|
|
Post by beachcoma on Dec 11, 2008 14:25:47 GMT
Hi all,
I recently bought a MF X10-D, with the idea of modding it. So far I have decided to upgrade the diodes to UF4004 types, and build a new external transformer for it. The transformer I'm looking at is the Nuvotem Talema encapsulated toroidial 25VA I believe.
Obviously I am also planning on upgrading the electrolytics, I am however stuck at the cap choice. I am going with Pana FM caps, because of the specs. Looking at the PSU schematic, I see 6 1000µF 35V caps. Am I correct in thinking these are for DC ripple removal after the rectifier stage? I'm not sure about this due to the transistors in the last stage of the PSU. I have only designed "purely passive" PSU's for tube amps.
Anywho, if I am correct about the DC ripple removal by these 6 caps, wouldn't it make sense to drop in 6 1500µF Pana FM caps, as these have the exact same diameter (12.5mm)? This would increase the ripple correction.
Also, as nothing is set in stone yet, would anyone suggest any other mods? Or maybe a different choice of parts? Just don't suggest a blue LED, it ain't gonna happen. PINK!
Edit: At the moment I'm using a copied, handdrawn schematic for the proper amp stage, and an annoying "redrawn by:" filled schematic for the PSU. Anybody have proper schematics?
|
|
|
|
Post by dejanm on Dec 11, 2008 19:06:20 GMT
Hi all, I recently bought a MF X10-D, with the idea of modding it. So far I have decided to upgrade the diodes to UF4004 types, and build a new external transformer for it. The transformer I'm looking at is the Nuvotem Talema encapsulated toroidial 25VA I believe. Obviously I am also planning on upgrading the electrolytics, I am however stuck at the cap choice. I am going with Pana FM caps, because of the specs. Looking at the PSU schematic, I see 6 1000µF 35V caps. Am I correct in thinking these are for DC ripple removal after the rectifier stage? I'm not sure about this due to the transistors in the last stage of the PSU. I have only designed "purely passive" PSU's for tube amps. Anywho, if I am correct about the DC ripple removal by these 6 caps, wouldn't it make sense to drop in 6 1500µF Pana FM caps, as these have the exact same diameter (12.5mm)? This would increase the ripple correction. Also, as nothing is set in stone yet, would anyone suggest any other mods? Or maybe a different choice of parts? Just don't suggest a blue LED, it ain't gonna happen. PINK! Edit: At the moment I'm using a copied, handdrawn schematic for the proper amp stage, and an annoying "redrawn by:" filled schematic for the PSU. Anybody have proper schematics? Here on this forum you have a schematic that you are looking for. Just look a bit down on a list of posts here ... I think that 1500 uF will be also ok. When you see the schematic it will be quite clear to you what you have to do. Input and output cap exchange is a must. Good MKP will do the job there. Also, change the tubes. |
|
|
|
Post by beachcoma on Dec 11, 2008 22:06:53 GMT
I have the philips Miniwatt ECC88 tubes in there, NOS  Are there any better tubes, as I only know that miniwatt tubes are highly sought after in guitar amps? Also, I'm not going for "ok", would 1500µF be a marked improvement? I'll have a look at the schem, see what MKP (WIMA?) caps I'll need, if it makes a big difference I'll exchange them all. |
|
|
|
Post by PinkFloyd on Dec 11, 2008 23:35:54 GMT
I have recently discovered that upping the input coupling cap value from 2.2uF to 4.7uF brings about a VERY welcome improvement in the signature of the X-10D....... Bags more clarity! Also increase the 100uF non polars (you know the ones.... black with NP written on them  ) to 470uF non polar.... you'll hear what I mean Strap a 470nF class X2 capacitor across the 12V / IN, fit some low impedance 1200uF in the PSU circuit and........... ENJOY the music  |
|
|
|
Post by beachcoma on Dec 12, 2008 20:09:57 GMT
I take it you mean 0.22 to 0.44µF as the input cap? Also, I can't find the 100µF NP anywhere on the schematic, got a Cxxx tag? I get what you mean about the 470nF cap parallel over the AC line, but what's an X2 class? And the 1200µF cap, is that parallel over the secondary xformer line? Loads of questions, hope you don't mind...great ideas so far  Edit: I've found the X2 class of caps at farnell, makes sense now ;D |
|
XTRProf
Fully Modded
Pssst ! Got any spare capacitors ?
