John Broskie's Guide to Tube Circuit Analysis & Design

21 December 2010

PCBs At Last!
I came to doubt that the boards would ever arrive, I must admit. But they are here and they look good. Many of the most popular boards are back in stock, such as the Janus and PS-1 regulators, the Aikido phono and Tetra phono stages, the mono noval and octal Aikido boards, the Aikido LV, the ACF, the Select-2 and Select AC switches, the H-PS-1 and PS-3 and PS-4 power supplies, and many new boards. I cannot release the Aikido phono and Tetra phono stage PCBs yet, as both boards are in Rev. A now and I must test them first, along with the new boards. I had twenty of most of the PCBs layouts made (and many more of the Select-2 and Select-AC PCBs), which made for big, heavy boxes on my doorstep last night; yet it isn't that many PCBs, considering the popularity of many of the boards, such as the Janus and PS-1 regulators. (In other words, I expect to run out of those boards fairly quickly.) While the PCBs are not "just in time for Christmas," they are just in time for post-Christmas and the New Year.


Split-Load Phase Splitter
After reading Stuart Yaniger's article in Linear Audio, "Taming the split-load inverter" and after designing my 6AS7/6080/6082 based OTL, it seemed to me that many more topological tricks remained latent within the humble but capable split-load phase splitter.  But first, let us do a quick recap on the circuit.

The split-load phase splitter (AKA cathodyne phase splitter) receives an unbalanced input signal and delivers a balanced output signal. It provides a differential gain close to 2 and a single-phase gain close to unity. It is actually just a variation on the grounded-cathode amplifier that stipulates equal values for the cathode and plate resistors. Both resistors are AC grounded at one end, at ground and at the B+ connection, and output is taken from the resistor's other end. Since both resistors are in the same current path, each must experience the same voltages swings as the other, which ensures this phase splitter's fabulous balance. Symmetry is all important here. Each phase output must see the same load impedance, otherwise the balance is lost. Another design issue is getting the idle current right. From blog post number 167 the following summery is worth repeating:

The split-load phase splitter works best with a balanced load and using it in an unbalanced fashion greatly diminishes its performance. Its output impedance can prove substantially lower than expected, when coupled to a differential load. In addition, its relatively limited voltage swing is only a problem if one needs big voltage swings. My recommendation is to use it early in a compound circuit, such as a power amplifier. So, instead of gain stage, split-load phase splitter, output stage, I would use split-load phase splitter, gain stage, output stage. By the way, if your design goal is to have the biggest voltage swings possible, then the following procedure should be followed. First find the maximum peak current of which the split-load phase splitter is capable with a given B+ voltage and plate and cathode resistor values:
      Ipk = B+/(2R + rp).
Next, set the idle current to half the peak current's value:
       Iq = B+/(4R + 2rp).

One potential problem we face when using the split-load circuit is its vastly differing PSRR figures at its outputs. The output taken from the plate holds almost all of the power supply noise, while the output taken from the cathode holds almost none.

Of course, if the B+ contributes no power-supply noise, then the dissimilar PSRR figures would not be an important design consideration. In the real world, this discrepancy must be addressed, as the power-supply noise will get amplified further down the signal path. Ideally, a phase splitter would completely reject the power-supply noise. Second best, it would present equal power-supply noise at each phase output, equal in amplitude, equal in phase; thus, the push-pull output stage could apply its CMRR to this common-mode signal and reject much of it from the amplifier's output.

One problem with this approach is that a push-pull output stage's CMRR is not as constant as most imagine. At idle, with both output tubes conducting equally, the output tube and output transformer work to reject what is common to balanced input signal; true enough. But when in use, playing music, when one output cuts off, the remaining on output tube simply treats the common-mode noise signal as signal, as something to amplify. (I am convinced that this a major part of explaining why class-A amplifiers sound so much better than class-AB amplifiers.)

Or what if the spilt-load phase splitter's output delivers signal to an unbalanced device, with the phase splitter only providing a choice between phases? In this situation, there is no CMRR, as the input signal is effectively unbalanced? For example, the following spilt-load phase splitter offers two-phase outputs and largely rejects power-supply noise.

The assumption behind this circuit is that the bipolar power supply develops equal but out-of-phase power-supply noise on both rails, which the spilt-load phase splitter exploits to null the noise from its two outputs.

I used a variation on this topology in my "Phase Selection & Zero Gain Line Stage Amplifier" back in 1998 at the GlassWare website.

The goal was a unity-gain line-stage amplifier (buffer, actually) that offered phase reversal. (A friend fell in love with my prototype and he still uses it to this day. I used a large-valued polypropylene power supply capacitor that bridged both power supply rails, which helps make this line-stage amplifier almost ageless. In addition, he runs it without an output coupling capacitor, as the output DC offset is low and his tube power amplifiers are indifferent to the few millivolts of offset voltage.) Here the decision was to add an input coupling capacitor so that the extra cathode resistor and its electrolytic bypass capacitor could be left out of the circuit.


