Separate names with a comma.
And class A
This would probably involve some mods and being very careful.
Okay, guys, you're getting this chapter a bit early...because tomorrow morning is toooooooo crazy and I'd rather it not slip to the afternoon.
For the TL;DR crowd: a real Sol first article will be at the Schiitr on Friday.
And tons of backstory.
And not the last chapter.
Yes, I'm gonna milk this one for a bit. Give me a while to format it, and then it'll go live.
Just a gentle reminder it’s actually two matched pairs of tubes, not one matched quad... but a single tube is way simpler.
2019, Chapter 3:
Here Comes The Sol
Or, How I Stopped Worrying and Learned To Like the Disk
A chapter about Sol?
Before you get yourselves too wound up, this doesn’t mean that Sol is shipping this month. Consider this chapter more of a “first in a series” about our upcoming turntable.
<sarcasm> Yes, more blathering and less products, I know that’s what you all want. </sarcasm>
Seriously, though, Sol has hit a milestone as of this week: we have completed our first Sol made with production castings, production motor, production sheetmetal, and production CNC parts.
This may not sound like much, but it’s a very big deal for a product that has like 36 mechanical components. Consider that our products to date have never had more than a single digit’s worth of mechanical components (most have 2 or 3 major mechanical parts), and you can see this is a much different product than anything we’ve launched before.
Actually, it probably wouldn’t be terrible to outline some of the things that have caused the development of Sol to take much longer than normal. Here goes:
This is, by far, the most mechanically complex product we’ve attempted—it has 10x more mechanical parts than anything else we make
Sol has parts with tolerances down to 2 thousandths of an inch—far tighter than any previous product
This turntable features two very large aluminum castings, which required the production of tooling that’s approximately the size of a semi truck engine for each part—the development and tweaking of the tooling was a not-insignificant task, both for the casters and ourselves
It also has a complex mix of aluminum, brass, plastic, steel, carbon fiber, sorbothane, and rubber components, several of which have to be glued together in precision jigs—a level of materials engineering far in excess of anything we’ve done previously
Sol’s production process is a complete unknown—we’re still breaking down provisional steps and deciding who builds what at the moment, and this might…er, will change radically after we’ve made a few dozen
In short, this is a biiiigggg project, and we’re, to be honest, a little scared of it.
Which, frankly, is good for you. It means we aren’t rushing out a half-baked design, so full of ourselves that we think we can’t fail. Here’s the reality: we’ve never built a turntable. There are a lot of other turntable companies out there. We really, really need to offer you something different—and better—and make sure it’s as flawless as possible when it goes out. So we’re taking it very, very, very slow.
“Yeah yeah, enough excuses,” you might ask. “When’re we gonna see this thing, what’s its claim to fame, how much’s it gonna cost?”
Well, you’re going to see the first true first article assembled from production parts at the Schiitrshow on February 8, if you’re in the area and feel like coming by. It’ll be pretty close to what we’re shipping, with the caveat that the plinth probably will be painted in some wonky spray crap rather than the textured powder coat we specified. Oh, and the caveat that the real-time VTA adjustment will feel a little rougher than we wanted it to, because we had to make a change to the machining. Future first articles will be right.
“No, I don’t mean first articles, I mean production. I also noticed you dodged any claims to fame, and didn’t tell us what it cost.”
Production? Let’s start first with some more first articles. I’m hoping that we can make 10-12 more first articles from production parts and send them out for people to beat up on a little, in case we need to make some tweaks we missed. Beyond that, if all goes well, April/May might not be a bad bet. Recall how my “if it all goes well” predictions have worked in the past. We’ll see.
An important caveat for when we start selling these, though: we’re going to make it as inconvenient as possible, in order to limit demand.
I hear the screams from here. “Inconvenient! Limit demand! Why do you hate us?”
Actually, we’re not doing it because we hate you. We’re doing it for a very pragmatic reason: limiting demand allows us to have a chance to keep producing high-quality turntables, rather than being overwhelmed with orders:
Here’s what we’re doing, specifically:
We’re not including any standard cartridge.Yes, that’s right. No standard cartridge. None. Nada. Zero. If you’re looking at Sol as an alternative to take-it-out-the-box-and-start-playing-immediately tables, this may be a dealbreaker. Because yes, Virginia, you have to choose your own cartridge, and set it up yourself, using a video Tony will produce.
