The Return of No Gain
Do you need gain? Signal gain? I don't. My DAC puts out enough output signal voltage to drive my amplifiers to full output. It's not alone, as most modern audio equipment delivers at least 1Vpk of output signal; even smart phones deliver 1Vpk of output signal. If that's the case, why use a line-stage amplifier (preamp, as some call it) that delivers 20dB or 30dB of gain? By the way, 20dB of gain equals an amplification of tenfold; 30dB, almost thirtyfold (x 29.54). That is a lot of signal gain.

In days past, a ready reserve of gain proved a feature, as it could overcome the large span of phono cartridge outputs and various signal sources, such as FM tuners, tape decks... At the most, I can see a need for a fourfold amplification (12dB) in some systems, as some tube-based power amplifiers require over 2Vpk to 3Vpk of input signal to be driven to full output. Well, if you don't need any gain, should you consider a passive line stage that only holds an input-signal selector switch and a volume control, but no active electronics?

Perhaps your confidence the famous FACT that less is more in audio electronics, compels you to contemplate the theoretical advantage of eliminating all those electronic parts and their power supplies, and to imagine the sweet, short path the signal travels to get to the power amplifier. Alas, I cannot join you on this reverie, as I have built many passive line stages, some holding the most expensive volume controls, switches, resistors, and wire. All proved unsatisfying in the end, as the resulting sound always seemed unanchored and insubstantial, merely a wraith or shadow of its substantial self. The theory that less must be best gave way to hard experience. My system didn't need signal voltage gain just current gain.

Back in the early 1990s, I designed and built my first No-Gain, No-Pain tube-based buffer. As I was still under the sway of the theoretical advantage of passive line stages, I had designed the circuit just because it was novel and interesting. In 1998, I detailed the design at my GlassWare website, in a section titled "Circuit of the Month," which was a precursor to the Tube CAD Journal. The design used a simple cathode follower loaded by a compliant-constant-current source controlled by a DC servo that eliminated both any DC offset and the need for an output coupling capacitor. Forgive the following GIF schematic from 1998:

The triode's idle current is set by its cathode resistor, just as if the resistor terminated into ground. The compliant-constant-current source below it then matches the triode's current flow and prevents a DC offset voltage. The 150k resistor to the B+ voltage simply provides a current path to the B+ voltage when the tube is either missing from its socket or still cold and not conducting. The negative 12V voltage regulator, along with the diode in series with its ground pin and ground, establishes the -12.6Vdc negative power-supply rail.

The LF412 OpAmp with an FET input stage cannot accept input signal greater than its positive power-supply voltage, i.e. 0V in this case, so the four 1M resistors are used to split the -12.6 voltage, thereby delivering safe DC voltages to the OpAmp's inputs. The tube's heater element was also powered by the negative power-supply rail.

You can read more about the circuit in my Post 102, where I explain the importance of tightly-matched 1M resistors in the design. (When I built my No-Gain buffer, I lived in Silicon Valley and had easy access to inexpensive 0.01% resistors at electronic-surplus stores.) In addition, you will see a nicer schematic:

Six years ago, I revisited the No-Gain, No-Pain buffer stage in my Post 520. I had decided to add some Aikido Mojo to the design:

Dang sneaky ("sneaky is my egalitarian—indeed, equalitarian—euphuism for "clever," which in contrast is inegalitarian and immodest in the extreme). Does a cathode follower really need Aikido-Mojo enhancement?

It really does, as a current-source loaded cathode follower exhibits a PSRR only equal to 20Log(1/mu), where mu is the triode's amplification factor. A 6SN7 offers a mu of 20, making for a PSRR of only -26dB.

By the way, the hidden assumption behind the circuit is that negative 12V power-supply rail is well regulated, so much so that in AC terms it is virtually identical a connection to ground. This bothered me a tad. But it was the need for the two capacitors within the DC servo circuit that bothered me more, as the two complicate the circuit design. Thus, my quest was to eliminate one capacitor and still inject Aikido Mojo.

