There is a number of people who asked me if I build Melquiades or if I know anyone who does it. I do not build them for sale and I do not know currently any technician who would offer his services to build Melquiades.

I know some folks who might be interested to have Milq and do not mind to built it but they are distracted by the Melquiades’ complexity and have fear that it might be too difficult for them. The irony is that Melquiades is ridiculously simple amplifier and with a proper approach in building the construction of the amps might be surprisingly painless.

In this thread I will lay down very simple, detailed, illustrative and very non-nonsensical steps and techniques to build Melquiades for the people who have very rudimental amplifier building experience, like me. Following this thread you should be able to make Melquiades as a well, spending 1-2 hours per day. Do not forget – I am very much NOT a person who is in position to provide advises about building electronics. I have as much as many of you very bare-bone skills and knowledge about building electronics. Still, I with all my ignorance and unawareness have successfully make a number of those amps, was able to make them to sound liberatevly good, using no special skills and no professional snobbism but nothing else then a common sense. I see no reasons why you can’t do the same and particularly because the Melquiades’ resulting Sound would make the entire project more then beneficial.

I divided the construction of a full-range Melquiades on the following steps:

1) Planning and ordering parts
2) Building Bais chain
3) Building First state
4) Building Second Stage
5) The amplifier tuning and calibration

It takes 7 days, a week to do everything. Then it takes months to analyze the mistakes and prospective areas of improvement and another 7 days to do it better. The process is recursive. I went over it a number of times and will s-l-o-w-l-y share with you my current view how I would do it as now, trying to keep a balance between quality, productivity, rationality and common sense, as least as I can see it. If some of you have more expertise, experiences or the alternative view in what I will be describing then please do not hesitate to voice your opinion – I like this amp and I’m opened to learn what else might be done with it.

drdna wrote:

Hi!  I have got these 6E6P tubes from Georgia (in the former USSR, not the home of sweet potato pie).  I have removed them all from their protective glass packaging, but I am unsure of how next to hook them up.

Here is the general specification for the parts that I used. The selection of the parts implies some liberty of course. It will be your amp after all, not mine - do whatever you wish. In some case when I feel that requirement is critical I will stress it. Please do not accept my words as recommendations of some kind if DIY specialist – I’m not and I have in fact very rudimental experience in the subject.  So, use you own common sense and your ears – it was what I used. Do not forget that Milq is not the amp that would need any extravagant parts – you just do not need to use bad parts and all the rest will be taken care by the sound of the topology. Currently there is two known to me built Milq– Dima juts finished his amp but has no OPT yet and a guy in Sweden I believe (he never passed at my site). I hope they both show up here and update about third experience or perhaps recommendations…

Power Transformers.  My own version Milqiades uses input choke filters everywhere but it is not what is necessary. The First Stage and the Bias Supplies do not need to have choke input. If you go for CRC, CLRC filtration or for tube rectifications then recalculate currents and voltages of your transformers accordingly.  The voltage and currants in my specific case are marked at the circuit: (TR1, TR2 and TR3). I do see a rational to separate the filament and first stage transformer but it is not mandatory. Whatever you do make the TR3 transformer to have as low impedance as you can. The Output stage must use SS rectification and input choke filter.

Output Transformer. This is the key element and if you have the transformer then the rest of the amp so very easy. An Amorphous core is absolutely mandatory. Make it to be able to care 250-350mA, depends the core size you will be able to find and how hard you will be wiling to drive your loudspeakers. The 200mA at 200V is already acceptable operation mode for 6C33C, particularly you have high compliance/efficiency loudspeakers. However, you might go up from there until you get 50W-55W on the 6C33C’s plate. With better speakers it is unnecessary thought. Keep the DC gap in your transformer large enough for at least 15% of reserved current (if you will be driving through the OTP 250mA then use 280mA gap) Your OPT will be from ~ 6:1 to 11:1. I intestinally do not give is the precise specification for transformers because it is “depends”. If I build it now then  I would go for the re-mappable configuration of 8 or more secondarys,  8:1 in the middle mapping, DCR primary <50R, 320-350mA, inductions enough to give 17Hz-18Hz at 20W. Prepare to pay around $1000 an Amorphous transformer.

