Tag Archives: custom audio equipment fabrication

Another Carbon Mic Power Supply

CarbonPS_TopCarbonPS_FrontHere is a one-of-a-kind unit using a re-purposed Thordarson transformer.  IT works great, and the vintage Shurite (made in New Haven CT!!!) DC voltmeter is a bonus.

CarbonPS_MeterCarbonPS_TerminalsOn the ‘business-end’ (topmost image) you can see XLR output jack (for connection to the input of your mic preamp), on/off switch, pot for controlling the DC voltage that mic receives, and at right the cable-exit for the seven-foot cable harness that connects to your double-button carbon mic.  At right: red and black wires connect one-to-each button, and the clear wire connects to the metal shell of the mic.

For an explanation of WTF this thing is, and why you might possibly need it in your life, see this previous post.

Marantz 7 Phono Pre/ Passive Line Preamp

Above: a  Marantz 7 stereo phono stage built for stand-alone use.  See this link for an earlier build of this same circuit, along with an explanation of exactly wtf a phono preamp is (for my 7 or 8 non-technical readers).

The major difference with this build is that I included a 2nd set of (passive) inputs and a volume pot.  This is to allow the user to connect both a phonograph and a 2nd line-level source, select a listening source, and control overall volume level ahead of a stereo power amp.  I also used a tube rectifier and a choke this time.  The piece sounds fantastic.

Stereo phono preamps are fairly time-consuming to build, and small differences in layout can have dramatic effect on the overall performance.  Here are a few snapshots of the process.

Hammond steel chassis, punched-out using Greenlee hand-punches

The underside of the unit following mechanical assembly

Initial wiring.  I always start with the ‘no-brainer,’ rote wiring tasks:  First, the 120AC wiring.  Followed by the B+ supply.  Followed by the Filament supply (if any; here you see the DC filament supply constructed at the left of the turrett board).  Finally, any passive audio-control wiring (the switch, pot, and Belden cable on the left). I wire up each one of these sub-assemblies and test each one; having 100% confidence that all this stuff is functioning properly makes it a lot easier to troubleshoot and vague performance issues once the piece has been fully wired.

…and done.  It’s hard to see how many components are mounted on the tube sockets, but trust me, it’s dense.  It never ceases to amaze (annoy) me how complex phono pres end up being.  The schematics look so simple!

Vacuum-Tube Output Stage for Shure Level Loc

Many of my regular readers will be familiar with the Shure Level-Loc.  For those unfamiliar, the basics:  the Level-Loc is a brickwall limiter made by the Shure microphone company for public-address-system use (podium mics, specifically) in the 1960s.   It uses discrete transistors and transformers in the signal path; it offers balanced mic-level i/o and an unbalanced consumer-level 10K ohm impedance output as well.  There is an input-level control (simply a pot that follows that secondary of the input transformer) and a switch marked ‘distance selector,’ which seems to me to be a threshold control.  That’s it for control.  It is fairly noisy (full-bandwidth noise), even after a recap,  and the transformers are not especially well-shielded.  It runs off of a 9V battery.  For more information on the Level-Loc, you might want to start here.

Anyhow…seems like a toy/piece-of-junk and maybe it is, but these things have become highly coveted for use in recording rock drum performances.  How much so? Well, how many other prosumer PA-system pieces are currently available as a plug-in, an API-500 series module, and a boutique re-build? (image source for above)

I recently picked up a clean Shure Level-Loc for a few dollars at a yard sale; after the aforementioned re-cap (and we’re talking about 20 capacitors here…), it was sounding like it was probably operating within its original design parameters.  I was intrigued, and figured it might be worth getting it into the racks at Gold Coast Recorders to see what it could do.   GCR is a big, live-sounding room, so there’s plenty of sound to get out of it with a squashed compressor.  The only potential problem: the Level-Loc offers only mic-level or low-level medium-impedance output.  I like to run my mic preamps directly into the Lynx convertors; so for the most direct signal, I would need a bridging amp to bring up the level and lower the impedance of the Level-Loc.  It would be nice to have an output level control too, and I wanted the piece to be as physically small as possible so that it could sit directly next to the Level-Loc on a 2RU rack shelf.  Here’s what I did to solve all of these problems and fold the Level Loc into the studio alongside all the other outboard mic preamps.