No-Gain, No-Pain Revisited and Improved
Two capacitors are needed in the DC servo circuit because the N-channel MOSFET inverts, forcing us to invert the inputs of the OpAmp. In contrast, if the MOSFET didn't invert, we need not invert the inputs, allowing us to use one capacitor. In other words, we need to use a P-channel MOSFET instead. The obvious problem with this solution is that the MOSFET's source presents a very low output impedance, unlike its drain. The workaround is to use both N-channel and P-channel MOSFETs in series; or as in the following example, NPN and PNP transistors.

Two OpAmps are used. The one on the left handles the Aikido Mojo, while the one on the right functions as a DC servo.

The two transistors are low-noise, low-voltage and low-wattage types. As long as the cathode resistor's voltage drop is not too great, say more than 5V, resistors R1 and R2 can match each other in value. Capacitor C1 and resistor R3 set the amount of power-supply noise to inject to create a power-supply noise null at the output. No output coupling capacitor is needed, but one could be used, such as a PIO type, to add sonic flavoring. In fact, two outputs can be offered, one capacitor-coupled for solid-state power amplifiers and one DC coupled for tube-based amplifiers.

The two OpAmps can be housed in an 8-pin dual-amplifier IC package. If the OpAmp cannot tolerate input voltage near or greater than the V+ voltage, they will require a low voltage bipolar power supply. My original workaround was to use four super high-tolerance 1M resistors. Here is an example:

Where do you get 0.01% resistors? One good source is the Vishay Dale thin-film conformal, single in-line, through-hole resistor network (about $10 each):

The worst-case scenario of the 1M resistors being either 1.0001 or 0.9999 off the 1 meg-ohm nominal value results in a potential DC offset voltage of no more than ± 2.13mV, assuming the OpAmp does not bring its own DC offset. If we used 0.1% 1M resistors, the potential DC offset voltage rises to ± 21.3mV. Of course, if you are driving a tube-based power amplifier or a solid-state power amplifier that does not amplify DC voltages (which is the vast majority of them, as DC amplifiers were a 1990's fad), then you need not overly worry.

Note that the 100µF non-polarized electrolytic capacitor prevents the amplifier from amplifying any DC offset at its input, as the amplifier ceases being an amplifier at DC and becomes a unity-gain buffer. If we pull back, we will see that the two-resistor voltage dividers need not exactly split the negative power-supply rail voltage, just lower the center voltage below the ground potential enough not to cause the OpAmp to latch up. For example, I recall the very long-in-the-tooth input-FET AD712's maximum positive input voltage being 2V less than its V+, which in this example would be 0Vdc, so -2Vdc is a suitable target voltage.

Note the increased capacitor values within the DC servo circuit due to the 200k resistors. With 0.1% resistors in the DC servo circuit, the worst-case DC offset is only ±8.16mV. With 0.01% resistors, the worst-case DC offset falls to 1mV. Here is a design example with a 6SN7 (or 6CG7 or 12SN7 or 12SX7).

By the way, some (but not all) OpAmps with transistor input stages can readily accept an input voltage equal to their positive power-supply voltage. The problem with these OpAmps is that the transistor-based input stage draws current at its inputs, not much, but enough to create a big DC offset voltage with high-ohmage resistor values.

Note that a bipolar power supply is used in both examples. One the left, we see that the FET-based input stage can simply ground the non-inverting input, whereas the transistor-based input stage requires the additional 200k resistor and capacitor. With the additional resistor and capacitor, both of the OpAmp's inputs see the same current-induced offset voltage, so they cancel. Nice. Here is how we create a dissimilar-voltage bipolar power supply:

The -12Vdc voltage is regulated, while the +4Vdc voltage is not; indeed, it is relatively dirty. Does this matter? Not really, as most OpAmps exhibit far better PSRR figures on their +V pins compared to their –V pins.

If you are still nervous, the +4V output voltage can be cleaned up with a 10-ohm RC resistor and 1kµF capacitor. Okay, how do you figure out the needed part values? Of course, you can use SPICE and machine gun your way to the optimal values—or you can use the following table.

