K&K Audio / Lundahl Transformers

Welcome! Need support, you got it. Or share youe ideas and experiences.

Return to K&K Audio / Lundahl Transformers

K&K Audio / Lundahl Transformers

Karna arrives!

66.167.253.37

[ Follow Ups ] [ Post Followup ] [ Thread: Display Email Next ] [ K&K Audio / Lundahl Transformers ]

Here's the latest web page with pix of the new Karna amplifiers and some quick comments about how they were built (many many thanks to Gary Pimm) and the design considerations that went into them. We all had a great weekend, with Gary Pimm, Gary Dahl, Bud Purvine, and myself listening to the new arrivals.

Warning to those with slow connections, the pix are pretty big, and take a while to load.

Karna Page (Open in New Window)

Follow Ups:

Re: Karna arrives! - cap type - AndyN 13:53:00 10/05/04 (0)
I love the circuit and I have a few questions. Particularly regarding the caps used. What type do you use for the 40mfd in the driver and output stage and also the 10nfd in the ECC99 stage.
Andy
Re: Karna arrives! - SteveBrown 10:38:51 09/15/04 (1)
Wow, great work! In looking at the schematic I see you used the LL1635 5ma IT. Why the 5ma version? In this application, could you have used the PP version? Just curious.
- Re: Karna arrives! - gdahl 14:51:25 09/23/04 (0)
  Hi Steve,
  The 5 ma version was chosen for this position because it reduces the effect of any residual imbalance current through the primary. Actually, I used the 1635/pp version in my Auroras (now converted to Amitys) and have had no problems--my ECC99's have produced extremely low offset.
  Anyway, the 5 ma version buys some offset tolerance, at the expense of a certain amount of primary inductance.
  Gary Dahl
  "So many tubes, so little time..."
Re: Karna arrives! - MikeT 04:39:53 09/15/04 (0)
Hello Lynn, the amp looks great.
I've always wondered why you don't shunt reg the output stage. Can you comment on that?
I'm also pretty close to buying a couple of 6W6's and giving them a spin......
Mike
Re: Nice work Lynn et al... - Jose R 13:16:30 09/11/04 (5)
the key is good sources -> Dave Slagle Autoformer -> 3 stage amp. No two stage amp with active preamp comes close...IMHO.
Hey listen, who did your power transformer? It seems to have dual CT secondaries.
With only a small input cap and a 40uF cap after the choke hows the ripple on the output stage?
And with all AC filaments, any noise on Gary's modified Heresys? I read in your write-up you wrote that if hum was an issue you would resort to DC, but I assume its a non-issue? BTW, have you tried it with your modified Chorus IIs?
I'm still working on my ideas for a PP design and would like to better understand you value/design choices.
After building my SE amp, I can't contemplate using under 100uf in the B+, or AC on the filaments, regardless of topology.
And building a separate chassis for the power supply is "de rigeur," but for a PP amp it means a four chassis design. Big, big, big...
Looks like a great group effort.
- Comments ... - Lynn Olson 15:37:33 09/12/04 (4)
  "Hey listen, who did your power transformer? It seems to have dual CT secondaries."
  Bud Purvine of O-Netics lives right here in the Seattle area. Bud's proximity to the project made it possible to discuss the design and performance requirements before ordering any parts. It was Bud that insisted on a separate heater transformer for the quartet of 6CJ3 damper diodes, which isolates the heaters from the very high peak currents and voltages in the B+ circuit. You'll see Bud's e-mail address on the schematic for the amplifier.
  "With only a small input cap and a 40uF cap after the choke hows the ripple on the output stage?"
