Dominic wrote:

I know there are millions of ways to implement a digital amp, but i have no idea how easy it is to switch from one implementation to another. Or whether they sound any different. I got to thinking. Maybe a Bottlehead kit of some variety would satisfy the basic requirements of above? Which might be most promising perhaps? There are a couple heathkit models out there in the wild that go pretty cheap at least might be useful for parts? I keep trying to do an amp for me but i keep ending up not doing it.

Dominic, yes, there are millions ways to implement a digital amps and all of them wrong.

If you consider building a SET then, without being falsely humble, I do suggest to go for Melquiades (if you have right speakers and right objectives)

http://www.goodsoundclub.com/Melquiades.aspx

It is simple amp and it has very high impact result. It is possible to simplify Melquiades by do not making positive bias and ruining positive bias off the first stage's B+. It will not affect sound but it would be even less parts. It is also possible to go for first stage's and negative bias' CRC filtration instead of LCRC. It will make the Melq even further simple to make. If all parts are there then to make the amp literally is one weekend project.

The benefit is that Melquiades is well-thought, it is free, it is  tested at multiple levels (all its complexity is within and have already taken care) and it is a completed amplifier that assure results, at least as I understated the results.

If eventually you decided to go for DSET then you will own the mastership over your amp and you might easily go inot DSET by updating your Melq.

Dominic wrote:

I've run into that yucky place where words get in the way of ideas. The mastership is what i'm after.

Yes, it is the yucky place where words get in the way of ideas. I did not use “mastership” in term of skills or craftsmanship but rather in term of ownership and responsibility for results.

Dominic wrote:

My problem has always been the electronic side of it. I bet i could design a driver that would get me as far as a driver could, but i'm always swimming somewhere out of my confort zone with the electronics side of playback, capacitors for instance have little intuitive use to me.

I do not see a problem here. I have no skills on electronic side as well. You might learn eventually that in order to build audio you need to know how capacitors work. However, in order to making Sound it requires quite different skills. The narrow technical skills could be always obtained or facilitated from outside. However, it is your responsibility to navigate and to apply the skills to benefit your results.

Dominic wrote:

I started reading about the Melquiades from the link and started to build up the idea into some kind of cure-all.

Absolutely not! It is juts an amplifiers and it means not more then juts an amplifier.

Dominic wrote:

Not a productive mentality, but i'll get over it. I did end up with two questions though: -What on earth is geoff's playback like if he can suggest such effective little things off hand?

Hmmmm, I do not think he has any playback at all, and I’m not quite sure that he ever had any, at least in his home. He also did not suggested anything specific in context of Melquiades. Howebe, during that time he did provided quite useful simulative influence, some ideas that eventually were used in Milq very inspired by some Jeff’s ideas.

Dominic wrote:

-In parts, roughly what should i budget for a melq build?

Will all desire to spend MAXIMUM it is somewhere well under $1.5K per monoblock. It is possible to go more if you go for “different” more expansive transformer.

Dominic wrote:

-perhaps since you say Used in past many others amplifiers, preferably SE maybe i should go for a bottlehead amp to familiarize myself with like things?

Well, I would not claim that I used “many others SETs. I had “Old” ML2.0 as my first SET and then slowly I brought probably a dozen other SETs in my listening room. Frankly speaking, hearing in what shit other SETs get turned next to ML2.0 I very soon lost interest to waste my time (I’m very lezzy and I hate to lift those things). Still I head a lot more then a dozen of SETs at “remote” locations but not in my own listening room.

Regarding the Bottlehead. Although I went a few years ago to thier meetings and I was bored to death I think they should be fine in terms cost per transition ratio (sorry for IT slogan). I think if the Bottlehead community offers two stages only, 15W SET, with grounded cathodes, no feedback, relatively high biased, linear driver that can pump 100V AC and cosul “sound” good at the same time, and if they have expertise to properly evaluate the sound of that amplifier then I would look what they offer…

Dominic wrote:

-as i mentioned to start off this thread i'm essentially starting from scratch. I'm loathe to ask this question, but i will do as i must: what should i be looking for from a pre-amp (with relation to prospective matcing with melq.)?