The 6J5, 12SN7, and 12SX7 share the same values as the 6SN7. Of course, a triode-connected pentode, such as the EL84 or EL86 could be used, as well as DHT tubes, such as the 2A3 or 300B, but it will be up to you to establish the optimal values. One trick when dealing with reality and seeking the best Aikido Mojo is to use a specially dirtied high-voltage power supply. Let's say that the B+ voltage power supply holds a 100µF reservoir capacitor bypassed by a 3.6µF film capacitor. We disconnect the 100µF capacitor connection to the output voltage, thereby relying on the 3.6µF capacitor to do the heavy filtering, which it can't, thereby creating an especially noisy B+ voltage, which makes for obvious PSRR-improvement adjustments. Think of it as an audio magnifying glass.

Augmented Cathode Follower
As triodes offer so little transconductance compared to solid-state devices, such as FETs and MOSFETs and transistors, a cathode follower never delivers true unity-gain output, as there is always some insertion loss. The workaround is to supercharge the cathode follower with the addition of an input stage and negative-feedback loop. Here is an example from my Post 186:

Triode T1 drives triode T2's cathode making T2 function as a grounded-grid amplifier, so there is no phase inversion at the output. Internal coupling capacitor C2 relays 100% of the output signal to T2's grid, creating a negative-feedback loop, so unity-gain results, along with lower distortion and output impedance. The hassle with this circuit is that triode T3's cathode is situated +100V above T1 and T2 cathodes, so the heater power supply must be referenced to +150Vdc, so as to split the voltage difference. In addition, capacitor C2 must be relatively large in capacitance.

Another example comes from my Post 212:

This hybrid augmented buffer requires a negative power-supply rail and delivers such amazing performance in SPICE simulations that I dared not post the results, as no one would believe them. This is definitely not your grandfather's cathode follower. I rather cheekily labeled this topology the Brains and Brawn Cathode Follower.

Last year's Post 616 showed another hybrid augmented cathode follower, which I named the Ultra-Low Zo cathode follower.

It uses the bastode configuration and uses resistors rather than constant-current sources, although they could be used, as in the following design example:

This variation employs Aikido Mojo to enhance its PSRR, which got me thinking about how to add some output gain to the circuit; not a lot of gain, say just 2X or +6dB.

Note the use of plain electrolytic and non-polar electrolytic capacitors. Why not just non-polar electrolytic capacitors? They typically stop at a maximum voltage of 100V. Also note that the 9.53k negative feedback resistor sets the gain of +6dB. The protective diode limits the maximum negative output voltage to about -4.5Vpk. An additional safety diode can be added to the input triode by bridging ground to its cathode with the diode. The PSRR is exemplary:

With 1Vpk of input signal at 1kHz and 2Vpk of output signal, the SPICE-generated Fourier graph reveals a lovely single-ended cascade of harmonics.

With no gain, the Fourier graphs shows a strong 2nd harmonic with the higher harmonics greatly suppressed.

By the way, to be down 80dB equals 0.01% distortion.

Sonic Miracles
Recently, I beheld two sonic miracles, both at my friend Steve's home. Steve runs an amazing tri-amped system with dipole subwoofers and Klipsch corner-horn woofers and huge direct horn-loaded Lowther fullrange drivers, with bullet horn super-tweeter augmentation. It plays loud and clean with just one watt of power, which is provided by a custom-built single-ended power amplifier. (The Klipsch corner-horns get 300B single-ended power amplifiers, while the double 15-inch subwoofers get 300W class-D amplifiers.)

A previous improvement was wrought by digitally time-delaying the subwoofer's input signal to bring it into time alignment with the midrange horns, as the subwoofers are nearer to the listener. This had the effect of moving the deep bass from being free-floating to being anchored in the instruments that produced the bass.

