i’m going to add a disclaimer at the beginning here: please read on with a grain of salt. as always, these are my opinions here, and i do it mainly for me, not you. to help me make sense of things… to organize myself. and to be honest about my shtick. if you pick up something useful here, it’s yours. if it pisses you off, then you can file me in the trash bin. never darken this door again. you got your money back in full. if you don’t agree with me, so what? you’re entitled to your own opinion.
its been my experience that this is a truly controversial subject, unlike the previous posts. and i intend to chip away at it in all my graceless fury. you don’t have to like it.
power is the beginning of everything. everything we know apparently began with an explosion of it. life came later… and how fragile life is in comparison. how easily broken or dysfunctional it can become when the power relations are skewed. a grab for power in one opportunistic moment can wreck a million lives; or create a mass extinction for other forms of life. what a strange kind of creation that is!
i have lived through all three (and a half) of the most recent big new york city blackouts… very interesting socially. power runs the city. pretty amazing what happens when you take it away! all the norms become inverted. the people on the bottom become the people on top. the powerful become powerless. good thing most new yorkers, at their bottoms, know what love is. or simply hates everyone else so democratically. imagine how it is in a place without that. how interconnected norms are to power?! how much power maintains the socioeconomic organization of the city. how vulnerable large systems are that have such a dependence upon power to function.
there are rare exceptions but power is behind almost every successful human effort, even when it’s upside down or sick. people tend to zealously insist it’s neutral or utterly evil, unless they have a lot of it. but it clearly does do good! if you can read this, you and i eat most days. we email one another. we build musical toys and talk about it across the world. it can give rise to or revive almost any work we know or perhaps can imagine but only when it’s tempered with love.
there’s no love at the end of a gun. only resentment and humiliation… if you’re still alive. i wish the US could get it’s head around that one. but i shouldn’t go there. i want to talk about power, in it’s most simple form…
i think life is like a war between power and love. freud said it’s all about power and sex… when you are dealing with humans. but, did he say anything about war? no. i like to think of myself as more evolved, but maybe he was right? maybe i’m devolved? in any case, my life is like a war between power and love… it’s often hard to tell who’s on top. power seems to have the upper hand always. it’s hard to say no to power. but life without love is no life at all. the only actual weapon love has in this battle is withholding itself, but it’s a sneaky little tool that. a subtle knife. the silent treatment is the basic strategy here. or the long quiet walk away in the opposite direction… i always appreciate the slow fade, too, as if the other was never there at all; and even, “ah ah ahhh, no nicey, no nookie”.
there are many ways to withdraw love, not all of them healthy. i have experienced most of them by now, i think? these days, i want to love and be loved and i am willing to share power to have it. it really isn’t all about me. there are other issues, too. if you give it all away… if it’s all love, then no movement. no making, no creation, no hunting or gathering. no desire. just dumb contentment. you are curled up in the fetal position with a head full of endorphin beholding the flame of love. weirdly alone. wait. is that love or morphine? religion! and, you can always get dumped in the meantime.
the most successful arrangements i have ever had have always involved a love that was just unavailable enough. that keeps you moving in between the bouts of bliss and misery. it’s more difficult to get picked off that way, too! it’s harder to hit a moving target… however, movement does put one at risk of bumping into others… and when the crash comes, the aftermath is not always merry. it’s a nice idea, to be so compassionate you cause NO harm to anything. of course you are sitting quietly in a cave, without eating. or, eating just what the novices beg for you. i think some kind of deal has to be made between the two. love and power. and it means we get a little dirty under the nails, no way to get around that. i think it’s ok to get a little dirty, and to wear it honestly. like at the end of an hard day’s work… i don’t trust soft hands, and clean suits. especially when they are presented in a way to suggest they have never been dirty. the filth has only gone indoors.
power for music reproduction is no different. it’s a touchy subject for many designers, believe me! and many others hide their “secret sauce” in the supply, too. more conflict is generated between the members of the DIY audio community around power than any other subject, except maybe the single ended/push pull thing, or the tube/solid state thing, which is actually meaningless in comparison. perhaps partly because most ordinary people normally take it for granted. not taking power for granted is one of the first big steps a “novice” takes. and largely misunderstands. but it’s a start nonetheless, and it’s a good one.