Diode Bridges – I used full valve bridge rectifiers if you wish to go for other topologies then redo the secondary’s of transformers accordingly. Use the ultra-fast and soft switching diodes. Schottky are highly advisable

 http://www.goodsoundclub.com/pdf/SchottkyBridge.pdf

but watch for the size of your capacitors. The Schottkys of that voltage can care <7A. The choke and the between-filter-resistors are a very good buffer for transient current but it is depends how ridicules you will go in the last capacitor values…. In the worst case you might put an extra soft start in the B+ of output stage.

DL1, DL2 – any time delay relays you like and was tested by you for sonic transparency. I used Magnecraft series relays. Also, you might use other delay schemas, not necessarily the off the shelf relays - it is up to you. I went for the Magnecraft series relays because it is ready to go and had no sonic problems that I was able to hear – I decided to do not reinvent the bicycle….

L1, L2, L3, L4 – the regular Hammonds do juts fine – they have low DCR and very nice made to be silent while handling the input choke ripples. The L2 should be particularly silent as it will care a lot of current with very high ripples. It should be well varnished and well baked. I used Hammond 193Q Epoxy “Ripley" version that is a regular choke double dipped in epoxy – before and after the final assembly. If you have money then I would go for a caned toroidal L2 – it will be even more silent. We are talking about a purely mechanical noise…

The V1/V2 tubes are 0A2 might be substituted with SG1P,STV150/30, StR150/30, 150C1, G150m 150C2, 150C4, CV10664, CV1832, CV2903, CV8161, HD51, M8223, QS1207, GD150/4K, QS1210, SR56. Use any brand you wish. 

http://www.goodsoundclub.com/LatestPosts.aspx?ThreadID=1339

V3 is 6E5P might be substituted with 6E6P or 6E6P-DR, no 6E5P-E or 6E6P-E must be used

V4 is 6C33C, 6C33C-B or 6C18C, any brand and any vintage.

All electrolytic capacitors I used Nichicons LK, VZ, NR, NX.  Use whatever you wish, besides Black Gates. You might keep the “last caps” C2, C8 and C9 as C10 as it market or you can go for higher values. The Super Melquiades has 6.000uF of C2, 2000 of R13. 13.000 uF of R14 (that is negate bias - very critical) and 45.000uF of C10. If you can make more absurd that I did then you will winn a nomination of being more certifiable then I am but it will not make your amp to sound better. Read the section about the Diode Bridges again.

C3 is Electovube 950B or 950D. You might use any best 400V cap you know of - it is a very critical cap. Keep it = 2uF or above

All low wattage resistors (with non-marked power) are Dale RN60 (CMF)

The R2 is 12W Mills weirwound non-inductive. 

The R22 is 30W-50W Vishay/Dale weirwound non-inductive. 

R11, R12, R17, R15, R16 I used Mills or Vishay/Dale’s CW, RW wirewound.

VR1, VR2, VR3 - any high quality, precision >5-turns trimmers. With the the VR2 I would go wirewound and the rest whatever you wish, with bulk metal-foil being my preference.

The tube sockets for small tubes - Connex Tube Sockets, for 6C33C - any you will find. There are expansive Johnson and Japanese Teflon– they are more the subject usability then sonic results.

Cables - absolutely nothing special. No silver, no high-tech stuff - a regular common reasonable-gauge hook up wires. Use >14ga for 6C33C filaments. It is preferable do not use any wires for signal path but use the natural parts leads – however it will be upon your layout. Use 6ga automobile grounding stranded cable for mains negative buss terminal of your amp.

Soldering – no high cont silver solder. The Wonder Solder as the similar with low silver content are fine.

Meters – you might use milivoltmeter or miliampermiter, once again it is upon your layout.

It is very important to visualize and eventually after you get the parts to plan,l the layout of the amplifier. The Milq circuit is drawing in the way that gives a very clean idea how it might be.  Here some basic tips:

1) Keep point “M” as close to R2 as possible.
2) Keep point “L” as close to R18 as possible.
3) Keep C10 as close to OTP as possible.
4) Keep R4 and R6 as close to grid as possible.
5) Keep R5 close to grid as possible.
6) Bring the C4 and C5 ground leads right to the gas tubes sockets.
7) Keep the gas tube very close to the first stage’s grid
8) Keep filament supplies far for the signal paths
9) Keep OPT as close as possible to the V4
10) Keep V3 and V4 as close as possible.
11) Keep V4 with an island where no parts will be closer to the tube socket then a reasonable temperature permitted distance.

