SAE 47, The Next Generation
I've been waiting all year to talk about this, wanting to make sure that the mics I'm currently shipping are all up to these new specs; and now the time has come to finally talk about it. This blog is going to be about two things really; the change in what the SAE 'Baseline Edition' 47 is and means, and also the overall updates to the design of our U47 and U48.
Let's talk about the updates to the design first.
So, we've been handmaking and selling these tube 47's, using mostly our own SAE 47 REV 1 PCB system, since about mid or late 2017. Time sure flies. This design sure served us well and sounds fantastic; but at the same time we knew there were improvements that could be made.
One subtle change has more to do with the philosophy of the 'dual tube 408a solution'. This is the tube solution we use, one originally proposed by the late Oliver Archut of AMI fame, which utilizes two selected 408a tubes set up in a very particular way. The audio input is run in parallel with the heater supply running in series. This solution, unlike any other tube solution out there, allows the 47 to have about the same gain factor (mu), plate resistance, and voltage parameters as the original 47. Every other '47' design out there more or less requires a more elaborate power supply design. That's perfectly fine; but the dual tube solution just allows one to get as close to historical accuracy as possible and retain the archaic 'single rail' power supply scheme of the original. For some time, however, I've realized that it may not be the wisest thing to force the 408a pair to 'live' in the shoes of the original VF14m, and maybe one could coax that last 1% out of these tubes by giving them exactly what their spec sheets are looking for in terms of their nominal heater voltage and bias, etc. These figures were VERY close to what they were already getting; but we have made some slight tweaks to give them 'exactly' what they want to see instead of what the VF14m was given. We also redesigned the way the vacuum tube daughterboard connects to the mains board, and its a bit more ruggedized than before, mating the board using gold plated mil-max right angle headers which are allowed to flex, so that there are no fragile solder pads on boards which could be snapped due to stress or shipping damage. This never happened before; but we wanted to ruggedize this aspect anyway.
In the 3+ years I have been making these mics, I had only one specific customer who lived near a radio station, complain about RF interference in the microphone. I was unable to resolve this, and to this day I currently own his returned unit as my personal studio 47, and have never been able to reproduce the issue. Nonetheless, it bugged the crap out of me and I wanted to offer a solution. Therefore, on this new design, we have built in a complete RF filter circuit (accessible by jumpers on the mic board) inspired by the RF filters used in broadcast on later editions of the U87 and other models. This filter ships in the 'off' position, and we don't see any reason to engage it UNLESS you experience RF issues... but it is now there for your peace of mind! Ideally, of course, the mic would sound the very best in the off position; but engaging it is barely audibly noticeable. Yes, we engineered something literally due to the experiences of one customer... but these things matter to me! In addition to the RF filter option on the audio output; we have added Ferrite rings to both power rails coming into the microphone (B+ voltage rail, and relay voltage rail), which will block out any slight EMI or RFI that may creep in down the cable from the power supply. These are inside the bottom bell of the mic, and are always present... and do not affect sound.
Let's talk about parts!
One of the first things I want to say, even though this is a bit superficial, is that hey... you only live once, and we are really proud of these new boards. So we have printed our new mk2 PCBs using one of the most expensive methods and via a very high quality provider known for delivery very pure materials and rugged PCBs. These new boards have extra wide traces with gentler curves wherever possible, and are made using the ENIG process of 24k gold immersion, meaning that every exposed trace and contact is corrosion resistant and gold plated.
We have now also developed our own custom wound BV8 type transformer based on a golden reference part that was my personal favorite BV8 transformer. This has been a long time in the making. The new, high quality, American made transformer has exactly the right amount of inductance and windings to perfectly communicate with our specific tube compliment, and it matches the girth, bandwidth, tonal qualities, and THD profile of my personal favorite specimen German part.
We are still working down from our private stock of NOS Western Electric 408a valves, still considered by me to be the 'only one that sounds right' for this application.
We are now using Sommer cable build the 7 pin cable, and have upgraded to the PSU end to a rugged, twist-locking connection. We still use Gotham cable for the 3 pin XLR.
We came upon a stash of the 'red' NOS paper in oil capacitors, which is used in the output coupling stage, the most critical position, in the mic circuit. Both the green and red Soviet era caps are mil-spec parts; but the reds were considered to be longer life, higher tolerance, critical application parts; and to my ears, have a slightly more authoritative bass response. We will use these sparingly, for as long as supplies last. To my ears, nothing made today still has the 3 dimensional soundstage of these PIO caps.
Speaking of caps, two of the caps in our new PSU are actually custom made electrolytic capacitors, designed and made just for Signal Art by SuperTech out of Taiwan. We wanted a 47uF electrolytic that was 10k-hour and +105 degree C rated, as well as better than 5% tolerance. We couldn't find anything on the market that quite cut it, so we actually had the part made. These are the 2'd and 4'th capacitors on the board; the other 2 being ultra long-life/high temp, high quality Nichicon parts.
