pmillett

Neither am I. (Sorry, this looked like so much fun I thought I'd throw some gasoline on the fire. I'll go back to the shadows now.) Pete

http://en.wikipedia.org/wiki/Coilgun http://en.wikipedia.org/wiki/Railgun

Actually, I'm gussing you're not far off the mark. Those are waaaay cool caps. I bet they cost thousands of dollars each. My guess would be some type of particle or laser weapon research project - or maybe even an electromagnetic gun. Both require massive peak currents. I worked at a particle accelerator for a while (another story) and to fire the linear accelerator we had banks of big caps... Pete

Not sure that this is totally applicable but it's just too cool... I like to use clamps like the "TH" series from CDE: http://www.cde.com/catalogs/capacitorhardware.pdf They're a little hard to find - Mouser only carries the TH25. Newark and Allied carry a few more. The other way I do it is to use a tie wrap with a mounting hole, like these: http://www.panduit.com/products/PartDrawings/078067.pdf Pete

Peter - Interesting thread. You should have emailed me sooner The amp does run very hot (temperature-wise)... not much to be done about that, I'm afraid. I'm surprised you say it burns through tubes quickly (or at least more quickly than any other tube equipment). Never had that issue, nor heard it elsewhere. What tube(s) are getting killed? And how do you tell when they're dead? If your AC voltage is high this may contribute to both heat and shorter tube life. Measure it; if it is much over 120V you could maybe get a transformer of some type to drop it. Clean or dirty (noisy) power probably won't have any effect. It's also possible that there is something wrong with the amp, maybe the electrolytic caps are getting leaky (they do get hot inside the chassis), which compounds things and makes things even hotter. Depending on how many hours of it's life the amp has been turned on, the HA-2 has been out there for a while so it's possible some of the components could be getting stressed. As far as tubes, the good old 5U4 is quite rugged, probably the best choice. An old US manufacture one is definitely better than Russian or Chinese as far as reliability goes; I doubt that it will affect the sound. For the output tube, the 6528A runs very hot and heats up the whole amp due to it's 5A filament. And yes, the transformer is rated for it (6A, if I recall, so 5A for a 6528A and 600mA for a 6SN7), but it does get VERY hot with that load. A 5998 tube is almost as good to drive low impedance phones, it's easier to find than a 7236, and I think it sounds better anyway. It is true that the HA2 was designed to optimally drive phones over 100 ohms. To tell the truth the lower-Z phones was a bit of an afterthought. But I've always thought that most low-Z phones (Grados, at least) sounded better with some damping (higher output impedance). So to me, an HA-2 with a 5998, 7236, or 6528A tube sounds OK. But it is a matter of taste. Along those lines: the reason you can substitutes any of these seemingly dissimilar tubes in the same circuit is just due to the topology. The output stage is a direct-coupled cathode follower, so the grid voltage is fixed by the 6SN7 stage, and the bias current is set by that and the cathode resistor (which burns ~ 20 watts of heat per channel!). So swapping tubes in and out does change the stage current a little, it turns out that all of these tubes are within a normal operating range under those conditions. I did do testing with 6AS7, 6080, 5998, 5998A, 7236, 6336A, and 6528A. The amp shipped initially with 6AS7, then 6080. The "Premium" version shipped with 7236 then 6528A tubes. I dropped the 7236 tube after I had one unit suffer heater-to-cathode breakdown - the resulting pop in the headphones was very painful (I had them on my head at the time!) You can estimate Zout by looking up the tube's voltage gain (mu) and plate resistance (Rp). Zout is approximately Rp/mu. So you can lower Zout by lowering Rp or raising mu. That's why the 6528A has the lowest Zout of the group; it has higher mu (9) than the 6AS7 (2), and an Rp of 245 ohms vs. 280 ohms for the 6AS7. The 5998 gets you a mu of 5.4 and Rp of 330 ohms, which is not quite as good as the 6528A, but better than the 6AS7/6080. I'll post this to the forum so others can read. Another forum... I can't keep track of them all any more! So feel free to email me if there's something I should read. Pete