Posts: 5,689
|
Post by XTRProf on Dec 18, 2008 9:40:22 GMT
Strap a 470nF class X2 capacitor across the 12V / IN, fit some low impedance 1200uF in the PSU circuit and........... Actually, since it's low 12 VAC we are talking of, an audio grade NP cap will be even better than a X2 cap. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 18, 2008 10:11:48 GMT
Seems to me that the best mod would be a pair of low capacitance interconnects for the output, in combination with a Nelson Pass B1 Buffer Stage. This current circuit would have a fairly high output impedance. SandyK |
|
XTRProf
Fully Modded
Pssst ! Got any spare capacitors ?
Posts: 5,689
|
Post by XTRProf on Dec 19, 2008 3:41:25 GMT
Seems to me that the best mod would be a pair of low capacitance interconnects for the output, in combination with a Nelson Pass B1 Buffer Stage. This current circuit would have a fairly high output impedance. SandyK Oh gosh, a solid state buffer on top of a tube buffer? This is an X-10D tube buffer. And you know why people are going for tube buffer. Heh, heh, heh ....................  |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 19, 2008 4:02:02 GMT
XTRProf Partly said tongue in cheek. A B1 buffer would retain the Tube sound, but ensure a considerably lower output impedance. I was almost going to suggest losing the X10-D and relying solely on a much higher performing B1 Buffer ! SandyK |
|
|
|
Post by dejanm on Dec 19, 2008 9:10:39 GMT
XTRProf Partly said tongue in cheek. A B1 buffer would retain the Tube sound, but ensure a considerably lower output impedance. I was almost going to suggest losing the X10-D and relying solely on a much higher performing B1 Buffer ! SandyK SandyK, I understand your point. B1 is a very simple, no gain, one stage buffer without feedback. X10-D is two stage, with strong feedback, tube buffer. And theoretically B1 should be another league in comparison to x10-D ... I do not see much sense in putting B1 immediately after or before x10-D but why not using B1 in preamp and X10-D before preamp ... for example ... Although the question is good: why to use X10-D at all if you have B1 ? I am working on B1 at the moment (waiting for ages for PCBs to come from Pass) and I have modded X10-D and I think that this small MF buffer is not bad ... not bad at all. It has its charm ... |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 19, 2008 20:15:17 GMT
SandyK, your continued and unprovoked attacks on tube gear is about to get you an smite! Mrarroyo dejanm didn't seem too upset by the suggestion. He saw the point I was trying to make, but didn't pursue for fear of outraging those who prefer their music through glowing vacuum filled signal "conditioners" SandyK |
|
|
|
Post by beachcoma on Dec 20, 2008 1:41:32 GMT
Lets face it, i'm listening to a lower-midrange denon CDP and surround reciever with KEF iQ3 speakers, the KEF and X10-D are the only parts remotely "hifi". I love it to bits, and if I can get it better by modding it.....why not? After all, I'm a poor student, otherwise i'd have my dream amp (the MC275) being all goopy and tubey  |
|
|
|
Post by dejanm on Dec 20, 2008 8:24:25 GMT
Allow me to add a couple of more comments to be able to explain my point of view concerning this discussion ... First of all, I do not believe that technology in itself (tubes vs. SS) brings advantages per se ... It is all about implementation. There are good and bad SS and tube implementations and the fact that something is tube or SS does not mean much - except that it is tube or SS The devil lies in detail .. as they say .. Very important fact (not the only one though) concerning implementation is the basic topology of the unit. I was putting some comments on the topology. The topology of B1 is an old one. It has been implemented before with tubes. The popular name of that topology was WOT and typical tubes used for its implementation were: 5842. There are lot of DIY examples on the net for WOT. B1 has very similar topology adding CCS and having no transformer at the output, but that's it ... So