erik2A3

Yes, notch filters, resonant peak filters, zobel networks (like L-pads, also more a component of the driver than the actual crossover) add loss, which understandably makes them not well suited for low power amplifiers. One starts running out of headroom comparatively quickly. I must say I appreciate our moderator's effort to remind us about civility, It is possible to constructively argue without personal insults. I can't help but think about the old days around here, though.... man, that got pretty crazy at times. And yeah. 2a3 refers to the triode, having built several amps using it as the output tube. Same with 300Bs and the 45, which is very much like the 2a3 (I have talked with elderly and extremely experienced designers that often referred to the 2a3 as a valve made up of paralleled 45s.

Dean, I think you mentioned something above along the lines of, and just approximately paraphrasing here, "...if a crossover network is balanced correctly, an L-pad isn't needed." If that is correct, than. I would like to respond: As we know, and as you and I and many others discussed over a decade ago, an L-pad is technically not part of a crossover per se' (aka frequency dividing) network. An L-pad (as I also know you know -- I'm simply setting a context) has nothing to do with splitting frequencies between two or more drivers -- which is the crossover's function. L-pads are in fact the means by which drivers of differing efficiency are matched to a desired output level. They may accurately be considered as 'fixed' volume controls. The variation on this is of course what are known (again, as you are aware) as 'variable' L-pads (we can correctly think of them as driver volume controls). These have the advantage of not only providing adjustable output levels of the drivers with which they are associated, but of course constant reflective impedance between the driving amplifier and loudspeaker voice coil; very much like the swamping resistors some (not I) like to use across the autoformer in the Klipsch networks of yore. Klipsch no longer uses the autoformer, to the best of my knowledge. My point is this: An L-pad is used where differing output efficiencies of individual drivers need to be matched. A crossover can be absolutely beautifully designed and balanced in terms of frequency distribution between drivers, but may still benefit significantly from the attenuation correction provided by a correctly designed and simple L-pad. They are in effect more a component of the driver than the crossover, reagrdless of whether the crossover is well-balanced or not. Wdecho: I'm in agreement with you on the AL. With some heavy power in front of it, not bad, but with the majority of amps I use (below 5 watts), it's performance is not acceptable.

wdecho, I follow precisely what you are saying, and we had traveled this ground very well many years ago on this forum. I had experimented with numerous options for attenuating the mid-horn on both our La scalas and Klipschorns -- virtually all of which were 6dB/octave slopes in order to minimize parts count and complexity -- which is most suitable, in my view, when one is using low horse power amplification. I also don't care for the sound of shunt elements, impedance equalization (zobel), swamping resistors (which I've used both on big Heritage horns, as well as back loaded horns using very expensive full-range Lowther drivers (which in the cabinet I'm using are a couple of dB more efficient than K-horns). So, a simple first order network with true midrange band-pass, and simple, inexpensive yet entirely effective L-pad on the squawker, sans autoformer. It was the most transparent sounding of any network I had used, including improved (recapped) klipsch Heritage Types A and AA. The autoformer in A/B testing, in my opinion (which in my listening room is what matters!) threw a mild veil over things resulting in slightly softened transient response, dulling of leading edges, and so on. NOT bad per se', but significant enough for me to prefer the subjectively better performance and exponentially lower cost of resistors.

Love it! very well done, and awesome help from wdecho too. Chassis top plate bolts/screws/washers might also be nice as black allen heads. Very well done! Kind of satisfying, isn't it!

Great! Heresies and a Dyna 70 are really nice together!

If there was an assembly manual that came with the modifications you did on the amp, have a look at it to see if it (just possibly) reverses absolute phase (aka polarity). If it does, you will need to reverse the polarity of the speaker connections at either the amplifier end or speaker terminals.

The ST 70 is an outstanding match for Heresies. In addition to the above suggestions, I would retrace every single connection you made in the modification of what was previously a very decent amplifier. I am not clear on whether you used the modified amp with the original crossovers -- meaning that it would be good to recheck your crossover connections, particulalrly in the low-pass section of the network. The 6SN7 really is not a "power hungry tube;" I have built many an amp and preamplifier with it without any such problem. Nor should the brand or country of origin necessarily be suspect in the anemic bass response you describe. Also check your speaker connections for correct phase, both on the input terminals on the cabinets and (more specifically) on the crossovers. A sub can always be a nice addition, but if you are not hearing the low end response you heard prior to the modifications you made, something else is amiss that needs to be addressed first.