I have no idea; there is nothing in Milq that would require anything special for preamp.

Dominic wrote:

But the main reason i started this thread was the realisation that what i really wanted was a solid basis for pushing my understanding of music.

I know where you going but I do not think that a playback is a direction to go in order to develop an understanding of music. I would suggest different avenues.  Play back is JUST playback and it takes time. Are you sure the you wiling to invest that time?

Dominic wrote:

I wanted a tool that i could conquer, at least given time, that would allow me to tailor according to my needs and understandings, the playback i got from whichever speaker implementation i had available to me. I figured, as i started to open this thread, that this could be something acheivable regardless of topology as long as the implementation was simple/smart enough.

Well, we hope thinking in terms of tool but end up counting the fucking decibels and measuring voltages. I know, you, and many others will disagree with me but I do not confuse audio objective and musical objective. Read my post about “Abstract Audio” and try to understand why I also do not feel any conflict between Audio and Music. Sure there is a large “melt” at curtain level between musicality and audio but I have my motivations do not talk about it. Also, audio as tool might be USED deferent. It might be a useful tool, it might be crutch and it might be a fifths wheel. Audio is like a gun. You might use it for horrible things - like shooting cops or bank empoiyes during an armed rubbery or it might be use for something nice - like shooting in their enemy heads those Bush’s republicans…

Dominic wrote:

Further I was asking if perhaps there were certain basic amplifiers that would allow my playback to eventually hit on at least those sonic points i listed, not that i'm certain they can all fall within the scope of the amplification section of a playback installation, they may still be things i'd like to get through my playback. And I had a sort of system architecture in mind that i tried to describe.

http://www.goodsoundclub.com/Forums/ShowPost.aspx?PostID=432

So, instead of listeing the mouths running by some Morons, me including, try to expose yours to different Audio experiences and see how you react to them. Then you might shape the necessary questions to yourself in more precise format.

First of all: I have moved this thread into the Milq Forum…..

RonyWeissman wrote:

I'm going to try and build a Melquiades type amplifier (with some help from friends) following the advice given above  to make it as simple as possible.  I could spend more money for "different" transformer, but I will ask you for advice on that later if you don't mind.

Actually it is quite complicated as I do not think that I rave reasons to advise anything. I did not try a lot of different transformers with Milq and my chose of Lundahl was very targeted: I researched a lot and consulted with many people that I thought were serious, however that chose was not based upon my own experience.  So, the positive result that I got with Lundahls was in a way accidental, still it was positive. I would love to have different transformers bought of custom built and try it with Milq to learn hot is sound would change and I would probably do it if I was interested to become audio builder. Since I have not needs to go into this direction a stopped with Lundahl as it did very well, really-really well.

Still, I would to point out some moments that I have managed to learn. I would like to do it earlier as you, or perhaps others, should have all cards on the tables.

Since you damp the grid of the driver stage with the gas tube the Milq begin to sound very soft and very “elastic”. You could make an experiment by applying -3.4V of a fixed bias directly from filter of negative supply and the Melquiades will sound like a solid stage amp. The gas tube buffer will change the Milq sound very dramatically with this driver but at the same time it will make the amp quite melancholic sort of speak… So, you might need a very “fast” transformer with very high core acceleration rate in order to be able to “publish” all that multifaceted glory of Milq’s driver stage outside of the amplifier. The use of amorphous cores or any other fast cores of your choose is highly advisable.  From my point of view the Lundahl’s amorphous transformers were an excellent chose. They have multiple sections (you do NOT want to use taps) that could be easily rearranged to accommodate any loading you wish. Sonically the Amorphous Lundahls did excellent in full range-Milq.

There are limitations of Lundahls though. They have fix core mass and therefore you can not get a lot of current out of them. My first Milq that had a limited task to “compete” with vantage Lamm ML2.0 was running 200V and 200mA. It is OK mode for 6C33C but I would like to see in there a transformer gaped for 300-400mA. You can not get it out of Lundahl as they have “default” core-size and increase currant handling will loose inductance and consequentially bass. Lundahl is Sweden Company and if someone locally would “incentive” them to make custom transformer, for instance LL1627 but at 350mA and 12-13Henry, then it might be VERY interesting. It was the direction that I was considering to go but then I went for DSET concept and I do not care about inductance or currents anymore. I still use Lundahls in Super Milq: upperbass and HF channels.

