointed out by Steve it's not a LPF but a BPF!! We also noted that the circuits in Fig 1 and 2 can't work as any kind of filter except the dreaded NPF (No Pass Filter) 

Not sure why the editors didn't catch this, as it's not a LPF and the Op-Amp inputs are reversed in Fig 1 and 2.

We decided to play around as this is an interesting Active BPF that can be "tuned" by a simple resistor (R2 in our schematic). According to the article this filter should be credited to the late brilliant Bob Pease from National Semi.

Here's some LTspice simulations using a LM358, which we show how "bias loading" the output can improve the Stop Band Rejection.

Also a quickly done circuit analysis for those readers interested, didn't spend much time on this so please check for errors.

Anyway, fun little circuit indeed

Best,

[mawyatt]

Couldn't resist kludging something together on a Protoboard. PNG 32 is without LM358 output bias resistor, PNG 34 is with such. Note the good agreement with the actual circuit and the simulations.

Best,

[mawyatt]

Also take a look at the other figures in the article, some detailed analysis and one realizes these are good examples of how NOT to design an active BPF.

https://www.edn.com/simple-low-pass-filters-tunable-with-a-single-potentiometer/

Edit: Added is simulation with R3 fixed @ 10K and just the Op-Amp model changed from circuit in EDN article Figure 3 (also graph Figure 4). Note the excessive frequency shift just due to Op-Amp change, and this is a low frequency where Op-Amp characteristics should be minimal in a well designed BPF

So IMHO this article is a poor example of BPF active designs, shame on EDN for publishing such

Best,

[RoGeorge]

Aside from the EDN article quality, I find the filter very interesting, thank you for bringing it!   

At a first look it reminded me of the Baxandall tone correction http://www.tubebooks.org/file_downloads/Baxandall.pdf yet it is different.  Tried to find the Bob Pease article referred by EDN, but couldn't find the corresponding Electronic Journal number:  "Sallen R.P., Key E.L. “A Practical Method of Designing RC Active Filters”. IRE Transactions on Circuit Theory, 1955, Vol. 2, № 1 (March), pp. 74–85." 

Fooled around with the BPF schematic a little (in LTspice for now, my try was with OP07, for no particular reason).



I was looking for the influence of
- stepping the values of R2 (schematic a) while R3 is kept constant, then
- stepping R3 (in schematic b) with R2 constant, then
- stepping both R2 and R3 (in opposite directions) so to keep the frequency constant and vary the quality factor

The response for Both V(out) and V(sum) nodes was plotted for all 3 (a/b/c) types of resistor(s) stepping.  As for the practical build, straight to the wish list again. 

Would like to use digital potentiometer(s) for R2 and/or R3, and try some audio effects for the fun of it, maybe adding more filters in parallel to make some sort of equalizer.  Or, maybe make it full analog, adding a few filters in parallel using a volume potentiometer for R2 and R3, to make some sort of graphic equalizer.

What I would like to try the most, would be to use this BPF filter as a sinusoidal RC oscillator, instead of a Wien osc.  Would be easier to adjust the frequency.

[langwadt]

Not sure why the editors didn't catch this, as it's not a LPF and the Op-Amp inputs are reversed in Fig 1 and 2.

why would you think the editors know anything about electronics? they are just placing filler between the ads
it's "Design Ideas" anyone can send in those and get paid the huge sum of $200 if it gets published

[mawyatt]

Agree it's an interesting little circuit, and thought you would be interested

This was the 1st we'd heard of this BPF circuit from Bob Pease, so we weren't aware either.

If you look at the Transfer Function in 1st post and take partial derivative wrt "s", thus d(T(s))/ds = 0 should yield the minima and maxima points.

By observation the minima are obvious at s=0 and s=infinity, after all it's a BPF

The maxima is more involved and should yield R3/(2*R1).

If you include op-amp parameters like GBW product, frequency compensation and such, then things get much more involved analytically.

Best,

[RoGeorge]

Meanwhile traced back the Bob Pease writing about the filter:
- hand drawn schematic and the tale - "What’s All This Absurd Filter Stuff, Anyhow?" - https://archive.org/details/Bob_Pease_Lab_Notes_Part_7/page/n67/mode/1up
- page 237 of his book Analog Circuits (World Class Designs) referred in the "What’s All This Absurd Filter Stuff, Anyhow?" - https://d1.amobbs.com/bbs_upload782111/files_17/ourdev_472160.pdf

[mawyatt]

Not sure why the editors didn't catch this, as it's not a LPF and the Op-Amp inputs are reversed in Fig 1 and 2.

why would you think the editors know anything about electronics? they are just placing filler between the ads
it's "Design Ideas" anyone can send in those and get paid the huge sum of $200 if it gets published

Well they certainly did when we were allowed to published stuff in EDN way back, in fact they had excellent reviewers including the late Bob Pease to make sure things like this article didn't get in!!

In fact Bob Pease reviewed one of our allowed articles on a Reactive Clock Driver and rejected it thinking it was non-functional. The circuit was a special clock driver for CCD devices which employed clock edge energy recovery (CCD device multiple phase clock inputs have massive capacitance and must be driven with large voltage swings). This circuit achieved the most efficient operation dynamically, however was very simple and upon casual appearance shouldn't work.

When we challenged Bob with "proof" (measured ~96% efficient) he backed down and it was published. You can probably find this article in the EDN archives with a search.

https://www.eevblog.com/forum/projects/reactive-clock-driver-circuit/

Certainly hope EDN hasn't resigned to non-technical editors without expert reviewers, as this raises question with anything they publish

Best,

[mawyatt]

Meanwhile traced back the Bob Pease writing about the filter:
- hand drawn schematic and the tale - "What’s All This Absurd Filter Stuff, Anyhow?" - https://archive.org/details/Bob_Pease_Lab_Notes_Part_7/page/n67/mode/1up
- page 237 of his book Analog Circuits (World Class Designs) referred in the "What’s All This Absurd Filter Stuff, Anyhow?" - https://d1.amobbs.com/bbs_upload782111/files_17/ourdev_472160.pdf

Great find, thanks

Bob Pease's BPF circuit was completely missed in our past career, so nice to find something new from the past 

We only had one direct encounter with Bob mentioned above, and recall the only time he backed down. His electronics common sense and intuition, along with Jim Williams, were amazing and likley never to be duplicated

Best,

[mawyatt]

Short term memory isn't very good

We actually discussed the Reactive Clock Driver here some time ago. Fun read if interested, part of which was discovered by accident!!

https://www.eevblog.com/forum/projects/reactive-clock-driver-circuit/

Best,

[schmitt trigger]

Meanwhile traced back the Bob Pease writing about the filter:
- hand drawn schematic and the tale - "What’s All This Absurd Filter Stuff, Anyhow?" - https://archive.org/details/Bob_Pease_Lab_Notes_Part_7/page/n67/mode/1up

Bob Pease was famous for its crooked hand drawn schematics, but this one must have achieved an Olympic record! 

[mawyatt]

And his famous Rats Nest circuits.

Recall one, think it was a Voltage to Frequency converter, or visa versa, that looked like a big 3D ball of leaded components soldered with full length leads. He wanted to minimize stay capacitance and this certainly placed the components as far apart as practical with nothing but air between them and well above any ground plane

Best