secrets of the dangerous music bax eq

[seavote]

andyP could you share theschematic you worked from? i have quite a few quality dual pots looking for a project.
i searched and found quite alot on the baxadall but notmuch on the james.can someone explain the differences between the 2? thanks

 

[strangeandbouncy]

HI,



  of course!


              http://makearadio.com/tech/tone.htm


you can even muck about with an online simulator at


              http://www.duncanamps.com/tsc/


pc only sadly, but worth hijacking the missuses pc for a while!


I think we are a bit off topic, since Im sure the Dangerous BAx is active. I can wholeheartedly recommend the Studer 169 for API 500 series that AUdiox designed a pcb for, and Gustav makes pcb. I built mine with only Hf and LF, and switchable frequencies. That IS BAxandall if I understand correctly.

  I could be wrong, but basically, James is passive, and Baxandall is active, with the filter in the negative feedback loop. Baxandalls original paper can be reached from


  http://en.wikipedia.org/wiki/Peter_Baxandall


                under external links



  kindest regards,



    ANdyP

 

[seavote]

thanks for the info. i dont remember but i dont think the stacked pots i have are in the 10k ,50k  area. not the 100k  500k i see being used in the baxandall circuit. probably be a while till i get around to doing this. i saw the studer thread but im looking do this as a  "throw together" project P2P.

 

[strangeandbouncy]

HI,



  you could use your stacked pots, and add extra resistors at either end, giving a smaller max cut boost, which could be quite useful actually! More suitable in a mastering environ, for sure.



    Kindest regards,


      ANdyP

 

[Joechris]

To get the treble boost flat in middle posission, there is a solution on this page:
http://www.sentex.ca/~mec1995/tutorial/xtor/xtor5/xtor5.html
Look at fig 12 to 14. He add a 1k and 10k to the wiper of the 100k treble log pot.
Since there is a James fever rising, or mabye we could call it a poor mans pultec, do anyone have a suggestions to a makeup amp that will fit this old baby?
I was thinking of something simple, mabye a two transistor class A amp like the output part of this preamp:
http://www.tcaas.btinternet.co.uk/jlhmodpre.htm
It can provide 20 Db of gain and 2V rms output, a little weak, but as a unbal insert thing, its maybe ok?

j

 

[strangeandbouncy]

Hi,


  let me correct you, this is no Poormans Pultec! for one thing, apart from the inductor, the parts count is pretty much thesame . . . a handfull of caps and resistors.

  For another, the slopes are much more gentle.

  Makeup is about the same at approx 20dB. i am thinking of using OA10 Neumann DOA from Gustav. At 20dB gain, it has a slewrate of 50V per us. Or Kingstons ecc99 tube makeup stage. or both, - switchable!


  thanks a lot for the sentex link.


  Kindest regards,

  ANdy(erm . . . .James) Parker

 

[Joechris]

;D, no its not a Pultek, but the bass boost migth be a bit similar, and inductors for peak treble boost is expensive so...
ECC99...hmm sounds good.
j

 

[abbey road d enfer]

thanks for the info. i dont remember but i dont think the stacked pots i have are in the 10k ,50k  area. not the 100k  500k i see being used in the baxandall circuit. probably be a while till i get around to doing this. i saw the studer thread but im looking do this as a  "throw together" project P2P.

As in almost any electronic circuit, values may be scaled to allow using existing parts. Basically all ressitors are increased or decreased in the same ratio and caps in inverse ratio. The limits to this exercise are that the input impedance must not be too low and the impedance seen by the make-up amplifier must be manageable. The typical James is intended to be driven by a rather high impedance source (tube plate) (although low impedance won't do no harm) and followed by a vacuum-tube amp with 500k-1Meg input Z.
A solid-state make-up gain should be specifically designed to present high input Z. Can be done by proper bootstrapping or using FET or Darlington. But high z and bipolar transistors are often mutually exclusive in terms of noise performance.
If intended to be used with modern solid state equipment, the pot value can be divided by 10

 

[Joechris]

I have been playing with this curcuit it Duncans simulator, and the only thing that will place the mid point in the middle of rotation is a log B pot in the simulator, witch I belive is not a reverse taper, but a true log scale
witch has to be done with a resistor string to get rigth. A lin pot has the middle with 10k on one side and 90k on the other, and that is a true log scale, am I rigth? With a 1k input imp and 100k "pots" and 100k load the curves looks great. At the far ends, the corner freq starts to move upwards as well. Looks like a it has some creative vibe.

j

 

[strangeandbouncy]

HI,



  creative fits the bill!


  ;D

  ANdyP

 

[PRR]

> A lin pot has the middle with 10k on one side and 90k on the other, and that is a true log scale, am I rigth?

No. A 100K linear, set to "center", is 50K either way.

A proper James needs 10% Audio taper pots. This suits the 10:1 nominal loss.

Duncan TSC calls 10% audio "type B", but makers use various letters, and 20% pots are far more common today.

Try it. Select linear pots, then turn for flat response: happens at 9%-11% rotation. With Linear this is 10K below the wiper and 90K above the wiper, We want a pot with 1K below the _center_ and 90K above the center. This is formed with two different track-inks, one 9 times juicier than the other. That's an extra step, ink juiciness and cooking times interact, 10% pots are more than a buck's worth of trouble to make, 20% taper is a bit easier and good enough for many volume controls. Nobody makes pots FOR James stacks any more.

James also needs zero source and infinite load and some 9%-odd cap values. Using standard cap values it can be pretty good with 5K source and 5Meg load. If you need FLAT response, bypass it: there's too many 1dB errors to ever get consistently perfect, not to mention 20% pot tolerance.

 

[Joechris]

Thanks PRR, that confirms what I thougth. I was a little unclear, the linar pot was in the simulator, and 10k/90K was the "flat" spot, not the middle of rotation.

j