The Headcase Stax thread

[jgazal]

I would like to understand this also.

 

There is no physical discontinuity between both ends of dynamic voice coil. There is some resistance to direct current. If a certain level of direct current flows through the voice coil, heat is produced and the voice coil can be damaged.

 

There is a physical discontinuity between both stators. Direct current cannot flow from one stator to another in a functional condition. Direct current through the stators means failure (dielectric breakdown of air is achieved and an arc is produced, damaging the diaphragm). This one I learned from Birgir.

 

There is some alternating current flowing in the stators. You can see such alternating current as flowing from one stator to another, but in the imaginary plane. This one I learned from Dr. Gilmore.

 

I like this explanation about capacitors (java required): http://micro.magnet.fsu.edu/electromag/java/capacitor/index.html.

 

The engineers will have to jump in if I said something wrong, but I think there is ground at the transformer primaries, but a bipolar power supply may push pull all the circuit.. edit: said something really wrong before...

Edited April 10, 2013 by jgazal

[spritzer]

One thing to remember is that we measure amps phase to phase, so stator to stator.  That is how you measure voltage swing so something like the GES has a B+ of 600V so it can swing 600V+600V aka 1200Vp-p.  The gap between the two stators with the diaphragm in the middle is what the amp "sees" so when one stators swings + the other must swing - (phase that is since we are talking about AC) to push-pull the diaphragm.  This is a system so you have to think of it as such, one stator pushing or pulling the diaphragm means single ended which means a lot of distortion... 

 

I can't get the Java stuff to load but ground is just what you make it.  Take the standard BH PSU and many other bipolar PSU's that do not use a center tap for example.  It's two identical + supplies essentially that give say +350V each.  Now run one PSU normally and the potential between ground and + is +350V.  Now take the other psu and attach the + of it to the other PSU's ground... what happens?  The potential is still 350V between the poles but since the + is now ground the potential has to be - so -350V. 

 

It's similar to when you stack PSU's so the power supply I use in my amps (B10, GES, WES etc.) is just two +300V supplies one referenced to ground, the other to the +300V output from the other supply.  The potential is still +300V but since "ground" is +300V as well it yields +600V.  This saves on caps since they only need to be rated for 350V in both supplies... 

 

Shouldn't it be the capacitance as seen by the amplifier when driving the load? Each amplifier is driving the 1 electrode independently of the other, or so it seems, no?

 

BTW, this is mundane, but while I see the "hot" wire of each amplifier connected to each stator, where does the "ground" wire connect to and is it common to the 2 differential amplifiers? 

 

There was a related discussion not so long ago on HF about the influence or not of distortion in the amplifier when driving the electrode in push-pull fashion using differential amps (after birgir's comment that the bhse had a tiny amount of distortion factored in by design). I could not understand why it would cancel out (so much for the efforts from JCX to explain it), but it's probably cause I have no clue how the load looks like from the amplifier side.

 

I must admit that the second harmonic stuff was just to annoy the shills on HF and to see if they would try to use it against the BHSE.  They were naturally stupid enough to do so...     What.... I can't be naughty at times...?    Kevin does add a tiny amount of flavor into his designs but it's tiny and it's more of a tiny tradeoff to get big rewards elsewhere.  I for one will take a controlled compromise over an uncontrolled clusterfuck any day.  Krell, Pass, Ayre and others seem to agree with this stance as well... 

 

As for the cancelling of distortion, in an ideal world and with fully complementary amps then this would happen but we have neither.  No electrostatic amp is truly complementary (well there is one but it has never been built... yet) so most are "balanced" amps with single ended output stages.  At these voltages it's the only way to do it.  You can do fully balanced with tubes but then you have the wide variance of the tubes to account with which means the two halves will not be able to be identical and thus not cancel each other out 

 

There is no ground in electrostatics since they are essentially not referenced against ground in terms of AC.  Just like an XLR cable or the power coming out of your wall the two poles are just referenced against each other and you only see the potential difference. 

[arnaud]

Birgir, Thanks for the clarification, it was much needed as I did not realize it was two poles connected to each stator without grounding. With the 3 traces at the connectors of the electrode (+,-,DC), it is understandable though (palm). I think I need to read the explenation a few times to fully get it, but that's the same as usual .

[arnaud]

In that case, how different is this from bridged amplifiers we see for driving PA systems? Looks the same (but again, what do I know eh )

[spritzer]

Well... there is no inductive load to drive here.    Bridged amps are just a cheap way of getting more power by doubling up the amp modules.  It's similar in practice to electrostatic amps with one pole driving each side of the transducer.  In fact some of the very high power bridged amps with high rail voltages can be used to drive Stax... 

 

All of this does remind me that I really should do some electrostatic articles for Tyll but I'd rather just spend my time building something. 

[arnaud]

I may stop reading and start making, but probably not until the kids grow up... Till then, I will leave in my ideal world of computer models where everything is linear and simple . Learning about this stuff from people like is absolutely enlightening, you shouldn't refrain from writing articles, seriously .