it takes power to make power, no way around that. no matter what you read on the internet, on this planet there is no roswell generator, or free energy machine. the greys pay for their juice too, they just get more for their money. and besides, there are many roads to mecca. i personally prefer the shaolin outcast style over the roswell thing. kung fu! when the mothership comes, the shaolins will do just fine. and the outcasts don’t have to wear saffron. the most brilliant ninjas really work on their power supplies, no matter how much power they need, have or are after. i always look closely at that. it’s love that tempers power.
why talk about power as if it was all the same thing? because it is all the same thing. yes, there are flavors… different wavelengths. kinetic tastes different from thermal, RF current is different from ultraviolet light, even though they are all very related. military power looks very different from a bouncer holding coke. but they can transform from one to another. they wield each other. they can change state. and most importantly for us, they can transform the immediate environment too. an amplifier isn’t some device that takes a small signal and makes it bigger… fuggeddabowdit. an amplifier is a device that converts a source of power into a load in order to accomplish some kind of work. it isn’t in anyway separate from the source of power. the supply powers the speaker, not the signal. and if that work is an “intelligent” cultural practice (lady gaga isn’t what i mean exactly, but sure…), it’s not simply an engineering problem. it’s a cultural problem. and the messy range of values possible in the human experience come into play. all of it. joy, misery, hope and pessimism, boredom and greed… sometimes all at the same time, if it’s well done. if you have a problem with that, your work will definitely sound like you have a problem with that. i think analog engineering is one of the hardest and most rewarding human endeavors. i treat it that way.
some think nuclear power is the most dangerous thing known… i think unexpressed emotions, fear – anger – despair, are the most dangerous power we face because all it takes is some psychopath with a resentment, or a shut down mama’s boy with a napoleon complex and some nuclear power to fuck it all. a nuclear bomb converts power into a load… just like an amplifier. someone took the time and care to make sure it does that in a focused and efficient manner. analog engineers, among others. i wonder about that… what kind of love guided those hands?
power supplies come in different flavors and styles, and they accomplish varying things within their capabilities. i am of the mind, now as always, that there are many different ways to do the same thing. a bare minimum of 144 perfect ways to do everything… 12 dozens seems like the right bare minimum. a shaolin baker’s number. anyone tells you there’s only one way to do things, my suggestion is to keep an eye on that shifty bastard. especially if the issue concerns power. but don’t look too hard into those weasel eyes. he’s no friend and when he pops you wanna be in a position to put some distance between you and him. also, keep your wallet in the front pocket. that, and do the world a favor and just don’t give a shit… indifference is a loving death for a creep. but do keep an eye peeled… fuckers will sneak up on you like a tick. anyway, “what are you actually trying to do?” you need to think about the answer to that question while you gather parts. i want to suggest a manner of thinking that builds on everything i’ve expressed so far, but especially the idea of tempering power with a somewhat withheld and considered love.
david berning’s PWM 300B/mosfet amps use RF electronic techiniques to convert RF AC into speaker audio… ken shindo does it as if sid smith, joe grado and dick sequerra were japanese and made single ended 300B amps, or F2a amps… cary audio does it like radio shack would have… in the 70’s, if they smoked crank and gambled on tractor pulls; etc. some are better than others (hahahahahh). they all dump roughly the same power into an 8 ohm load on paper. but, for example, berning’s and shindo’s 300B amps are weirdly WAY more alike into a speaker than different, even though the supply and amplifier couldn’t be more different. at least not when i heard them. berning’s had better bass than shindo’s but shindo beat him on top. they were close, really close.
if we are working on a linear model, no matter how we do it, the closer it is to linear, the more alike they all have to be. it isn’t at all strange that berning and shindo wind up close to each other… not at all. but, define linear…
the most commonly used unit of measure for power in electrical terms is the Watt. you do use the joule sometimes… i think in calories for fun. it’s provocative to think of power in terms of oysters or porterhouse steak. while most of you are very aware of that, it’s good to have some other kind of comparable measurement just to stay honest. to keep your horizons wide. one horsepower is equal to 755 watts, just about. think of the hay it takes to power a horse. horses are a damn sight more efficient than most amplifiers are!
