Monica USB: Lack of "speed" and "timing"?

[assendor]

Hello Martin,

You'll have to give it some time to brake in, Even these JFETs seem to need some break in time.

Reversed polariy:
   I was just testing the polarity reversal again.
On my speakers and in my system, the sound image falls down some when they are connected "correctly"
Also the bass does not sound as pleasant. Maybe this has been the case all along, still without this buffer?
I am really sure we have done some real thorough check on this as it was, but I could remember wrong.

The Magic part:
   I am pretty sure that all the magic is back now after it has been playing since the JFETs where mounted in.
The overall performance is much better. The positioning of the singers are more defined, instruments are clearer.
The depth of the room is deeper and taller.. Love it. The lower frequency instruments, OH they are so defined and beautiful.
The mans voices, those with a lot of power in them, like Johnny Cash, Elvis Presley, you can feel the power in the chest.
Womans voice Norah Jones, Oh Oh it is both very powerful and so smooth and airy like :) Love it.
Rasmus said it in a way that explains pretty well how it sounds "It feels like the singers stands closer to the Microphone now".

Like I stated before, this tweak here reminds me a lot of the performance improvement of the "Common Mode Filter trans!"
To be sure of the difference:
The best thing would of course be to compare directly with the Original Monica3, it is so much easier then.
It's just that I like best  to solder all cables directly, so it will take some time to change it all.
Three power cables, the analog signals and the digital is actually via a plug.

Another little thing with the voltage:
I do feel that when the power supply voltage into the Gainstage is 26+ volts it can sound a little tiering.
When the battery is discharged some, it sounds better.. I am pretty sure of this, I have to test some more to be 100%
There is something, but I can't exactly say what it is.
OK, I was out again, testing some more, Now that the batteries has gone from Charge 27.5volts, down to 26.04volts, Much better.
(Does the battery make more noise when it has just been charged, maybe it needs some time to settle?? )
I plugged the power-supply back in, Harsh, pulled it out, after a few minutes smoother??
Does anyone else experience this?

Could it be that this is the limit for the Gain-stage all together, that it gets more hard sounding when the voltage gets above 24-26volts?
Wasn't it Martin when I think about it, that mentioned he did not like the Sound when the voltage got too high?
Could it be that this Buffer stage unveils also this harshness some more, as the voltage gets higher..`?'

-Jonas

[martinbls]

Hi Jonas,

this is interesting! Yes, I was the guy who said that 18v supply sounds better to me than 24v. My smps can deliver both voltages via changing a small resistor plug. Now it comes up that I use 24v all the time because of two things: first, I?m sick of adjusting these trimpods every few hours, it?s quite difficult to match them to a nearly perfect voltage like, say, 11,98v and 11,97v. Tooks about 20min every time. Secondly, and that?s important, I don?t know if I could power the fet-stage with 9v, can I???

If so, I will try 18v supply again, if not, I will go for 24v and forget about it ;-)

A quick note for Wong:
This simple diagram shows how the connections of the BGs and signal have to be made after inserting the fets as shown on Yeo?s photo. Well, it?s made with paint, and as you know, there?s "pain" in paint...

martin

[Paul N]

Hi  
 
Jonas- The FET mod- which is just a simple source follower, with a constant current source- doesn't invert the phase. But the absolute phase of recordings is random, so it's perfectly possible that inverting the phase for a particular recording has a beneficial effect my new CD transport has a phase invert button on its remote control- the difference is easily audible- if your system's good enough!
 
Nelson Pass is a guy who knows his onions. His buffer circuit's design is trivial, but I'm sure he will have selected the recommended FETs carefully for optimum performance, so I've stuck with them
 
Martinbls- I certainly prefer 24V over 18V. I prefer 24V to 35V too- but here I'm not comparing like with like- my 35V supply is just the simple linear one discussed here, whereas my 24V comes form a separate 24V shunt reg per channel. BTW the input voltage the I/V stage output provides is the bias voltage- not the supply voltage
 
I noticed that the mod- while better in many ways than before- did lose some magic, as I said at the time, but that this improves with run in. I still feel that the sound's perhaps a little greyer. I'll have a think. My new CD transport is the Cyrus XT SE, made here in the UK. See Cyrus' website for details. Superb. What you feed Monica matters too!
 
