Pass Labs XVR1 crossover

Nelson Pass

The one and only
Paid Member
2001-03-29 12:38 am
The first run of XVR's is in production and will ship at the
end of the month. Lots of information will be out soon,
as I am having to essentially write a book for the owner's
manual, and we'll be putting it up on the site.

I've been working on this piece for almost 3 years, and
I believe the results are pretty spectacular.
The piece looks beautiful! It will be nice to have a decent crossover available...

I've been a lurker for quite some time and this is my first post...

Thanks to Nelson for his generosity and patience with those of us out there attempting to tackle aspects of this stuff ourselves... Thanks for never losing the love of the hobby!

I currently own Threshold equipment and now am eyeing the X-1 as a replacement preamp for my next step... It will take a while to get there but it will be well worth the wait... :D

Thanks also to the other members for their constant help and advice!!

I have a question regarding the circuit design of this unit. On the Pass site it mentions "The discrete circuitry for every stage consists of matched low noise JFETs driving bipolar output transistors, all biased single-ended with constant current sources." Why the departure from Mosfets in the output stage?

Thanks and again, beautiful work!

Steve Majerick

Nelson Pass

The one and only
Paid Member
2001-03-29 12:38 am
In this particular circuit with very low gain,
the bipolar outputs worked better.

The high/low pass project will have a vague similarity,
but is not far enough along to say for certain.

In the meantime, parts 2a, b, and c of the Zen
Variations is almost done, with an ETA of April 1,
and the Claw is a large pile of wood.
The manual for the crossover has been posted on the Pass Labs site.
They'll be publishing a <i>Crossovers For Dummies</i> book in the near future for those who are unable to cope with the psychic trauma induced by having to think for themselves...
The poor darlings will be forced to choose between Low, Medium, and High Q (the manual doesn't specify, but I'm guessing .5, .707, and 1), in addition to crossover points (a bakers' dozen, not counting multipliers), slopes (1st, 2nd, 3rd, and 4th order), and levels. The only thing it doesn't have is infinitely variable phase, but that doesn't implement well in the real world short of manipulating things in the digital domain (yuk).
Interesting reading. If you keep your wits about you, certain deductions can be made about the circuit.
Looks like a fun toy.


Nelson Pass

The one and only
Paid Member
2001-03-29 12:38 am
I think one of the most interesting features is the
ability to independently adjust the frequency of
each pole.

By the way, the simplified schematic is part of the manual.
The capacitors are equal value, and the Q gain figures
are 0, 3 dB, and 6 dB.
I just downloaded and read the XVR1 owners manual from This device absolutely rules! I've never had so much fun reading an owner's manual. Looking forward to the DIY version. I was never all that enthusiastic about bi-amping until now. That means I'll need to build 2 Penultimate Zens. Really cool work NP!

Incidentally, how's that discrete mosfet version of ENIAC coming along?
20,000! Wow. There's a great article in this month's Invention and Technology about Steve Wozniak. In it he describes a 10bit adder he built when he was young from all discrete components. I thought of your project, but 20,000! That's major. He also talks about how they used to send schematic packs out with their computers with code listings, etc. because people actually wanted to know how the things worked. That was the golden age of computing. I can still remember writing games in 8K of RAM! Best of Luck.

Given any thought to designing a discrete equalizer? I never liked those either, but if you designed one, I bet it would change a few minds. And just think of all the knobs to adjust!
Board Layout

Thought I would jump in on this one. Nelson and Wayne do the layouts in house. Have pity on us; Nelson does this without a reference schematic...... flawless.

Nelson leaps from mental concept to finished layout with nary a bit of documentation, and the just the odd bit of calculation on paper.

Such was the birth of XVR1, the prototype had one resistor value which needed changing, otherwise the board was flawless.


This is not the intended use for the XVR1 but it certainly is possible. Having 39 frequency steps per pole, one to four poles per filter, and variable "Q" for each pair of poles is a bit overkill for subwoofers.

You can sum the outputs and drive a single subwoofer.... or not sum the outputs and drive dual voice coils on a single subwoofer.


Nelson Pass

The one and only
Paid Member
2001-03-29 12:38 am
You forgot the 3 frequency multipliers. ;)

Actually though, I find that degree of adjustability most
useful for subwoofers, being that they usually need some
equalization also.

I occasionally put 1 or 2 poles down at 22 Hz or so to really
flatten out the response and then additional poles to get
rid of the woofer above its usable frequency. It may only
result in two octaves of response, but it's flat.