the basic parts you use are transformers/inductors for distributing power from the mains and filtering (two completely different jobs), energy storage, and then distributing it to the loads (i am not going to give the SE/PP argument any attention as i think it’s a distraction); capacitors for energy storage and filtering (two completely different jobs) and possibly distributing further to the loads; rectifiers (tube, solid state) for AC/DC switching and for voltage reference; amplifiers (tube, fet, bjt) for use as switches, modulators, regulators, oscillators and references; voltage and current references (weird devices that defy ohm’s law), and last but not least, wire.
ok, what are you trying to do? my guess is you are playing back music through loudspeakers… probably. you could be driving a cutting lathe, or a shaker table for watching michael bay movies…? no. not likely. what kind of speaker? statistically, that isn’t the first question most of you think about when you are thinking about power. hopefully some will be a bit more comfortable with the idea in a few minutes. it is strange to sort out the speaker after the electronics… but it is the fashion. everyone wants a sexy 2 watt 2A3 or a rockin 300B amp. or a 1000 watt krown psychlomatic floss Z chip amp. then go looking for a speaker. good luck with that. the speaker is always the hardest part. here is what doing it the other way can get you….
you know, dave slagle (of intact audio) has had a long relationship with the lowther driver. i fucking hate lowthers. they suck donkey ass. no bass no matter what and nasty out of phase highs. i get pissed off when they are described as “full range” because they are anything but. but slagle, god bless him, never gave up. he started there and fucking built a thousand different schemes to make those pieces of shit shine. and he got some place! he even got rid of the magnet! this is deep. i try to remember this, although i will ALWAYS loathe lowthers. love can do things power never will be able to.
let’s say you like the dynaudio BM5 (i am not endorsing it. it just seems like a neutral consumer choice, it’s affordable and it can be bought anywhere). i don’t like them much, but i saw a pair last week in a recording studio. they sounded like high quality LS35a’s. no DIY i know uses them. they aren’t horns or “full range” speakers, so that would make you unique. it is apparently popular among some recording engineers for basic monitor work. it is specified as having a sensitivity of 87 dB/watt. that’s 10 watts for 97 dB at 1 meter. 100 watts for 107 dB at one meter. i’m not sure when that speaker starts to fry? my guess is somewhere around 100 watts… maybe 200. that’s only 3dB louder. i would guess the minimum power you would want for this speaker would be 25 tube watts, based on the efficiency and the small size. that will be compressed and gently distorted for classical music and old jazz, where the peaks can be enormous, unless you listen very quietly. but, that speaker can’t do the dynamic range of an orchestra either… it’s a nearfield monitor. it is meant to be listened to close up and not so loud.
actually, you would probably want 50 watts if you listen to a lot of piano and you’re demanding. but, let’s say you like tube amps, and you don’t give a crap about classical music. for pop music, which is stomped on so hard with every kind of compressor known to humanity… 25 tube watts will probably be enough in a small apartment where you sit close to them… there are some who could live with half that. 75 watts would be better quality-wise and the minimum for a solid state analog or digital amp (so horrible in clipping so you want more headroom). but you won’t notice it if the amp and the speaker are doing away with the dynamics because there are no dynamics in 99% of pop music. what that means is that small sounds are made bigger and big sounds are made smaller in the construction of pop. the range is compressed. and then boosted in level to be as big and loud as can be played on the radio legally. used to be +4dB, but now… i think +18 is the legal limit? a speaker that can reproduce tiny differences and also big peaks will be a realistic speaker. efficiency is an issue here because the driver has many losses that chip away at the little stuff. pop music reduces the requirement. some cd players put out almost 5V peak… because the supply is +/- 5 volts by the way. pop recordings will be clustered around that when they are cranked up. that’s entertainment! tube amps clip so benignly with the right circuit, you’ll hardly know what you’re missing. a comforting thought. the challenging stuff is the audio verite. for that, you’ll need more power.