Another possibility is to install a volume control directly on the amp chassis, itself, which I have done in the past on the Dyna 70. There are a couple of ways of doing this, though slightly different depending on whether or not the grid leak resistor on the input to the first stage (which establishes the input impedance to the amplifier) is kept or removed. However, if you have more than one source, one can very easily make a simple switch box that switches two or more sources between a single output. For some reason I always find the majority of variable ouputs on digital sources to be sonically inferior to fixed line-level outputs. That's just my experience, though, and if there is one constant in this hobby it's that it's all completely subjective in terms of what sounds better or worse than something else. This essentially is the passive attenuator to which both above posts from Maynard and Craig refer - I'm in complete agreement with them. An onboard attenuator is also an effective way of variably knocking the gain down a bit if one insists on using a high ouput active line-stage. I performed this exact modification to a Dyna 70 (along with a pair of Heresy IIs) I gave my dad several years ago. He HAS to keep his linestage preamp because of the need to boost HF response as much as possible, but complained about the lack of volume control rotation on the preamp. The slight clockwise nudge was WAY too loud. I drilled a carefully placed hole on the deck of the amp, installed the pot he had on hand (he is a tube electronics wizard who taught me to solder when I was 5) and ran the inputs to that. Problem solved.

Further: The DRD topic is of interest to me, and we discussed it extensively in 2-channel so many years ago it amazes me. I, along with a couple of members had built the DRD Welborne kits, and posted pictures of parts of the process. All those images are on a flash drive somewhere, though maybe the posts are archived here someplace. I also wrote a lengthy bit of text outlining the output stage of the circuit to which the designer referred to as "Ultrapath." Far too long to repeat again, but it did show up here awhile back attached to another forum member's post who referenced it back when I was posting under my full name (beginning on this forum in about 2002). An important feature of the Ultrapath connection (as it was called) was initially the use of a high quality motor run capacitor, which as chance would have it, was also used by PWK and company in the original heritage crossovers. DRD deaigner found the motor run caps to be particulalrly effective and sensitive at lower voltages, and that they proved at the time to be more sonically transparent, allowing for transmission of small but important details. I, as have numerous other DIYers, have also put good motor run capacitors to very satisfactory use in smoothing filters in B+ (high voltage) power supplies and output stage cathode resistor bypass positions. I'll look for the pictures of the DRD kits I worked on (not for myself, I built them for others). It's really a very clever design, and one where all sections, as Elliano explained, "shake hands with each other."

So here you go: http://www.electra-print.com/schematics.php

Don't think anyone ever said OTLs are inherently or universally better except on a person by person basis. Were that not the case, I don't think designers like Ralph Karsten. Bruce Rozenblit and others would have gone to such lengths to design extremely well-reviewed OTL circuits. There are obviously all kinds of listeners, Maynard, and we all have personal preferences. I would say that OTLs are arguably more challenging to build however. Same with caps, regardless of quality or cost. There are those designers who would just rather do without them if and where possible. Doesn't mean one is better or worse than something else; it's absolutely and entirely a personal preference sort of thing and nothing more. Jack Eliano's very first DRD is also available in online VTV articles (I think!?) I was fascinated with it when I received that issue, and it's not difficult to build. Moreover, even with AC filaments on the output triode, the things are inherently and amazingly hum-free.

How I wish I did have that amp, configured of course as monoblocks. Of the single-ended transformer coupled amps I own, it was the one, along with the Moth si2a3 close behind in terms of performance, that simply did not sound like a SET amplifier. I even experimented with a design that combined some of the best features of Eliano's DRD circuit with the Horus, which, while quite good in its own right was admittedly mildly out-performed by the other two as originally designed. However, there was one important feature of the DRD that I ultimately retained in the Horus, which I suppose transformed it into a sort of non-Horus Horus. The signal flow of this thing at the output stage became complicated, but the end result was something I won't forget because of its sound quality - for me. The stock Horus parafeed amp is truly extraordinary, but not inexpensive to build. The MagneQuest parafeed OPTs, output stage plate choke, and 5687 input/driver grid choke are an investment, but the quality of theae critical parts is very high. I remember some criticism of parallel feed amps in general at the time, where some thought of them as inferior because of their comparatively smaller output transformers. What that revealed, however, was a lack of understanding of how these amps work. The last thing for which the Horus could be blamed is a deficit in frequency response. I sold the amps to another forum member years ago, and to this day remains a big regret. The Horus was inspired in part by the work of other Parallel-feed designers such as Dan Schmalle (Bottlehead) and Gordon Rankin (Wavelength Audio), but the Horus without a doubt has important traits that are unique. Bottlehead sells very well-put-together kits, which make a step toward parafeed much easier. You might have a look on their website to see what they have available right now. Bottlhead also used to be a distributor for MagnaQuest, but don't think that's the case now. Anyway! if you are interested in SET, parallel-feed designs are one of the very best ways to do it...IMHO.

