Rerurn to Romy the Cat's Site

Melquiades Amplifier
Topic: More power is less heat and more efficiency. Do direct impact to sound.

Page 1 of 1 (13 items)


Posted by Iraschwips on 10-05-2017
I'm thinking about trying to put together a Melquiades amplifier for myself. I printed out a copy of the schematic (thanks for sharing). I have a simple early question. Both the 12.6 and 6.3V windings for the filaments are shown as being center taped. My question is whether it would be acceptable to use a non-center tapped windings, at least for the 6E5P (6.3v). Looks like the resistors used with the center tap are optional components.

I attached a photo of a transformer I was considering purchasing in anticipation of starting this project. The cost would be $39 x 2, or $78.

Posted by anthony on 10-05-2017
Looking at that diagram, it seems as though Romy has made an artificial centre tap for a conventional filament transformer.  You can either get a centre tapped transformer, make an artificial centre tap as shown or not bother with a centre tap at all:  all those approaches are valid.  I use centre tapped transformers but they are not necessarily required.

Your photo is no longer attached.

Posted by Romy the Cat on 10-06-2017
Iraschwips, sorry did not see your post. Anthony is basically right, it does not matter for this specific application of you have center tap or not. It is also not really important if you connect filaments on 6C33C in series or parallel. There is a catch in here. If you connect parallel then you have a freedom to drive the first stage and second stage from the same fully loaded transformer or to drive a half of 6C33C if you would like to. There is a catch in this catch as well. If you chose to add or remove filaments at will then the voltage drop as you load the transformer more might think your voltage well below of what you need to have. From what I observed the audio filaments transformers are real crap.  You buy let say a transformer that should give you 6.3V and 10A but if you begin to load it more then you might discover that the current rating is very not really accurate. It might give you 6.3V at 6A but as you draw let say 9A then the voltage will drop to 6.0V. If you have just one channel then it is not a problem but then you have many channels and you have an option to wire the thing differently than you do want to have a lot of extra VA transformer. So, if I need 500VA then got for 1200VA transformer and have a lot of spare power in there. This is just a direction t think that I share. You should be of for one channel with a direct power matching..

Posted by Iraschwips on 10-09-2017
Anthony and Romy, thanks for taking the time to respond to my question. I want to take on this project as a learning experience both in terms of simply building an amplifier from a schematic and for my edification on sound reproduction.

I’m starting out by doing some planning and picking out the chassis mount parts. Based on what you’ve said I think I found a filament transformer which will work. I also from the same supplier picked up some Johnson 7-pin (1 large) sockets, which I’m sure I overpaid for.

I’m going to attempt to build just the full range schematic, as mono blocks, not any kind of multichannel version. The filament transformer I ordered (x2) is a Freed potted transformer model No. 28130. I wasn’t able to attach an image but you can search for the image online. It has two secondary windings, a 6.3v and 12.6V. The former rated at 1.8A, the latter rated for 5.5A. The 12.6V is center tapped. I think for my application the current handling should be sufficient. Looks like the tube data sheet for the 6C33C shows that it requires 12.6V when heaters A and B are wired in series. A parallel heater connection halves that number. I guess I will heat both tubes from separate windings; the 6E5P from the 6.3V winding and the 6C33C in series from the 12.6V winding.

The next big decision to make will probably be which output transformer to buy, whether I should go with the Lundahl LL1627AM or something else with a 300-350mA gap so I can drive the tube harder. My speakers right now are four-way dynamic driver units rated very optimistically at 97db. It’s probably more like 92-94db.

Posted by Iraschwips on 10-09-2017
The following information might be helpful for neophytes like myself with questions about heater wiring.

http://www.valvewizard.co.uk/heater.html

I now understand what Anthony was saying when he spoke of an artificial center tap.

Posted by Romy the Cat on 10-09-2017
Iraschwips , sorry that your post was deleted. The site was hacked this morning, I fixed it and your post was the only casualty. I restored it under my name and since you reported it we all set for now. 
 