Most of the RCA Receiving Tube manuals have a schematic for some sort of ‘audio input amplifier’; I wanted one that would provide about 20db of gain as well as a very low-impedance output so that I could drive a 15K:600 output transformer easily (I used an NOS UTC Ouncer to save space).  Based on this i selected the circuit above RCA manual # RC-24.  One 12AU7 tube.  Simple, easy.  The physically smallest plate/filament transformer that I had on hand was an NOS Stancor 120v:120V/6.3V, so I used the voltage-doubler B+ supply circuit as-found in the Altec 1566 and 438:  here is that schematic via Tangible Technology:

power_2b  For once i actually did not bother wiring up a DC filament supply, since the gain of the unit is pretty low.  This was the right choice, as i can hear no hum at all in the finished unit.  I added a 500K pot at the input jack and voila.  The whole thing fit inside one of those aluminum Bud Boxes that some folks use for DIY’ing guitar effects pedals.  I left the power transformer bolted outside the unit on the rear; it’s always a good idea to keep power transformers away from audio transformers if you can.   Here’s the interior of the unit:

…and below you can see the finished piece.  Not my finest piece of industrial design but it does the job.  I put the 1/4″ TS input on the front of the unit so that I can use it as a DI input for Keyboards ETC if the need arises.  The circular grill on the top surface of the piece is a heat-vent positioned directly above the 12AU7 tube.

The unit performs well, especially considering that there are only two stages of filtering in the power supply.  I always use four stages of filtering in the equipment that I build for customers, with at least one choke; I was curious this time to see how the basic Altec B+ scheme worked, though, and it seems just fine!  People love their 1566s and 438s so fukk it.  Good enough is sometimes good enough…

Above, the two units side-by side.  So how does it sound?  TW and I were putting down live drums on a track at GCR and here’s the result we got.

First, the drumbeat: close-mics only: CloseOnly

Here’s the same mix, but with the Level Loc signal added: in this case, the Level Loc was amplifying a figure-8 ribbon mic 20 feet from the kit, with the null of the mic facing the kit; the waveform was then re-aligned to eliminate some of the delay: withLevelLoc

And finally, the Level-loc signal only: LevelLocOnly

The Level-Loc is aptly named.  Regardless of what you put in – a baby’s breath or an atomic blast – you get the saaaaame level out.  Zero dynamics.  It’s pretty uncanny.  And a great sound for heavy rock drum beats.  This is the 2nd track that I have used it on in a week and I think it will continue to get a lot of use at the studio.   The output of the balancing amp is a little low – even with the input attenuator all the way open it cannot quite get to full level via the Lynx convertors.  It’s good enough, but it could stand to put out a few more DB.  If you build one of these devices for use with your Level-loc you might consider using a 15K:60K interstage transformer at the input to get a little bit more level out of it; or re-bias the two stages in order to use a 12AT7 instead of the 12AU7.

Audio Transformer As Signal Processor

Sometime in the past couple of years, Tape Op ran a short piece by Allen Farmelo titled “Using Transformers to Transform Audio.”  You can read the piece in its entirety here.   My reaction at the time was ‘it’s about time!’  Audio transformers are a crucial part of what we think of as an ‘old-school’ or ‘vintage’ sound.  My clients at the studio often ask me what makes tube-audio gear desirable, or ‘better,’ and I am always quick to relate that when vacuum tubes are operating in a linear (IE., not-distorting) way, you shouldn’t really ‘hear’ the tube – it should be amplifying, nothing more, nothing less.  Of course once you push a tube into breakup the effect can be quite different than a distorted FET or transistor but you get the idea.  A clean tube signal should sound… clean!  So, anyhow, the next point that I will make is that tubes are rarely very far from audio transformers, at least in pro-audio equipment, owing to the usefulness of ‘free-gain’ at input stages and the necessity of plate-or-cathode-matching at output stages (if this sounds like jargon to you/// basically/// tubes need transformers in order to play-nice with other pieces of gear).   The point: what we think of as ‘that tube equipment sound’ is really due to the transformers as much as the tubes themselves.

