Antonio J. wrote:

- Regarding C2 the 2000uF/450v cap, he suggests to replace it with 5 x 200uF/200V by Mundorf (he can get them for a good price, probably better those five caps than a single 2000/450). I'm not sure about this replacement, it could change completely the sound or maybe not, but he says 450V is quite a lot and maybe not really necessary. My knowledge is 0 so advice is needed.

Nope. This cap should be 450V as it will be 400V in it during the amp operation. Get any 450V cap of any brand and do not worry about the “quality” of C2.  What defiantly do not do is peruse the audiophile grade caps like Black Gate and the rest garbage. Do for any commercial grade: Cornell Dublier, Nichicon, Nippon Chemi-Con and many others. Do not worry about the cap brand at this point

Antonio J. wrote:

- C10> He suggest using three caps of the same type, the Mundorf HV 200uF/200V.

I would keep the C10 above 250V. it all depends what operation voltage you select. Keep the C10!50V above the plate voltage and you should be save. I do not know anything about the Mundorf caps, it presumable you European brand.

Antonio J. wrote:

- C3 (electrocube) to be replaced for a Mundorf silver in oil 3.3 uF. In his opinion the Mundorf is the best cap for that application.

The C3 is superbly important cap and it will highly affect how the amp sounds. I personally do not like any oil-based capacitors. They might in some way behave OK in speaker level placations but form my point of view they are not useable as coupling capacitors. Most likely he will charge a fortune for this silver in oil cap. I did pay that fortune trued them and I do not use them. Still, I never tried the Mundorf brand.

Antonio J. wrote:

- He can't find wire-wound non inductive Vishay/Dale CW or RW resistors. He can get some other non-inductive resistors, but not exactly those. He not either can find the VR1 trimmer which he says is very "exquisite". I would be very thankful if you could address us to some reliable provider of the Vishay/dales or suitable substitutes.

It is fine, use whatever resistors he/you fine appropriate and convenient. The parts selection is suggested but some parts are no critical.

Antonio J. wrote:

- Delay Timer There's nothing specified about the circuits. He has his own design, but he doesn't know if it would meet the requirements. Hence some description or schemes about the timer's circuit are needed.

No requirements in there. Let him use any reliable 10-15A relays. I personally us the Magnecraft time delay that I tested and found perfectly acceptable. If you wish I might hive you the model number.

Antonio J. wrote:

- Binding Posts = He suggests using some heavy Monacors since those are the ones he used in some other projects, but maybe you have experimented with others and find them more adequate.

I do not know the Monacors . Years ago, When I built external crossover for Trios I bout all existing in US Binding Posts and it turned out that the most famous hi-fi brands were the most horrible. The best were the “Superior Electric” the cheapest looking but the most transparently sounding binding posts.

Antonio J. wrote:

- RCAs = he uses Monacors too, but maybe other make are better.

I do not know. I do not think that this is critical for this amp. Let him use whatever he trusts.

Antonio J. wrote:

Thanks a lot, I've forwarded to him your answers, let's see what he says. I have tried Mundorfs in some critical places in the integrated tube amp I'm using now and they are pretty good, expensive, but not at the level of those Auricaps, Hovlands and Cardas caps. They're made in Germany. I'd say they have no sound of their own, nothing like the "highs silkenization" the teflon caps do, or the muddiness some polypropilene add. I've not tried other "in-oil" caps so I don't know what could they do you might dislike. I have suggested him to use the Electrocube and when the amp is working, trying if possible the Mundorf in C3. By now we're at the "price quotation" point, he's finding the materials and we'll see how much the thing takes to be made.

Audio people unfortunately behave very Moronic while they assess and evaluate components for their audio and partially it related to the capacitors. Most of audio people use initially faulty and dead topologic decision and then the try to ornament the dead sound of their components with idiosyncratic sounds of the specific passive elements. Thos freaks would tell you day and night about sound of different pars or the detractively of the conductors or about the thousands other extraneous things. Ironically if those people see an element that in context of a proper application demonstrate neutral sound then those people find it “not impressive enough”, without the understanding that the real impressiveness should not derive form the sound of passive elements. The  Auricaps, Hovlands, Cardas (especially the Hovlands) and many other caps, although they are the famous hi-end brand in reality the are horrible capacitors that are very far form musical or sonic neutrality. My suggestion of the Cube 950 was not accident but my chose among over $2.500 worth bundle of capacitates that I have trued. You may built the milk around whatever you wish but if you ask me then I would suggest the Cube 950.

Yes, if you need the "price quotation" point or the leads to source whatever you might need for the Melquiades then I will be happy to assist.