Still if you wiling to “spend more money for different transformer” then by all means please do so. This is EXACTLY MY INTEREST the others to try Melquiades: to see what else is possible to get out of this amp. Still, I would highly encourage you to stay with very fast cores: amorphous, cobalt or whatever…

I would very highly advise you to talk with Tribute Transformers

http://www.tribute-audio.nl/

They do amorphous cores and with them you can specify any specifications you want. With 200V on plate you will not drive your tube more then 300mA (it will be less actually), so if you get from Tribute their amorphous thing with around 350A and over 10H then it will be fine. Target to get inductance enough to get 19-20Hz at full 60W of plate dissipation. The Tributes are very slightly more pricy then Lundahls. The only problem with Tribute is that I heard that it takes forever to make those transformers. It was why I did not go with them initially as I was not wiling to wait. I do not know if I was right and since I dived into DSETs I would never answer this question. Now it is your turn guys… BTW, the long lead-time that Tributes has is one of the reasons why I write it now and not “later”.  The only concern that I have I would my wondering what kind if initial plate loading to select for my loudspeakers if I have no flexibility to change loading…. (I do not buy the tap idea)

RonyWeissman wrote:

It will drive  either my Tannoy Gold's full-range, or my 3-way system (Altec 515/Vitavox S2/EV T-350 or JBL 2450).  I could also use it just to drive the Altec 515 and use my current 300B to drive the MF/HF of my 3-way but that is something to play with after.  I will be adding the dedicated sub-woofers that I can use with both systems (i hope?)  later this year.

The 515/S2 are fine. The Milq will drive them, though I hope you do not have a very large room or have 515 horn loaded. The 515 (Do you have B-class or D-class?) are very fine drivers but they have limited excursion and in large room you need to pile them up. (an array of 515B should be very cool). From a different perspective if you have your 515 horn-loaded then you do not need to go for any crazy LF… It is all should be viewed on context of everything else and in pursuit for an abstract “higher” investment might be ….a limiting factor of deign. (The very same like in the case with Lamm ML3) You would need “unload” bass form 515 eventualt… so why invest efforts into a higher inductance transformer?

You will be less successful with Tannoy Golds – you would need Reds to use Tannoy with SETs. The Golds are later drivers that looks like meant to be uses for PP and SS amps and they need a lot of power dissipate on them to light those drivers up.  From what you described I see not needs going for "different" transformer” and regular amorphous Lundahls of 200mA would do…

RonyWeissman wrote:

 I have the 515-Gs If I remember correctly, and I am trying to find suggestions on what is proper horn-loaded cabinet still.  And I agree my tannoy golds so far have worked best with early 70's mcintosh 2205 (200w/channel). Otherwise I am afraid that I don't understand your advice above.  Let's say I find proper horn-loaded cabinet for 515s, are you saying that I probably won't need to go for seperate sub-woofer set-up, and then i'm afraid I can't figure out in which case you think it's worth going for the more expensive transformer and in which case it's not really going to be worth it...My tannoy's are in fairly large room (500sq ft plus), the horn set-up in room maybe half-that with very high sloped ceilings and two-feet thick stone walls.

http://www.ae-europe.nl/audio_engels.htm

What I meant suggesting that you might not need a larger output transformer? I mean that you might not need to spend money and peruse for as high bandwidth Milq as possible… Forget about "Golds" and Milq and let look at your horns...