I won’t go into all the various effects that transformers create, as Farmelo does a very good job of explaining it in his piece.  Suffice to say: it is a very real, and very subtle effect.  Audio is a game of inches, though, ain’t it.  So when a regular customer of mine recently ordered a custom piece to allow him to use some high-quality transformers as a subtle signal processor in his studio, I was ready to go.   Here’s what I whipped up:

A single-rackspace unit – two 4PDT toggle switches on the front offer clickless true-bypass for each channel.  The switches are beautiful Japanese made units; each can handle 12,000 (yes twelve thousand) watts of electricity.  They should last…

On the rear we see Neutrik XLRs (my price/performance favorite) and…  a pair of 600:600 FREED output transformers pulled from some Scully 280 electronics that were too far gone to rebuild.  The transformers themselves are flawless, though, and they sound great; I have many of them at use in my own studio for various tasks.

Inside it’s just a buncha wire… Belden 9451… and at the rear you can see the heavy copper ground buss with a single chassis-contact point on the left.

Overall the transformers introduce a 1db loss in level to the program.  The effect is certainly subtle at reasonable levels, but I notice a more ‘organized’ sound to the extreme low end – it seems less vague while still retaining the full extension in the subwoofer.

Stereo Audio Power Amplifier (Home Use)

Above, a recently completed piece.  It is built on one of two identical NOS cabinets that I pulled from a Milford CT basement some years ago.  Aside from the unusual enclosure, this is the same “iPod” stereo amplifier that I have built before; see here and here for other examples.

The meter (0-30V DC) is a bias meter for the output tubes.  The switch below selects left, right, or off position.  This will help alert the user as to an appropriate time to replace the output tubes.  The circuit is very simple: RCA jacks on the back feed a 100K dual-pot, then on to a 6SL7 (one half per channel) and then to a single 6L6 per channel.

Inside the cabinet are a very large EDCOR power transformer, two paper-wrapped output transformers, a choke, and a whole bunch of filter caps. Rectifier is a 5U4 and filaments are AC.  Electronically it is nothing special but visually it is one of my favorite pieces yet.

Christmas 2011

Despite the hectic nature of the holiday-season, I am inevitably able to produce one piece each year as a gift for some person or persons near+dear to me.  This year’s recipient of the prize-winning entry in Connecticut’s Got Free Time! will receive this charming set.

I’d been thinking a bit about V. Gallo’s comment in Sound Practices #1…  see the previous post for the full details…  basically, the idea that ‘good hi-fi should sound like the best radio you’ve ever heard’ (as opposed to some supposed verisimilitude to an actual acoustic event).  Well, here’s an attempt at the best-sounding radio you’ve ever heard.   At the front end, two RCA inputs mix via fixed resistors to a single 100K pot, and then onto the grid of the voltage amplifer stage (6J5, the ‘single’ triode iteration of the more common 6SN7); this feeds one 6V6 via a gigantic paper coupling cap; in the power supply, a 5Y3 rectifier and R/C filtering with large (50uf) caps and an extra filtering stage before the output transformer for a n extremely quiet, stable signal.  The circuit is very similar to what you would find on the back-end of most transformer-mains-isolated AM/FM tube radios of the 40s and 50s.

The speaker is a 1950’s extension speaker, maker unknown; it has extremely nice finish work and detail; it is unusual in that both the front and back of the enclosure are open via grille clothe, creating a ‘bright’ and ‘dull’ side; position the speaker as-you-like for a very unconventional tone control!

