Active Going Passive

[Hoerninger]

There are different ways to perform a crossover with different advantages and disadvantages.
The mostly used passive crossover can be pretty expensive with good capacitors and inductors, especially when going charge coupled.

In comparison an active solution is not expensive anymore. But some do dislike the often used feedback with opamps. This can be a real issue with higher frequencies and exceptional filter characteristics.

A solution might be to use passive filters between buffers. The buffers may be discrete with high performance, but inductors are more often than not necessary. The capacitors will have lower values which show up in minimized costs. The use of electrolytics is avoided. The inductors will have a much smaller construction form but their value (in Henry) will be much higher. The latter might be a restriction as you will not find commercial offers.

The only company which I have found for suitable inductors is FASTRON:
http://www.fastrongroup.com/products...p?category=007
Costs are pretty low.
(Another possibility is to do it yourself with SIEMENS Sifferit - quite an effort)

Any thoughts or experiences?
____________
Peter

[yggdrasil]

I have been doing some simulations on passive line-level networks. Worked out spreadsheets for both hi-pass and lo-pass circuits.

The work was initiated by a post Nelson Pass did on diyaudio.com regarding the use of B1 buffer in a line-level crossover network. I believe he is going to publish a crossover, B2 or B3.

Anyway there are questions, like e.g. If using the B1 or similar buffers there would have to be two separate circuits for balanced operation. There would be no single ended to balanced, nor the other way around. Of course one could add buffers at each end to accomodate this. Then this would raise the question of single ended or balanced internal operation.

Have you read Nelson Pass's article on diy opamps?

[Mr. Widget]

Any thoughts or experiences?

Over the years I have given it a lot of thought but never taken any action. I suppose the sound quality will depend a fair amount on the quality of your buffers. I have used a couple of passive preamps (line stages) and generally didn't like them as well as going active. In both cases the same preamp had a passive and active option.


Widget

[jerv]

Have you read Nelson Pass's article on diy opamps?

Very interesting. Do you have a link?

Espen

[yggdrasil]

http://www.passdiy.com/pdf/diyopamp.pdf

[Ian Mackenzie]

I think a Canadian diy supplier has such a design.

Manley have some studio gear that is passive (equalisers)

Pass talks about the pro's and con's of passive versus active attenuators and in some respects the same issues apply to crossover filters.

If you only need a low order filter you can use the input of a high impediance power amp (valve or fet) and drive it from a low source impediance (preamp)

I think the problem with using passive filters is that they would require more current from the buffer (due to the impediances/ capacitance involved) and as a rule this in turn means the buffer needs the application of feedback to lower the output impediance to keep distortion down. So its a catch 22.

[yggdrasil]

In both cases the same preamp had a passive and active option.

In the case of a crossover, passive circuitry does not necessarily mean there is no active part. It means there are no feedback with the use of op-amps.

There can be active buffers at the output of each section.

A passive pre-amp on the other hand is normally just a gain controlling device, with no active component. This construct usually gives you low input impedance and high output impedance, which is far from optimal. If you add an active buffer with no gain to this construct you will get higher input impedance and very low output impedance...

[Mr. Widget]

In the case of a crossover, passive circuitry does not necessarily mean there is no active part. It means there are no feedback with the use of op-amps.

There can be active buffers at the output of each section.

I realize that and understood Peter's first post to imply that type of design.

A passive pre-amp on the other hand is normally just a gain controlling device, with no active component. This construct usually gives you low input impedance and high output impedance, which is far from optimal.

Widget

[Ian Mackenzie]

Hi Peter,

I think this is what you have been thinking about

Its a simple LC network with potted chokes in the 1-2 henry region.

No doubt it would be excellent with the caveats below covered.

http://www.marchandelec.com/xm46.html
http://www.marchandelec.com/ftp/xm46man.pdf

*Note the input impediance is 1k ohms and mininum output impediance is 10K ohms.

This implies the preamp must be able to drive a large voltage swing into 1 K ohms (inductive / capacitive). That rules out valve preamps without a line stage coupling transformer.

What they dont tell you is Not all preamps (event solid state) will meet this spec and most would only do so with rapidly rising distortion below 2 kohm load. Therefore you would need a high current buffer / driver to get the best out of this arrangement.