Welcome, Guest. Please Login or Register
Decware Audio Forums
03/29/24 at 12:12:30 




Pages: 1
Send Topic Print
CSP2 Design notes (Read 7567 times)
Steve Deckert
Administrator
*****


If the 1st watt
sucks why continue?

Posts: 6234
CSP2 Design notes
06/19/06 at 00:47:50
 
CSP2 Designer’s Notes


The CSP2 is one of those things that comes from constant R&D. It started as a project one night to see if I could come up with a headphone amp that could replace the MLB.  For a dedicated headphone amp, the MLB was at a price point that kept people who were not into headphones from trying it.  I wanted to do some less expensive circuit designs and see what was possible.  You see, I’ve heard magic in headphones - it requires a lot more then a cheep pair of headphones and a headphone jack.  I know my customer base, and what they like.  I know if they heard this magic I’m talking about they would embrace it.  By doing so they would expand their audio experience in many wonderful ways.  It’s a fact that it would improve the sound of their two speaker system because it’s such an excellent tool for hearing what the music is suppose to sound like.

I know why there are not more audiophiles into headphones.  I was one of them - and being the type who is really taken by good sound stage depth I never particularly cared for that in your head imaging of headphones.   Headphone enthusiasts will often remind you that “good” headphones are typically not like that and while partially correct, most fail to realize that it is in fact the headphone amplifier that makes the most difference.

In designing our first headphone amp, the MLB, I wanted something that was really top drawer - something that made those who heard it suddenly take headphone listening seriously.  To do that I choose the more expensive approach of using top quality output transformers to drive the headphones.  After listening to several candidates I choose Sowter Transformers from England.  A lot of effort and work went into the design of the MLB and it turned out great.  It demonstrated that headphones don’t have to have that in your head sound.

So back to the bench… I’m basically thinking that by eliminating the expensive Sowter transformers we could get the price way down, and that basically leaves OTL.  I choose not to go with OTL designs when I did the MLB simply because I only knew of two ways to accomplish it, and both had little appeal to me.  The first way is to run several tubes in parallel to get the output impedance down low enough to drive headphones.  With every tube you parallel, you loose transparency.  I just can’t get into the sound, too spoiled.  The second way is to use a super large value capacitor in place of the output transformer.  This second method would ideally hit the goal of reducing cost, but I have never considered it seriously.  

This is one of those defects of too much thinking.  I can say that now because I built one thinking it was going to be inferior in almost every way.  For years I would never waist my time trying it.  I listened to a couple OTL headphone amps and concluded that the extra large cap just couldn’t compete.  I was wrong.  In the one I built, I was completely shocked at first listen.  Even fresh of the bench it was so pleasing to listen to that it made the MLB sound almost hard.  The ambience was so prominent that it sounded as if I were at the venue, not wearing headphones.  The bass was unbelievably low and organic.  Music that draws your ears to it rather then music that penetrates or sometimes violates.  I was amazed that it sounded so much better.  Had I tried this from day one, things would have been different.

Still fascinated that this sounds good, I wanted to hear it in the context of a preamp driving our amplifiers in my reference room so I added some line level outputs to the circuit and gave it a listen. After about 20 minutes I realized this may be as good as the CSP preamp so I warmed it up and A/B’d the two.  It was better then the CSP.  Slightly more defined and focused / a bit more powerful sounding but otherwise almost identical.  This was another surprise.

And that’s how it happens.  Serious now, it changes from a project to consuming all of the R&D time and my full attention for awhile.  

This comes at the same time as our new Anniversary platform which trades the black 10 x 6 x 2 inch steel chassis for a 2mm thick T6 Aluminum plate that has been stretched to 14 x 6.5 and mounted into a solid walnut base.  The higher mass chassis are less effected by resonances and offer an extreme high quality appearance and other advantages.

The final prototype when viewed as a replacement for the CSP has several advantages over it besides sounding slightly better.  The CSP circuit consisted of a gain stage with low plate voltages driving a single cathode follower with a trim pot between the two stages for calibration.  Due to the lower then normal voltages that in part made it sound so good, the input stage had to be adjusted to match different input voltages.  It was well adept at offering lots of gain to lower input signals of 2 volts or less, but had to be adjusted internally with the trim pots for voltages higher then 2 volts.  

The CSP2 circuit consists of a gain stage with higher plate voltages driving a SRPP stage similar to the original MLB design.  The stages are direct coupled.  The SRPP stage drives the output capacitors.  The line level is taken directly off the outputs via a 100K precision trim pot.  The output impedance is about 10 times lower then the CSP so using line level outputs has no audible effect on the sound of the headphones.  