Sadly, the Klipsch corner-horns rattled the Lowther fullrange drivers, occasionally resulting in their sympathetic resonance. Steve's first thought was to separate the Lowther fullrange drivers from the Klipsch corner-horns with some Resonix sheets between the two, and it worked, no buzzing. Steve then told me that he had a bunch of marble left over from some home improvement work. I told him to avoid it, as marble rings like a bell; instead, he should get some gray cast iron, as that is what is used in machine shops to prevent vibrations. He wasn't sure about trusting me about the marble (or the iron), so he called other audio luminaries; they said: marble rings like a bell. He then asked Gemini AI (he pays for the fancy version) about gray cast iron. It said:

For high-end audio and vibration eradication, Cast Iron is the clear winner scientifically, while Brass is the choice for organic tuning, and Steel is purely a budget option.

Cast Iron (The High-Damping Structural Powerhouse)

The Physics: Specifically, you want Gray Cast Iron. Under a microscope, gray iron contains millions of tiny, interconnected graphite flakes suspended within the iron matrix.

How it works: When low-frequency mechanical energy from the 15" woofer travels up through the stone and hits the cast iron, these graphite flakes act like microscopic friction points. They instantly scatter the vibrational waves and convert them into micro-thermal energy (heat).

The Advantage: Cast iron has a "Specific Damping Capacity" (SDC) that is 5 to 10 times higher than steel and far superior to almost any other structural metal. It is completely dead acoustically (which is why heavy industrial CNC machine beds and engine blocks are always made from it). It will not "ring" or sing in sympathy with your music.

 

The Final Ranking & Recommendation

First Choice: Gray Cast Iron. If you want absolute, textbook vibration eradication, use a cast iron plate. It matches the "deadness" of the ResoNix perfectly and acts as an ultimate mechanical dead-end for the Klipschorn's energy.

Second Choice: Brass. If you are willing to pay a premium for custom-cut metal and prefer a slightly richer, more musically organic midrange bloom, brass is an exquisite choice that provides excellent weight.

Third Choice: Steel. Only use steel if it is a matter of cost or immediate local availability. If you go this route, make sure the ResoNix sheet covers the entire surface area between the granite and the steel to ensure the plate cannot ring out like a bell.

So, Steve ordered some thick gray iron sheets from out of state, cut to conform to the Khorn's top profile, along with some synthetic marble and six used Stillpoints isolation footers; he already had some of them under each Khorn. When I found out how expensive the gray cat iron was, and that just the shipping of the gray cast-iron slabs cost $200, I began to worry that it would not pay off. I was so wrong.

The cast iron and synthetic marble and Resonix and Stillpoints have utterly transformed his system. (It turned out that he had his StillPoint footers upside-down on the Khorns, which have been up righted.) Every goddamned thing improved, the subwoofers, the Khorns, the Lowther horns. The bass is taut and coherent now. It's insane how much tighter everything sounds now. His system went from grade B+ to grade A. Our shared friend, Glenn, and I listened in disbelief, as we had both written off the Khorns as the weak sonic link. Sadly, our also shared friend, Ken, left to return to California just the day before the sonic transformation.

Second sonic miracle: When Steve told me that he had just bought a $1500 network switch, I was glad that I didn't have the money to do such crazy impulse buys. It goes between the CAT-6 wall connector and his streamer.

Glenn and I were leaving Steve's place, when the Matrix SS-1 Pro switch arrived, and Steve insisted that we stay to hear it. We had spent the morning running test tones and reading SPL meters in an attempt to figure out why his tweeter, which claims to go out to 30kHz, falls like a brick past 15kHz. No music, in other words. Measuring is hard work. Hell, serious listening is also hard work. Okay, I was impressed by the SS-1 Pro's build. Solid, heavy, beautiful to behold.

Steve first played four tracks that Glenn and I know by heart, without the Matrix, streaming from Qobuz. Sounded flat, boring, a bit dull. (At least twice, I have subscribed to Qobuz, as I quite like its clean layout and capsule reviews and biographies—but I could not stand the constantly recurring micro dropouts, which have only occurred twice with Amazon Music streaming service in the last 6 years or so and about twice a year with Presto streaming service.)

