|
A few people have commented on "rats-nest" wiring, and gluing-in of capacitors. I have found, through experimentation, listening, and measuring hum & ripple VS overall transparency in amps, that where ground points are located VS other components in an amp are critical. As many here know, I use as little capacitances and reactances in an amp power supply as I can comfortably get away with, and you're likely to find these components mounted in odd places all-over the amp.. This isn't rocket science-- a battle has been raging for decades over whether a vacuum tube cathode should be bypassed or not, and if so, how much bypassing can be gotten away with before the capacitors literally devour all of the music, since even the world's best caps are electronic garbage-- they're non-linear filters and energy traps. So are inductors, and that is why I try to minimize their values and maximize their energy pass-through capabilities. People have commented on load VS measured distortions in vacuum tube loading. TRE has commented-- and he has reason to do so. It has been noted that lighter loading allows vacuum tubes to produce less overall distortion-- up to a point. It has been noted that a 2.5K load on a 2A3 tube, for instance, produces more measured distortions than a 5K load, etc. With resistors, this is certainly true, and is why I like large resistances for plate resistors-- to allow good voltage swing without having to heavily-drive the resistor and to obtain lighter loading on the tube. Inductors are not resistors. They have iron and many turns of wire in them. They are also (like capacitors)-- at best-- they're non-linear filters and energy-wasting traps. Upping the turns-ratio in an output transformer-- if all else is equal-- greatly lessens the quality of that device. While it will load the output tube less, and the tube will show better distortion figures, it will also trap-out more musical information because of its increased turns-ratio, and usually-- with smaller diameter primary wire-- the smaller this wire is, the lower the quality of the device if it has to handle tube plate current-- only up to a point for sure, but it IS a factor that must be considered. Does the 2.5K transformer outperform the 5K, all else being equal? It sure does, and it is a large factor, but in some cases, one might prefer the 5K. Why would one guy want higher distortion figures, but enjoy better dynamics, musical flow, and immediate musical "attack" response-- not to mention greater transparency-- with the 2.5K? Well, he probably likes marching bands, good Rock bands, Country & Western, Grand Piano, Big Orchestras, etc. He can still get voices "perfect" by careful wiring and superior parts choices, and placement. Another may prefer the 5K. Why? If he likes quiet music, devotes time to appreciating smooth renditions of voices, and simple Jazz bands, he will appreciate the more laid-back feel of all that extra wire in the output transformer primary, the transformer's increased inductance, and the blunted attacks of more lively music's transients when he does play it-- which may not be that often-- the more relaxed 5K transformer may be something he LIKES-- as long as he doesn't try to replicate the Eagles Blu-Ray-- their "farewell" in Melbourne, Australia. Someone trying to do that might load the 2A3 at 2K! Yes, the tube LOVES it.... The point here is that there are reasons for what people believe and do, and for them, they are valid. When I look at an amplifier-- I think it should be a perfect "slave" component. What I want it to do is play anything I can throw at it perfectly-- musical taste/preference has zero influence on me. I want it all-- everything better than ANYTHING.. That's where rat's nest wiring comes in in a big way. It's those grounds and circuit paths. In general, they're better if they're shorter, but not always. And when that occurs, you have a compromise that you have to figure out before you can proceed any further. The question is which conductors need to be short, where are the ideal grounding points, and what compromises must be made in this "ideal" landscape, in order to accommodate minimal hum-- many of the things you do to reduce it will impact musical performance-- and NOT favorably.. Which compromises will you tolerate and which ones will you rule-out? Longer wires COST-- in many ways. They can also aid in balancing-out hum at the expense of true musical transparency. If you can't stand hum, how far are you willing to go to reduce it? If you love musical attack, transparency, dynamics, etc., how much hum will you tolerate in order to create ideal circuit paths? If you change a grounding point by 1/4 inch, then it's YOUR design! The Rat's Nest. WHY? It's because if you parallel capacitors (such as on a cathode bypass), all of them must ground-- and connect at their "hot" ends-- at the same exact points, NOT 1/4 inch-- or more-- away from where you started connecting those caps together. WHERE are you going to put all of them? Well, now we're into deep chassis and 3-dimensional wiring and component placement, not neat-looking pictures, as those are only 2-dimensional, and cannot show how one can neatly separate components vertically as well as horizontally-- allowing adequate room to mount them where they belong sonically. And that brings us to G.E. Silicone 2. A comment was made regarding gluing-in capacitors. The first thing to understand is that this need WILL arise should one decide to do 3-dimensional wiring. All those component mounting devices-- while neatly military-looking-- are stealing a LOT of your music. I know-- I spent years in military electronics and I know what "neat" is. Unfortunately, military-style construction, while pretty, super-serviceable, and making nice on-line 2-dimensional pictures-- all things that the military-trained technician needs-- will absolutely steal-away most of your music. You're going to need to mount critical musical components where they belong magnetically, statically, and separation-wise from wiring and other components. If you hope to do this by using a lot of hardware, it will look good and sound deficient! G.E. Silicone 2 is the rescue. I've used this in over 250 chassis to date since 1989. NOTHING has ever come loose-- even with FEDEX and UPS. It is important to mention 3 things here: (1) Silicone 2 is TOXIC. It is not "vinegar" based-- it is based on a toxic, somewhat odorless solvent. Use it, and then put that chassis outside for 3 days-- let it cure. DO NOT breathe it! (2) It just doesn't "let go"-- if you wish to remove a component glued with it, you'll have to cut it with a sharp knife. (3) Silicone 2 should be RUBBED INTO the pores of the two surfaces you're going to join with it. (don't use your bare fingers-- toxic!) THEN, if you're mounting capacitors, etc., with it, build-up a "pad" of it for the cap (or whatever )to sit upon. Twist the component around a bit-- to make sure the glue connects well to both surfaces which are already glue-prepped... You need about 1/3 inch separation of the component, so don't squish-out the glue stack. This method is how you can place components exactly where they belong in 3-dimensions. It won't take pretty on-line pictures, but it will allow maximum performance and fairly easy component removal, should repair become necessary. Remember that this kind of construction can take place only in a deep chassis, but the musical rewards are a'plenty. ---Dennis---
|