If the lowest channel that your prospective new SET will care will be a channel around 515G drivers then use 8Henry Lundahls and it will be absolutely enough. My 200mA does 21.5H at 12W (all my measurements everywhere are into 12Ohm). You can make it 300mA and get 30H that still is perfectly fine for a channel around 515G.  The 515G is a very good drive. It has 37Hz air resonance and it usually used on horn or 45-60Hz. It is important do not drive those 50Hz horns with very low signal - it kills sound of upper bass and it is what the Morons out there do typically.  The LF of the half octave under the horn’ rate MUST be unloaded from the horn, otherwise the excessive LF will pressurize horn’ bell but will not be exiting the horn leading to “tight” and “congestive” upper bass.  Usually for 50Hz horn (with no back chamber) a high pass 25Hz filter is a must. So, what is the point to invest effort into a transformer that would do more LF if you will need to intentionally limit the LF?

Well, it might be an argument that a transformer near its boundary begin to distort and to cut signal with filter is preferable. However, you will be using a first order filter and it is very nice when a filter just begins to cut and then a transformer kicks in. It will make a sort of shaper LF dying…

Rony,

After reading my post about the undersized core (for the given current) of Lundahl transformers you might be an impression that Lundahl are some kind of second choice. The undersized amorphous Lundahls forum is second chose for an ultimate full-range amp. The amorphous core “dies” in saturation fast therefore it needs a lot of mass-reserved. However it does not look like you are wiling to do a an ultimate full-range SET (whatever it might means, this I do not believe into the notion)

I think the only reasonable judgment would be not “how to make it abstractedly better” but rather “how to make is as good as it necessary for the given application”. Any further “abstract improvements” are compromises. Let me give you some facts. The 200mA amorphous 1627 Lundahl and the Milq at 210V are very moderate settings. However, even at 180mA and 210V and in context of “open bottom” speakers the Milq demonstrated a phenomenal bass. I really do not what to sound like I am trying to solicit people for Melquiades by demeaning Lamm ML2.0 by even with 1627A the Milq domesticated WAY more interesting bass then ML2.0. Do not forget the ML2.0 is not juts a stunning bass performer but it is the ump in context of wish I lend how reproduced bass might sound. Generally the subject of “comparing” the ML2.0’s bass capacity and Milq bass capacity is incredible interesting and I spent a lot of time to research it and experimenting with it. However, there are no audiences for this subject as there are very few people who used ML2.0 with over 100dB sensitive, properly gone, “open bottom” LF sections and who have any remote idea of what Lamm ML2.0 is really cables in bass.

Still, the 200mA of amorphous Lundahl was enough to defeat all ML2’s bass advantages (there are many reason why, and ML2 feedback is one of them). Surely, you will not get out of 200mA amorphous Lundahl a lot of power, think of 12W but you also with your 515G will not drive it deep. Do not forget that power in SETs transformer is a property of frequency and current. If you do not need with your 515G to get anything lower then 30-35Hz then Lundahl 1627A might be exactly what I called  “how to make is as good as it necessary for the given application”.

Do not forget another VERY important thing. Most of the SETs are “feedbacked SET” and they less sensitive for load (they pay some penalties for it by equalizing “quality” again different loads but they do not deliver the best “quality” against the specific load… good luck with that ML3 :-). So, the Melquiades has not feedback and therefore it is hyper-sensitive to the amp load. Those Lundahls have 8 sections of secondary and you will have access to them. They are sections not the idiotic taps that screw up everything 9even if then are implemented “properly”). So, by remapping those Lundahls sections your will be able to tune your 515G drive so precisely that no one “feedbacked SET” could even dream on. However, it might be a problem as well as it is very possible that the best loading for your 515G will not be the best loading for your S2 driver. Generally my expectations would be that with S2 you would need load the output stage at 500R-600R but that 515G with twice higher. If it is not the case then you lucky and enjoy the enjoyable…. If it is the case (you will pick some unnecessary gain with S2 and will loose so necessary gain with 515G) then you might reread my DSET concept… :-)

Paul S wrote:

Romy, you have not recently spoken as much about the FR version of Melquiades as you have about the "multi-channel' versions, albeit for obvious reasons; but have you tried the FR version on, say, a ~"FR" 3-way system that makes bass?  Or does that simply not make sense?  I am not asking for myself (again, for obvious reasons) but because it seems like others may be in (or nearly in) that situation, and it may give them a way to start from nearer where they are now.