spritzer

Could be but in my experience it is the need to keep things simple that is the ultimate deciding factor.

Fake it 'till you make it works for me... Some proper simulations would be cool and it might be a good idea to contact DuPont and see if they have any data on how mylar behaves while under such tension and heat annealed. The newest sets use some other polymer but mylar would be a good start. A proper simulation could also dispel the age old myths that electrostatics are just capacitors that require virtually no power and are just voltage creatures.

I'm sure Quad would have tried to build something like that but they also never bothered to hire production engineers or look at how much some of this actually cost. The diagram tension is essentially fixed (though there is some variation) and it is then put in an oven to rearrange the polymer and lock that tension in place. The tension is very high though and there is next to no excursion even at silly levels. The diaphragm certainly can never hit the stators as you can't push it that far with a finger, let alone using the very weak electrostatic force. What the diaphragm can do is vibrate while being locked in place thus creating the very low distortion output. The bias voltage sets the film to stator distance or rather it's the other way around. If you use a voltage that is too high then the driver will arc and burn up the film. Low voltage will lessen the grip the stators have over the diaphragm and lower the sensitivity of the system. In the middle is the sweet spot which will allow enough control while allowing plenty of voltage on the stators. Now lets discuss that a bit. The electrostatic force is by definition very, very weak and it's strength diminishes by distance squared. That's why the normal bias Stax have a 230V bias voltage of a 0.3mm D/S gap but the Pro bias is 580V for 0.5mm. The gap is up by 60% but the bias is up by a factor of 2.5. Stax could have pushed it further but they clearly wanted the same sensitivity for both standards so they could be used together on the same amps. While the SR-Omega was being designed they tried going even further (1100V or 0.6-0.7mm) but decided against it for some reason. Now the extra distance could give more excursion but in reality is just gives the ability to put more voltage on the stators without damaging the diaphragm. The max voltage any electrostatic driver can see is bias*2 which is why the SRD-7Mk2 boxes have 450V zener diodes forming a clamping circuit for the normal bias output. The transformers are capable of crazy voltage swings so Stax tried to limit it somehow but they would never reach the 1200V needed to damage the Pro bias drivers. More voltage means more control and that the drivers can go louder without damage. Anyway, enough from this lowly baker in Iceland as Sachi put it...

Yeah, ineffective dustcovers and the cable entry is too cramped leading to arcing issues. Still I want to buy one and tinker with the amp a bit...

Given the distances involved it would be very hard to use only two stators. Stax pro bias diaphragm to stator gap is 0.5mm so if you want to use two diaphragms then the bias has to go down quite a bit and with it, sensitivity and the "grip" the stators have of the diaphragms. The breakdown voltage of air is around 100V/mill (not mm mind you and can be much higher depending on external factors) so there isn't a lot of room for error here. Now if we were to just layer one diaphragm on top of the other with just the thickness of the glue separating them then that will make the driver less linear (the diaphragm is never exactly in the center) and just add distortion. I also wonder how much an effect the vibration of one diaphragm will have on the other.

The isobarik was designed to compensate for the inherent non linear nature of piston drivers but will naturally bring it's own issues. No use for this with electrostatics as the issues with dipoles are more to do with baffle size and the bleed between the sides. The idea of using essentially a tweeter and woofer is intriguing but I'm worried about the backwave of the smaller driver collapsing with the back driver and cancelling out part of the range. Only real way to deal with that is a small time delay but even that isn't a perfect solution. Quad's PJW's Egg speaker springs to mind as a neat way of stacking diaphragms but we would need a time delay there as well.

Plenty of preamps use the ECC83 but many designers think that with our low gain preamps and high output sources a tube with a mu this high can only lead to issues.

Pretty much anything is possible but the normal electrostatic wiring setup limits all of this somewhat. I'm not sure many would like a 10 conductor headphone cable needed to feed different bias voltages and time delayed signals to the drivers. I do think that operating two full range drivers sort of misses the point though so one should be designed for higher frequencies while the other for lower. A simple way to do this would be via diaphragm thickness and bias setup.

Lovely tube but an odd choice for a linestage.

Stacking diaphragms is certainly possible and both Sony and Koss have done that in commercial products. It does bring up the idea of how do you drive them, three stators and two diaphragms with them working in tandem though a time delay or run the out one out of absolute phase. It would also be possible to simply ground the outer most stator (so it wouldn't be running at full power) for a semi passive radiator. The other approach would be to stack two different drivers but then you have interference issues plus how to handle the backwave off the front driver which will bounce off the second driver. One could use ports and a bass reflex design (like the 4070) but then it becomes very complex. Let's just say I've been thinking about this stuff for some time now... Wasn't it Linn who make the IsoBarik speakers?

The Stax amps use just a standard relay but the AC voltages are far lower. I'll check out the part number when I do final checks on this SRM-007tII I have here.

I'm trying to get some trade info on this line as I'm curious as well. Something like the 6922 alternatives spring to mind as the tube of choice. It's clearly a dual triode and normal noval size. I'd sure like to see the circuit as Samsung has an army of engineers so they could come up with something cool.

Stax switch the secondaries. Simple DPDT relay each half switching one wire.

All my HV transformers are 100mA+ and no problems at all aside from some buzzing. I'd be all for some R-cores but we'd probably have to go to Japan for them. The Chinese ones are rather poor...

That is an awesome show.

Thank you all. Love the cakes but sadly I had none today, too busy to make one. I'll make up for it in an hour or so by visiting the most expensive and best restaurant in Iceland and they may have some cake...

Happy birthday fellow Baker boy!!!

Thanks y'all. It is actually tomorrow but that's all good. I'm freaking out a bi though as it's the big 3-0...

Happy Birthday Greg!!!

I hope you aren't using the TRS plug with the t-amp as they really don't like a common ground. Same reason why the WEE should never be used with a bridged amp...

The 2SA1831 is 1/3 the price and should also work. It's not isolated but no heatsinks needed here. Since this is the "something different" thread I've stumbled over an even simpler and crappier way of doing the Darkstar and it's all on one chip. Stay tuned for what will probably be the worst amp I'll ever build...

I'm always a fan of using up some of these 2SA1968's...

Whether the transformer unit is self bias or mains bias should have effect at all. The power needed to energize the diaphragms is so tiny that a bit of leaching of the audio signal won't matter. That brings us back to the age old issue, amp vs. transformers. The amps have their own set of challenges but high ratio transformers built to a small price will always be a problem. The cores are too small, the winding setup is too simple and the layout inside the Stax boxes isn't doing the sound any favors. Amps all the way for me but there is also something to be said about the simplicity of transformers...

Given that none of us can agree on attenuators I do think it is best to keep them offboard. I certainly wouldn't be caught dead with an RK27 in this amp... When I have a bit of time I'll order boards from me and KG so we can test the beast.

Happy birthday buddy!!!