I do not advise anyone about the chassis – it is purely your chose. If you do no go for moonoblocks but for a stereo amps then you might integrate some power supplies. It would require some recalculations of transformers secondarys. 

Planning step is very important. Get the largest parts, make a cardboard box or 1:1 plan and try to position the elements inn a way you like and in a way that will maximum benefit the currents flows. It is always a good idea to keep PS off the main amp chassis. If you go for it then pay double attention to the quality of the cable and connectors between power supplies and the amplifiers. If you go for a separate power supplies then keep the “last” caps of the PS chain on the amplifier chassis, niot on the PS chassis.

I do not spend a lot of time talking about planning - it is your job. If you do it well then your following success will logically flow from a good planning.

felixx wrote:

I wonder how much will cost a pair of amorphous transformers (350mA) on Pieter from Tribute....

Felixx,

I do not know. The amorphous transformers that I ordered from Tribute’s Pieter were for a single stage amps and HF DSET configuration. They were 60mA and with very limited inductance.  I do not remember exactly how much I paid but I remember that it was not bad, in fact at that time that when Pieter gave me the price I felt a revelation as I expected to pay more. So, it was very reasonable. I am sure that it is was a full-range amorphous  transformer for 350mA and inductance good to drive the 6C33C down to 17-20Hz at full power then it would be much more expensive. If I was in a market for such a thing then I would expect somewhere around $1200-$1500 for a pair of really good transformers. But it is my expectation, not the Pieter’s price….

serenechaos wrote:

I too plan to build Milq, but the DSET version.  Nonetheless, I am following this thread, and any information will continue to be appreciated.  I'm in no rush, and not even starting actual construction yet, as I want to build horns first, and get Sound "baseline" before trying to change and adjust too many things at once--

Robert, it is not that I refused to continue, I volunteered to do it to begin with. I will continue it when I see that somebody use it, mean building the Milq instead of talking or thinking about it. Sure I will describe each step as I intended initially but at this time it is nether among my attention nor among somebody else need. I promise to finish the Melquiades For Dummies™ sequel as the need arise.

decoud wrote:

I am getting the components together for the full range version, and I was wondering if you would be kind enough to consider making available the drawings and the manufacturer for the enclosure.  I rather like your design, and it would naturally make it much easier to use an enclosure that has already been optimized for its contents.

Decoud,

I do not think that I have all drawings of my original Melquiades design. I was looking in my drawing in my old paper drawer and that was all that I found. They are more evolutionary drawing and I do not even know if they were finals. Do not forget the my fist Milq was my first amp I built myself, so I do not think that there is a lot in there worth imitating.

http://www.goodsoundclub.com/Site_images/Milq_Sketch_1.JPG

http://www.goodsoundclub.com/Site_images/Milq_Sketch_2.JPG

http://www.goodsoundclub.com/Site_images/Milq_Sketch_3.JPG

http://www.goodsoundclub.com/Site_images/Milq_Sketch_4.JPG

 From my today point of view I would make a few things different for a full range amp. Still, some of the things in my original design I do like even now. I do like the shape of the amps. I like that the power transformer are sexy extended outside of the chassis and cover with an angled slope. I like a lot the layout of the signal – I do not think it could be better. I also do like that I use the “dumpster” method – cheating with silicone glue instead of mount the caps and few other elements by drilling of precise holes.

If you with you might contact John at Par-Metal NY, USA who made for me the metal work and he has all in AutoCAD.

http://www.par-metal.com/

He might press the metal for you very cheaply as it is all set on his computer. If you embrace my “dumpster” method then it would be an easy and inexpensive solution for you. I would only suggest do not make provision for a fan on back. A single 6C33C with sufficient amount of holes atop and at bottom do not need any forced cooling. At that time I did not know it…

Anyhow, it is very possible that my dimensions and my chassis for Milq would not be optimum for you as my design is bound to the parts I used. For instance the bottom past of the Milq has the heights of big power choke I use for 6C33C plate, the heights of the top part is the heights of the power transformers, You might go for less current in plate choke (it is what I would do now) of for a toroidal power transformer and then you will end up with much smaller case.