Paul

[assendor]

Hello Martin,
Mr.Nelson Pass has noted voltage supply to be.
Taken from Mr Nelson Pass PDF-File:
"The buffer uses an external power supply from 18 to 24 Volts DC."

So you should be perfectly fine with a lower voltage YES!!
Please try it, so that I don't have to ha ha ha ha  :)

I really can crank that volume up now with this new Buffer in place. The Big drums really hit hard.
The Dynamics is incredible, I love it.
If it gets any better now, I think there must be something seriously wrong with my head....  ;D ;D ;D
It is already unbelievable so.

It's strange when you just sit there and listen with eyes closed, then after a few songs open them...
Oh am I here, Where is the stage? Where are those beautiful chorus girls I heard just beside me here??
Oh Yes that's right, I was just listening to some music.
It is so realistic it gets scary sometimes.

I just have to connect this up to some big TV or something.
Watch a movie with this sound, that would have been strangely interesting.
(It's sad with those LCD,s and Plasma's, the image shiver and jumps whenever there is a movement, I can't watch that. Not with this sound :)   :)  Hope they invent some better TV Screens soon)
Maybe some of you Can DIY one, so that it can compete with this sound, in the same realism category :) :)

-Jonas

[martinbls]

Hi Jonas,

18V to 24V, eh? So, what a coincidence! Ok ok, I?ll try, but it will take some time, I guess I have to switch to 18V, listen to a few songs, than back to 24V, same songs again, and so on, perhaps 2 or 3 times, to make a LAST decision. Then I?ll never have to adjust them pots again...

You know what would be really nice? To solder out the pots, solder in a socket in which resistors fit, calculate which resistor values are needed for 18v and 24v, buy them, buy some more and match them, and just plug the ones in which are needed, depening on the current supply voltage! But that would be some work again...

martin

[assendor]

Hello Paul,
Thanks for the Information.
Then this polarity obviously have been wrong all the time in my system. Just that it was so much clearer now.

Totally agree with you, Mr Nelson Pass does know what he is doing.

   The 2SK170 BL has been ordered so that it can be tried. 25 of them actually, hope they are Genuine. from Hong Kong.
(Hey a friend of mine have a bunch of them, forgot about that in the enthusiastic moment.)
Think I'll try also the 2N4416 and match them also this time, just to see what the difference is.

   Should have had three Monica3's he he,, and a remote to switch between them. :)
But then again I'm a little against the A/B Switching test to be able to hear the difference.
Often tell people that if they can hardly hear any difference even with a direct compare, then what's the point of an upgrade??

Want to solder some OS-Cons over my 2x12volt SLA Batteries to see if it makes any difference...
The batteries are suppose to make some noise, so maybe this can improve this theoretical noise?

The Resistor adjustment scheme:
Yes Martin Like mentioned above, there are many things that would be interesting to be able to just Switch over and listen...

-Jonas

[Paul N]

Hi

When you've settled on 24V  ;) it's certainly a good idea to replacethe trimmeresxwith appropriate fxed resistors- but hard wired

The optimum output V for the I/V stage is actually slightly higher than the power supply mid point- say 13V for a 24V supply. But the optimum forthe FET stage is half. This is in both cases to allow maximum symmetric voltage swing. But's it not critical- anythig around 12-13V is fine.

There's been somediscussion off the FE's input capacitance. This low enough not to matter. In fact, it's a good idea to put a high quality polystyrene cap across your I/V resistor- 330pF for 5k6. This provides a little first order filtering of hf garbage. It also helps smoth out the steps in the DAC's output. Normally tthe capacitance of hteinterconnect etc would do this, though I've always used the cap too. But with the FET in place the additional parallel cap may make more of a difference. Try it!

Paul

[assendor]

Hi Paul,

Thanks again.

