spritzer

Using the regulated output is a good idea and what I did in the Liquid Lightning. The zeners will just add noise though.

Let's just start off with one thing, nobody on HF that they value can read a schematic and make any meaningful judgements from that. There is a whole lot more to it than to just know what the symbol for a capacitor or a resistor is. Also assuming the lemming mentality that is the norm applies to the Stax mafia is just nonsense. We post the data for people to make up their own minds. That's the sole reason for this thread, educate people, and Alex is certainly amongst those as he clearly hasn't got the slightest clue what he's doing. Same reason why we did the Single Power stuff, the KingSound and why the EC Electra is next. People need to make informed decisions based on facts, not the shill drivel on HF or faked measurements. As Amos so routinely posts, most amps are based on the same designs (not true anybody who knows anything about electronics knows that) so why don't people post the designs? Why won't they post internal pictures? Because they have something to hide. If anybody doubts these images then I'll gladly supply the RAW's from Lightroom. As for the safety of the amp, of course it isn't safe. There are no output resistors and the ballast resistors will fail when asked to do their job. Think of it this way, you have a car with weak breaks which function ok under normal circumstances but what happens in an emergency? They won't do their job and it is the same case here. Under normal circumstances the amp works, sounds horrible but works but say somebody is using the headphones with wet hair. Not recommended but a drop of moisture gets into cable entry and shorts one of the conductors to the housing. The user then grabs the metal as something sounds off and boom, instant conductor to ground with nothing to limit the voltage/current. That is not a good feeling... What has been the standout of all of this for me is the constant cry from the shills and other people of ill refute about the wast difference between the LL1 and LL2 yet nobody has chimed in as to what said difference is. No internal pics naturally save for the one that shows the fucked DRV134/BUF634 phase splitter and the magical input wiring which is mostly just the same old cheap 1oz copper trace on the PCB. As for other makers entering the market, would anybody doubt the design and build quality of a Stax amp sold by Schiit or Pete Millet? The difference is that they know what they are doing, Cavalli has no clue.

Don't use those zeners on the bias. Just adds noise and it is already adjustable.

Fuck that... just wake up and drink some more!!!

It was just pointed out to me that I forgot to post this pic. The volume knob or rather the back of it: That's over spray from when it was painted. No primer or anything...

Here is an update on the next step of the mods. Now removing the PSU is an epic pain given all the wires soldered into it. Seriously, if people don't like terminal blocks then what about push on connectors? For fucks sake... All of the cleared bit is the bias supply. All those zeners add a huge quantity of noise and as a whole it's just unstable. Trying to fix some of this clusterfuck by adding connectors where I can. Here is the new bias supply. Kept things simple and used the same supply we've used on the KGSS for the last 15 years. It's an AC referenced voltage doubler but the input voltage is fed through a voltage divider. The input is taken directly off the V+ output so far lower noise than than the stock unit. I could have gone fully regulated but it would have been a lot more work so I just couldn't be bothered. I also found some 10uf/600V caps that fit at the output of the PSU which should help matters. If the protection circuit starts acting up then I'll just remove it as it is useless anyway.

Yup it was me and I'll post some pics. Speaking of which, this just arrived yesterday... This is a super rare, early SR-1. Housing material is completely different from all the later versions and models, it is smaller and the earpads are supposed to be made out of foam like you can see in the early Stax ads. Haven't tried them yet but I need to open them up first. All that rubber is perished as can be seen on this unique plug: Never seen one with screws and the full SATO Parts number on the front. Next to a SR-3 that I know was bought in 1971. Still looks like new though... Not the best picture but the SR-1 is a bit smaller, a whole lot lighter and the housing material flexes a bit. Good news is that the SR-1 housings are still sealed shut so they have never been tampered with. If I had to venture a guess I'd date them from 1960-1962. Condition is awesome for a 50 year old headphone... Also got one of these: The first full size Stax transformer box which debuted in 1965. Completely mint...

That's a lot of energy to burn off in the pass transistors so I wouldn't do it.

430ma??!!??

The 600V caps are going in the LL so we'll know soon enough.

The prices should go down as the yields go up.

Keep at it but bear in mind that the high voltage really wants to escape to that lovely ground plane. When in doubt, lift the component up.

Max cost for this amp in terms of components with the volume control is 500$. Everything is the cheapest they can get away with. We use 1/2W resistors for reliability but even the 1W resistors used here are smaller than 1/4W units. The input wiring is crap but I have no doubt the JPS stuff is worse. At least this is Belden... Next up is the power supply but I'll also remove the input fets and replace them with 2SJ109's. Remove those stupid diodes on the +/-400V inputs and replace them with caps for a small local reservoir. Finish up the CCS so it's fully KGSSHV styled. What does bother me is the lack of caps after the regulator circuit in the PSU. 0.44uf is nothing for something that runs this hot.

Yeah, this is very nicely done and please don't compare it to the Studio Six. Anything that uses 54$ output transformers (for the price of 2 units, not dozens) and is sold at that price can't be taken seriously. I also have my doubts about all this TKD 2CP-2500 use as the 601 is being overlooked. Just as good and with less issues. Plus it is tiny and PCB mountable.

Nahh... the MOAR AWESOME version.

Now what do we have here: http://www.diyaudio.com/store/parts/lsj74-to-92-jfet.html Fucking finally...

I just bypassed the bias supply with one of my adapter cables so now the headphones are fed from a SRM-252S. Most of the issues above have now disappeared so how fucked is the built in bias supply?

One standby trafo on the PCB too.

You don't mess around with 600VDC and it's tendency to arc.

Exactly. This is how it should be done. Just look at the parts used in that thing, 3 transformers!!

Powered it up a few hours ago with zero drama. The offset and balance were all less than 10V off base so a quick adjustment and were are good. I'll probably have to go in again and change the current for the BJT's as they weren't designed to run this hot. As for the sound, well holy shit sticks this is different. Now we have deep bass, full soundstage and high frequencies. The amp isn't 100% though and it still has too much of a "character" for my liking. The bass slams a bit too much and the top end is not as linear as it should be. Also some slight phase related issues. Next up, fix this fucking bias supply. Not only does it make no sense it's also just an accident waiting to happen. In every Stax amp we have a voltage doubler off the filtered V+ which is then cut down with a divider and further filtered. So why on earth wasn't the regulated V+ used here? This cost a bunch more in parts just to do something stupid.

That would be a plan. A lot of high end studio gear is single ended with transformers on each end so this makes sense. Same reason the ECP L-2 should have an XLR output, we now the difference but 95% of HF have no clue. I'd love to take that amp apart and show how 5K$ amps should be built.

Same thing here, lasted until mid season 3 and gave up.

This is it.

We are dealing with a lot of raw DC voltage here so the only option is to use the isolated diodes. Any other route may work but it probably won't.