  Very low. Remember, the first cap *is not a filter*. To build a lowpass filter, the first element must be a resistance or a choke. Rectifiers don't qualify. So a conventional CLC pi-filter power supply only has two elements of low-pass filtering - the LC formed by the choke and following capacitor. The first cap, if any, shapes the peak currents and voltages flowing through the rectifier. This is very evident using PSUD II, where you can examine the peak currents and voltages going through the rectifiers and power trans secondary, and see the effect of gradually increasing the first cap from 0.1uF to 20uF. Very illuminating, and less dangerous than measuring the real thing.
  Gary Pimm has measured the noise and distortion spectrum, and the overall noise floor is below -120dB. The hum-nuller circuit (a joint venture between GPimm and myself) reduces 60Hz by 30dB, leaving behind a 3 millivolt residual of 120Hz (as measured at the speaker terminals). GPimm and I have no tolerance for system buzz, which is usually a sign of a ground loop somewhere and associated unwanted capacitive coupling.
  "And with all AC filaments, any noise on Gary's modified Heresys? I read in your write-up you wrote that if hum was an issue you would resort to DC, but I assume its a non-issue? BTW, have you tried it with your modified Chorus IIs?"
  Not yet. GPimm's new speakers are pretty amazing - the only improvement remaining is increasing the slope of the tweeter crossover to a low-Q (about 0.5) 12dB/oct filter, which should substantially improve the Heil power-handling, make the midrange sweeter and give even more dynamic range. I can say this, no Klipsch speaker ever sounded as good as the combo that GPimm has right now.
  "After building my SE amp, I can't contemplate using under 100uf in the B+, or AC on the filaments, regardless of topology."
  SE follows a different set of rules than Class A PP triode. PP noise cancellation gives a 26-30dB reduction of power supply noise, very welcome to be sure, but power supply colorations are more annoying since the overall amp distortion is much lower as well. The remaining PP/supply colorations are more noticeable since they no longer have the sweet sound that SE colorations tend to have - what's left in a PP amplifier is a nasty-sounding residue of what didn't get cancelled by the PP balance.
  This means the power supply has to be very carefully designed. What might be kind of grainy and mushy sounding in SE is harsh and brittle in PP, since that's the character of the uncancelled residue - also, the 2nd harmonic masking isn't there to help smooth out the colorations and make them more agreeable. 2nd harmonic distortion can conceal a multitude of sins elsewhere in the amplifier.
  But the problem isn't insoluble. You just have to carefully avoid anything that generates hash and HF noise - most especially, solid-state rectifier bridges that look directly into high-capacitance "filters". These generate a *lot* of HF hash that is radiated into the air, AC power cord, and the rest of the circuit.
  The main B+ supplies are also carefully designed to have a minimum of HF content, which is much more audible than 60/120 Hz hum, especially in a all-PP amplifier, which efficiently cancels most of the 60/120Hz hum. Both PSUD II and a bench spectrum analyzer or computer FFT analysis come in handy, since your goal is to keep HF hash and noise to a minimum.
  Although I'm trying to minimize amplifier coloration overall, I also want the inevitable residue (which will be there no matter what I do) to be agreeable sounding. I'd rather have a very slight 120Hz hum than any amount of hash or buzz.
  As it is, with GPimm's 92dB/metre speakers, the hum is only audible with your face right in the 12" speaker diaphragm - in fact, the faint mechanical buzz from the power trans is usually louder. Bud's working on the remaining buzz, which although very very quiet, is not quite zero yet.
  Lynn Olson
  - Karna Amplifier (Open in New Window)
  - Re: atachments coppied and pasted that I tried to send previously - George Louis 19:01:27 11/07/04 (0)
    Dear Mr. Olson:
    
    George S. Louis
    1573 Kimberly Woods Drive
    El Cajon, California 92020-7261
    Telephones: 619-401-9876 & 888-588-9542
    Fax: 619-588-5524, E-mail
    