I'll do the adjustment to a little above 13v, it is a little under now. 26volts supply.
390pF/500 Silver mica''s are soldered directly over the output before the coupling caps..
My I/V resistor is 4.75K.
I am thinking of changing it to 1K, I have too much gain right now.
What value on the Mica would be proper then, do you know?

-Jonas

[Paul N]

I'd put the caps directly across the I/V resistors

It's easy to work out the cap value. The output impedance of the I/V stage is about the same as the I/V resistor. So with 5k6, 330pf gives a low pass first order filter with a 3bB point of 1 / 2 x pi x RC = 86kHz. So, to keep this cutoff the same, for 1k C needs to be 5.6 x 330pf = 1800pF

[yeo]

wow! can see you guys having fun now!

i've compared couple of times but still prefer 24v over 18v. i have compared 30v as well and prefer 30v over 24v. so that's what i'm using.

hi martin

you don't have to set the trim pots to exactly half of the supply. setting to half is best but 1 or 2v off isn't going to be a major problem. after all, voltage swing is only 1v. even though i'm running on 30v now, i think mine is still set on 12v... as i said, if it ain't broke, i can be so sloppy with my own stuffs.  ;)

btw, it's nice to hear that you don't need the naim preamp at all!

hi jonas

i don't feel like the magic is lesser. it sounds better to my ears. everything sounds better. i'm not critical on phase and as paul as explained, the buffer stage does not invert phase. yeah, could be that before this you were listening to the wrong phase. but as i said, i could be doing the same too as i'm never bothered by this.

hey, sounds like i'm not bothered by a lot of things?! why am i into this hobby?  ;D

hi fswong

which part isn't clear to you? will try to explain better.

yeo

[fswong]

Hi, is this correct?

[assendor]

Hello Wong,
That schematic must be correct, YES. Quite simple and easy.

Hey Paul,
Did try a fast change of my I/V resistor, paralleled the one already there, just to try it. 680ohm par on the appr 5k.
Will be approx 500ohm, after adjusting the Bias Heavily the JFET's got so horribly hot ?:o
It is probably a little low with 500ohm I/V resistor?
Could not hear any difference in sound though. JUST Had to try it for some time, even if those JFET's got burning hot.

Silver Mica 390pF/500v:
 ? Did move those caps over to the I/V resistor. Just as this was to be done, suddenly it got clear that they actually where in the wrong place after the Buffer stage was put in. (Did not move them after putting this in place, forgot about it.)
Pretty sure that this helped some on the detail level of higher frequencies. Some smoothness factor.
Now the Buffer stage don't have to be bothered with those unnecessary high freq noise that much, maybe??

But again Paul,
after listening quite a lot to this new upgrade/tweak with the Buffer in place...
Even if the total impression is better, there is something that is different, something missing that I want back? Even though: Bass is much better, voices are even more defined, it is more in the upper top end I think there is something not as good??
Is there any possibility that matching of the JFET's could have anything to do with this? ?:-?
Could it be that a extra stage of sand could miss out on some "magic" ??
Could it be that the "magic" simply just is Imaginary, so that if the circuit is put back to original state, we will be shocked by
how good this improvement with the buffer really is?? ?
 ;) ;) ;)
It is so easy to hear how something was, compared to hear how good it gets. It's the case in any situation really :)
It's harder to actually notice an improvement, than a reduction of performance, quality or benefit.
Like a Friend of mine put in that Common Mode filter, he said he heard some improvement, but not really shocking.
After a couple of weeks he took it away.....Then He called me, said he could not understand how he ever could have imagined his system to ever had anything close to high end performance, before he got that filter in place ;D.

Is there any possibility that this Upgrade with the Buffer-Stage have improved performance in a way that my speakers lack of
performance are unveiled??? ?It's so difficult to tell exactly what gives this sense of lack Magic....