the real comparison would be between B1 and some WOT implementation. My remarks were B1 topology vs. X-10D topology. X-10D topology is also very well known, but I believe B1 (WOT) is a better one, because: - it doesn't have feedback - it has only one amplification(buffering) stage - this is a problematic statement, because again well designed two stage amplifiers/buffers could sound better than bad implemented one stage ... Using JFets in this topology does make sense because the output impedance is very low, THD also and the linearity is high on a very wide band. And one thing more: using JFets in the way as they are used in B1 does not cost much. Implementing the same thing with tubes (WOT) is not a cheap excersise ... But all this should be verified by comparing the units of interest. At the end: all comes to the answer on a very simple question: how it sounds ? And that is the reason why I asked if somebody had the opportunity to compare B1 with X-10D or with WOT ... That would be very interesting to hear. I hope I would be soon capable of reporting some experiences ... |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 20, 2008 9:08:56 GMT
dejanm
I also have reservations about capacitive coupling between valve stages. There used to be a lot of information on the drawbacks of this method. I do remember that earlier model B & W TVs used DC restoration to overcome these limitations, but later models went the cheaper way of using capacitive coupling, and the on screen quality, particularly of high contrast scenes was much poorer. I also remember modifying an old TV to DC restoration. ( I have forgotten how I did it at the time !) I am unsure how this capacitive coupling affects LF audio in particular, but I have a feeling that it would be better without it, as would other performance aspects.
An interesting exercise would to be to construct a B1 buffer, using a dual regulated adjustable supply in combination with a JLH, which would be better than using those big electros. Adjusting the -VE rail voltage by a small amount,should also permit removal of the input and output capacitors, again further improving overall performance.
SandyK.
The extract below, is in reference to Colour TV, but should help explain my misgivings.
No DC Restorer. Col-R-Tel's color-difference amplifier is C5A, one triode section of a 12BH7 tube. The plate of this tube couples to the CRT grid through 0.5 ìF capacitor C25. (On some TVs, the CRT cathode couples to C25.) Since Col-R-Tel has no DC restoration circuit, the CRT receives a floating signal. Between color pulses, a DC restorer would maintain a constant, average CRT grid (or cathode) bias. The effect would be consistent rendition of large, static, solid objects. The contrary effect occurs in RC coupling. RC coupling arbitrarily averages the color saturation of large objects. That is, a bright or dark object floats to some "medium" saturation level. Remember that NTSC TV only depicts large objects in color. For this reason, DC restoration is very important. Also, without DC restoration, supply drift could cause large objects to vary in color. DC restoration would also blank the colors during retrace, eliminating chances of colored retrace lines.
|
|
|
|
Post by dejanm on Dec 20, 2008 13:34:58 GMT
SandyK,
Your concerns about usage of interstage cap are justified. This cap is charging and discharging all the time with different time constants for charging and discharging. This is the source of distorsion. Therefore, sometimes other coupling topologies are used: transformer coupling or direct coupling (there are some other techniques but these are the most often used). Both of them are seen as advantageous compaired to interstaging with cap ...
But on the other hand - there are a lot of (especially tube) amps with interstage caps that are sounding extremly well ... Ongaku for example ... So it is not so obvious and straight-forward ....
Anyway, B1 is a very simple, very cheap design for implementation with the promise of great sound. And we simply have to check that ...