Broskie Spilt-Load Phase Splitter
This phase splitter not only provides power-supply-noise-free outputs, it offers a solution to the two-triode problem. What's the two-triode problem? What shall we do with the second triode in the 6AQ8, 6BL7, 6BX7, 6CG7, 6DJ8, 6SL7, 6SN7, 12AU7, 12AT7, 12AX7, 12BH7, 12SX7, 5687, 5751, 6072…holds. Say you are building a mono power amplifier and you use a 6CG7 as the spilt-load phase splitter tube, what do you with the second triode? In the Broskie spilt-load phase splitter, you use the redundant triode to null the power-supply noise from the outputs.

The left triode functions as a simple spilt-load phase splitter; the right triode, as a noise canceller. When the right triode sees voltage perturbation at its grid, it takes that input signal and inverts at its plate, thereby nulling the perturbation from the phase splitter's inverting output. The phase splitter's non-inverting output also benefits, as the right triode never sees the power-supply noise, so it cannot pass any of it along at its non-inverting output. Clever, don't you think?

Once you have the concept down pat, you can move on to the actual circuit implementation shown below.

The two-resistor voltage divider (300k & 100k) quarters the B+ voltage nicely and the diode protects the right triode at turn-on, when the tube is cold and not yet conducting; but its grid is at the B+ voltage, while its cathode is at ground potential. Why is a 9.6k cathode resistor used for the right triode and not a 10k resistor? If the right triode provided true unity-gain at its cathode, then 10k would be the right value, but with a 6DJ8/6922 triode, the gain will be only about 0.96; thus, the 9.6k cathode resistor, as cathode follower gain against Ra is the right value.

Doesn't this topology unbalance the split-load phase splitter? Not really; certainly a little bit, but not that much. The 10k plate resistor is effectively in parallel with the following impedance: rp + (mu + 1)Rk. We can re-balance phase splitter by the non-inverting output a slightly lower load resistor after the coupling capacitor. For example, with a 6DJ8/6922, the inverting output would get a 100k resistor, while the non-inverting output would get a 76k resistor.

The downside to this topology, other than added complexity, is a somwhat limited output swing from the outputs, as now the plate resistor drops twice the voltage that it would normally. One workaround is to use a lower DC grid-bias voltage for both triodes, say 50Vdc.


Next Time
Back to OTL amplifiers and 72-ohm speakers and a new phase splitter. And be sure to check out the GlassWare Yahoo store for the back-in-stock PCBs. In fact, I plan on stuffing my Christmas stocking with quite a few boards.









Kit User Guide PDFs
Click image to download

BCF User Guide

Download PS-3 User Guide

Janus regulator user guide


E-mail from GlassWare customers:

Mr Broskie,

I bought an Aikido stereo linestage kit from you some days ago, and I received it just this Monday. I have a few things to say about it. Firstly, I'm extremely impressed at the quality of what I've been sent. In fact, this is the highest quality kit I've seen anywhere, of anything. I have no idea how you managed to fit all this stuff in under what I paid for it. Second, your shipping was lightning-quick. Just more satisfaction in the bag, there. I wish everyone did business like you.

Sean H.


Hi John,

I received the Aikido PCB today - thank you for the first rate shipping

Wanted to let you know that this is simply the best PCB I have had in my hands, bar none. The quality is fabulous, and your documentation is superb. I know you do this because you love audio, but I think your price of $39 is a bit of a giveaway! I'm sure you could charge double and still have happy customers.

Looking forward to building the Aikido, will send some comments when I'm done!

Thank you, regards,

9-Pin & Octal
PCBs & Kits

High-quality, double-sided, extra thick, 2-oz traces, plated-through holes, dual sets of resistor pads and pads for two coupling capacitors. Stereo and mono, octal and 9-pin printed circuit boards available.

Designed by John Broskie & Made in USA

Aikido PCBs for as little as $24

Only $12.95
to keep track of your
tube and part collection

TCJ My-Stock DB

TCJ My-Stock DB helps you know just what you have, what it looks like, where it is, what it will be used for, and what it's worth. TCJ My-Stock DB helps you to keep track of your heap of electronic parts. More details.

Version 2 Improvements
   List all of your parts in one DB.
    Add part Images.
    One-click web searches for part information.
    Vertical and horizontal grids.*
    Create reports as PDFs.*
    Graphs added 2D/3D: pie & bar.*
    More powerful DB search.
    Help system added.
    Editable drop-down lists for location, projects,         brands, styles, vendors and more.

     *User definable

Download or CD ROM
Windows 95/98/Me/NT/2000/XP

For more information, please visit:


To purchase , please visit our Yahoo Store:           Copyright © 1999-2010 GlassWare           All Rights Reserved