so, you have made an “informed” adult decision on behalf of your speakers for 25 watts. if you plan on using a class A amplifier, the output stage will be drawing 4 times that continuously. class B will be roughly half that. that’s because the efficiency is about 25% for class A. 40-66% at full output for class B. let’s keep it simple and stick to the class A. a bit more for recovery headroom on clipping peaks… a bit more to drive it and a prudent reserve. let’s make it a sensible 110 watts so far. if you plan on using tubes, you have to heat them all up. that can be another 30 watts or so… more if it’s a transmitting tube at the output. that 25 watt rating, is an RMS rating. that means the peak drawn from the wall will be 1.414 times the RMS. the total is in the ballpark of 200 watts of peak power consumed in order to deliver 25 watts of tube audio power continuously to the speaker. i wonder how much cheescake you need to make 200 watts for a record or two? well, you can see why so many engineers sneer at vacuum tubes. it probably looks ridiculous to the young ones. especially when there are class D amps that are 98% efficient. and that’s fabulously green! you know the greys listen to class DX, radiated straight into their brainstem from the mothership. funkadelic! there are some guru’s who can do that too, apparently. ahh, if it was only so simple.
ok, back to the hypothetical design. the power transformer must be rated for a minimum of 200 watts per channel without overheating. trannies for guitar amps are usually specified for 55 degrees C at the rated power. i think that’s too hot for children and animals… not hot enough for solder heads, though. won’t keep your coffee hot. your mileage will vary. i like 35C for year round use near critters. cool is not warm.
this process can be repeated for any speaker. the level of detail can be increased, but the trend is the same. it’s a comfort to know what you need. to provide it with consideration. to know where you are going. start with the load and lovingly move – backwards.
after the power requirements are estimated, you can start looking around at your options. you find a power tube or tubes that will get you the 25 watts you want… maybe you already knew what you wanted? you figure out how to drive them. you are ready to design the supply. let’s just say you need 450 volts for the output, 400 for the driver at 110 watts for the two, 6.3 volts and 5 volts at 30 watts for the two for the heaters… and you have found a trannie that will distribute the 200 watts in various voltages from the wall to the various loads without catching on fire. you will want to rectify, filter, and store a prudent reserve for those moments of abrupt change of state. let’s take this one at a time.
for linear supplies, the power trans will have several ratings of interest beyond the power and thermal considerations. the DC resistance of the primary and secondary will have a bearing on the regulation of the trannie. more R, more voltage drop when lots of current is demanded by the loads. this can be worked with if you know you must. this is also a feature of certain kinds of supplies, such as guitar amps where the supply adds to the compressor-like quality. you don’t want a power trans with good regulation in that case… there is a thread that runs through the DIY diaspora that attempts to reduce the supply series DCR, and hence supply output impedance, to the lowest possible. Medwin… need i say more. this does improve the frequency extremes and the transient response in an amplifier that can be improved. i myself spent a good deal of energy exploring this. it is interesting, but in the end, i found that there is still 143 other ways to do the same thing. you can regulate passively and actively… and there are amp circuits that are also highly immune to whatever the hell you put into them. sort of like an old john deere.
a word about materials and format: E/I trans are by far the most economical and flexible of all construction methods for power transformers and inductors. they are also the noisiest cores because someone stacks them (the windings can be quieter than say toroids simply because there is some slop in the lams…). there’s a fair amount of magnetic attraction and opposition, i.e. mechanical vibration, in a trans and that is part of the deal… if the lams can move, then the wire doesn’t as much; and vice versa. if you remember your right hand rule you know this, but the problem is that the tricks manufacturers use to keep trans small and affordable can make this worse. it is “solved” occasionally by dipping the entire thing in epoxy. this solves nothing. but it can actually improve heat transfer because there are epoxies that are good at that. it adds dielectric, though, and that does change a power trans… not always for the better. as of writing this, the dielectric absorption of most epoxy resin is relatively high. anyway, both the lams and the wire can get loose and move around at the frequency of your local AC mains supply.