With the Matrix in place, we listened to the same tracks. Instantly, and I mean in the first 3 seconds, we heard a truly huge improvement. Increased dynamics, better imaging, and seemingly louder playback. Wow! Check out this review. The Matrix connects through optical cable, which galvanically isolates the data stream from the signal source ground. (This might be a large part of the secret sauce, as I have wanted to hear Steve's system with the network cable disconnected from his system for a long time now.) It also includes a fancy femtosecond clock and the best and latest and finest electronic goodies.

In other words, further digital domestication.

In fact, the stream from Qobuz sounded better than the SSD in streamer, a sonic first, as the SSD had always sounded better than any stream. With digital, unlike much of analog, newer is always way better. In short, $$$ well spent. If nothing else, the Matrix is super well-made and heavy and attractive. Steve's system has advanced to grade A. Hell, Steve's system now deserves an A+.

Matrix Audio also makes a simpler and cheaper SI-1 Audio Grade Network Isolator, which might deliver much of the same sonic goods at a far lower price, for those who do not need multiple network outputs.

Semi-Disposable Suppressor
In my Post 643, I argued that what the world needed was a gun suppressor (aka silencer) that held replaceable cheap baffle inserts. Well, it turns out that a 22-caliber suppressor has enter the market with replaceable 3d-printed carbon-fiber extensions (for a mere $18). Next Level Armament makes a fascinating two-part hybrid suppressor, the NLS 22 Varmint Suppressor for only $149.99.

From their website:

Designed specifically for the backyard shooter and the .22LR varmint hunter, the NLS 22 Varmint suppressor was born from an idea that effective shouldn't’t equal expensive.  Shooting .22 suppressed provides greater enjoyment while shooting, however, the steep cost for a small suppressor creates a barrier to entry for most people.  Therefore,  Next Level Armament  created an affordable, yet effective suppressor- The NLS 22 Varmint.

Each 22 Varmint features an aluminum threaded mount and blast chamber, finished with a hard coat anodize for durability.  Since the blast chamber endures the majority of the pressure, this allows a  light-weight design for the rest of the suppressor.   Therefore, we 3D printed each Modular Baffle Stack in carbon fiber for a precise fit to the contours of the mount.    *Although they are relatively durable, the modular baffle stacks cannot be installed directly onto a firearm without the mount and blast chamber.  Any attempt to modify them for use without the mount will render it unusable and can lead to immediate failure of the stack due to high pressure.

Fully assembled, the entire suppressor measures only 6 inches long, just over an inch in circumference, and weighs in at a very light 3.2 ounces.  It’s so light, you barely notice it on a rifle, and it cycles reliably when attached to your .22LR semi-auto pistol without the need for spring assists or special adapters.

Did we mention it's quiet? It gets even better when you add subsonic rounds to the mix.  Protect everyone’s ears while having fun with friends and family.

 

Serviceable:

Need to change out a worn or damage baffle stack module?  The modular design makes servicing your suppressor quick and easy.  Remove the mounting screws to remove the baffle module from the mount.  Contact us to order a replacement module and send us the old module (we need the old one back in order to complete the exchange).  Once we have the old module, the replacement will ship back to you.    When you receive the new module, slide it onto the threaded mount, insert the mounting screws, and you are back in business!

NOTE:  We will not replace any module without the original module returned to us, along with proof of ownership for the suppressor (currently a copy of your approved Form 4/tax stamp).   Cost of replacement module is $18.00 (includes shipping & handling).

 

Specifications

• Aluminum threaded mount and blast chamber with hard-coat anodized finish
• 1/2-28 threaded mount
• ONE replaceable carbon-fiber baffle stack
• 1.18-inch diameter
• 6-inch overall length
• 3.2 ounces

 

Caliber Ratings

Rated for:

• .22 LR standard velocity
• .22 LR subsonic
• .17 HMR

Okay, this is brilliant, and I hope they make millions. Needless to say, I want one.

https://www.youtube.com/watch?v=5EpASitmjsY

Music Recommendation:
Chris Jones' and Steve Baker's Smoke and Noise
Audiophiles love Chris Jones, as his 2003 album, Roadhouses and Automobiles, as it definitely counts as demo quality, as does his 2000 album, Moonstruck—just listen to the track "No Sanctuary." Sadly, this great American blues guitarist died in 2005. Steve Baker is a British harmonica virtuoso. These two in this live recording make an amazingly big sound. To be honest, I came late to hearing this album, because I feared it would lack the studio albums exemplary sonics. I had nothing to fear; it's a great-sounding album, definitely demo grade.