[Tachikoma]

Has anyone else encountered signifcant amounts of corrosion on electrostatic stators? I've been wondering whether it would be worth it to disassemble it and scrape that shit off. The drivers with corroded stators always take a while to charge up; having said that its usually ok after I give it enough time to do that.

Edited April 17, 2013 by Tachikoma

[chinsettawong]

The problem isn't the corroded stators.  The coating on the diaphragm often deteriorates over time.  If you want to fix it, you'll need to recoat your diaphragms.

 

By the way, how can you tell that the stators are corroded without opening them up?

 

Wachara C.

Edited April 17, 2013 by chinsettawong

[spritzer]

I've seen spots on the utterly fubar Stax sets but that's about it.  No large levels of corrosion and it wouldn't have any effect on the sound of the drivers.  The drivers could be completely covered such as ML does. 

[Tachikoma]

I've opened up a few Sometimes I can see it from the outside as well.

 

I remember experimenting with one particular pair of drivers that had an imbalance, with the weaker side taking time to build up some charge, I don't think swapping the diaphragms between the drivers made any difference.

Edited April 18, 2013 by Tachikoma

[TMoney]

The US Dollar is now worth 100 yen.

 

If anybody wanted to ever order some Stax that time is now. It's scary how cheap the 009s are in US dollars. Pricejapan has them down under $3,400!

 

http://www.nytimes.com/2013/05/10/business/dollar-breaches-100-yen.html

Edited May 9, 2013 by TMoney

[Grahame]

He warned last year that the country had "crossed the proverbial rubicon" as it moved to devalue its currency by increasing the money supply, a strategy which he says, could see a US dollar - which currently trades at around 98 yen - worth more than 120 yen over the next two years.

 

 

http://www.telegraph.co.uk/finance/economics/10045933/Japan-is-adding-a-Ponzi-scheme-to-a-Ponzi-scheme.html

[guzziguy]

The last time I was in Japan, the exchange rate was 360Y to the $.  God am I old.

[Tachikoma]

So... I've found that the ECA-80 transformer box has overload protection circuitry built-in, and it kicks in far too easily when I'm trying to drive my O2. I want to direct wire the audio transformers to bypass that, but I've found that it has 4 input wires on each transformer; since each transformer drives one channel, two of these should correspond to the positive and negative audio inputs, but what would the other two inputs do?

 

I'm under no illusion that these transformers are even remotely suited to the task, they aren't any larger than those in the SRD-7.

 

Picture: http://i.imgur.com/XZ5nkyT.jpg

[n_maher]

So let me get this straight, you want to mod something that you knowingly admit is ill suited for the task at hand and on top of that bypass it's built in overload protection? Are you gunning for the audio equivalent of a Darwin award?

[Nebby]

I agree, it seems rather ill-advised to bypass overload protection without knowing definitively that it's not tripping for a reason.

[Tachikoma]

Ill-suited in the sense that I'm not going to get amazing results out of it. Its not different from the SRD-7 direct wire mod as far as I can tell.

 

The multimeter indicates that the amp isn't putting out ridiculous 100V spikes or anything like that. The overload kicks in around ~14Vpp

Edited May 24, 2013 by Tachikoma

[spritzer]

We have no clear idea why the transformer has extra taps or even what Sony were doing with this design.  The Sony transformers have very high ratio so 14Vp-p might be enough to drive them to arcing. 

[Tachikoma]

Ok, 5Vpp on the input tuns into ~250Vpp, so its a 1:50 transformer :0

[spritzer]

That's twice what Stax were using and that was really for the 6W tube amps of the 60's.  What is the bias voltage in the Sony box? 

[Tachikoma]

Can't be sure since my multimeter probably doesn't have a high enough impedance. I'm reading 122V on the Fluke 75.

 

The date on the box says it was manufactured in 78.

 

Guess I'll look elsewhere for my "semi-decent" transformer solution.

[n_maher]

Once again I'm struck dumb by someone who found a way to buy a $2k pair of headphones and on the flip side of that coin has no plan for how to drive them other than with something that may very well be damaging the headphones. 

[Tachikoma]

Well, got it for less than $1k several years ago Pretty sure the O2s can handle something like 1K Vpp anyway, and with the ~200V bias its unlikely to blow up that easily.

 

All transformer solutions have the potential to destroy headphones, by that logic.

Edited May 24, 2013 by Tachikoma

[spritzer]

All Stax Pro bias sets will handle 1200Vp-p in an ideal world but humidity is a huge factor. 

 

Can't be sure since my multimeter probably doesn't have a high enough impedance. I'm reading 122V on the Fluke 75.

 

The date on the box says it was manufactured in 78.

 

Guess I'll look elsewhere for my "semi-decent" transformer solution.

 

I read somewhere back in the day that the bias was 200V so you could be close to it.  Low bias with high voltage swings on the stators is certainly one way of doing it... 

[Tachikoma]

I'll get that bias board you sent me shoved in at some point, when the electronics workshop people find the time for it. Not doing it myself because I'm not confident that I can do mains wiring safely.