UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > General Vintage Technology > Components and Circuits

Notices

Components and Circuits For discussions about component types, alternatives and availability, circuit configurations and modifications etc. Discussions here should be of a general nature and not about specific sets.

Closed Thread
 
Thread Tools
Old 6th Jan 2022, 5:43 am   #21
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,799
Default Re: Active filters - Inductors instead of resistors.

The oscillator Bill Hewlett developed was actually the first amplitude-stabilised Wien Bridge. The Wien Bridge already existed, what Bill added was automation of level control.

Like all oscillators the Wien bridge relies on complex numbers and phase shifts in order to work.

The circuit called the phase shift oscillator is quit different, it is a loop of an amplifier with three concatenated RC meshes, and not a bridge at all.

A long time ago I sat in a packed conference room while Bill himself spoke of how he'd come up with the idea. He spoke with a very slow drawl, but it was worth waiting for each word.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 6th Jan 2022, 9:58 am   #22
Ed_Dinning
Dekatron
 
Join Date: Jan 2004
Location: Newcastle upon Tyne, Tyne & Wear, UK.
Posts: 8,171
Default Re: Active filters - Inductors instead of resistors.

Hi Gents, as pointed out J operator led to an understanding of phase shifts in AC power circuits; A operator and the conversions between A & J operators allowed the analysis of unbalanced 3 phase circuits.
The application of these maths operations lead to much better protective systems for high power 3 phase lines (imagine a fuse for a 500KV/ 2000A power line). In the early days they used quite simple LCR (filter) circuits to operate protective relays.
Bias systems then allowed the development of protection against through faults where only the line "block" with the fault was isolated.
Mertz and Reyrolle in Newcastle pioneered the application of these systems.

To further make the brain hurt, and based on A & J operators, we have symmetrical components. This analyses a 3 phase waveform, under unbalanced/ fault conditions into positive, negative and zero frequency waveforms (probably akin to reflected power at RF); again simple passive filters are used

As a final thought there are also a range of other operators that are all prime/ odd number roots of -1. I have no idea of their designations, but they are now being used for the analysis of triplen (3N) phase systems in power supplies.
Airbus use 18 phase and above, derived from 3 phase alternators on the A400M to reduce harmonics and increase waveform purity.

I'm immensely glad I retired when I did

Ed
Ed_Dinning is offline  
Old 8th Jan 2022, 1:50 am   #23
Synchrodyne
Nonode
 
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,943
Default Re: Active filters - Inductors instead of resistors.

Returning to the original theme, it may be noted that active RC filters were developed for applications where passive LCR types were difficult to implement in respect of the inductance part. Thus they were a means to avoid the use of inductances altogether.

The opening part of the 1955 Sallen & Key paper was:

“In the frequency range below about 30 cps, the dissipation factors of available inductors are generally too large to permit the practical design of inductance-capacitance (LC) or resistance-inductance-capacitance (RLC) filter networks. The circuits described in the following pages were developed and collected to provide an alternative method of realizing sharp cut-off filters at very low frequencies. In many cases the active elements can be simple cathode-follower circuits that have stable gain, low output impedance and a large dynamic range.”

The prior (1950) work by Shumard of RCA had a similar basis:

“It has been necessary to use band=pass (B.P.) filters with relatively very low cutoff frequencies. Three ranges were required: 0.005 to 0.5, 0.01 to 1, and 0.03 to 3 cycles per second. To build filters of the conventional L-C type would require impractically large if not unrealizable values. On the other hand a filter using a combination of R-C networks and a direct-current (dc) amplifier with feedback, it has been found, results in a relatively small, practical unit.

“The fundamental principle involved is operation of the dc amplifier near the 1,0 point of the Nyquist loop, resulting in controlled regeneration without oscillation. In conjunction with external filtering the combination is then made to give a pass band comparable to the L-C filter used at higher frequencies.”


(The article, “Design pf High-Pass, Low-Pass and Band-Pass Filters using R-C networks and Direct-Current Amplifiers with Feedback” was carried in RCA Review 1950 December.)


I wonder if subsequently someone found situations where it was beneficial to use active LR or LCR filters?


Unsurprisingly, both Sallen & Key and Shumard used operator j as an integral part of their calculations.

I am not sure when active RC filters of these types were first used in audio application, but Baxandall seems to have been an early user (based upon the RCA work) in his WW 1955 preamplifier, with which there was probably a long gestation period, given that it was intended to accompany his 1948 power amplifier. The pre-amplifier included high-pass (30 Hz) and a low-pass (5 and 7.5 kHz) filters wrapped around the same valve. For anyone of an anti-j disposition, the formulae shown did not include j (unless any of the square rooting involved negative radicands.)

