• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Purity and simplicity of amp design

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have been dreaming about tube amps for more then 10 years but never owned one until this year. My original idea was to build my first amp using a simple recipe. I liked the Decware SE84C and was planning to build it. I started sourcing parts last December and was looking at speaker options, speaker cables and interconnects options and diy audio stuff.
Soon, I now had projects to build the amp, the speakers, voltage regulator and all the cables. Too many projects, too little time.
I knew I would not be able to build it all and have the time to enjoy music before 2012.
In an attempt to accelerate the process, I decided to purchase the 1st amp, a simple source and the speakers for my speaker project. So I ordered the Decware Taboo amp, an Oppo DV-980H SACD player and 2 Fostex FE207E drivers.
I am now listening to music with my FE207E drivers installed in Rubbermaid bins. There is very little bass, but the mids and high are much, much better then what I was able to produce with my original Sima PW3000 and Paradigm 5Se speakers.
I have been working since December on the final cabinets for the Fostex drivers. The only time I have available to work on this is usually on Saturdays. The Cabinets are huge :bigeyes: I am building a non-folded BLH that should go to about 30Hz. 54" high, 50" deep, 12" wide :D I hope to be finished in the next 2-3 weeks.
I will post pictures :) Hope I like the sound ... after the tuning of the compression chamber that is.

And that's not all ...
I am also working on the CAT5 braided speaker cables ... and the braided silver interconnects.

More ...
I have purchased 2 Tubelab SE boards and sourced the components to build one. All I have missing are the 300B tubes.

I have one month vacation planned soon, and I think I know what I will be working on.

I am already planning the projects for 2009 :idea: My goal is to build a high quality 300B amp. I know I will have to spend a substancial amount in tubes and iron.

What do you think about the Axiom 300B ?
http://home.att.net/~chimeraone/allianceaudioprojects.html
http://home.att.net/~chimeraone/axiom300bschematic.html

The Leonardo ?
http://www.lennartlabs.se/projects/leonardo/leonardo1.htm
http://www.lennartlabs.se/projects/leonardo/leonardo3.htm

The WECO 91 and similar ?
http://www.fortunecity.com/rivendell/xentar/1179/projects/legacy/Legacy.html
http://www.bonavolta.ch/hobby/en/audio/300b_1.htm
http://www.diyzone.net/diy/references/tube/we300bse.pdf
 
Let me start by saying my knowledge in audio is very limited. I understand most of the circuits but did not enough experiment to have my own opinion. I feed from internet information and make my own conclusion. A great part of that is my belief that simpler/natural things are better. It is as simple as the sound of knocking on wood compared to plastic. For that specific reason, I refuse to paint or varnish speaker cabinets, or use plastic anchors or Tyeraps in my amps. I prefer point-to-point over circuit boards.
Again, it is not a rule, but more a guideline. That's why I am ready to try the Tubelab SE and make my own opinion. I like Georges way to try new things and share with the DIY community. Maybe one day, I will be able to contribute my share :)

Power, distortion, source impedance ?

Not power. I believe in the 1st Watt and willing to sacrifice power.
I prefer to limit the number of components, the purity of the material used in the components. 300B with 5K OPT is ok.

Distorsion ? I like the 2nd harmonic in moderation. So I prefer SET over push-pull. Also willing to sacrifice power to reduce distortion.

Source impedance ... I did not collect enough knowledge and information to completely understand impedance and it's effect on a design. Any good reference reading on this subject ?
 
After the first watt

Hi Lowtherdream,

I think it should be changed from "first watt" to "first power increment giving 93db" or something like that, because the sound of that first watt depends a lot on the efficiency/sensitivity of the loudspeaker.

Paul Joppa suggests enough power to achieve 102db peak at your listening position. My own experience bears this out as a minimum for a wide range of music. If you like listening to Santana etc, at impactful levels you will want at least 3db more headroom.

http://www.audioasylum.com/scripts/d.pl?audio/faq/sets.html

If you listen to one watt at 93db, 102db gives you 9db of headroom for dynamic peaks in the music. That would require an 8 watt amplifier.

