Pendulum 6386
Tube Limiter
 by Eddie Ciletti - EQ December 1999

   In the world of vacuum tubes, nothing is new. It should be no surprise then, that "Tube" Limiters have existed since the early days of radio, becoming popular in recording studios through the fifties and early sixties. Altec, Fairchild and Universal Audio were building tube limiters then; Manley and Tube-Tech make all-tube versions now. What differentiates the Pendulum 6386 from the rest is its hybrid design. The front-end consists of an input transformer followed by a 6386 tube (the same dual-triode that was used in the Fairchild). The rest of the circuitry consists of high performance opamps, biased for Class-A operation to lower the distortion below that of standard Class-AB. All connectors and tube sockets are gold-plated. Layout is clean and straightforward, with special attention to internal grounding so that external interference won't be a problem.


   As mentioned in previous articles, Fairchild built disc-cutting systems and their stereo limiters interfaced directly with their cutting amplifiers - hence the "Lateral" and "Vertical" control designations. Unlike our "modern" 45/45 system, Fairchild created a matrix, combining left and right into mono (the lateral signal) and subtracting left from right to make the vertical groove component. To do this required many transformers. Pendulum's designer, Greg Gualtieri, wanted the functionality of the Fairchild without all of the 'iron' clouding the signal.


   I tend to be very heavy handed when reviewing compressor / limiters. After cranking up the 6386 and getting those meters swingin' wildly, I found that it remained clean, open and airy as expected, despite the abusive settings. However, I felt I hadn't crossed the threshold into where it really began to have a "sound" Well, in fact, I had passed it.

   One of the beauties of reviewing products made by small companies is the ability to communicate directly with the designer. therefore, after my initial tests, I picked up the phone and called Mister G. our chat resulted in a personal sonic revelation.

   I know that the slow response of a VU meter inhibits its ability to register transients (drums, percussion) while more accurate needle movement occurs with average material such as vocals, bass and crunchy guitars. Despite the geek degree, I did not transfer this knowledge of VU meter characteristics to its inability to display transient amounts of gain reduction, Until now...

   You would not record a tambourine at 0 VU on an analog tape machine because the transients are 10-dB to 15-dB higher. I got the sound I was looking for with the 6386 set to FAST mode, with no control options except Threshold and no more than 1 dB of meter movement when viewing the amount gain reduction. In Fast mode, the 6386 is doing much more than the 1-dB of meter movement indicates. Much more!

   The 6386 has a three position Mode switch, labeled Fast, Presets and Manual. There is no Ratio control and the 6386 vacuum tube doesn't need a dominatrix to get it to move. The way it reacts to the FAST preset is impressive. It simultaneously turned on the mental light bulb while humbling me with its almost hidden power. The first four presets are modeled after the Fairchild, while the last two presets -one faster than the other - are both program-dependent to minimize "breathing."


   I daisy-chained the 6386 and Pendulum's OCL-2 (optical tube compressor) and installed them in a rack along with a Great River transformer-less mic preamp and the TASCAM DA-45HR 24-bit DAT recorder. The rig was schlepped to a room to make a stereo recording of a rock band in rehearsal mode. I set the band up around a pair of spaced omni mics (AKG C300 bodies with CK92 omni capsules).

   Off in another room I listened through Fostex headphones while experimenting with both limiters, using one, then the other, sometimes both and sometimes none (taking detailed notes). Up until this point, I only used the slowest Attack and fastest Release settings to play it safe. More aggressive tweaks were planned after the recording.

   Back at the ranch I listened on two systems, one not so great, to prove the original purpose of compression - to squeeze a wide dynamic range into a smaller dynamic space. As deranged as some boom boxes can be, most only have between 5 and 10 watts of power. A single guitar amp has way more than that, not to mention the bass amp and the drum kit.


   Listening on the compromised system - computer speakers with a subwoofer - I began with the slowest Attack and fastest Release settings, which in this case are 40mS and 0.1S, respectively. Attack time was gradually increased until the small amplifiers stopped distorting, taking maximum advantage of their limited dynamic range and making a DAT copy for later comparison.

   On my real system, which features Dynaudio BM15A self-powered monitors, I quickly realized that reducing the dynamic range only confirmed what my ears (in the original recording space) had first told me - that the band's playing was a bit too sloppy and no magic box was going to polish their performance. It is not always possible to hear compression on small speakers. Always check an alternate set of monitors, preferably a more accurate system, to make sure you can stand to listen to all the details, as well as the compression artifacts. (Wink-wink. Nod-nod!)