The Cat

Antonio J. wrote:

Regarding the C2 thing, He said that I misunderstood his proposal, what he says is using five double 200uF+200uF cans (400 uF each cap), which are rated for 500V (not 200 as I misunderstood) to get the 2000uF needed for C2, but having lower ESR and ESL than using a single 2000uF/450 or 500V cap. I'll do whatever you advice, using smaller caps paralleled to get a certain value has some advantages compared to using a single one, but this might have other advantages that you'll probably know better. He says the cost of a single big cap is higher than using his solution, but again he'll do what I (you) say. The same goes for C10, he would use three 200+200 (400uF/500V) to get the 1200uF/450V needed.

Sure, if the 500V rated then he can do whatever he wants.  What is important for sound in this chain is do not have C1 larger then 100uF.  The amount of C2 I feel still be too little with 400uF but he can measure the ripples and add the values accordingly if it necessary and if the noise will be the issues. 

Antonio J. wrote:

I think we'd need some supplier to find the VR1 trimmer and some of the other resistors. As I said, he can use some low inductance ones instead of the Vishays to keep the budget low, but he can't find all the values, being the most difficult the trimmer.

Di not over abuse yourself with the VR1, it is also not really critical, put a chap trimmer in there. You always after you make the amsl will be able to replace it.

Antonio J. wrote:

Regarding the delay timer, he says that using the Magnecraft pre-built timer is overkill at 120 USD and he can design and build an equally suitable circuit for way less money. I suppose he's skilled to do that.

Sure, if he wishes to design and to build the 12A/220V time delay circuit with heavy-duty dual silver sealed contacts, with led indication (very comfortable to use those LEDs externally), unlimited time adjustability, play and play applicability, protection circuit, with no sonic influence (tested!!!)… and addition all of it for $35 then certainly let him do it.

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

BTW, if you wish to turn it on manually then you would not need any delays… It I were you then I would do it without delay to test that that I like how the amp sounds….

The Cat

Antonio J. wrote:

but the builder thinks that performance wise it's better to use several smaller caps paralleled

If he thinks this way then let him to do whatever he wishes. Still, I assure you that Sound of the Milq dose not derive from this way of thinking.

Antonio J. wrote:

Some chokes will be wired by a local builder, Lundahl refused to make them.

Make sure that the chokes will be able do not buzz with the high ripple of the input choke application. Most of them do, and partially at high currents. Also, do not use the Lundahl chokes for input. They did not sound good when I tried them.

Romy the Cat wrote:

I bought the Lundahl chokes for L1 and was not able to us them and they were insultingly sharp. For the L1 you would need more henries and more DCR, the higher DCR will sound better and the “bad” and “cheap” Hammond chokes worked very fine.  You might go even for 20-25H of L1 and do not use the R23 bleeder. Whatever you do will be fine, juts keep input choke with at least triple critical inductance, a small following cap, the plate supple ~400V and ripples after the C2 responsibly low. If you Luis (the builder) is wiling to crate a federal case around his section and experiments with C2 then it is fine. But you will be eventually very surprised how irrelevant it really is. Put ANY cap in there and let it work. Then if you wish you can excrement with better or kinkier cap or play with bypassing it…  Really, Antonio, this not the subject that you need my help as the C2 is not something that responsible for the Milq sound. Do whatever you and your builder feel comfortable. The Cat

Antonio J. wrote:

The builder regards VR1 more like an "offset to 0" adjuster than a bias control for the 6E5P since its working range is quite narrow. He asked me to know from you if along the 6E5P's lifespan it needs further adjustments at VR1 or if a single initial adjustment is enough. He wants to know it in order to make the VR1 easily reachable from the top of the chasis as VR2 should be, and also make the meter able to measure DC offset at VR1 or just making it simpler by forgetting about VR1 once is correctly set. At this point I believe he's drawing the plans to fit the parts into some print board. He's going to use a digital display since it's cheaper and more available than the nice analog meters you sported.

The VR1 has nothing to do with the 6E5P lifespan. The VR1 cancels out DC voltage at input. You will need to adjust it as frequently as you change the gas tubes, means practically one in a few year of even… never. Sometime the DC voltage will be running away within 10mA but it is ability irrelevant and not important. In the Super Milq I put the VR1 under the chassis behind a miniature hole that if I want I would be able to reach by a small screwdriver. Last time I measured my DC at input it was July… I have no needs or motivations to do it again. Pretty much if I you any DC voltage at input then you will be hearing the switching of volume in your preamp. Since the switching in preamp I not auditable then you should not worry about the DC at inputs and you will adjust it one when you bult the preamp and then only if you change the gas tubes.

