• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Building a valve crossover

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Actual crossover network is quite crude, and tubes just act as buffers.
WAY better crossovers can be built busing Op Amps (or discrete transistors if you wish)..

Or using tubes, but then OP will need about an dozen of them.

Whatever´s easy and practical with Op Amps quickly becomes very hard to build with tubes.

Not sure why tubes were chosen to build this, they are certainly more of a handicap than something providing enhanced sound and performance.
 

PRR

Member
Joined 2003
Paid Member
I think there is a typo in the value of the 3rd filter-cap in the woofer side.

...Something like a spice sim...,

I simmed three ways:
12AX7 per plan #1
12AU7 per plan #1
A Sallen-Key using the same buffer design (to check JM's assertion of "a dozen").

The difference between Mu=100 12AX7 and Mu=20 12AU7 is so small it would be hard to measure. Use ANY general-purpose triode. While the current demand does change, neither is any problem. (12AU7 might like more than 227r cathode bias.)

FWIW: the use of 12AY7 marks this as a late-1950s design. 12AY7 had only a short span of popularity.

The 2nd stage is just a buffer, no effect on response. To save sim-work I emulated with a gain of 0.9 ideal buffer. Stray resistors damp sim-check complaints.

The plan #1 looks like 3rd-order, but does not get near a 18dB/Oct slope until way-way down and out.

Since 1950s good essays on active filters were available in obscure books; since mid-1970s Lancaster's book made them easy to cook-up. Active filters can work JUST like the classical L-C filters of the big good talking-movie systems, not the slow and soft "cut-off" of a cascaded R-C filter. Lancaster's mid-1970's book made them easy to cook. Sites like (Sample)Sallen-Key Low-pass Filter Design Tool - Result - even easier.

My "SKey" plot is the same tube-circuit working as an Active filter. While it is "only 2-pole", it gives a "better shape" through the crossover zone. Ratios of values can give variations other than the one I picked (Q~~=1). While the result with tube buffer (gain not quite 1.000) is not as the online calc predicts, it isn't far off. You CAN make a sweet active filter with tubes. (No real shock to the people who worked on WWII radar systems; but modern math tools reduce the drudgery.) Note that my active form has -less- parts than the cascaded R-C form.

Oh: the active form has lower loss. I plotted "SKey-1" to get the bass-gain similar to the others.
 

Attachments

  • victordascalescu54-1.gif
    victordascalescu54-1.gif
    17.7 KB · Views: 299
  • victordascalescu54-2.gif
    victordascalescu54-2.gif
    10.1 KB · Views: 282
I have built two stuf with the diagram I sent. Both start with 6 dB and then modified on 12 dB. For a clone of the Voice of Theatre ( with original speakers) They works fine and the tests are good. I suggest the use of Ecc88 or 6H6/6N6 ( sister of ECC88). With the CF, id the load is high we havent any changes on dynamic load line, of course the swing is not so high in this use. But if we use a filter where the value is comparable with the cathode resistor some changes can happen so in this case the use of (relative) low Rp helps. The 88 I set at 6-8 mA with 200 Vdc. Now I am preparing another one to drive a pair of B&W 800D Nautilus with four mono amps with 4 x KT150 each one.; 12 dB and 600 Hz of crossover.

Walter
 
I built one of these not too long ago, and used the 6N1P. Very clean, transparent sounding (as it should be, when done right!). It works very nicely for a friend who wanted to do some biamping on some rather large open baffle Lowthers, and crossed to a Modulus-86 for the low end.

Tube Active Crossover

My last version that I really liked was based on the design from Elliot Sound Products, and based on his two way design, I simply kept all the same parts values and scaled them (up for resistors, down for caps) and used 6CG7 as buffers, with long tail resistors to a negative rail instead of CCS tails.

Linkwitz-Riley Electronic Crossover

I'm drawing up a three way version using the 6N16B subminiature tubes now, and want to borrow the DC servo from the Heretical linestage by Sy, that way I can reduce the number of coupling capacitors. I will also incorporate shunting relays on the outputs to reduce turn-on/off thump.

If you've got an itch to build a project to use up a stockpile of small tubes, the crossovers are a good start :)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.