The output level is adjustable and the trim pots are accessible from the outside, one per channel.  With a 2 volt input, the line level output can be adjusted smoothly between 0 and 36 volts with virtually no distortion on the scope.  Factory setting will be a 4 volt output giving it plenty of authority with our amps.  High power amplifiers and or amplifiers with input sensitivities of 1 volt or less will benefit from adjusting the voltage to around 1 volt.  Lower power tube gear that sometimes requires 5 volts or more to drive it because it lacks enough gain stages, will be best served by settings between 5 and 8 volts.   This insures complete synergy and compatibility with virtually all amplifiers.  Having adjustable gain allows the serious audiophiles to run the volume control wherever they feel it sounds best while maintaining the desired level of output at all times.  Independent adjustment on each channel also allows for channel balancing.

Back to top
 
 
WWW   IP Logged
Steve Deckert
Administrator
*****


If the 1st watt
sucks why continue?

Posts: 6234
Re: CSP2 Design notes
Reply #1 - 06/19/06 at 00:48:21
 
NOISE

Having the line level output variable means that any noise in the system will be reduced rather then amplified.  For example, the CSP2 has a noise and hum level of 0.4 mv with the volume turned all the way up, and the trim pots all the way up at the full 36 volts of output.  That alone is by far the cleanest thing I’ve ever built, but since the average user will have the output level adjusted to around 4 volts.  That means whatever residual noise floor there is will be dropped by around 8 times.  When I set the units up for a 4 volt output the level drops to the limit of my meters at 0.1 mv  Part of this comes from the decision to implement a DC power supply for the tube heaters.  These improvements will make the preamp far more forgiving when it comes to tube quality and should invite more NOS tube rolling with better results.

The general layout and configuration has a few improvements also.  There are still two inputs and a selector switch, but the locations have been changed resulting in half the wire being needed to interconnect the input jacks to the first gain stage.  Now a simple non-shielded run of silver/ Teflon wire can interconnect the two.  The output besides having the obvious headphone jack, also has a 3rd mono jack that can be used to drive a subwoofer or mono center channel etc.

As with all the products built on this platform, we are using our more expensive world voltage compatible transformers.  These are significantly larger in physical size to accommodate the extra voltage windings and meat CE ratings.  These can be wired to any voltage requirement.

One of the biggest secrets behind the original CSP was the fact that the power supply is so large that it has no variance in output just like a regulated supply, but does it naturally and without a lot of heat.  The CSP circuit draws around 30 ma and the power supply can deliver 150 ma.  Power supply size and quality are always the first place corners are cut in the world of mass produced hi-fi gear.  This is why we do the exact opposite.  A lifetime warranty is just a pleasant bonus made possible by this type of thinking.

In the end we have a product that looks 10 times better then both the MLB and CSP that it replaces and sounds better then both for less then the price of the original MLB.  Occasional production glitches from delayed English transformers are also eliminated.

If someone who was new to the whole tube world wanted to hear it at it’s absolute best, they could get the CSP2 and a decent pair of headphones without breaking the bank.  They could then experiment with one or two CD players to get a feel for how much improvement in sources is possible and in the process would indeed be able to hear the magic I so often go on about.  Unlike if someone buys just a preamp or amplifier and introduces it into their system, this way is a sure shot.  

And as I mentioned, it can be the worlds best tool for evaluating cables, sources and becoming familiar with how recordings actually sound.  Armed with this as a reference you can EASILY hear huge differences between sources and cables whereas in your listening room those same differences are often unheard due to other things handicapping the system including and especially the room and speaker locations.  It allows you to hear frequency response that is actually flat without the peaks and holes.  You never hear the music get sharp, dry or offensive.  The bass is completely flat all the way down to around 5 Hz.  Bass with no holes or peaks down to any frequency is impossible in any room and in even the best rooms true 20 cycle notes are mostly fantasy.  Like I said, it will let you hear what it was intended to sound like - and that reference alone will cause you to improve your system by your knowing exactly what you want it to sound like.  This is so much easier then being blind with no true reference and just trying to make something sound good.  It ultimately means you make far fewer mistakes when purchasing gear and spend far less money as a result.

 



Back to top
 
 
WWW   IP Logged
Steve Deckert
Administrator
*****


If the 1st watt
sucks why continue?