E/I construction, is still king for a reason. it always works, and can be easily adjusted. silicon steel is common for lams, but improved power trans can be found using M6 (smaller and cooler), M4 (insane compact military radios) and even nickel! (dave, you know where you are, i think?). “C” core , powdered toroid, tape toroid, “D” core (a kind of “C” core), are all more expensive or specialized cores that are generally smaller, more efficient, radiate less, or don’t heat up as much, or a combination of improvements. they are less flexible in terms of manufacture as they usually can’t be adjusted so easily by adding or subtracting core. that isn’t always a problem. these cores often provide an advantage in equipment that must be compact… but, there is no overwhelming sonic advantage to any of this! i have some russian power trans made from the worst steel in the world. they work fine… hot, but happening. whoever made them had no choice and did a good job! it all depends upon how it is connected to the rest of the supply. you can have a sintered nickel/cobalt “D” core hooked up to a bunch of hand made oil caps and it hums like coffeeshop waitress on diet pills. and the windings make as much mechanical sound as the music at low volumes! and no bass to boot! and all that just to be hooked up to a pair of lowthers! there is so much more to this than core material, but it is one piece of the puzzle. the first piece. knowledge isn’t love, it’s power.
“chokes” (the colloquial 1930’s slang that stuck), are obvious cuzzins to trannies. these are electromagnetic storage devices. inductors to the snide young ones coming out of school, who think of them as those nubbins you order out of a catalog and hope your switcher doesn’t burn. the old school knows them as inbred hormone mutated sterno spiked koolaid guzzling kin of transformers… you stub your toes on them and curse. divorces have been initiated over chokes. want your partner to go? pile them up in random places in your shared home… these are one of the reasons you don’t notice what the power trans is made out of. you are adding this in series with the current moving into the load… it has a much bigger effect on the “sound” than any but the wimpiest transformer behind it. all the previous construction caveats have less importance than the materials and the winding characteristics do. including the wire. very interesting swapping chokes. it is subtle, but improbably obvious. that is an A/B test to force upon the “objectivist” pinheads. fuck those angry emotionally arrested morons up. there is no best. but the differences can push things in better directions if you know where you are going. think of a spice rack, or even a gun rack. 12 gauge kill you just as dead as a .22… but the hole is really different. do you reach for a shotgun when you are shooting at a cat? says more about you than it does anything else. not only that you shoot cats.
chokes store power in an electromagnetic field that is built up around a length of wire or foil that has current moving through it. this energy can be concentrated by winding into a coil, and adding a core that has less resistance to magnetic flux than air does. a layered coil around an iron/nickel/cobalt and even rare earth element/s in the core can increase the inductance a very great deal over an air core. this isn’t a perfect scheme and is very analog, too. the type of core has a significant impact on the character of the choke. one of the best ultra fi winders is dave slagle at intact audio. that is an endorsement and my opinion. his website/forum can help you understand a great deal more than i can briefly summarize here. but, any and all chokes can come in handy… even old weird transformers can be reused as chokes… the key features of this ability to store power, is the TIME it takes to charge and discharge, the slope of the delay and the release, and the amount of energy. one interesting possibility is negative regulation, within a certain range. more about this later.