For sure you can do whatever you want but frankly with 92-94db acoustic system I am not sure if Miq is something that I would advise you to be involved unless you are listening in a closet. I would look at different amplifier.

Posted by Iraschwips on 10-09-2017
 Romy the Cat wrote:
Iraschwips , sorry that your post was deleted. The site
was hacked this morning, I fixed it and your post was the only casualty. I
restored it under my name and since you reported it we all set for now.</div>


That’s not a problem. Thanks for reposting.


The conventional wisdom would say that the speaker is a poor match for a low powered amplifier. If I needed more power, couldn’t I just add a second 6C33C as an additional stage after the first?

Since I have been pairing the speaker with a SET amp using a 300B driver tube I feel that even without an additional stage it may still be worth experimenting with this amplifier or other low powered amplifiers. I can say that the current amp has no problem outputting way more volume than I need in my room, nor does it account for any noticeable frequency response shifts. The details of a recording while maybe not displayed with the microscopic precision of a great headphone setup are there as well, not obfuscated greatly by inefficiency. I don’t have any way of testing but the speaker is supposed to present a relatively flat 10ohm load. The shortcomings seem to be that the speaker can sound overly polite and distant, and can struggle to keep up with complex musical passages. Peak to peak dynamics don’t begin to replicate real world live acoustic music, though the dynamic shadings of individual instruments help flesh out points of emphasis.

Higher powered solid state amplifiers don’t necessarily correct these problems which I would say are indicative of an incompatibility between amp and loudspeaker. More power also just seems to drain the performance of its subtlety and much of the emotion of whatever is playing. The problem with the setup as is has to do with the underlying musical intent or direction not always being clearly presented. It can be hit or miss even with the same recording played at different times. There are a lot of variables at play and certainly the listener plays the biggest part in the equation. I just feel that as a tool used to place you in that meditative state of being absorbed with a piece of music, it’s a pretty inconsistent device and too incompatible with differing recordings.

I would be interested to hear how the Melquiades amp presents musical content even facing the limitations of the speaker. The limitations I find which I assume are related to an imperfect match between speaker and amp are certainly distracting but they’re not deal breaking for me, at this point.

Posted by anthony on 10-09-2017
 Iraschwips wrote:

I don’t have any way of testing but the speaker is supposed to present a relatively flat 10ohm load. 


This is an excellent tool for measuring the impedance load of your loudspeaker.  I have one and use it all the time.  You can also test caps, inductors, resistors etc..

http://tempuri.org/tempuri.html


If you do go down this road I would look at not only the raw impedance values, but also the phase angles.  If the phase angles are less than say 30 degrees then the speaker should not be too difficult to drive and may possibly be suitable for the Melquiades.

Personally, I would start with the known Melq design and then perhaps experiment with paralleled tubes.  Perhaps.  First though, I would investigate further to see if your speaker is in fact a relatively easy 10R load.

Posted by Iraschwips on 10-09-2017
Anthony,

I was not aware of this DATS device. It would be one of these things that would be nice to borrow. I know someone who does speaker design, maybe he has one. Right now they are sold out on parts-express. I put an alert in to be notified if it's restocked. Thanks very much for the suggestion. I'll try to get my hands on one. Seeing these measurements would be informative.