   I then dug up a DAT remix of a 24-track project originally recorded 20 years ago and recently re-mixed on a Soundscape workstation. The song, "I Move Easy," written and produced by guitarist T.J. Tindall, had a slow almost N'awlins-style groove. The session players comprised the main rhythm section for many of the Gamble and Huff projects that came out of Philadelphia back in the Seventies. The vocalist was Willie Chambers of the Chambers Brothers.

   In the re-mix, the breakdown was too soft relative to the meat of the track. The OCL-2 was first in the chain, using preset four (the faster program-dependent setting) and a 5:1 Ratio. The goal was to raise the breakdown up to the level of the main section of the song. It worked, though it would have been more perfect had I added side-chain EQ. (A low frequency roll-off would have desensitized the OCL-2 when the bass came in about midway through the breakdown.)


   With one problem solved, I was still wishing for more Release time to bring out more of the natural room sound. The track was very precisely mixed, with a punchy kick and snare up front (the latter with an early Kevlar head giving it a monstrously dead 'thwomp') but the real room ambience was a bit too far back.

   Instead of asking for more from the OCL-2, I added the 6386 in the Fast mode. Reducing the Threshold - until the gain reduction meter was just beginning to show signs of processing - instantly made the track open right up. I felt like Moses conducting the Red Sea! After finding the magic settings, the tracks became "one" and more alive without losing impact and without side effects.

   Previous attempts at using FAST mode failed because I was looking at the amount of gain reduction instead of listening. If the meter could be more sensitive in FAST mode, and/or perhaps augmented with a multicolored LED, it would give the necessary user-feedback, the knowledge that work was actually being done.


   The luxury of using both units together was not lost on me. I am very aware of the price for each, which is comparable to other low-volume pieces of quality outboard; together, they're still about one-third to one-quarter of the street price for the beloved Fairchild. The combination of Pendulum OCL-2 and the Pendulum 6386 kicked sonic ass, and, as a new dad I can comfortably say it wiped it clean and changed its diaper! I was completely comfortable using both with no fear of signal degradation.

   In all cases, I was so thankful for the BYPASS switches, which are not common on vintage products (and even some modern units). Having both units in the signal path allowed me to compare identical settings. With slow attack and fast release the response was similar, but in FAST mode, the 6386 is more responsive.

   The OCL-2 has the absolute minimal signal path with enough speed if need be (certainly faster than any other optical processor). Though not as quiet as the 6386 it has gobs of headroom. (Maximum out is +35 dBu into a hi-Z load, +24dBm into a 600-ohm load.) Faster yet is the 6386 with more support circuitry but no detectable difference in the sonic texture. (Maximum out is +27 dBm in to a 600-ohm load.) Maximum gain reduction is only 12 dB for the Variable-Mu design, while the optical is capable of 27-dB. I did not test either of these units on individual tracks - consider this more of a mastering-style review.


   In my conversations with designer Gualtieri, one comment that really stuck was his description of the physicist's philosophy as being, 'in search of the truth.' In scientific terms, it means the ability to repeatedly conduct an experiment and get consistent results. In audio terms, this translates to a product's ability to consistently deliver the goods - something the 6386 certainly does. The Pendulum 6386 Tube Limiter is good sound, clean science.


EQ Lab Report

MANUFACTURER: Pendulum Audio, Inc.; PO Box 339, Gillette, NJ 07933; Vox: 908-665-9333 Web: e-mail: info@

APPLICATION: Audio Dynamics Processing

SUMMARY: Stereo (or Dual-Mono) compressor / limiter with vacuum tube front-end followed by high performance opamps, biased Class A for low distortion. I/O connections: transformer-balanced inputs, active-balanced outputs, both on XLRs. In addition, 1/4 inch TRS side-chain access. (Transformer-balanced outputs are available, but not recommended.)

STRENGTHS: Three Response modes: Fast, Preset (4 are "normal" plus 2 "program dependent" settings) and Manual (front panel Attack and Release controls); in addition to a Bypass switch, the output is normally bypassed until the unit warms up and voltages are stabilized.

WEAKNESS: The Gain Reduction Meter needs (and may get) expanded range to more accurately reflect the amount of processing when in FAST mode.

PRICE: $3995.00