Battery paralleled with CAPS:
 ? Soldered those 2X Sanyo Os-Con 16v/270Uf sp parallel with the 2x12volts SLA batteries,
Now the sound is not harsh and strange just after the charger has been switched off. Obviously those batteries make some more
noise just after charging.
This means that it is not when the voltage gets above 26volts that the sound gets worsened.
I will leave the project with the batteries on MONICA3. Can't hear any improvement on the sound using them instead of my relatively heavily equipped regulated power supply cards.
Not after the Common mode filter had been put into operation.

Cannot do any more experiments for 14day's, and just when it was getting all that fun ?:-/ :-/ :-[ :( :( >:( :'(

-Jonas

[martinbls]

Hi Jonas!

I totally agree with you that with the fets in the feeling of a lack of "magic" into the sound of Monica appears. I couldn?t describe it better than you already did: "It's so difficult to tell exactly what gives this sense of lack Magic....". But in the last few days, listening to music as much as I could, I had an idea of what could possibly cause this "lack": I think Monica is still Monica, it?s the same circuit, the same components, the same magic. But this magic now might somehow become "covered" by this new dynamic and transparency with the fets. Magic is still there, but less obvious. What do you think, could this be an answer? Sometimes I think it?s a great pity that I heaven?t got a second stock-Monica for a/b Testing... Perhaps I should buy one ;-)

Next logical step to me is to build this clean power supply. Isolationtransformer with 30VA 230V / 115-0-115V is no problem, but 30VA 230V / 18V isn?t in stock. What do you think about a ring core transformer as second trans? There are some nice 30VA 2x18V types in stock. Or are there any technical restrictions for the use of ring cores in this circuit? And, is the bleeder resistor necessary or optional? (Unfortunately I don?t know much about "bleeding resistors"...)

martin

[Paul N]

Hi

I posted this in the wrong thread -as it's still relevant-
 
Re possible grey sound of FET follower mod. I?d forgotten that on Sunday morning I?d replaced the FETs and some other components in one channel to try and get rid of a little, very low level, intermittent noise (sounds like cathode crackle). Not sure that it did- the problem?s elsewhere I think- but the new components needed time to bed down
 
I've just bought a Cyrus XTSE transport (?800 in the UK). You can read about it here, and it's causing considerable excitement in the UK press-
 
http://www.cyrusaudio.com/product.asp?ProductID=96
 
Listening last night, I have no reservations at all. The Cyrus transport- which is truly excellent, and, if not exactly cheap, is still a bargain- and the upgraded Monica together produce music to standard which is way beyond anything I?ve heard from red book before- we?re talking good vinyl!!
 
Extraordinary


I think the "magic loss"- at least after a few days run in- is an illusion. At first, there is a certain greyness but this goes completely in a day or two. Try playing a wide range of music, and I think you'll agree Monica is better in all respects. I often find, after a major upgrade like this, that some favoured recordings now sound bad, but this is because the improvement is allowing me to hear defects not apparent before. But, overall, you should find most recordings sound better

One of the defects might be the quality of the signal Monica's receiving- see above! My previous transport was a DVD player, but Tent reclocked and some othr improvement too. Cyrus trounced it

Paul

BTW talking to Paul Hynes etc I'm more convinced than ever than some FETs make much better source followers than others. Try 2SK170- or, I suspect better still, LSK170

[yeo]

hi jonas, hi martin

i have been so lucky to be able to work from home for the past 2 days so this also means i managed to sneak in more time to listen.

the fact is that we now have an additional stage between monica and the rest of your system. meaning there is something extra that wasn't there before.

surely, it sounds a bit different now. but then notice this, pauln has less to complain and he uses 2sk170. i believe martin, jonas and me, we are all using 2n3819. perhaps the 2sk170's higher transconductance could explain our different experience here.

yes! it's hard to put in words!

but i like the current one better.  the previous could be a bit more "heavy" in the mids. now, there is more of everything now so in essence, i feel that there's more balance.

but the ultimate test of a tweak is always undoing it after a few weeks, then listen for a few weeks, then redoing it and listen for a few weeks.

whichever you could live with the longest, to me is the most "correct" one.

at this moment, i'm more inclined towards the sonics with the jfets in place. feels like it helps the system "disappear" better!

yeo