There are no ease-of-use accessories. Although yes, Sol will have a cueing mechanism, it’s not going to have a dust cover. Yes. Again, if you’re comparing Sol to unbox-and-use stuff, it’s probably not a good comparison.
So does this mean we’ll never have standard cartridges or a dustcover? No, but it means we won’t have them to start. Nor will we start up with 18-tube phono preamps or 1/10000 Hz adjustable motor speed control with 24-bit sine wave generator. We’ll see how we keep up with demand before we expand. Let’s walk before we run.
Because, you know, we’ve been at this a lonnngggg time. And with that, it might be best to flash back and look at one of the earliest Schiit turntable documents, written by Conrad Hoffman, and see how the design has changed…
(And yeah, guys, I know, I haven’t gotten to selling features or price. Let’s save that for a future chapter).
And with that, let’s flash back to 2013. Yes, 2013.
The Conrad Manifesto
Sounds like the title of a B-grade Netflix film, doesn’t it? Well, relax, it isn’t. It’s just my tongue-in-cheek moniker for a 4300-word, single-spaced treatise that Conrad Hoffman gave Mike, around the time the first Schiit turntable design was complete.
This is the actual title:
Design Philosophy & Technical Reference
Last edit: 12/29/13
To give you some context, Conrad Hoffman is the turntable designer Mike worked with to develop the Schiit turntable, after Mike had shopped his grab-bag of crazy ideas (er, I mean "gotta haves") to a bunch of other guys who said he was nuts to try to do that on a an affordable turntable, or nuts to do it, period.
To provide some more context, when Mike first mentioned doing a turntable to me, I said something like, “Cool, sounds good, go ahead,” because (a) I’m not really much of a turntable person (see link to this chapter) and (b) Mike will pretty much do whatever he wants to do, no matter what I say. Remember House, M.D.? Yeah, like that. And yeah, Mike's usually right, too.
Anyway, several months after Mike said he was gonna do a turntable, he had one spinning on his kitchen table when I came for a visit. It wasn’t a real looker, and it had only a passing resemblance to what we’ll soon be introducing, but it was a real, functional, playing turntable, which I had to admit was pretty cool. Mike and Conrad went back and forth with the design for a while, and it sounded better and better, and then, around the end of 2013, Mike sent me Conrad’s document.
“What do you think?” he asked.
“I think we should use it unedited as the product description of the turntable.”
“Har har,” Mike said. “No, seriously.”
“I am being serious. It’s wayyyy more honest and forthright than anything else out there. It’s also very technical. Who would argue with it?” Plus, I didn’t really feel like writing copy…I get like that sometimes, regardless of product.
“Well, I hope we have copy when we launch it,” Mike told me.
As in, sell a turntable.
At that moment, it kinda sunk in. And at that moment, I figured I might want to have a look at the design. And in the intervening years, in fits and starts, we took Sol from Conrad’s original design draft to something quite a bit different—and quite a bit better, I think.
Now, I’m not going to put Conrad’s complete document up here, but I think it might be fun to look at some key sections, and discuss how things have changed.
We’ll assume familiarity with the various styles of both vintage and modern turntables. This turntable was designed as a low cost introductory unit, but avoiding “engineering 101” blunders like motor noise coupling into the arm or platter. Components are turned whenever possible, that being less expensive than mill work, mill work generally having six sides to contend with, rather than 3 for lathe parts. The assumption is that production volumes will be low compared to the heyday of vinyl, preventing the use of cast or molded parts. Parts are expected to be produced in a modern CNC shop and have been designed and specified accordingly.
Actually, the assumption that cast or molded parts were a no-go is the central difference between the Sol That Was and the Sol That Will Be. When Mike figured out we could do aluminum castings, literally everything changed.
Popular platter materials include cast aluminum, acrylic and even steel. It’s desired to have as heavy a platter as possible and one that doesn’t ring. Traditionally aluminum was used, but the current cost of the raw blank, plus machining, makes this an expensive solution. Without damping those platters have a tendency to ring like a bell.