The cabinet is 1×3 solid cherry with internal bracing.  Keeping with the ‘best-radio’ theme, the tubes, transformers, and brown Bakelite sockets are pulls from various Ancient Radios that have wandered into the shop over the years.  And yea it sounds great…



Above: The BRDCSTR, a two-stage vacuum-tube microphone preamp inspired by classic 1940s broadcast audio preamps.  Single-ended design with a 6SJ7 and 6J7 tubes, massive output transformer, and only one capacitor in the signal path.  I have built more of these than any other design; people really seem to respond to the sound.  It’s pretty much the least ‘stuff’ that you can put between a microphone and your A/D convertor.  All the attitude of ancient tube studio equipment without the hum and the noise, and with much better frequency response.

The Old ’76

My latest microphone preamp design is completed and sounding very cool… The Old ’76.

The Old ’76 is a novel three-stage design; it’s not based on any past or current production microphone preamp.  That being said, the circuit is nothing revolutionary: an input transformer (in this case a UTC O-1, as I was temporarily out of Jensen 115s) connects to an RCA 76 tube, biased in textbook (or, at least, RCA-Receiving-Tube-Manual-book) fashion; volume pot follows and then on to a 6SL7 tube with the first stage plate directly coupled to the grid of triode 2; the output is a cathode follower with a Solen cap and onto an Edcor output transformer.   The pushbutton switch (with associated pilot lamp) activates 48V phantom power.  Overall gain is approx. 60db and response is flat 50hz – 18khz.  I am betting that the slight high and low end roll off is due to the UTC O-1; the next build will use my customary Jensen input transformer and we should see 25hz-20khz flat response.

76 tubes were often used in ancient console radios with a large tubular shield positioned around them; the reason for this became quiet clear once I had finished my piece; before I added the above-depicted aluminum plate between the 76 and 6SL7, I was getting some unpleasant ringing on very high frequencies.

I had wanted to build a mic pre with some of these very old two-digit designation tubes for some time now.  Hi-Fi fans seem to love the 76 tube, so I figured there was probably something worth investigating.  76 tubes are readily available and pretty inexpensive; it is a bit of a challenge to find the 5-pin bases that they require though.  AES sells only one 5-pin base, and it’s expensive and oversized.  My prototype unit here used an older Amphenol socket that I dug up somewhere.

Some other miscellaneous design notes: B+ rectifier is a 6×5 tube; hammond 15mh choke was used in the B+ chain; DC filaments, as usual; the phantom power supply has a slow ramp-up when activated; the three-pole phantom switch 1)connects the 48v supply to the main power supply; 2)connects the phantom indicator lamp to the 6v supply; and 3) connects the 48v to 6.8k resistors that actually connect to the mic input jack.  I find a three-pole switch necessary in these instances because there is some ‘drain’ time involved when phantom is turned off, even with a bleeder resistor; disconnecting the 48v right before it hits the mic jack provides 100% assurance that you (I) won’t melt that BK5 ribbon (again).

An A/B test of The Old ’76 vs. the usual API 512 will be posted here in the next month or so.


Sound From A Glass Box

This design project began with the goal of crafting an entire amplifier that echoed the form of a vacuum tube itself.  See here for example of the intial execution.  Thanks to cabinet-maker N.N. for the beautiful walnut frames.

The 22277 is a two-channel audio amplifier for home music-listening.  Power output is approximately seven watts per channel.  Each channel uses 1/2 of a 6SL7 twin triode and one 6L6.  The rectifier used is a 6AX5.

Volume control is provided.  Inputs are via twin RCA jacks and speaker outputs are via 1/4″ jacks.  The relatively high gain of the 6SL7 tube allows the unit to be driven to full output from any line-level source (E.G., radio tuner, DVD player, iPod, etc).