    PROFESSIONAL EXPERIENCE
    
    Sales Manager, Customer Service Manager, Director of Software Engineering, Product Management, Marketing, and Technical Writer
    DIGITAL SYSTEMS & SOLUTIONS, El Cajon, California 1994 � August 2004
    * Project manager for the development and manufacture of Finyl�, a transparent coating that
    improves data recovery from laser read media. Finyl is in worldwide distribution and received
    a "Golden Triangle Award" from a "High End" European high fidelity standards organization.
    * Wrote and illustrated technical instruction manuals, repair manuals, promotional literature, and
    "The Finyl� Word a White Paper on the Art and Science of Optical Impedance Matching of
    Laser Read Media" that was instrumental in securing a major international distribution contract.
    * Project manager for the design and manufacture for of Blue�, Micromat Gold�, and
    Musicmat Gold�, constrained layer dampers, that further enhance the playback of CD media.
    * Project manager for the Virtual Vault�� Cyber Vault� data base and the Cyber Net�
    communications network. Managed the12 software engineers that work on the project.
    General Counsel, Contract Administrator, and Sales Manager 1990 -1994
    UNITED STATES PHOTOVOLTAIC, CORP., San Diego, California
    * Incorporated the company.
    * Negotiated successfully with Kyocera America Inc. for the company�s seed money.
    * Managed the company�s 35 software and firmware engineers.
    * Designed and sold components and integrated on-line power systems as well as performed
    customer service.
    * Wrote system setup instructions, operation manuals, and component descriptions.
    Contract Administrator and General Counsel 1985 -1990
    ELECTRON INCORPORATED, Escondido, California
    * Negotiated, wrote, and administrated contracts with suppliers and customers.
    * Re-negotiated contracts with customers including the FAA to resolve performance disputes.
    * Managed the company�s 44 software and firmware engineers.
    * Handled the company�s legal work.
    * Authored system setup instructions, operation manuals, and component descriptions and performed
    customer service in person and on the telephone.
    
    George S. Louis, Page 2
    Manager 1980 -1985
    UNIVERSAL FINANCIAL, CORP., San Diego, California
    * Inaugurated the two San Diego offices and managed 36 sales and clerical staff employees,
    overseeing loans on real property and the selling of trust deeds. They were the
    two most profitable of the company�s 40 offices over 93% of the time.
    * Loan officer with the highest personal production out of all 244 in the company.
    * Performed extensive customer service in person and on the telephone.
    Attorney 1972 -1980
    LOUIS & DuVAL, San Diego, California
    * General practice covering: criminal, business, real estate (Licensed Real Estate Broker),
    contract, corporate, domestic relations, mediation, and personal injury law
    (member of the California bar).
    * Arbitrator for Superior and Municipal courts.
    Special Agent Intelligence Division, IRS 1969 -1970
    Internal Revenue Service, San Diego, California
    * Special Agent Intelligence Division.
    * Investigation of income tax fraud.
    * Successful prosecution of tax fraud.
    Student Intern 1960 -1963
    NAVAL ELECTRONICS LABORATORY, San Diego, California
    * Mathematical and statistical analysis.
    * Physics and engineering.
    * Computer analysis of ASW data requiring security clearance.
    EDUCATION
    Software Engineering Certificate, San Diego State University, January 1998,
    covering: PSPICE circuit simulation and filter analysis, UNIX, C, C++,
    object oriented programming, JAVA, Visual J++, Windows 95, Access 97,
    HTML, and embedded systems programming.
    Certificate in Basic Electronics, Brunswick Technical Institute, 1985.
    Valedictorian.
    J.D. in Law, University of San Diego School of Law, 1969.
    Book Award in Medical Jurisprudence.
    B.A. in Mathematics, San Diego State University, 1963.
    Discovered new recursive power functions for generating large prime numbers.
    MILITARY SERVICE
    United States Air Force. 1959-1965.
    Personnel Specialist with security clearance. Honorable discharge.
    COMMUNITY INVOLVEMENT
    Executive Board of the San Diego Chapter of the Audio Engineering Society.
    Life Time Chairman of the San Diego Audio Society.
    Pro Bono Law Practice.
    George S. Louis, CEO
    Digital Systems & Solutions
    1573 Kimberly Woods Drive
    El Cajon, California 92020-7261
    Telephone: 888-588-9542
    
    Summary of My Background in Audio
    
    My love of music began in early childhood and has grown steadily over the years. In grammar school I studied the violin, guitar, and piano and was concertmaster of my high school orchestra. By 1962 my devotion to hi-fidelity turned me into an avid audiophile hobbyist. I worked my way through law school, buying and selling new and used audio and video components. I�ve assembled, setup, and designed, speakers, crossovers, and complete sound systems for both friends and customers.
    