Acrylic has become popular, though it’s a rather unpleasant material to machine. We’ve chosen Delrin® (also available as a generic POM), a highly stable engineering plastic with a density of 1.41 g/cm3. That’s a bit better than acrylic at 1.18 g/cm3, though far shy of aluminum at 2.7 g/cm3. Though Delrin is not inexpensive, the ease of machining will somewhat make up for the cost of the blank.
Delrin is pretty insanely expensive, which made it problematic as we got to higher planned production numbers.
Why not do machined aluminum? Holy hell that’s super expensive. Conrad was wayyyyy downplaying the cost of a machined aluminum platter.
We’re now doing die cast aluminum, damped with Poron foam. This is waaaaaaaayyyyyyy better than Delrin in terms of weight, and ringing is negligible with the foam.
The platter bearing is a critical component as it can inject noise and rumble into the system. The most common traditional (and modern) bearings tend to use a shaft the same diameter as the record spindle (0.281”), mounted to the platter or sub-platter and running in an Oilite bushing. Though obviously successful, this design has several shortcomings. At a speed of 33 1/3 RPM the low surface speed makes it difficult to develop a complete oil film that prevents metal-to-metal contact. Add to that the failure of the user to periodically lubricate the bearing and you get noise and wear.
The heavy duty broadcast tables did better, using a 0.5” diameter shaft, but still needed periodic lubrication.
Some modern tables use an inverted bearing, where the shaft is stationary and the bushing rotates with the platter. This has the advantage of not cantilevering the shaft if the platter is driven at the perimeter. The side load is within the bushing or bushings. There are two downsides. First, long term lubrication is difficult because the lubricant runs out of the bearing unless special measures are taken. Second, wobble and run-out of the platter can be harder to control because more tolerances stack up in the design.
Our design takes advantage of recent developments in plastic bearings, making possible an inverted bearing with no lubrication issues. The heavy duty 0.5” diameter shaft has been maintained for rigidity, running against Igus bushings and combined with a ball and Igus thrust plate that requires no lubrication and should be free from wear for many years. These parts are compatible with lubrication, which does reduce the already small noise contribution slightly, so if a user decides to oil the bearing no harm will come of it and there may possibly be another 1 or 2 dB of signal to noise ratio obtained.
It should be noted that the bearing performance is almost entirely dependent on the surface finish of the shaft. This component requires the utmost care in manufacturing and subsequent handling. The bearing is the closest tolerance assembly in most turntables and should be considered a critical QC item in production.
This part of the turntable is largely unchanged. The insanely large-diameter, long (almost 3”) inverted bearing is a key part of Sol, and a big differentiator from other turntable designs.
The big difference is that now the inverted bearing shaft is inserted into an aluminum plinth, rather than wood or MDF.
The plinth is a T-shaped design, currently made out of wood. Unit #1 was Butternut, stained with Minwax Red Mahogany and finished with Formby’s Satin Tung Oil. Careful thought should be given to the appearance of production units. A contrasting plinth of white finished maple, combined with black feet and black anodized tone arm components could be striking. In any case, though easy to machine, Butternut is expensive and a bit low in the mass department. Heavier woods or even a cast plastic loaded with mineral filler would be a good choice.
Various feet were tested, including Sorbothane. Compliant feet caused resonances of the plinth and didn’t isolate footfalls particularly well. The best foot solution is close to rigid. Unit #1 presently has small stiff rubber feet as are typically used on electronic chassis. They have a nub that presses into a hole. The feet should prevent sliding on a smooth surface because of the separate motor design used. Note that rubber can stain painted surfaces. Stick-on feet (3M “bumps” and squares) are, IMO, cheap and tend to fall off, but stick-on felt pads might do well. Shallow cone feet would not be a bad choice for this design, if one has a liking for such things.
We talked about making the plinth originally in painted MDF, but abandoned that when we realized aluminum was feasible.
The plinth is now a Y-shaped design in die cast aluminum—a pretty darn big die, producing a part about 10 x 16 x 3.5” in overall size. It’s solid aluminum, too, not hollow.
We’ve added a small sheetmetal pod to the plinth for arm wire input, power I/O, power switch (on top, not on back), and RCA and ground outputs.
The tone arm is a unipivot, and that will be the most controversial aspect of the design. First contact with a unipivot design is disturbing at best. A unipivot optimized for playing records will have tremendous wobble while cueing. The impression is of almost no stability at all, and a tendency to rock side to side like a pendulum, which is exactly what the arm is. The instability can be reduced to some extent by lowering the center of gravity, but that introduces sensitivity to off-center records and causes the tracking force to change excessively while negotiating record warps.