chokes can be combined with caps to create filters for hum and noise, but also energy storage. there are two basic arrangements and the cheaty in between, which gets poo-pooed alot by various dipsticks. you know, it’s like the reaction straight and even some gay people often have for openly bi people: freaked out. it exposes us to the possibility that we are all capable of everything… that we can create ourselves. that we can change. change our mind, anyway. not everything is a choice but… no one is strictly straight or bent. they might switch at any time… it’s either that, or we’re all gay somehow. i think it’s time for a run of lesbian presidents. they can’t possibly do worse than the assholes we usually get, the current one excepted (although i’m not happy about where he’s at…). it comforts me that the possibility of that is terrifying to the republican leader of congress. i’d love to see pix of that prick waking up with a guy in an airport bathroom. god bless him! i’d respect him more and he’d be out of office where he can’t do anyone any harm. anyway, where was i? oh yes, choke input and “pi” filter. pi filter works and is ubiquitous… no need to elaborate. but choke input is still interesting, even after all this time. interesting because it offers the possibility of “passive regulation” or negative regulation. you often need regulation and this has no moving parts. one thing i like to do is use the chokes on the AC side of the rectifiers. this has several beneficial features. it doesn’t have to be gapped; it can be smaller in size and value for the same amount of filtering; it reduces the inrush current on turn on; and if you use two of them in both legs of the secondary, and two more in a pi filter on the DC side… and couple them. the mechanical noise can be somewhat canceled… you can have a very baddass supply, without the dunker factor going ballistic. (footnote: thomas dunker is a mad norwegian who will go down in infamy for proudly representing the poorest weight to watt ratio of anyone… hence “dunker factor” – pound/Kg per watt )
the next component is the capacitor. it is an electrostatic energy storage device. georg von kleist and musschenbroek are credited as having independently invented it and ben franklin famously used them to discover that electricity has 2 flavors: + and – . there is such a choice of groovy caps to be had now… it’s not at all like it was even in the 70’s. yet, caps create some of the biggest dumbest power fights there are in the audio world. i am going to give you my version of it. since this is a one way communication, i don’t have to care what you think about it. caps are like buckets, sponges, and also very like batteries. it takes time to fill them and time to empty them. that fact is terribly useful, within reason. they come in various arrangements, but can be split into three rough categories: dry/wet plus insulator, wet electrolytic, solid electrolytic.
dry/wet foil and insulator type caps are 100 years old now. their physics and material science are well understood. not much improvement has occurred since the second world war. slit foil, low inductance connections, and teflon and polystyrene are the main improvements. for power, i use and still like aerovox poly and oil and dry. but there are many many high quality sources. the biggest changes have been developments for wet and solid electrolytic capacitors. these are revolutionary compared to the old mallory FP multi units of the 60’s and 70’s, any awful tantalum solid, and the better quality high voltage wet formulas that replaced those in the guitar amp business.
your supply must provide the energy needed to be converted into the load, but sometimes a prudent reserve of current is needed for when things fall apart: clipping, pulses of large magnitude and complex harmonic content. the prudent reserve can help the amp recover after the catastrophe, if it is possible. this has a great deal to do with the design of the amp. it can also burn or melt things. sometimes it is also clever to go the other way, to limit the reserve so that things collapse in a predictable manner if a threshold is exceeded. the amount of capacitance required increases proportionately with the amount of current. it can be limited by using less. tube amps are high impedance, and do not usually require millifarad sized reservoirs like 200 watt transistor amps do. but they do need more than you may imagine. especially AB amps and B amps. it is possible to provide this with film caps although the physical size of the supply expands to a remarkable degree. this is the way power was done in the 20’s and 30’s. it is brute force and always works. film caps are not perfect. but they are closer to perfect than any other type of cap. i am not that particular… i am not into paper though. not anymore. sounds ok but just too unreliable. also too much dialectric absorption. they are good for toning stuff down. they smear nicely and kill fine detail and digital nastiness in a sweet way. but, this is not the only way… if you want large values of capacitance and don’t have a 7 foot rack to hold 400 2uF oil and foil caps, you will be using electrolytic caps. i have no problem with them.
electrolytic caps do not work perfectly either… none of them. they are quite analog. they rely on electrochemical design to increase the storage capacity. chemistry is slower than electricity. that can cause some undesirable side effects. all of the materials used in between the stators of a capacitor create connections between them and so leakage “resistance”, equivalent series resistance, galvanic storage, heating of the chemistry by the ac current across the various impedances, voltage variable qualities to the previously mentioned aspects… etc. my friend komuro used to make fun of me at how many caps i have blown up. often in my face. the classic situation for me is heater supplies. i was in a phase where i really tried to use the cheapest stuff possible. very anti audiophile. very tim dP… don’t laugh! have you tried it? then step off mf! high current and lots of caps. using cheap stuff or not quite enough makes them get hot… boom! still, a rack space of film can be replaced with 4 caps the size of an apple! a technical marvel actually. i have used lots of everything made by now… black gates were ok, but not worth the money. also very inconsistent. they change. crappy caps can be used too, but you have to parallel a butt load of them and bypass with film and even snubbing networks with resistors and ceramics. not weirdly enough, you end up at the same place no matter how you get there.