Recorded music used to have a dynamic range headroom from "average" to peak of about 14db. That was based on the capability of magnetic tape and the vinyl system. Nowadays it is much less (like 3db) due to extreme compression employed in the production process. Since I don't like to listen to compressed music, I like 9db to 12db headroom to provide a sense of openness and space. Otherwise, the amp and speakers do the compression for you.

This can be carried to extremes :-0

http://sound.westhost.com/dynamic-range.htm


PS A corollary: If you only listen to highly compressed music, a lot of headroom is not needed and the 8 watt amp will not sound much different from the 1 watt amp at a 1 watt average level.
 
Re: After the first watt

Michael Koster said:


If you like listening to Santana etc, at impactful levels you will want at least 3db more headroom.


Thzat's nothing.The notion that all CD's are level compressed is a complete illusion. In reality the levels are all-over-the-place. Listening to Mike Oldfields "Amarok" with an 8W amp SET amp with high sensitivity LS's will create a simply depressing distorted performance. And I don't want to design an amp with a level overload indicator.
I don't know where one is not going to find uncompressed music material about....I bet the London promenade concert broadcast played live for those audiophiles is also somewhat compressed for FM radio, even digital somewhat equating to CD classical medium.
For those who like somewhat expanded music loud, the standard 50W p-p has all the salts to offer at a thd far lower than a SET design can offer.
Long live push pull.:D

richj
 
Yo lowtherdream,

A few things come to mind.

First of all you've decided to follow a specific "philosophy".
You'll get results comensurate with that "philosophy."
That being, in short hand, the Fostex full range, with SE tube amp and DIY cables.

I've heard a Fostex system sound surprisingly "nice" to listen to on two separate occasions - at the homes of two very well known "personalities" in this hobby/biz.

Neither time I thought that the presentation was of the sort that was any better than pleasing and non-irritating - not that this is an achievement to be skoffed at!

It's not a "high resolution" sound, nor is it going to play Led Zep rockin' out, nor a big choral & symphonic piece! It's not going loud, although the smaller the room, the louder it will seem.

Having said all that... the Axiom is an advanced, sophisticated circuit. The other two are simple circuits. That means absolutely nothing about how they will sound.

How they will sound will depend on both parts and execution. And especially the transformers.

I personally greatly prefer anything run in A2 over A1 - that means DC coupling of some sort to the grid of the output tube so that the grid can draw current - a cap coupling will not permit A2, only A1.

The simple advice is to build something now and as fast as you can with whatever you can get your hands on quickly! Forget about getting it "right" at all. Build something that works. You can improve upon it, or build something else once you get going with it.

Within reason, and assuming you don't care that much about aesthetics (at least to start) you can wack out an amplifier in a weekend or two at most - some folks have built on upside down cake pans, works fine, looks like hell.

Btw, I do not like CAT5 cables at all, but maybe they sound good with fullrange speakers like the Fostex... doesn't matter. Just do something, and get results. Don't wait for making a CAT5 cable if you have everything else, or just SOMETHING to listen to, use phone wire if that is all you have on hand today.

One last bit of free advice - don't be seduced by the flowery words in print. Sure try something because people wax poetic over it, but keep your mind open and find out what works for you - heresy is fine.

Btw, do not skimp on power, given the choice. Within reason almost any Class A SE is lower in distortion at lower powers than at higher power... so paralleled 300Bs SE will likely be better than a single 300B... and have more power... you might end up prefering PP, but maybe not if you are using the higher output Z (lower DF) to "tune" the bass end of your FR speakers...

It's all yin/yang, and one tradeoff against others.
You have to start by doing something and then after time you'll start to see what is what, and make some choices about the tradeoffs you can live with vs. those that you can not. :D


_-_-bear
 
ambience exists said:
So... If I pair my 8w push pull with massively efficient speakers, say, 102db, I should get quite the range eh?
I clipped a 50W class A PP tube amp into my KHorns and own designed fronthorn system on occasion. Big room though.
Michael Koster said:
That's the theory... I'm going to test it myself with my 300B PP amps at 16WPC thru a friend's Klipsch LaScalas (105db supposedly). Problem is the LaScalas have no bass so I guess I'll need to bring my SS powered sub.. is that cheating?
Why? Dunno why anyone would bother with tube under 100Hz or so.
 
bear said:
Btw, I do not like CAT5 cables at all, but maybe they sound good with fullrange speakers like the Fostex... doesn't matter. Just do something, and get results. Don't wait for making a CAT5 cable if you have everything else, or just SOMETHING to listen to, use phone wire if that is all you have on hand today.