//JRB

Adobe AI Summary:

Reasoning and Approach

To summarize the key points of the document, I focused on extracting the main technical arguments, design insights, and practical recommendations presented. The document covers both technical aspects of audio electronics (specifically zero-gain tube buffer circuits and passive line stages) and practical audio system improvements, as well as some related product recommendations and anecdotes. Below is a structured summary of the key points, with examples and explanations for clarity.

Key Points Summary

1. The Case Against Excessive Gain in Modern Audio Systems

• Modern audio sources (DACs, smartphones, etc.) typically output enough voltage (at least 1V peak) to drive most amplifiers to full output.
• Traditional line-stage amplifiers often provide unnecessary gain (20–30dB), which is rarely needed today.
• Passive line stages (selector switch and volume control, no active electronics) are theoretically appealing but often result in unsatisfactory sound—described as unanchored and insubstantial.
• Current gain, not voltage gain, is often what is needed in modern systems.

2. Zero-Gain Tube Buffer (No-Gain, No-Pain) Designs

• The author describes the evolution of his tube-based buffer circuit (cathode follower with a compliant-constant-current source and DC servo), which provides current gain without voltage gain.
• Design improvements include:
• Use of Aikido Mojo to enhance power supply rejection ratio (PSRR).
• Replacement of N-channel MOSFETs with a combination of NPN and PNP transistors to simplify the DC servo circuit and reduce the number of required capacitors.
• Emphasis on using high-precision resistors (0.01% or 0.1%) to minimize DC offset at the output.
• Example circuits and a table of recommended component values for various tube types (e.g., 6SN7, 12AU7, ECC99).

3. Augmented Cathode Follower Circuits

• Standard cathode followers have some insertion loss and limited PSRR.
• Augmented designs add an input stage and negative feedback loop to achieve true unity gain, lower distortion, and lower output impedance.
• Hybrid designs (using both tubes and solid-state devices) can achieve ultra-low output impedance and improved PSRR, with options for modest gain (+6dB) if needed.
• Example: The "Ultra-Low Zo Cathode Follower" circuit with and without gain, showing improved harmonic performance and PSRR.

4. Practical Audio System Improvements

• Vibration control:
• Gray cast iron is recommended as the best material for vibration damping in audio systems, outperforming marble, brass, and steel due to its high specific damping capacity.
• Example: Adding gray cast iron plates, synthetic marble, and Stillpoints footers to a high-end system resulted in significant sonic improvements (tighter bass, better coherence).
• Network switch upgrade:
• Installing a high-end network switch (Matrix SS-1 Pro) between the wall and streamer led to immediate and dramatic improvements in dynamics, imaging, and overall sound quality.
• The switch uses optical isolation and a femtosecond clock, which may contribute to the sonic benefits.

5. Product Highlight: Semi-Disposable Suppressor

• The document briefly discusses a new .22-caliber suppressor with replaceable 3D-printed carbon-fiber baffle stacks, making it affordable and easy to service.
• Key features: lightweight, quiet, modular, and cost-effective.

6. Music Recommendation

• Chris Jones and Steve Baker's live album "Smoke and Noise" is recommended for its excellent sound quality and musicality.

Example Implementation

• For audio DIYers: Consider building a zero-gain tube buffer using the provided schematics and component recommendations to improve current drive without adding unnecessary gain.
• For system tweakers: Use gray cast iron plates and proper isolation footers under speakers or equipment racks to reduce vibration and improve sound.
• For digital audio enthusiasts: Evaluate the impact of network isolation and high-quality switches in streaming setups.

This summary captures the technical, practical, and anecdotal highlights of the document, providing actionable insights for audiophiles and audio engineers.

Did you enjoy my post? Do you want to see me make it to post 1,000? If so, think about supporting me at Patreon.