all electrolytic caps do create extra stuff you don’t want, but less so today than ever before in the history of their use. this is partly because of one company: panasonic. yes, panasonic/matsushita electric corp. the world’s most advanced manufacturer of electrolytic caps. nichicon, is the close second. sanyo, also deserves a place ere because of their work with solid-electrolyte capacitors. no other companies come remotely close. many companies you know copy or pay a license to use their shit. no one else spends as much on chemistry and material science r&d… no one has as many patents in the design of electrolytic caps. no one is as ripped off unsuccessfully as them too… as the computer world, including apple, can rightly attest. they cost more and are worth every penny. as far as electrolytics go, you can’t go wrong with panasonic. there are quite a lot of new tricks… for high voltage energy storage, i like Pana TS-HA, and T-ED, EB, and T-HA. for low voltage stuff type FM… OSCon and conductive polymer solid electrolytic are fuckin amazing… Nichicon LE and LG types are fab. i use them for coupling in certain applications!
other manufacturers of note are F&T (germany), who make a fabulously reliable 500V cap. these are hard to break. they have triple, or more, ESR and leakage of TS-HA. but they cost less than half! if you aren’t making a savage current dumping supply, you don’t need that. F&T work great. sprague and mallory still have their old formulations for sale. they sound vintage and are reliable. i’m not into vintage, but i am way into reliable. in any case, some basic things to keep in mind: you can parallel caps to get increased values and lower impedance; ESR is lowered by parallel connection and ripple current handling is increased…
rectifiers are, of course, the engine of the supply. they slice and dice, they chop and lop, no they are not made by ronco… at least not yet. there are actually quite a few ways to extract a pulsating DC from an AC source. my favorite one of all and the biggest pain in the ass is “synchronous rectification”. you can look it up cause i’m too tired to explain it. this can be done a few different ways, but i like mosfets best and thyratrons next best. you need a small dual 9 – 12 VAC secondary trans to add to the high voltage secondary. i wish IRF would make a module with a pair of isolated and heatsunk high voltage mosfets in one package for synchronous rect…. next, hydrogen thyratrons are number two (5C22). i only did it once but it was great! pain in the ass though, and huge and the heater is a waste. and third, call me weird, but i really like 1N4007s. they are cheap, low impedance, and good for 1000 PIV. i am fucking sick of GZ34s and WE274s. not that there’s anything wrong with them… but they are just motherfucking rectifiers. if i never see or hear of them again it’ll be too soon. audiophiles destroy everything they lay their sweaty hands on. they should be really all gone soon and we’ll be done with that. you keep em. don’t tell me about them. i’ve had enough. you all know i like damper diodes… and i like them together with 1N4007s. look, if there’s a great way to rectify that the DIY community hates… trust me, i’ll find a way to love it. i like tube rectifiers that are not popular and slow solid state stuff. hexfreds are fab! but, mos… that is invisible and the lowest impedance of all. and NO NP/PN junction! that is love! just a pain to wire it all up. my hands are shaky now and i need an intern for that stuff, but it’s unlikely. mercury tubes are cool too, but i have a kid and ain’t bringing any home. all good.
ok, i’ll leave this with a completed example. its the basic power supply for the RI-25. yes, i HAD to use shit eating GZ34s because goddamn audiophiles can’t imagine anything else. if you can overlook that, you’ll see the simplicity and the complexity of this arrangement. no, the AC side chokes are not wound on the same core… they could be if you used a Ccore and two separate windings on opposite sides of the core. that or the sides of an E/I core (not the window). there’s 1600 volts AC between them in order to get just under 1200 DC. gotta watch the insulation. these need to be low DCR too, because there is so much added Z in the bridge arrangement. that’s entertainment!