Baxandall puzzle

[joaquins]

  I just don't know... but, isn't interaction between controls what we like about vintage stuff, or is it the only thing we hate about them but we love all the rest? Jus't saying, is such a problem that interaction?

  In the other hand, as you seen the compensation with the output impedance, using a dual fader, where one would be used as fader and the other as compensation isn't an option? It would add a little complex in a much 'better' (with less interaction) and clever design I guess.

  There are other options in passive eq, with more or less interaction with source, but you already know that, plus I really love how a baxandall reacts in a 12AX7 feedback, it's ones of the options I would like to implement in my mixer, but for me silicon is acceptable (only discrete, I also have my mind problems) and I won't have this problem. In my project there is a 22pin slot with a HV rail for this kind of applications, far away from the original pins of the 51X, don't cry! so no DC-DC converter needed.

JS

 

[ricardo]

.. all my existing passive EQs can be driven from a 10K potentiometer, usually a channel fader.

Ian, I'd really like to see your passive EQ whose response is unaffected when fed from a 10k fader.

Can you post a schematic ?

I am sure you would. I believe I never made such a claim.

My interpretation of your statement about "existing passive EQs" is that you claim, unlike a naked Baxandall, they are unaffected by the fader position.

But if your "existing passive EQs" change their response with different fader positions, then all is right with the world and no unicorns were sighted  ;D

 

[abbey road d enfer]

  I just don't know... but, isn't interaction between controls what we like about vintage stuff, or is it the only thing we hate about them but we love all the rest? Jus't saying, is such a problem that interaction?

The question is : "How much interaction is too much interaction?".
For me, having the boost varying from 9 dB to 21 dB depending on the fader position is too much. Some may think it's acceptable.

In the other hand, as you seen the compensation with the output impedance, using a dual fader, where one would be used as fader and the other as compensation isn't an option? It would add a little complex in a much 'better' (with less interaction) and clever design I guess.

Ian expressed his will to "making decent tube gear affordable", which leads him to condone the use of a 2nd tube. I guess adding another fader is not really going in the right direction.
And anyway, I don't know of a fader that's capable of providing negative impedance. 

I really love how a baxandall reacts in a 12AX7 feedback

What do you mean by "reacts"? The whole Baxandall EQ principle is based on the network "reacting" with the active stage, in a way that's appropriate for the task at hand. Do you suggest there's something more?

 

[ruffrecords]

I think it would be useful to differentiate between different types of interaction. First there is the interaction of the overall EQ with what drives it and what loads it. My poor man's EQP1A for instances, assumes it is driven by a low source impedance and that it is loaded by a relatively high impedance. I have even documented the effect of different load impedances on its response.

The REDD EQ assumes a source impedance of around 2500 ohms to achieve the correct boost and cut values and if you drive it from a low impedance there is provision for fitting a 2500 ohm resistor. Driving from a variable source impedance like a fader will affect the boost and cut values. Boost will be slightly reduced and cut will be slightly increased. The Helios 69 EQ is very similar.

The other interaction is between the controls themselves. If the responses of the different bands overlap then some interaction is inevitable. However, when you set all the EQ controls to flat, you would expect the response to be flat no matter what the driving impedance. All my passive EQ designs achieve this goal. However, the initial simulations of the Baxandall EQ indicated that this was not the case unless the driving impedance was relatively low. My first goal was therefore to see if this interaction could be eliminated or at least reduced to an acceptably low level. The last circuit I posted achieves this reasonably well.

The other issue with the Baxandall is the degree to which the response is affected by driving impedance. At present this is unresolved.

Cheers

Ian

 

[ricardo]

However, when you set all the EQ controls to flat, you would expect the response to be flat no matter what the driving impedance. All my passive EQ designs achieve this goal.

Ian, I would very much like to see a passive EQ which achieves this goal.

Please could you post a rough schematic?

 

[joaquins]

Ian expressed his will to "making decent tube gear affordable", which leads him to condone the use of a 2nd tube. I guess adding another fader is not really going in the right direction.
And anyway, I don't know of a fader that's capable of providing negative impedance. 