Posts: 6234
Re: CSP2 Design notes
Reply #2 - 06/19/06 at 00:49:36
 
The above is a preview of the design notes for the new CSP2.  It should be available on the site in a few days.  :D
Back to top
 
 
WWW   IP Logged
JLM
Ex Member



Re: CSP2 Design notes
Reply #3 - 06/19/06 at 01:48:35
 
Thanks for the heads up Steve.

Looking forward to more details.

Will you be offering different bases, in different materials?  (Everything in my house is cherry.)
Back to top
 
 
  IP Logged
Giorgino
Ex Member



Re: CSP2 Design notes
Reply #4 - 06/19/06 at 23:34:33
 
Thanks for info Steve. Have you been able to establish the price of the new CSP2? Presumably they'll be the same price whether USA or World voltages - that's even better!

BTW - I like the new tube amp catalogue layout. Cool!

Regards

George
Back to top
 
 
  IP Logged
Hotsauce
Ex Member



Re: CSP2 Design notes
Reply #5 - 06/20/06 at 01:04:24
 
Is there enough clearence between the rectifier tube and the trafo for some of the fatter tubes like the GZ32 or 274B?  It looks really close in the pictures.

John C.

Back to top
 
 
  IP Logged
Steve Deckert
Administrator
*****


If the 1st watt
sucks why continue?

Posts: 6234
Re: CSP2 Design notes
Reply #6 - 06/20/06 at 02:00:17
 
The distance between the power transformer and rectifier was sometimes a problem with the old platform when larger bottle tubes were used.  In the new platform the distance between the two has been increased by 1/2 inch.
Back to top
 
 
WWW   IP Logged
JLM
Ex Member



Re: CSP2 Design notes
Reply #7 - 06/22/06 at 18:55:07
 
Steve,

Could you comment on the advantages/disadvantages of the choice between 6N1P, 6DJ8 or 6922 signal tubes as applied in the CSP2?  

Thanks
Back to top
 
 
  IP Logged
nblis
Ex Member



Re: CSP2 Design notes
Reply #8 - 06/24/06 at 17:20:08
 
Steve:  Can you supply some details on the power supply?  For example, what is the rating of the power supply transformer, and what is the power supply capacitance?  I would like to buy this unit, and I am wondering how it compares to the singlepower mpx.  Thanks,
Noel
Back to top
 
 
  IP Logged
Steve Deckert
Administrator
*****


If the 1st watt
sucks why continue?

Posts: 6234
Re: CSP2 Design notes
Reply #9 - 06/25/06 at 04:08:50
 
[quote author=JLM  link=1150678070/0#7 date=1150998907]Steve,

Could you comment on the advantages/disadvantages of the choice between 6N1P, 6DJ8 or 6922 signal tubes as applied in the CSP2?  

Thanks [/quote]

The advantages & disadvantages are exactly the same as the original CSP.

6N1P is usually the warmest and often the most musical. 6922 is a bit more dynamic and focused.  6DJ8 have less drive giving a slightly leaner bass and often boast the most detail.  They are also sometimes microphonic.

Back to top
 
 
WWW   IP Logged
Steve Deckert
Administrator
*****


If the 1st watt
sucks why continue?

Posts: 6234
Re: CSP2 Design notes
Reply #10 - 06/25/06 at 19:41:29
 
[quote author=nblis  link=1150678070/0#8 date=1151166008]Steve:  Can you supply some details on the power supply?  For example, what is the rating of the power supply transformer, and what is the power supply capacitance?  I would like to buy this unit, and I am wondering how it compares to the singlepower mpx.  Thanks,
Noel [/quote]


The power supply is covered in the Designer's notes that started this thread.


Back to top
 
 
WWW   IP Logged
Giorgino
Ex Member



Re: CSP2 Design notes
Reply #11 - 07/05/06 at 23:56:51
 
Hi Steve

I remember reading a long while back your white paper on the advantages of (Output transformers - July 2002):
https://www.decware.com/paper34.htm and it's contribution to "why most people think tube amps sound better than solid state.  The transformer couples the output stage to the hostile impedance of the loudspeaker in a synergetic way not possible without one."

In light of your latest experiences, and design notes on the CSP2, I was wondering how this might affect your overall thoughts on Output Transformerless designs for amplifiers? Is it that, for you, OTL designs is best serving pre and headphone amps with relatively benign loads or are you considering applying the principles you've recently gleaned to your power and integrated models?

Just wondering what the relationship might be between your paper in July 2002 and your recent discoveries.

Kind regards

(Curious) George  :)
Back to top
 
 
  IP Logged
Pages: 1
Send Topic Print