That terrifying introduction out of the way, let’s talk about this specific unipivot arm. The longer the arm, the lower the tracking error and the less anti-skate correction need be applied. The arm has an effective length of 278.8 mm (10.97”), longer than most arms and even capable of playing a 16” disk in a pinch. A true transcription arm would be yet longer but the overall dimensions of the turntable get excessive, and almost nobody will actually play a 16” disk.
The arm tube is carbon fiber, using the very accurate tubes also used in hunting arrows. The carbon fiber tube is extremely rigid, given its light weight. It’s also electrically conductive, not enough to provide shielding, but enough to drain away static charges.
The target alignment for the arm is Lofgren A, also known as Baerwald. That gives the lowest overall tracking error and is strongly recommended over all others. The headshell is aluminum, with slots long enough to use alternate alignments like Stevenson and also mount most other cartridges. The headshell is a bit oddly shaped for several reasons. It’s critical to have perfect lateral balance with a unipivot arm, lest the arm list to the side. The rear of the arm is inherently balanced, but the headshell has to supply the offset angle and carry the cartridge with the stylus held on the axis of the arm. The headshell outline was adjusted to achieve lateral balance once all the geometric issues were resolved.
There is tremendous adjustability hidden in this design. The support rod can be moved up and down in the plinth (held by friction at the moment) to accommodate both changes in bearing cup position, and to provide VTA adjustment once the bearing location has been decided on.
The counterweight is a brass cylinder, 1.5” OD x 0.545” thick, threaded (1/4-28) onto a steel stub at the rear of the arm. The counterweight has a mass of 131.2 grams. The ideal counterweight will be positioned as close to the arm pivot as possible. That means it should tend towards a large diameter of minimum thickness, as opposed to a long skinny counterweight, or a counterweight out on the tail end of the stub. The counterweight is easily adjusted on the relatively fine thread, and locked with a knurled Delrin nut.
The arm mechanism hasn’t changed very much from this preliminary description.
I left out some discussion of a collar to limit the wobble of the unipivot arm. In the end, we decided that an unconstrained unipivot was best—after all, there is a standard cuing mechanism on Sol.
The biggest change is beneath the arm, where we’ve added real-time-adjustable VTA. You can adjust VTA with a small knob while the record is playing, a feature (to use Conrad’s rather muted language) that is extremely unusual on a turntable at Sol’s price.
Most damped cueing mechanisms rely on close fitting parts, coated with a heavy silicone oil or motion control grease. This one is no different, using a ¼” pin in a close fitting hole, coated with Nye Corp. Nyogel 779. An adjustable spring on the lower end of the pin controls the decent rate. The size of the pin/hole probably needs some fine tuning, and features should be included to center the spring- hopefully that will make it to the prints, if not unit #1.
What makes this unit different from most is a very accurate anti-rotation feature, and a huge arm platform with inner and outer stops. The overall precision and unusual distance from the arm pivot means you should get a very clean set-down and pick-up when cueing individual tracks.
Most cueing mechanisms use a control rod, but the free air nature of this table makes a simple knob a better solution.
Not much has changed here. We went from an expensive machined lift arm to an inexpensive stamped arm, that’s about it.
But to give you an idea of how much impact an expensive part can have on a complete assembly, that single lift arm part was almost 5% of the total cost of the turntable. Now you know how you can get blindingly expensive products, blindingly fast.
Motor & Control
A top priority for this table was keeping motor noises and cogging out of the audio signal. This is a flaw of many modern turntables and the world doesn’t need another. The current solution is a small DC brush motor commonly used for capstan drives back when tape decks were common. The target rotation speed is about 1566 RPM. Drive is via a heavy cotton thread to the perimeter of the platter, though a urethane rubber or other belt can be used. Silicone cord can also be used. When making drive cords, use the “Micro Seiki knot”. Hold the ends of the cord parallel, as if they were one, and tie them in a three loop knot. That is, just make one extra pass through the hole compared to a regular knot. The knot will look physically large, but the cords will exit the knot on one end, and smoothly go around the pulley and platter.