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 20, 2008 20:14:10 GMT
dejanm
I heard a > AU$100,000 Ongaku at a HiFi show in Melbourne many years ago (went down specially from Sydney) From memory ,it even used silver wire in the transformers. It sounded very good on "Fuddy Duddy" type material, with quite a good soundstage. Then when the exhibitor wasn't watching, somebody shoved into the CD player, a recording from Madonna in "Q Sound".
It S.H.I.T itself !!
Just like most other valve amplifiers that I have ever listened to, it didn't like a lot happening at once.
I revisited valves about 18 months ago, making a Silicon Chip designed valve preamplifier , using the latest techniques,and used it with mainly DTV sound. It had a fair amount of "warmth" , One way to describe it would be to mention the difference between studio announcers on AM radio, compared to FM radio,but sounded quite good on most TV material, especially on a documentary about WW2. However , when a friend visited, we tried it with some very dynamic, and well recorded CDs. IMD was very evident. The preamp was stripped and case reused !
SandyK
|
|
|
|
Post by dejanm on Dec 21, 2008 7:28:53 GMT
SandyK,
I had the opportunity to hear Ongaky on one show several years ago, but it was rather brief listening with lot of people present ... Therefore, I cannot really comment on that. But I heard that argument before - you are not the first one claiming that Ongaku is not capable of reproducing Metallica as it should ... And one of the obvious suspects for that is the usage of interstage cap ...
Please though bear in mind that many SS designs also have interstage cap. So again, we should be talking about topologies and not about technologies. There are also valve designs with interstage transformers (like them very much) or direct coupled (excellent if properly implemented) .... they do not have problems with Metallica. Definitely not ...
Talking about caps ... B1 has an input and one output decoupling cap. Not an ideal thing as well, isn't it ...
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 21, 2008 7:41:55 GMT
dejanm
That is why I suggested using an adjustable dual regulator in combination with the JLH PSU addon, to supply B1. This should enable removal of those input and output capacitors ?
However, I have recently fitted a 4.7uF Auricap polypropylene capacitor in the input of my SS 15W/Ch.Class A amplifier, and it would have to be the closest that I have heard to no capacitor. However, it doesn't sound anything special for a few days, and is expensive at around AU$31 each !
SandyK
|
|
|
|
Post by dejanm on Dec 21, 2008 8:10:07 GMT
dejanm That is why I suggested using an adjustable dual regulator in combination with the JLH PSU addon, to supply B1. This should enable removal of those input and output capacitors ? However, I have recently fitted a 4.7uF Auricap polypropylene capacitor in the input of my SS 15W/Ch.Class A amplifier, and it would have to be the closest that I have heard to no capacitor. However, it doesn't sound anything special for a few days, and is expensive at around AU$31 each ! SandyK I have a great experience with Auricaps, too. I simply love them ... I bought them for these input and output decoupling purposes for B1 as well ... And I already installed them on the output of my X-10D ... so it would be interesting comparison ... Sandy I am not sure that I understood your remark about supply and that what you propose ... Can you be a bit more specific (actually can you be much more spoecific ;D) .... Or maybe it would be better to open separate topic about B1 ? |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 21, 2008 8:29:21 GMT
dejanm
I was thinking of using dual regulators, but with the -VE regulator voltage adjusted to ensure that DC offset could be virtually eliminated. Jeff Cowley also sent me information on how someone else proposed to do similar. I will email you the details.
Please check your SPAM Filter !
SandyK
|
|
|
|
Post by dejanm on Dec 21, 2008 16:09:48 GMT
dejanm I was thinking of using dual regulators, but with the -VE regulator voltage adjusted to ensure that DC offset could be virtually eliminated. Jeff Cowley also sent me information on how someone else proposed to do similar. I will email you the details. Please check your SPAM Filter ! SandyK Thanks a lot SandyK. I've got your mail. It looks really good ... One question though: this thing at the output, marked with: 12V, 700R - what is that ? A relay ? |
|
) to 470uF non polar.... you'll hear what I mean 