Maynard mentioned the DRD amplifiers from Welborne Labs and richieb suggests Electra-Print transformers on the Eddie Vaughn amp. If you don't know and/or just for the record, Jack Eliano was the one who designed the first 300B Direct Reactance Drive amplifier, the circuit and description for which were published in Vacuum Tubs Valley magazine a long time ago. So, here's the relationship: Jack Eliano is also behind Electra-Print transformers, and thus wound the power, output, and interstage transformers for the DRD amps sold at that time by Welborne Labs. What an intersting and knwledgeable gentleman, as well, not to mention an outstanding artist. He is responsible for many of the great vacuum tube audio cartoon drawings in VTV magazine. Moth Audio also used EP transformers...because they are built to a very high standard.

Lol! but my teenage nephew would beg to differ. He won't even touch the most amazing mashed potatoes in the world (lovingly whipped up by my wife and perfectly balanced with salt, pepper, butter, and a glug of whipping cream) unless they are covered with Ketchup, which to me is just, well....whatever. But he loves them, and that's what matters.

Steve, No doubt! That sounds like a very fine setup. GG is a killer linestage for the price. At some point you might consider the 300B headphone amp/preamp. A more involved build than the GG to be sure, but a truly spectacular product. BTW: As good as the TS OTLs are, there are of course other, perhaps even more widely known brands that employ the same approach -- Atma-Sphere and Tenor Audio being two that immediately come to mind. Getting one from either of those takes something of a larger investment to get into one's component rack, however. I've looked at the big Atma-Sphere amps from a distance for a long time. That's what makes TS such a truly unique company. Straightforward, functional aesthetics on the outside; extremely competent engineering on the inside - and all at a price approachable to most who are really serious about this pursuit. It is far more difficult to design exceptionally good sounding and stable (as in circuit stabilty) amplifiers that do not use transformer coupling than to design one that does. Whether one is inherently better than the other probably just depends on who you ask. When I was playing around with modifications of our K-horn crossovers years ago, it bothered me in an academic sense to enjoy the advantages of OTL amplification on the one hand, only to send its signal into an autoformer on the loudspeaker on the other. So, while maintaining the same basic crossover in terms of order of slope and crossover points, I plucked out the autoformer and replaced it with a fixed L-pad and true band-pass for the squawker, as well as separate termination for the tweeter - a cap calculated for 6Khz. I also had a bunch of good magnet wire on hand at the time, so wound my own air core chokes for the band pass and low pass sections. Those were soaked in varnish and allowed to dry a long time.

True enough about popularity of mayo on fries -- very common in the Netherlands, as I've seen the many times I've been there. BTW: Surely one of our national favorites (at least it's one of mine!) potato salad is much better with the mayonnaise in most recipes than with the same amount of Ketchup!

To you.

Ski Bum, Whether one component of ANY type can "out do" another is always going to be entirely subjective, including each and every statement I have made above. All components will also add their own specific colorations in more or less obvious ways, and it seems the best we can do is choose our own preferred sets of compromises and colorations. That someone is able to build a component - preamp, amplifier...whatever - does NOT by any means qualify that same person to be able to declare what's better or worse -- accept what's better or worse to them. They are making judgements specific to their own aesthetic sensibilities and are thus nothing more than opinions. My wife likes Ketchup on fries, I prefer mayo.

Not in agreement on the idea that parafeed (as Doc B. used the term) do not have the clarity of traditional SET. Quite the contrary in my experience, actually. I built Lessard's Horus....what was it, something like ten years or so ago now? And discussed my observations about it's absolutely stellar performance in terms of those very elements - extreme clarity compared to the other SET amps I had been using before I built the Horus -- and then a modified version of the same circuit. The other amps were the classic Welborne Labs Moondogs, which I still own. I also bought the Moth Audi si2a3 from Moth Audio's Craig U. (now of Eddie Current) in the form of a chassis, transformers, and box of parts. The Moth is direct coupled, and came closer to the sound of an OTL, but not so much as the Parallel feed Horus. DRDs are also very fine, and I built both the 2a3 and 45 versions for users in other countries. DRD combines direct and interstage coupling, along with the so-called Ultrapath connection on the output. That was designer Jack Eliano's term for it, but it is not new (nor is parafeed, for that matter). Both go back to the very early years of tube audio. I mentioned all of this years ago here, when writing under another handle.