Still, it should not be forgotten that the use of complex numbers was in fact a simplification exercise. To quote from the introduction of Steinmetz’ 1893 paper, Complex Quantities and their Use in Electrical Engineering:

“In the following, I shall outline a method of calculating alternate current phenomena, which, I believe, differs from former methods essentially in so far, as it allows us to represent the alternate current, the sine-function of time, by a constant numerical quantity, and thereby eliminates the independent variable "time" altogether from the calculation of alternate current phenomena.

“Herefrom results a considerable simplification of methods. Where before we had to deal with periodic functions of an independent variable, time, we have now to add, subtract, etc., constant quantities—a matter of elementary algebra—while problems like the discussion of circuits containing distributed capacity, which before involved the integration of differential equations containing two independent variables :”time” and "distance". are now reduced to a differential equation with one independent variable only, "distance", which can easily be integrated in its most general form.”



Cheers,
Synchrodyne is offline  
Old 8th Jan 2022, 6:25 am   #24
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,799
Default Re: Active filters - Inductors instead of resistors.

One of the best books never written:

There was a series of excellent articles on active filters By F E J Girling and E F Good in Wireless World in the early/mid 70's.

WW's publishers missed a major opportunity in not issuing these as a collection.

One I've used rather seriously: https://en.wikipedia.org/wiki/Leapfrog_filter

It's a derivative of the biquad and the Tow-Thomas, the latter of which you'll find as an example in LTspice. Don't dismiss these things as too complicated, look on them as examples of how far people found it worth going to avoid inductors. Compared to some methods of synthesising inductorless bandpasses, the leapfrog filters are low on criticality and quality as house-trained.




You can't make an all-resistor filter. Resistors live firmly in the present, insensitive to time, and time is 100% of what filters are all about. Inductors and capacitors involve time.

Can you do a filter without signal-handling inductors or capacitors?

Yes.... digital filters, but they too need a time input, a clock. And the clock comes from an oscillator, and the clock involves.... time-related components. L, C, or maybe a quartz crystal.

j-notation is simply one way of expressing the time relationship of these components. There are others, but it is the easiest to relate ot other things and a great convenience. There are other ways, but they are a lot harder for no delivered benefit.

Once you'v picked up the language and techniques, you are off and away. There are lots of places you can apply it and watch old difficulties crumble.

Complex numbers, s-planes, poles and zeroes have a reputation for being hard, so they discourage most peoplle. But if you learn to handle them, then your knowledge puts you at a premium over those who ran away. Ironically, they aren't hard. They're easier than the alternatives. They're just very very weird. If you can believe in a few impossible things before breakfast, then you'll find you can do a few impossible things.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 8th Jan 2022, 10:48 am   #25
kalee20
Dekatron
 
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 7,059
Default Re: Active filters - Inductors instead of resistors.

It's well noted that for very low frequencies, an active filter may be the only practicable way of realisation.

Quote:
Originally Posted by Synchrodyne View Post
I wonder if subsequently someone found situations where it was beneficial to use active LR or LCR filters?
Active LCR filters - very much so! After all, that's what reaction is in an RF amplifier - a method of sharpening-up the response curve where the 'R' is the parasitic loss resistance of the 'L' and the active amplification allows this to be overcome.
kalee20 is online now  
Old 8th Jan 2022, 1:00 pm   #26
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,799
Default Re: Active filters - Inductors instead of resistors.

Look up "Q-multiplier" There's a long history.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 8th Jan 2022, 3:57 pm   #27
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: Active filters - Inductors instead of resistors.

Quote:
When building a 3rd order active filter or any order filter come to that like a Sallen Key or Butterworth is there any advantage to using inductors instead of resistors in the filter circuit?
I've never tried using inductors in an active filter that uses op-amps but I would think it would be risky. There will be self capacitance within the active devices so you always get some capacitance for free, probably enough to cause resonance problems. Also, there could be voltage magnification due to resonances and this might stress the active devices and also this could make the design prone to instability.

Obviously one can make an active filter using L and C parts and a good example is a tuned RF amplifier using a transistor as the active device.

The nearest I've got to using inductors in an op-amp based active filter is to embed a passive elliptic LC filter in the active loop filter of an RF synthesiser in order to suppress reference sideband energy on the synthesiser output. This worked but was very prone to noise pickup. The Kvco was about 10MHz/V so even a tiny amount of noise pickup could modulate the synthesiser. The choice of inductor technology helped a lot here but the final result was still less than perfect. This was about 30 years ago and to solve this issue I helped develop a current pump circuit that had much lower reference sideband energy and so the horrible LC filter was no longer needed.