The DC motor needs feedback to stabilize the speed. This is presently being designed, but will consist of a pattern around the motor pulley, and a photo pickup to read it. A simple F/V converter or possibly a PLL will then control the motor voltage, typically 6.8 VDC on a 24 VDC motor.
The motor housing sits on two Sorbothane hemispheres- this is an excellent place to use them so vibrations are isolated from the table top. The hemispheres are located on the inboard side of the base. The weight of the motor and housing is then applied against the drive cord, making the system self tensioning. Though the wall transformer cord is light weight, one still has to take care not to have it pulling on the motor unit.
Here’s where you’ll find the biggest changes. The completely separate motor pod stays, but pretty much everything else changes.
We’re no longer using a DC servo motor. Why? Because the silent one we had was obsolete—and the company that made it said they couldn’t make them silent like that anymore (!) So, we’re using a Hurst synchronous motor instead.
The motor tilting, self-tensioning system? Nope. That didn’t work. Not for real. So you get a motor pod with three feet—still completely separate from the turntable. Much more reliable performance in real-world use.
The tie-it-yourself drive belt? No, we’ll be using a welded rubber belt like most sane people. If you would like to commune with monks who have passed down the Secret Of The Perfect Knot for 37 generations, then have fun with that.
Now, Conrad wrote a lot more—specifics pertaining to the first prototype, important things to call out in machining, preliminary specifications, etc, but you get the gist: Conrad is a master turntable guy, and, together with Mike’s oversight, he’s produced something unique and (we believe) very interesting to people who want very high turntable performance—without the very very very very insanely holy-crap-you-gotta-be-kidding-I-could-buy-a-car-for-that price.
However, it’s still a long road to production. So I’m going to get back to it, and work with Tony and Mike to see if I can keep Sol on track.
Until the next chapter, when more will be revealed…
Mike Moffat adds:
I have repeatedly told Jason that building a new class of product is like being screwing a pit bull; specifically, you are not through until the pit bull is. I see that Jason still entertains notions that he can keep products like this on track. (Maybe kinda if you are willing to age prematurely.) Well, if I can be compared to House, M.D., a person whose listener is broken to marginal at best, it just proves that our mutual business admiration and partnership is yet based on deafness and stubbornness. May it continue forever!
Pictures of the first Sol? Pretty please.
Also, as someone working in manufacturing engineering at a car plant, it would be cool to see what kind of special assembly solutions you are using with Sol, or in general. If you are willing to share, of course.
I'm hoping you inquired about why they cant make a quiet motor anymore. I fear the answer is "yeah we aren't as good at this as we used to be. A bunch of old guys retired, and no one left knows how to do it"
Due to still being a bachelors student wary of careless roommates I'll keep my Technics SL-10 for now, but I'm definitely excited for Sol when I have the room/privacy!
It sounds interesting, awesome, and a bit scary. I look forward to watching the setup videos.
I'd love to read the full manifesto Conrad wrote.
This Schiit turntable project has got me intrigued. I like the idea of a cast aluminum plinth in a 'Y' configuration. This seems to indicate a tripod footing, which should be good for wobble-free stability. Also appealing is the use of a conservatively oversized/rated shaft and bearing. Real-time VTA adjustment sounds like a pretty novel (and useful) feature too. I'm actually a bit surprised to hear they could still find someone in the U.S. who can/will do quality metal casting at an affordable price.
I hope they indeed get a few of these rascals operational soon. I'd be keen to take a little drive down to the Schiitr for a look and a listen this Spring.
I was hoping that you would take a more radical approach regarding the drive system... Not to be derisive, but another belt drive ? There are hundreds out there at all price levels...
Take your pick... DD, rim drive, idler drive....
DD are pretty common again these days and would normally involve buying in the drive unit, as far as I am aware rim drives and idler drives are basically different names for the same thing and generally went out of favour decades ago although there is continuing interest in the vintage professional level Garrards and Thorens tables.
I don't know what other types the ellipsis might imply.
Having read Mike's opinions on sound quality I imagine that is his reason for selecting belt drive for his deck.
Thanks Jason for an exciting chapter. I can hardly wait, I have a brand new cartridge just waiting for a new home.
Medical experimenting is in my blood..
So today my mini-vibrator came in to battle a mild tinnitus.
Three pressure points behind the ears for 30 secs each and Bob's your uncle.