Antonio J. wrote:

He has been looking at the circuit and believes that in the case some DC offset appeared for whatever reason at that point, it is going to be delivered backwards to the signal source. In your case and mine, the preamp. Have you experienced some problems for that? Does your preamp have any "backwards" protection to avoid any DC coming back from the Melq?

Nope it is not a big deal. Your preamp is capacitance coupled or transformer coupled … my is direct coupled and even so it does not care about the DC at output. Believe me or not but I operated it even without the V1 and it was fine. Sine you set 0 zero Volts at input then you will be juts fine.

Antonio J. wrote:

Just one more question, did you remove R23 from your Melquiades?

It does not mater. Since you have more then triple critical inductance then R23 is not necessary. I believe I do not use it in Milq. In the super Milq where C2 is 5.600uF and C1 and C2 sit on the different chastises I believed I used the R23 to shunt the large caps and to let it to discharge the bi cap. Still, it is has no relation to sound and I have no idea why keep asking me about it. Your technical guy knows very what bleeder resistor does in this case and let his to make this decision.

Rsg,
The Cat

AnonymousUser wrote:

I am thinking of launching into building or rather having buitl by a techie friend a pair of melquiades. Could you tell me more about the cost of building a pair of these amps? then how do they compare to other amps? I mean are they really better than such well-known amps as the lamms, wavacs and tenors to name but a few as rommie clearly states?? I am located in france and would appreciate help/feedback before leaving what I have for this project.

I understand your frustration. I do not think that anyone would be able to answer your question about the Melquiades’ performance vs. other amps. I have at a dozen or so people who heard Melquiades. The perception was different, mostly positive (when it was in my home) but frankly speaking none of the people who head Melquiades I might considers were serious listeners who were able for sensible audio critique and analyses.

I would not be trying to convince you that Melquiades is worth to pursuit as in order to do so you should be familiar what is the weight of my words. 

Also, looking at the circuitry you or your technical person would not be able to assess the sonic potential of this amplifier as the amplifier designed according slightly different methods and technical folks frequently are no equipped to understand them.

If you have sensitive enough loudspeakers or relatively small room and if you contemplate to use 6C33C then make your own SET. No mater what you do you would need a good out transformer. If you get the Amorphous Lundahl LL1627 for instance then it would be very versatile amps that you would be able to use in many applications with this tube. You might use any other amps of your choosing but be prepared that the cost of the OPT will be a half the amplifier cost. (The Lundahl LL1627A is ~$600).  If your have a transformer then it would be very easy for you to assemble a Melquiades prototype, driving it from lab power supply (still do use the Melq’s gas tubes – this is essential, also be very careful with negative supply as it should be superb quietly). If you do not like what you hear then you would not invest many efforts or money into the project. At In this case your would be able to write at this forum that Melquiades is crap and I am a pile of BS and then go for another 6C33C as your have most of the compo net are ready to be reused.

I personally do not think that you would be able to find such a phenomenally sounding driver as 6E5P or would be able to push our of 6E5P more interesting sound then it is done in Melquiades but I have open mind, will be learning about your views, and do not mind to help you in your evolution in this forum if you give some limited fate to the Melq.

Rgs,
Romy the Cat

Anonymous Gilles wrote:

I am not fgrustrated but trying to gather information before making a decision. I sen you an email from france and I own electrovoice patricians in a large room, about 100 square metres my question is is your amp real better than say the lamms or tenors and other much praised amps in the audio community

I did not think that me semantic Romynism “frustration” you would take too seriously, although to manage sound in 100 square meters room should create quite a few frustrations in audio life.

EV Patricians are quite fine loudspeakers, they are boundary loaded with all following consequences and around 105 dB sensitive, I do not know how “live” your room but still I would have difficult time to assess if it would be fine for 100 square meters room. Melq is 15W and approximately 20dB gain. My concern is not with advising you how the Melq would perform vs. Lamms or Tenors (I know those amps very well) but if you have enough power to drive a 1700 square feet room.  You see, Melquiades uses fixed bias  (for sonic reasons) but the 6C33C with fixed bias might behave peculiarly at high grid voltages. To drive1700 square feet you would/might need to drive the 6C33C quite hard. So, I do not feel comfortable to suggested you anything that I have no reasons to suggest …

If you got for prototype and I proposed you above then you would be able to assess the result yourself. In case if you feel that you more power then your could always replaces the Melq output tubes with more powerful triode (211, GM70 or whatever you prefer…) The Melq driver stage will drive those tubes and if the output stage and it’s power supply would be done properly then the Melquiadenization effect should not be hurt, as a dominating factor in the Melq sound cames from Melquiades’ driver stage

Rgs,
Romy the caT