Marantz 7 RIAA Phonograph Preamp

For my non-technical readers: a phono preamp is a device which does two basic things:  1) it equalizes the program that the phono cartridge picks up off the LP record, basically by boosting the low-end and cutting the high end, with this action centered at the fixed frequency of 1K hz (for full details on the ‘RIAA compensation curve’, and why it is necessary in the manufacture of LP records, read here); 2) secondly, a phono preamp must amplify the signal of the phono cartridge to roughly line-level (IE., the level that would come out of a CD player or VCR) and also deliver this signal at a low enough impedance such that it can be in-put to any receiver or consumer amplifier that it might encounter. Here is my attempt at a self-contained version of the phono pre-amp section from the Marantz 7 hi-fi preamlifier.

(web source)

I’ve never heard an actual Marantz 7c in action.  Considering that this dude, who does not seem like a total flake (based on his sales record) is selling one for $4,999 on eBay, I had to assume that it sounds fine (at least).   Marantz is one of the legendary American Hi-Fi brands from the ‘Golden age of Hi-fi,’;  their original line-up of products (before Saul Marantz sold the company) are widely lauded for both their sonic and aesthetic traits (see here for previous Marantz coverage on PS dot com).

(web source)

Above, the schematic that I used for this build.  It is a simplified version of the Marantz 7C, omitting such features as pre-RIAA disc compensation curves and bass/treble controls (for the complete 7C schematic, see here). The only change that I made to the audio section is that I added a 1K resistor between the signal output (the .47 cap) and the output jack.  Couldn’t hurt, right?  The phono pre-amp section consists of the three triode stages that you see at the left.  The three stages to the right are additional make-up gain that follow the overall balance and volume controls.  Important to note:  the output of the phono stage is a cathode follower.  This means that the signal is derived from the cathode of the tube rather than the plate.  This results in very little voltage gain (in that particular stage) but…  also…  pretty low impedance.  Which is what I wanted.

Above, the RCA RIAA phono preamp as published in several of their “Receiving Tube Manuals.”  I have built this RCA circuit before; while excellent-sounding, it does require an additional stage of amplification (I used a cathode follower circuit) to lower the output impedance of the device if you want to be sure it will ‘play-nice’ with all yr other kit..  Notice the note at the output leg of the circuit: “220000 ohms minimum.”  Wow!  That is very high impedance.  The only devices that this thing should feed are either the grid of another tube or possible a FET.  This condition makes the RCA circuit (on its own) insufficient as a stand-alone device.  And RCA did not really intend this to BE a stnd alone device; rather, they intended that you would build this into an amplifier where the circuit could directly feed the input grid of a preamp tube.  The Marantz 7c circuit, with it’s third cathode-follower stage, does not have this limitation.

Above, the interior of the completed piece (audio chassis).  Note the French SOLEN coupling caps (my favorite due to good performance, reliability, and small size) and single ground buss (the piece of copper that runs along the lower edge).  RCA used the single ground buss technique in most of their broadcast equipment, and AFAIK, this grounding style cannot be beat for performance and ease of manufacture.

Above, the interior of the power supply.  I built the unit into two seperate chassis: audio and power supply: this was an aesthetic preference of the customer, and it also makes good sense when you are dealing with the miniscule audio voltage that emanate from a phono cartridge.  Sure enough, the finished piece exhibits no (z-e-r-o) hum whatsoever. I did not follow the Marantz 7C schematic for the power supply; I just built what I though was neccessary: a DC supply for the filaments and a B+ chain with 4 stages of filtering (no choke).  I used an NOS RCA power transformer that was a lil’ bit too exuberant for the 280V B+ requirement, hence the large-ish 15k ohm resistor near the bottom (the Marantz 7c schematic called for a 3.9k ohm in this position).  When all was said and done, after experimenting with a three different resistors in this position, I was within 1% of the voltages specified in the schematic.

So how does it sound?  Very good.  Compared to the RCA phono pre, it rejects WAY more RF;  it is very rich, extremely quiet, and the low-end response is so, so much better than the phono pre amps that come built into modern stereo receivers.  I QC’d the piece with a good cartridge and a clean pressing of TUSK and it was “as  Lindsey Buckingham intended it to be heard.”  If you are thinking about making a tube phono preamp, give this one a try; build cost is very low and it went together very fast with no hassle.