    In January 1994 I was able to convert my favorite hobby into a highly profitable business by founding Digital Systems & Solutions to market some of my best audio and video enhancement products. At D S & S I invented, manufactured, promoted, and secured worldwide distribution for audio and video products that were considered so innovative that they defined an entire new category of audio video enhancement tweaks and have spawned many copy cat products. Two of my products, Finyl� and Micromat Gold� won critical acclaim and numerous accolades including the Golden Triangle Award from a prestigious European hi-fidelity standards organization. I am currently developing new products including ClearSine Technology�, ClearBit Technology�- UltraLog�, and the PowerClear Sineverter� that results in higher fidelity to the recording and reproduction of audio and video media.
    I am chairman of the San Diego Audio Society and a member of the Executive Committee of the San Diego chapter of the Audio Engineering Society. My views on audio were the subject of an hour long interview on the weekly program "In Fidelity" which aired on KPFK FM in Los Angeles. The "L.A. Audio File", "The Audiophile Voice", and "Stereophile" magazines credited me for discovering the "Louis Effect", a phenomenon that affects the playback of laser read media.
    For thirty-eight years as an audiophile hobbyist and entrepreneur I�ve been dedicated to excellence in the reproduction of music and sound and sharing that experience with my fellow audiophiles and customers.
    
    Sincerely,
    
    George S. Louis
  - Re: Lynn ... - Jose R 20:51:03 09/12/04 (2)
    many thanks for your generous reply.
    "Bud Purvine of O-Netics....Bud's e-mail address on the schematic for the amplifier."
    Should have looked more closely at your schematics...was distracted by your circuit ;-)
    "SE follows a different set of rules than Class A PP triode."
    Yes, I keep reading that, but having difficult time wrapping my head around the less aggressive power supplies used in PP designs. I took a brute force appoach with my SE amp, and have convinced myself the same should be done in PP regardless of the inherent noise cancellation.
    "But the problem isn't insoluble. You just have to carefully avoid anything that generates hash and HF noise - most especially, solid-state rectifier bridges that look directly into high-capacitance "filters". These generate a *lot* of HF hash that is radiated into the air, AC power cord, and the rest of the circuit."
    Yes, true. I have 5 Schottky rectifier bridges (neg bias, filaments) and a pair of Hexfreds on each HT rail with tube rectifers. (Insert rolling eyes here.) I have no hum on 101dB Cornwall IIs, but noticable hash/buzz when I put my ears right next to the tweeters. From a foot away no problem. Been trying various "fixes" over the past months...
    "As it is, with GPimm's 92dB/metre speakers, the hum is only audible with your face right in the 12" speaker diaphragm - in fact, the faint mechanical buzz from the power trans is usually louder. Bud's working on the remaining buzz, which although very very quiet, is not quite zero yet."
    