I agree, braiding cat5 is a waste of time. I bought some 18g mil spec wire as hook up wire, its $0.18/ft so I got 100 ft in 2 different colors and twisted a star quad for speaker cable. I really like the results and it was cheap, quick, and looks good. Braided ICs are really bad, too much capacitance... look into a LOW capacitance design for ICs. I like the anti-cable or counter-rotating helix geometry... both of which use 3-D geometry to keep the signal at a 90 deg angle with the ground wire.

Dave
 
Thank you all for the reactions and comments. My native tongue is French and I suspect my thoughts are not always well reflected by my choice of words or the structure of the phrases. Sorry for the spelling mistakes :xeye:

300B at 5K why?
The 5K load line is in a more linear region (on paper).
I have never tried a circuit with a DHT and I was just really saying that most of the SET 300B designs use a 3K OPT and I am ready to try a 5K OPT to gain linearity, reduce harmonics, at the cost of power.

CAT5 braided why?
I already have CAT5. Many say that the results are very good for the price, not counting labor ;) There are so many expensive cables on the market and I think most of them do not deliver the value. Let's not talk about Monster cable please.
I am willing to try CAT5 or 6, and get my own results. Of course, I base my tests on something like
http://www.audioholics.com/reviews/...eoff/diy-cable-faceoff-results-and-conclusion
The problem, is that I can only compare with what I have now:
Regular 14AWG flexible multi-strand verdigris covered Monster cable, and 3 x 18AWG teflon covered mil-spec copper braided cable. Any other good value options ? 100$ for 2 x 8 feet is my current psychological maximum.

SET, simplicity of design and part count ... why?
It is really a mixture of taste, and what I have heard at audio shows for the past 10 years.
Let's start by my musical taste. All over the place, but I prefer jazz and old rock. I love jazz female vocals, Stacy Kent, Anne Ducros, Sarah K, Norah Jones; also Pat Metheny, Avishai Cohen, Jaco, Charlie Haden; ELP, Genesis, Yes.
I also love Under Byen and Within Temptation, but the recording quality is not there, so I listen to it in the car. Electro-acoustic music !... but not on a SET. Anyway ... all over the place :rolleyes:
(like my previous statements :D)
When I listen to my favorite tracks at the audio shows, I rarely hear a better sound then my low powered SET. When I do, it is often a low power SET from a small company, or the exceptional Nagra system http://www.nagraaudio.com/highend/index.php
The price is out of my range. Most of the other systems / components at these shows are overpriced, poorly matched and aimed at people with a lot money. Kef muon ... what a joke. I did not like the sound or the price. Maybe the looks if I had a 1 million dollar house to put them in.
To recap, with my musical taste, experience and available $, SET has the best to offer. Simple designs, limited number of components, and quality of the components, is how I explain it.

Speakers ...
Same here, I prefer the sound of the fullrange, single driver speakers. What I usually like is BLH with Lowthers or Fostex drivers. + DIY is fun :)

I did not hear it all, I did not hear everything in the same context, I do not work in an audio shop ... So the only reference point I have is my experiences at audio shows, stores and at home. And I keep learning everyday :)

To return to the circuit examples, The axiom has no resistors, no caps. Only tubes and iron ... an interesting option. Of course, the quality of the iron and the multiple tubes add their own coloration to the resulting sound produced. Is it a good option? Only the people who really tried it or something similar can tell if it works, unless there is a design flaw.
 
#14

lowtherdream said:


300B at 5K why?
The 5K load line is in a more linear region (on paper).
I have never tried a circuit with a DHT and I was just really saying that most of the SET 300B designs use a 3K OPT and I am ready to try a 5K OPT to gain linearity, reduce harmonics, at the cost of power. ...