It is possible to buy chips designed for active filter use from Maxim that only need resistors to be fitted. I remember designing some critical AF filters like this way back in the 1990s. The capacitance was already built into the Maxim chips. I think the MAX274 is such a chip although there ought to be better devices available today.

Quote:
It's well noted that for very low frequencies, an active filter may be the only practicable way of realisation.
Agreed. LC filters can be realised down at AF frequencies but I've never tried to design one lower than this. The passive LC filter in the PLL loop filter had a cutoff of about 2kHz and a zero at 10kHz. I can't remember the design impedance. In terms of passive 50R lowpass filters I guess it's possible to build something with a cutoff below 100Hz but I've never tried. A cutoff of 1kHz, 3kHz or 10kHz should be much easier for a 50 ohm design impedance.
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 8th Jan 2022, 4:08 pm   #28
Ed_Dinning
Dekatron
 
Join Date: Jan 2004
Location: Newcastle upon Tyne, Tyne & Wear, UK.
Posts: 8,171
Default Re: Active filters - Inductors instead of resistors.

In the 40's when electrolytic production of Aluminium was ramping up there were serious problems with control of alternators used in this process due to the sub power frequencies involved.
One alternator in the power station was usually designated as the "control alternator" for this purpose and had a fast acting control system. This was, at the time normally based on L&C filters, but some systems used a type of servo control.
Frequencies were normally below 2Hz.

In the 60's when very long HVAC lines were being built, it was found that sub harmonics of the power frequency were being developed (akin to RF reflections in a line). This would affect the system stability; again L C filters were used.
I don't think I've ever seen this type of filter implemented as an active type.


By about 2010 these were being implemented in DSP, which is the state of things at present.


One of the local IEE members, John Cure did a lot of work on these problems and published papers for Mertz & Macellen (Electrical consulting engineers) and later from his own company on these topics

Ed
Ed_Dinning is offline  
Old 8th Jan 2022, 4:54 pm   #29
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: Active filters - Inductors instead of resistors.

Quote:
Frequencies were normally below 2Hz.
Wow that is low... the LC ratio goes down with lower design impedances so this is one way to prevent the inductor values from becoming huge at really low frequencies.

The lowest frequency (passive LC, 50 ohm) lowpass filter I've designed had a cutoff at about 5kHz. I also looked at making a 1kHz version but never got around to buying the parts for the 1kHz version. I still have the 5kHz lowpass filter in a small enclosure and it does get used fairly often.
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 8th Jan 2022, 4:59 pm   #30
Ed_Dinning
Dekatron
 
Join Date: Jan 2004
Location: Newcastle upon Tyne, Tyne & Wear, UK.
Posts: 8,171
Default Re: Active filters - Inductors instead of resistors.

Hi , yes Z values on these equipments would be pretty low, of the order of an ohm or so on the current transformer side as fed from 5A CT secondaries

Ed
Ed_Dinning is offline  
Old 9th Jan 2022, 1:18 am   #31
kalee20
Dekatron
 
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 7,059
Default Re: Active filters - Inductors instead of resistors.

Quote:
Originally Posted by G0HZU_JMR View Post
I've never tried using inductors in an active filter that uses op-amps but I would think it would be risky. There will be self capacitance within the active devices so you always get some capacitance for free, probably enough to cause resonance problems.
I haven't either- though now I see it as a challenge...

Of course, some op-amps have internal inductors as part of the internal frequency compensation for stability. The Philbrick P2 is one. But as a 'black box' it doesn't really count, and it's invisible as long as you use the op-amp well below its unity-gain cross-over frequency.

Quote:
Originally Posted by G0HZU_JMR View Post
Obviously one can make an active filter using L and C parts and a good example is a tuned RF amplifier using a transistor as the active device.
True, though to me that's just a passive filter preceded (or followed) by an amplifier. Whereas an active filter has feedback paths within the filter itself - such as the Q multiplier that David referred to.

A band pass filter formed by a parallel LC circuit, with feedback by a variable-gain amplifier, is interesting. The transfer function can have two real poles - as gain is increased these come together and further gain increase causes them to become complex. Still more gain sees the imaginary part of the poles stay constant, but the real part decreases, passes through zero, and moves to the right-half-plane. A right half-plane pole, of course, indicates an oscillatory response which increases with time - you've made an oscillator!
kalee20 is online now  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 5:35 pm.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.