But if the output is connected as you posted in reply nº12 and as the Zs in the feedback loop the second wiper of the pot with low side connected directly to the output and the other connected with a resistor equivalent as the original source impedance that would compensate for it, of course I'm not having in consideration the output impedance of the tube stage, but assuming it low enough it will work, in case it isn't low enough correction may be added at the input to make input impedance high enough. I don't have any scheme soft here to draw this to explain better, sorry. Other option is to use the eq pre fader but maybe I'm missing something why it's after.

What Im saying about I like this, is that I like the behavior of this circuit in certain sources as the distortion or some 'mojo' to use as an effect, not a scientific affirmation but a taste matter.

JS

 

[emrr]

However, when you set all the EQ controls to flat, you would expect the response to be flat no matter what the driving impedance. All my passive EQ designs achieve this goal.

Ian, I would very much like to see a passive EQ which achieves this goal.

Please could you post a rough schematic?

This is simple: pre-load the entire system.  All of the early American passive EQ's are flat when set flat under any external condition, because they are generally fixed 600 ohm systems using loaded T attenuators which naturally go into bypass at zero position, and have little variation from expectation with wild external variation anywhere other than at extreme settings. 

Also not cheap to build; actives make it all cheaper, even with tubes.  This is why Pultec eventually won the market, they were a fraction of the price with their active amplifiers. 

 

[joaquins]

The schematic I posted on the universal eq topic didn't get any equalization with an input Z change, (assuming input z resistive) I don't know what happens with the curves if eq is being applied, wouldn't be hard to know.

JS

 

[ricardo]

The schematic I posted on the universal eq topic didn't get any equalization with an input Z change, (assuming input z resistive) I don't know what happens with the curves if eq is being applied, wouldn't be hard to know.

Got a link?

 

[joaquins]

The schematic I posted on the universal eq topic didn't get any equalization with an input Z change, (assuming input z resistive) I don't know what happens with the curves if eq is being applied, wouldn't be hard to know.

Got a link?

Yes, look for universal passive eq en the searching engine on the top right, the topic with that name (just 4 pages) at the first page appear as the #10 reply... here I's done it for you: http://groupdiy.com/index.php?topic=50484.msg640488#msg640488

JS

 

[ruffrecords]

However, when you set all the EQ controls to flat, you would expect the response to be flat no matter what the driving impedance. All my passive EQ designs achieve this goal.

Ian, I would very much like to see a passive EQ which achieves this goal.

Please could you post a rough schematic?

http://groupdiy.com/index.php?topic=51030.msg651561#msg651561

Cheers

Ian

 

[ricardo]

Ian, I would very much like to see a passive EQ which achieves this goal.

http://groupdiy.com/index.php?topic=51030.msg651561#msg651561

Thanks for this Ian .. though I have to admit a preference for Joaquins'

http://groupdiy.com/index.php?topic=50484.msg640488#msg640488

This has been an eye-opener on evil passive EQs for me .. but I still think you should use a single triode for the Baxandall and the other half to buffer the fader 

 

[PRR]

> I still think you should use a single triode for the Baxandall and the other half to buffer the fader

Or do it right: triode-pentode. The original calls for a "high slope" pentode and after decades of sloppy 12AX7 Baxes, I think he was right.

 

[ricardo]

Or do it right: triode-pentode. The original calls for a "high slope" pentode and after decades of sloppy 12AX7 Baxes, I think he was right.

Maybe ECF82, 6AN8, 6BR8, 6U8 or 7199

Anyone know availability & price for these alternatives?

We can then have a zillion posts over sound of Classic Baxandall (triode/pentode) vs evil 12AX7 implementations vs Ian's evil modern CC/CF Baxandall stage. 

