dsavitsk

Indeed. As close as you can. Each board has a servo that should keep the output very close to ground. In a pinch, the servo opamps can have their offset trimmed, so you can solder in a trimmer across two of the pins (I think Colin left them unused) to zero this out. It probably won't be necessary. But, the transformers really do not like any offset.

1 as well, but it surely doesn't matter here. Since it is balanced, any induced ground noise will be cancelled. And, the preamp this is feeding has input transformers, and the circuit is truly differential and has no ground reference. What matters more is keeping the offset between the + and - pins to absolutely 0. Even a mV will interfere with the amorphous cores.

Hard to believe that the anodizing could build up so much as to make the sides no longer mate, or the PCB not fit. I think the best bet is to probably sand it down -- use a dremmel on the built up parts. These anodizing and powder coating shops are largely slop shops. They work on volume, and don't care so much about quality. I ran into the same issue when building my prototypes. Most places I called said they wouldn't do runs of under 100,000 parts. I eventually found a friendly local anodizer who said he's be happy to drop my stuff in with another run for a small fee. I spent hours sanding the aluminum only to have him smack it around, and then deny he'd done it (and get really nasty when I suggested he be more gentle). Fortunately, I got it back before it went in the bath so the scratches and dings were not permamently etched in. But, in the end I have discovered that to get case parts finished the way I want that they need to go to a specialist metal finisher who understands "high end audio" and who will work with small volumes. Such places are few and far between, and their work on this case would likely cost as much as the whole kit. As for a replacement -- it might not be as easy as you'd hope. It would leave us with an extra pcb and transfomer set. However, you could order a blank one from Context (not sure of the cost, or I have a blank from a different manufacturer that will work -- IIRC it was about $40) and send it to FPE to have it machined. Probably run about $50-60 for the machining. Not sure the one-off cost at Context to do that.

Hey, I grew up at Purdue -- dad was a prof there. Home of McCord Candies -- try the olive and nut sandwich. Indianapolis airport is a lot closer, but has the downside of not being in Chicago

The HD-800 is a very well damped headphone. If you drive them with an amplifier with a very low output impedance, they will be over damped and sound bright and a bit thin. On the other hand, if you drive them with an amplifier that has some output impedance, the damping factor will be less, and they will sound fuller, warmer, and bloomier. Contrarilly, something like the HD-650 is less well damped and thus does much better with an amplifier with a very low impedance. This, however, has little to do with tubes vs. solid state. It is true that most tube amplifiers have a higher output impedanace than most solid state amplifiers. There are a variety of semi-inherent reasons for this -- for instance, tubes generally have lower transconductance. But, when comparing something like the Eddie Current amps vs. say a B22, it has more to do with amplifier topology than anything else. However, it is not the case that tubes amplify different frequencies, or sound much different than transistors, or anything like that. Indeed, the differences are really overblown by many people. Instead, it is that tubes tend to be used in certain ways, and transistors tend to be used in different ways, leading to many gross generalizations about how tubes sound like this, and transistors sound like that. So, what I am trying to say is that I think you are ascribing traits to tubes and transistors that are actually traits of topologies, and not the devices themselves.

This ^^^

Happy Birthday, John! Have a great day in the snow.

Wow! I bet it sounds awesome.

It's a thing in as much as things my brother invented* are things -- he is a psychology professor if that gives him any authority (it probably shouldn't) *Looking around, I don't think the term is unique to him.

Post Christmas Letdown (PCD). Happy Birthday!

NYE and Halloween are my least favorites for the above reason, and for the enforced fun aspect ... ... and from the year that a kid, after eating too much candy, vomited all over our front porch. Then his mother, after getting some candy from us, showed us the vomit but didn't stick around to help clean it up.

In the U.S., too http://en.wikipedia.org/wiki/Implied_warranty

Happy Birthday!

Happy Birthday, and Happy New Year!

Happy Birthday!

There's no IP protection for circuit designs unless they are patented. PCB layout and schematic art is protected by copyright, but otherwise, all the "For DIY use only" scribbling on schematics is nonsense.

A cathode follower, which is what the 12B4 is, has a gain of slightly less than 1, and an output impedance (Zout) of 1/gm. The 12B4 has a gm of around 0.006 S. This means that the Zout from that stage is ~170 Ohms (the white paper you uploaded says 500 Ohms, but this is incorrect). A transformer is an impedance and voltage conversion device. The transformer used has an impedance ratio of 600:8 or 600:4 depending on the tap used. Since the 4 Ohm tap is used for Grados, this is equivalent to a 4800:32 impedance (or 150:1). This means that a 32 Ohm load on the secondary will appear as a 4800 Ohm load on the primary, and that the 170 Ohm output from the tube will appear as a 170/150 or just over an Ohm. Additionally, because transformers are not perfect devices, the impedance of the copper adds to this, which makes the Zout somewhat higher -- by at least a few Ohms. To know how much higher, one would need to measure the transformer. However, the quoted 7 Ohm number is based on the incorrectly calculated 500 Ohm Zout of the 12B4 stage and the 8 Ohm tap. It also fails to account for the losses in the transformer. So, the fact that the actual Zout is probably pretty close to that is a bit of a coincidence. Transformers also transform voltages, and they do this as the square root of the impedance ratio. So, the 150:1 impedance ratio means a ~12.25:1 voltage ratio. Thus, a 12V signal on the grid of the 12B4 will result in less than 1V at the amplifier's output. The driver stage has a lot of gain. A tube's mu is the gain, and the 12AX7 has a mu of 100. However, inefficiencies in the circuit keep this down -- let's say to 60 or so. This means that a 1V signal on the amplifier's input will be an estimated 60V signal at the grid of the 12B4 and close to 5V on the output. Thus an overall gain of ~5. Additionally, the input will clip at 2Vp-p which is 10Vp-p on the output, or 3.5Vrms (which is what is stated in the document as the power limit). Sound like the sort of thing I might say. I would doubt Blackgate caps would be better in this regard, but I really have no idea. 20 years is probably pretty optimistic for any electrolytic caps. There was a pretty nice post a few days ago from Tangent in the DIY section at HF regarding capacitor life expectancy. http://www.head-fi.org/t/583878/capacitor-lifespan

The last page of the file you uploaded has a pretty thorough description of how the circuit works.

I've not heard the combination, so I can't even give a biased opinion. There is a 300R tap on the transformer, so it is designed to pair well. There are also technical things I do not like about the Bottlehead, which I have voiced elsewhere. However, everyone has a different opinion as to what sounds good, and there are certainly people who like what the BH offers. So, there is my bland, noncommittal response.

I can't find where I wrote down the measured values, but gain is approximately 2 into the 32 Ohm tap and 5 into the 300 Ohm tap. I'll see if I can get a more precise value later tonight.

This is how it starts, isn't it? I'll throw in a Frosty.

For the record, I am willing to trade french fries for shilling ...

This one's pretty good, too http://gawker.com/5868761/christopher-hitchens-unforgivable-mistake