I think the change of the output resistance of the triode
and its influence in sound is an important aspect of triode sound.
A 5KΩ load gives an output resistance of only
1R6Ω at the 8Ω speaker terminal.
Thus you will loose this unique aspect of triode sound.
The ratio of speaker impedance to output resistance of the
triode amp should be 2 to max.4.("Damping factor")
A higher ratio provides you more power at higher
anode to cathode voltages (!).

If you want a high "damping factor", extremely low harmonics
and excellent linearity you should go and get a solid state amp.;)

Kind regards,
Darius
 
Re: #14

lowtherdream said:
I mean no caps or resistors in the direct signal path (Axiom). There are caps in the PS section to regulate the DC of course.

The Axiom reminds me of my very first amp doodle I did while
"looking out the window" in 6th grade (as the teacher used to call
it...)

I had just read RC-30 or RDH4 or something and thought I could
design the ideal amplifier by directly coupling a bunch of stages and
stacking the power supplies (only in my naive state I thought I could
make a 1000 watt amp by stacking maybe 5 or 6 stages!) I never
told anyone about it before because I was embarrassed by my
naivete'. Later I calculated the voltages involved and realized my
mistakes.

The Axiom needs a 900V power supply... quite reasonable by
comparison. I guess I was just 40 years ahead of my time ;-)

Also, the goal of removing all resistors and capacitors from the signal
path is questionable. Particularly as Darius says replacing linear
resistors with nonlinear diodes out of a quest for "purity".

I have found over the years that a quest for excellence yields
better results than a quest for purity. Extreme purity is achieved at
he expense of balance in design, and balance in design usually gives
the most excellent overall results.

Having said all that, I just built a DC 300B amplifier with stacked
supplies and have already though of a better, simpler way to
achieve the same results with a simpler topology (on paper...)


oldeurope said:


I think the change of the output resistance of the triode
and its influence in sound is an important aspect of triode sound.
A 5KΩ load gives an output resistance of only
1R6Ω at the 8Ω speaker terminal.
Thus you will loose this unique aspect of triode sound.
The ratio of speaker impedance to output resistance of the
triode amp should be 2 to max.4.("Damping factor")
A higher ratio provides you more power at higher
anode to cathode voltages (!).

If you want a high "damping factor", extremely low harmonics
and excellent linearity you should go and get a solid state amp.;)

Kind regards,
Darius

True, but I also find it depends a lot on the speaker and room setup.
I have a multi-tap OPT and have tried a number of settings with my
JBL 4320s. At low damping factor (~3) I hear boominess at the
woofer Fs and the box/port node. It might be perceived as "bass
extension" by some but it sounds like flabby bass to me. Things
tighten up nicely at DF = 5 to 8 with no loss in emotional impact
and only a small loss in sound pressure level.

I have yet to do the same experiment with OBs or horns, but I would
expect different results with different loudspeakers.

Cheers,

Michael
 
Re: After the first watt

Michael Koster said:
Paul Joppa suggests enough power to achieve 102db peak at your listening position. My own experience bears this out as a minimum for a wide range of music. If you like listening to Santana etc, at impactful levels you will want at least 3db more headroom.

http://www.audioasylum.com/scripts/d.pl?audio/faq/sets.html

If you listen to one watt at 93db, 102db gives you 9db of headroom for dynamic peaks in the music. That would require an 8 watt amplifier.
I agree that 102-105dB peak at the listening position is a good target to aim for. However that link is for speakers at 1m I would guess based on the context of what I read. You'll need to add about 9dB to get listening position figures.
 
lowtherdream said:
I mean no caps or resistors in the direct signal path (Axiom). There are caps in the PS section to regulate the DC of course.

Lowtherdream,

Have you ever built a tube amp?
My guess is no and it seems you have a adopted a philosophy and seem intent upon carrying it out, based upon little experience.

If you want a well thought out 300B amp with no caps in the signal path, then build this:
http://www.nutshellhifi.com/triode1.html

I have built an Amity, a Grand Amity (813 triode outputs - the 50W amp I mention in an earlier post) and many other tube amps. If you wish to follow the no-cap line, you will not do better easily. At least not with a lot more knowledge than you seem to have.

For an SE 300B design build Thorsten's Legacy (google it). A modern approach to the WE91.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.