Posted by Romy the Cat on 10-09-2017
Iraschwips, I do not think that it has anything to do with impedance at all. You might have sub 1R impedance of your loudspeakers and perfectly compensate it by wiring a custom transformer that will perfectly work against lower impedances. I would not use Milq for anything with lower sensitivity then 100Db, perhaps very high of 90s in case you have a very small room. It is not even about gain or power but some sort of energetic or sort of “enthusiasm” that you will not have with low powered SET and low sensitivity drive. There are panty of other options for low sensitivity acoustic systems. I do feel that if you like the sound of 300B with 92-94db then you need to listen more… or louder... Do not also forget the 300B is DHT tube and if you have enough current on the driving stage then the tube might be driven to the right side of graph and work with grid currents. Better of worse but all DHT will do it and this is what they are OK in A2 and they do very soft clipping. Some of them do not clip at all and will run like crazy until the driver stage has a current clip. The 6C33C is very different beast. It is IDHT and it absolutely not able to enter right side of graph and as soon the swing of the grid voltage approach the bias then tube is super hard clipping, producing horrible bass and generating very hard UHF harmonics that damage you twitters, lower IQ of your kids, fade your wife cactuses and make stupid audio reviewers to experience multiple orgasm during listening. So, do not experiment with indirectly heated triodes and low sensitivity acoustic system.

Posted by Iraschwips on 10-10-2017
Hmmm.... I'll have to speak with the future Dr. Schwips about intermodular distortion lowering IQ's in children. If it's just another risk factor on top of faulty genes then I think I can live with it. That cacti would suddenly turn pale and the basement would overflow with middle aged men getting their jollies courtesy of my toys, is highly disturbing. For the love of g-d, steer clear of the pinball machine!

Could the circuit be modified to accept a higher powered Directly Heated Triode like a GM70 tube? If the topology doesn't work with the speakers I have, where does my amp building project go from here? Specifically what other circuits are worth exploring or experimenting with given what I have said in my last post?

As I mentioned I think the speaker sounds polite or distant, which may be what you're referring to by lacking enthusiasm. My simplistic understanding of speaker loads seems in line with what you've said. You can customize the secondary of the transformer to suit whatever load you need. However I think if the load varies a lot you'll see nonlinearity of the frequency response with single ended designs.

Posted by Romy the Cat on 10-10-2017
 Iraschwips wrote:
Could the circuit be modified to accept a higher powered Directly Heated Triode like a GM70 tube?
   
Yes, there are version of Milq out there where the same front-end in SRPP configuration runs GM70. First of all if you are new to build amps then I would not advise you to start with GM70, high voltage is dangers. Second, I never hear it and I do not know how good or bad it is.
 Iraschwips wrote:
If the topology doesn't work with the speakers I have, where does my amp building project go from here? Specifically what other circuits are worth exploring or experimenting with given what I have said in my last post?
  Iraschwips, I am not an amplifier specialist and in fact I know very little about the subject. I built my acoustic system and I for a while multi-amped it with a very good SET amps (vintage Lamm ML2). I was looking for DSET multi-amping flexibility and an opportunity to go away from Lamm world. I initiated the Melquiades project and was able to come up via collective efforts with the design that in my view very much compete and in some aspects defeats ML2. I wired my playback with multiple Milq DSET versions. That is all extend of my familiarity with amplifiers. I know Melquiades design, implementation and utilization in playback but this is pretty much restrict the level of my expertise in “designer” region. I never fancy myself as a specialist in amplification and any other amps that I know or might know I am familiar only as a user. You need to find other people to ask what other circuits are worth exploring or experimenting. I truly do not informed and you do not want me to pull out of my ass ideas. Also, I got my ML2 in 1999, then I got Milq and in 2004 I got DH Milq version. So, for a good 18 year I am out of amplifiers world. What I am sing is that I am not the best person to ask.

Posted by Iraschwips on 10-10-2017
I have to be extremely careful with the original design as well. I wouldn't want to make a mistake in either case. More voltage and current atop of already potentially lethal levels may not change the picture all that much. However, it's a point well taken.



Sorry to bother you with these questions. Thank you for taking the time to read what I've written and help me out. I understand you've found both an acoustic system and amplifier system to pair together which you are happy with and haven't explored this topic personally in some time. That makes perfect sense. Since this is a public forum, I was hoping perhaps someone else might have some suggestions for me. Anthony, thanks a lot for your comment. I also appreciate you taking the time to read my questions and provide some guidance to me.


Page 1 of 1 (13 items)