Steve, disregard the negative commentary. I have followed Rozenblit's work since the early 90s, and around that time built from scratch his very first line stage -- a very decent SRPP, dual output (hi and low output impedance) line stage preamp. It had been published in Glass Audio magazine, which, along with Speaker Builder, was a subscription that dealt primarily with DIY vacuum tube audio electronics. I subscribed to both and now very much miss them. The majority of builders on this forum, including myself, picked up some of the basics from others kind enough to take the time to teach and share. Bruce has both the engineering background and creative insight to both conceive of and realize OTL (for those not familiar: output-transformer-less) circuits that make conventional output transformer-coupled designs seem (to me) rather simple by comparison. The output transformer in more common designs is needed in hollow state audio electronics as a means of coupling the high impedance of vacuum tubes with the much lower impedances of modern loudspeaker voice coils. To some ways of thinking, the output transformer is considered a sort of necessary hindrance -- one more source of possible coloration and signal degradation through which an audio signal must pass. The output transformer must also contend with not only AC audio signal components, but also the power supply - which is not ideal, and something dealt with by a number of designers through the use of shunt or parallel feed output topologies. The was an output circuit that separates power supply artifacts from audio through the use of a dedicated output transformer (only) for the power tubes, and a large choke connected to the plate of the output tube. I have built more than one such design using common big triodes, and the sonic dividends over conventional transformer coupling are in my view considerable. OTL is the logical next, but significantly more difficult step to doing away with the output transformer altogether. I was just reading an article/interview of a well-known deaigner yesterday, who (and I paraphrase) said essentially that there is no capacitor that is better than no capacitor at all -- something to which I also strongly agree, but it takes more specialized know how to design direct-coupled circuits where there are no coupling capacitors needed. And that is a very similar view of OTL designers. No output transformer, whether wound with pure silver or gold, is as good as no output transformer at all. As someone who owns several examples of both topologies (including both single-ended and push-pull) I am in the position to easily compare their differences. I have built, repaired, and own very good sounding transformer coupled amps (again both SET and PP), but if in the position to have to choose on sonic attributes alone, the OTL in general and single-ended OTL in particular would get my nod. Bass control, in particular, is truly outstanding. All my oipinion. I'm with you on Transcendent Sound. Some designers develop one or two products, and occasionally slight variations on those essential designs. And then stop. Bruce Rozenblit is constantly and successfully developing entirely new circiuts within his chosen realm of the OTL amplifier.

And to Matt of TS: You referred to the importance of dual secondary windings on output transformers - one for loudspeakers, the other for driving a subwoofer amp. Doubtless you're already aware of this, but just tossing it in here for the purpose of discussion: Another possibility is to use the existing lower impedance secondaries through a very simple voltage divider to drop the output down to a reasonably strong line level. There is yet another, even easier way to drive a sub plate amp that still avoids the use of subwoofer's (possibly somewhat inferior) internal speaker level crossover that some use to drive the main L/R speaker pair.

Thsese examples are some of the most aesthetically pleasing hand-built amplifiers I've seen on this forum in the fourteen years I've been here. As a builder myself, I certainly appreciate the amount of work that goes into not only the actual wiring of a component, but even more the manual labor involved with machining a chassis. Truly outstanding work! Very surprised, I have to say, that more of the other builders here haven't chimed in.... So, while attention to aesthetic detail (type of wood, chassis plate finish, etc.,) is all interesting, as a builder I tend to be more curious about actual circuit design and development; which, after all is what is most important when it comes to reproducing music. Can we have some description of input/ouput stage topology, type of interstage coupling, and so forth? Even the kinds and/or brands of passive components used are arguably much less significant than the design of the circuit in which they are used. Again, very admirable work!

We discussed this topic often some ten years ago or more. I was using La Scalas at the time, and decided one Saturday morning to build the P. Klipsch's little box that dropped R/L speaker level feeds down to a combined, strong line-level output to a third monophonic amp and Heresy I. I honestly thought it to be the single most gratifying modification to our system I had ever made. I became very interested in surround sound at the time, which led to the purchse of a used Lexicon DSP for much improved control and manipulation. Having already built several valve preamps, the Lexicon DC-1 became very much the 'brain' and control center of our system. One of the criticisms at the time, if I recall correctly, had to do with what was described as inadequate timbre matching between the La Scala and Heresy, however I found no issue whatever with that. What had become noteworthy, however, was a mild bass hump with the Heresy, and as a remedy for that I used an in-line, non-polarized electrolytic cap on the Heresy woofer to limit its response to 100Hz or so. This helped even things out tremendously. I was posting under a different handle at the time; maybe I can excavate some of those old posts. Klipsch's little box would make for an excellent first scratch-build project for someone interested in getting into a bit of audio DIYing - akin to rebuilding one of the low order Heritage crossover networks in terms of requisite wiring and soldering ability.

Very much missed.

Right. I first joined the forum in 2002 and was the not-particularly-proud author of many thousands of posts before venturing away for a bit to focus on building and repairing components. I've been rather used to such viewpoints; the old days here got pretty amazing at times... My point was really just that I find certain so-expressed opinions unjustified, particulalrly in this case where an extraordinarily talented designer is not coveting and hiding the specifics of his inventions but rather overtly and very generously sharing them at arguably very reasonable price points. I also have enough technical knowledge of hollow state audio electronics to be able to apply the description of "extraordinarily talented designer" with confidence. Transcendent Sound is not a new topic of discussion here. It's been brought up numerous times in the past, and there have been a number of past and present forum fellows who have owned and greatly enjoyed TS equipment.