 

[ruffrecords]

This has been an eye-opener on evil passive EQs for me .. but I still think you should use a single triode for the Baxandall and the other half to buffer the fader 

Everyone seems to wax lyrical about the need to drive the input of the Baxandall from a low source impedance and completely forgets the need to drive the feedback end from a low source impedance too. One difference from the input case is at least you have a reasonably fixed driving source impedance even it it is not low. The other difference is that the increased load on a non buffered 12AX7 or whatever reduces the open loop gain and puts a heavier load on the anode circuit both of which increase distortion. In an effort to reduce the load the circuit values tend to get scaled by a factor of 10

Cheers

Ian

 

[abbey road d enfer]

> I still think you should use a single triode for the Baxandall and the other half to buffer the fader

Or do it right: triode-pentode. The original calls for a "high slope" pentode and after decades of sloppy 12AX7 Baxes, I think he was right.

Would it make a serious difference? With the values in the original article (Ra=68k), an EF86 would have only about 3dB more open-loop gain than a 12AX7.
Simulations show that maximum boost with the pentode is marginally higher (+0.4dB) than with the triode. This difference may as well be chalked up to imperfections in the models.

 

[abbey road d enfer]

This has been an eye-opener on evil passive EQs for me .. but I still think you should use a single triode for the Baxandall and the other half to buffer the fader 

Everyone seems to wax lyrical about the need to drive the input of the Baxandall from a low source impedance and completely forgets the need to drive the feedback end from a low source impedance too.

Peter J. did not seem to consider it a necessity. But I tend to agree with you. The load on the anode tends to decrease the open-loop gain. Including a CF is an honest mean of increasing the open-loop gain.
However it is not a necessity for the good operation of the Baxendall EQ. Indeed simulations show that the differences between a single triode and a triode+CF are negligible in terms of response.
OTOH driving it with a stable impedance IS a necessity for the correct operation of the Baxendall EQ. Neglecting the output CF makes only the maximum boost less and sensitive to load. That's all. Failing to drive the circuit from a stable impedance makes the "neutral" response not flat and makes the boost and cut non reciprocal, which is a bigger sin IMO.

One difference from the input case is at least you have a reasonably fixed driving source impedance even it it is not low. The other difference is that the increased load on a non buffered 12AX7 or whatever reduces the open loop gain and puts a heavier load on the anode circuit both of which increase distortion.

In most cases, the biggest load will not be the EQ components, it will be the actual load to which the circuit is connected. That's what really motivates the existence of the output CF.

 

[PRR]

> Anyone know availability & price

In North America:

http://www.thetubestore.com

http://tubedepot.com

$5 price means they have truckloads. $20 tubes are readily available at that price but will probably run low in years to come. Oddly, 01A which has not been made in 90 years is listing at $40-$50 (but it was THE tube of the 1920s).

 

[ruffrecords]

In most cases, the biggest load will not be the EQ components, it will be the actual load to which the circuit is connected. That's what really motivates the existence of the output CF.

I am not 100% sure of this. Referring to the last schematic I posted with a 50K pot and 5K resistor in series at the cut side; when maximum cut is applied, the feedback impedance tends to 5K. Assume the reactive part of the feedback is also 5K. If the virtual earth is working then the load on the output is 10K. At maximum boost it drops to 55K or so. A CF will be fairly happy with these changes in load. A CC 12AX7 will not.  I know I have simplified the issue a lot but I think it is clear the load the stage sees varies a lot. Hence my comment about scaling component values. I'll do some sims to check this out.

Cheers

Ian

 

[ricardo]

> I still think you should use half a dual triode for the Baxandall and the other half to buffer the fader

Or do it right: triode-pentode. The original calls for a "high slope" pentode and after decades of sloppy 12AX7 Baxes, I think he was right.

Ha!  Great Guru B's own implementation is as I suggest.
http://www.douglas-self.com/ampins/wwarchive/wwarchive.htm#baxinpre57

Note his sensible choice of tube, not as a cathode follower .. but a high gm, low u triode in common cathode  .. also his comments about output levels.

But anyone using valve gear isn't doing it for 'accuracy', THD, bla bla  .. but for mojo & sound.  I'll now claim that NOT following GGB's exact recommendations leads to a definite roughness in the mid range and may make Glenn Gould's added embellishments to Bach more prominent.  ;D

I'm testing this (da Gould) now.