    Exactly, although very quiet, it should be dead quiet. Good luck.
    For grins, here's the schematic of my PP amp. Whether I build it is another story. I've gone through various ideas, especially after an e-mail exchange with Gary Dahl a little while ago. Its not up your alley, probably, but trying Allen's fully differential ideas might be worth a listen.
    - SE and PP design differences - Lynn Olson 02:59:32 09/13/04 (1)
      It's worth emphasizing that (well-designed) SE and PP have a fundamentally different character sonically. The error function (distortion + noise) of a SE amp is heard directly, unaltered by an cancellation mechanism. Thus the primary task is arranging the inevitable residue so it is sonically pleasing, or more accurately, able to fade into the fabric of the music without drawing attention to itself as an artifact of electronics.
      The error function cancellation of PP is both a blessing and curse. It's a blessing to get 26 to 30dB of overall reduction, but a curse in leaving an ugly sounding, HF-emphasized residue. So the truly worst-case PP amplifier would have incomplete balance over the waveform (not Class A), solid-state rectification without any attempt to snub the HF hash, feedback mechanisms that are too slew-limited to address Class AB crossover errors, low-grade dielectrics on the circuit boards, even lower-grade electrolytic B+ caps, and tubes that have lots of 3rd harmonics.
      Of course, this describes 95 to 99% of high-end tube amps in the stores, so what people think of as "PP coloration" amounts to little more than poor design optimization - which is a nice euphemism for cost-cutting in what are already overpriced boutique products. Sorry for the harsh picture, but I calls 'em as I as sees 'em - and I don't have to answer to a marketing manager these days.
      Since the inherent cancellation properties of PP emphasize HF colorations from distortion and noise, it's worthwhile to get 3rd harmonic as low as possible, and chase buzz and noise out of the supply, and avoid altogether sources that generate hard-to-filter switchnoise artifacts. Damper diodes come in handy here, since they have a very soft on-off transition (spread over 15V) , yet are the most efficient tube rectifiers (low "on" resistance). The transition characteristics of the rectifiers, choice of B+ filter system, and power supply layout all have a critical influence on the shape of the HF noise spectrum - which is also going to be program-modulated, depending on current draw of the output stage.
      Keep in mind that practice borrowed from the 1950's, or even worse, modern high-end amps, may be the *worst* in terms of a well-behaved noise spectra. Most audio engineers wouldn't even know how to turn the knobs on a spectrum analyzer, much less minimize RFI and EMI emission from the power supply.
      - Power supply layout - arend-jan 13:00:59 09/16/04 (0)
        
        In Reply to: SE and PP design differences posted by Lynn Olson on September 13, 2004 at 02:59:32:
        
        Lynn,
        You seem to be stressing more and more the importance of the power supply in your PP designs. Would you mind sharing a pic of the inside of the power supply, so we can all take a hint in good layout techniques? There does not seem to be much information about this available on the internet, so it would be much appreciated! (or at least by me :-)
        -- AJ
Re: Karna arrives! - KevinC 17:12:08 09/10/04 (1)
You NW guys have too much fun!! And that even more remarkable when you consider the gray winter you have to endure!
My fully push-pull 6SN7GT voltage amplifier, 2A3 driver, KR32BX PP amp is still going strong. I continue to be pleased with the authority it commands driving the Ariel Mark IIs. It also has been transparent enough to provide the goods with every improvement in the K&K Audio phono stage and RAKK dac.
A few more of you DIYers should take a header and try a push-pull amp like this or the Karna. This is the not the push-pull sound of the 50s, 60s, 70, or 80s. Get push-pull right and you can have much of your sonic cake and eat it, too!

Kevin Carter
K&K Audio
www.kandkaudio.com
- Actually, winter and coffee keep us going! - Lynn Olson 03:43:07 09/11/04 (0)
  With the sparkling summers and fall, not much audio gets done - everybody rushes outside and plays! But during the winter it's time to hang out at bookstores, coffeeshops, go to concerts, and build audio.
  This summer was unusual, especially the freakish heatwave we got for a couple of weeks. Poor bro' Pimm got oven-roasted without any A/C in 90+ degree weather and the development version of these amps in his living room.
  I was 10 degrees cooler in the shadow of the Olympic mountains (the Nutshell pix was taken right here in Silverdale) but ampless until last weekend.
  What's interesting is Kevin's experience with thinned-out bass on his all-trans-coupled 2A3/VV32B amp didn't happen here. The bass power on this thing is downright frightening - it really could fill an auditorium with the right speakers. I suspect the fairly extreme power supply gets some of the credit for the stunning dynamics, although DHT drivers have a dynamic power and intensity all their own, leaving IDHT's in the dust. It's true, the driver in some ways is *more* important than the output tube.
  The amp is also a testment to *not* having an all-in-one tube for both input and driver. I can assure you, this amp is more transparent - and far more colorful and dynamic - than any two-stage amp.
  - Page One (Open in New Window)

This post is made possible by the generous support of people like you and our sponsors:

Parts Connexion

You can not post to an archived thread.