As you may have noticed also, we are now using carbon composition and carbon film resistors in some positions for tonal reasons, along with standard metal film resistors in the higher value positions.
The new power supply board is a radical departure from our previous design, and, we believe, a radical departure from most others as well. This new board produces WAY less heat, is more than twice as energy efficient, delivers more available power, powers up quicker, stabilizes INSTANTLY, and remains locked in to my calibration settings literally FOREVER, unless changed. Utilizing a high voltage MOSFET regulation system and more diodes and better filtering, this board delivers clean, ultra-stable power, and plenty of it. The old PSU system worked just fine; but was a space heater and thus just not as eco-friendly. It was also devilishly hard to calibrate. Not to worry, because I put in the time and effort to do it right... I burned them in for DAYS and calibrated them at the final stage with their mic and final tube compliment connected, and using a regulated AC power supply and VARI-AC. But, it was tricky to tune them exactly perfect in a manner that would be stable if the user moved countries, moved to a different power grid, changed the tubes out with a different brand tube that had a slightly different current draw, etc. It was an Achilles' Heel of the old design... but no longer. This new system is literally bulletproof. You can change the tubes all day long, and move all over the world with it (switching 115/230 switch on PSU still prudent!!), but other than that it literally stays locked in to the perfect settings and doesn't care. The design is so well regulated that it measures the same even whether the mic is connected or not! (but please as a safety precaution do not power on without the mic connected). I can finally rest easy, knowing my settings will remain 'in stone' regardless of tube changes or territory changes over the course of the microphone's life once it leaves my hands. This is a great feeling, and we worked really hard to achieve this.
As advanced as the new power supply design is, it still gives a knowing nod to the past. We were hesitant to move on to a more modern 'switched mode' (also called switching) power supply, even though many (including Neumann) have already done so; preferring instead to accomplish the desired outcome using a more traditional approach. We also did not want to 'throw out' most of George Neumann's original passive PSU design. I still say, that if the mic 'doesn't get hot, then its not a 47'; and our mic 'still gets hot' (within safe parameters, of course). We still use the original heatsink-mounted resistor inside the microphone to divide off the H+ voltages and so the latter half of Neumann's ingenious 'single rail' passive power supply scheme still lives on, it simply lives behind the protective wall of an extraordinarily well regulated front end, no longer subject to changes at the wall. I still believe that some of the audio 'gold' lies in this type of design, and I didn't want to change it. It's truly the best of both worlds, and we're really delighted to finally share this info with folks.
Now on to the Baseline Edition...
I've really decided that I hate compromise. I don't like having 2 standards, even if those standards are good. At the end of the day, it literally takes the same amount of time and effort and TLC to build one of these mics, whether its the baseline edition or the premium edition... so it just made no sense to me to continue putting in the same work to build a second tier of products. I made the same profit on both versions, so from that perspective it made no difference to me... there's no reason to 'upsell'; I just ultimately want to make only the best mic I can make and not have to use up energy trying to explain the very subtle differences between the two models. Our imported K47 capsule was great, and our more economical BV8-style transformer was great... but its just easier for me to stock only the best parts and to create savings for folks in other ways. Ways that have zero impact on audio.
So, from now on, every 47 has the same M7 capsule made by Eric Heiserman (Heiserman K47 on request), and the same premium US made custom wound BV8 transformer. And all the same other internals, too. No sonic difference.. NONE. What is different is that the baseline edition is going to be a cardioid only 47 (and obviously there won't be a baseline 48). It saves some effort and cost to remove all the pattern select circuitry and just 'hardwire' the mic for cardioid. In fact, some might say that simpler is better in this regard. If you mainly do vocals, then this is ideal and it sacrifices nothing other than some parts you don't need. What is also different is that the Baseline Edition will forego the imported German wood box. It will sit in a softer foam insert, inside the cavity of our new and improved custom shipping packaging, inside the custom corrugated cardboard giftbox. The baseline edition will also have a standard 'off the shelf' AC power cable instead of the one I made by hand. This, IMO, also doesn't affect sonics... just aesthetics. and lastly the baseline edition will not ship with an extra set of tubes, as the premium one usually does (but we'll tell you where to find spare tubes).
So there you have it. We've simply eliminated about $150 worth of parts and trimmings to deliver the same exact quality of microphone for $300 less. I feel like that's a better deal in the end, and less confusing than having a different line of products. You can pick up a Baseline 47 and know that it will perform exactly like the premium one, just without the pattern feature.
I look forward to hearing from folks about this change, and if you agree it is the right way to go.