Jump to content
The Klipsch Audio Community
Sign in to follow this  
Jeffrey D. Medwin

KT88 Direct-Coupled Design

Recommended Posts

34 minutes ago, Jeffrey D. Medwin said:

Look, not two-dimensional, but a three-dimensional SE amp-build is developing

 

… o O (Gallium Arsenide chip's on a grand scale ?)

Share this post


Link to post
Share on other sites

OK, I will go to the standard type / font size at your suggestion, and see how it feels.   I have only had this new Weller D650 gun since 3-19-19, and this will take a bit of practice for me to get it really good.  You are viewing my third and fourth joint on that choke, using the new-to-me  D650.   Its quite a bit different soldering two 12s and a 14 AWG, , or in this case with the choke,  two 16s to a thin stock lead wire, as this called for.  I knew someone would chastise me for that solder joint, and you did so nicely.  Thanks.

 

 

Share this post


Link to post
Share on other sites
6 minutes ago, windashine said:

 

… o O (Gallium Arsenide chip's on a grand scale ?)

I don't know what that is or means.  Am I missing something ??  What am I missing, sir. ?

 

Share this post


Link to post
Share on other sites

… o O (crystal structures laser laminated between overlaid foil path's maybe) 

Share this post


Link to post
Share on other sites

This is not the first time I've seen or heard about the use of such chokes, Jeffrey.  There is another designer with whom I'm familiar that has a similar philosophy, and prefers to avoid higher values of inductance and DCR for the same reason:  improved sonic performance.

Share this post


Link to post
Share on other sites
29 minutes ago, erik2A3 said:

This is not the first time I've seen or heard about the use of such chokes, Jeffrey.  There is another designer with whom I'm familiar that has a similar philosophy, and prefers to avoid higher values of inductance and DCR for the same reason:  improved sonic performance.

 

Who??

Share this post


Link to post
Share on other sites

I would prefer to just leave my post as it is, Jeffrey.  It was many years ago, and mentioned in relation to a single-ended triode amplifier.

Share this post


Link to post
Share on other sites

P1010028.JPG

 

Okay... now you're just trying to pull our leg

Share this post


Link to post
Share on other sites
On 4/7/2019 at 11:37 AM, Schu said:

P1010028.JPG

 

Okay... now you're just trying to pull our leg

 

 

No, thats legitimate.  Do you DIY build??  

 

My soldering skills, the second and third joint I solder with a brand new-to-me 300 Watt iron,  are NOT important at all in this thread, IMHO Schu.  Why belabor it??  Why divert, criticize, throw the thread off-course???

 

Whats really important in that picture,  is particularly (a) below, also b) and c).

 

a)  to recognize a previously overlooked but better way to transfer pulsed-energy efficiently with minimal loss, inside an audio amp,  (  which is what music consists of, not sine waves,  pulsed energy,  odd waves. )

 

b) Also " nice " I think, is the idea of showing on-line  ( perhaps, likely ) audio's first-ever direct coupled KT88  build.  ( For sure, with a LSES supply ).

 

c) Also of interest will be new amplifier  " design ideas  " shared,  ( such as 6 Ohm DCR chokes ) to positively influence DIYers, and hopefully enable some to build better amps for themselves, Schu.

 

I think the above, a) through c),  is the guts of this.

 

Jeffrey

 

 

Share this post


Link to post
Share on other sites

Jeffrey,

Thanks for sharing your built here.

I  also designed and built my own amplifiers.

i choose to use 2 parallel 807 that I first operated as single ended pentode, and later converted to pseudo-triode mode.

 

What I like is your decision to operate your KT88 at 62% of max plate dissipation.

I choose 75% in my case, and i am close to 7000 Hours on my Russian 807, and still going strong.

Using 100% of max plate dissipation, to extract the last watt, is heavily paid in short  tube life.

 

What is your computed output impedance?

 

dan

Share this post


Link to post
Share on other sites
5 hours ago, Danieln said:

Jeffrey,

Thanks for sharing your built here.

I  also designed and built my own amplifiers.

i choose to use 2 parallel 807 that I first operated as single ended pentode, and later converted to pseudo-triode mode.

 

What I like is your decision to operate your KT88 at 62% of max plate dissipation.

I choose 75% in my case, and i am close to 7000 Hours on my Russian 807, and still going strong.

Using 100% of max plate dissipation, to extract the last watt, is heavily paid in short  tube life.

 

What is your computed output impedance?

 

dan

 

Hi Dan,

When you relax the dissipation of a Finals tube, you beneficially eliminate a nasty-sounding-to-me " thermally stressed sound ".   Unconscious designers, and some amateur DIYers, will tell me " the hotter I run the tube, the better it sounds".  Which it does initially IF you have a "typically poorly designed" tube amp in the mix, with a normal power supply and poor wiring.  The designer needs to " hotrod" the Finals tube, to "make up" for everything else in the amp, that is usually highly mediocre.  

 

But I always like to ask myself...for how LONG will this amp, with Finals operating at or near Maximum rated plate dissipation, sound good??  The answer is " Not too long at all ". 

 

How long would my MiniVan last if I ran it continuously at 105 MPH, versus 62 or 65 MPH ??? 

 

The TRICK in really GOOD amp designing, is to get a Finals tube operating at close to a Golden Ratio dissipation, and have the amp STILL sound fantastic, because of superior power supply design, parts choices, wiring, and general layout.  In the last dozen years, when DIYing amps, this has been what I try for..... and what I hope to show in this build thread.  

 

Daniel, I do not even know how to compute output impedance.  Hows that !!!

 

Many of the Formulas that an EE will spout out verbatim, and rely upon, mean very little to me.  When I take the time to answer Maynard's post on Power Supply Chokes, you and many others will see that a LSES supply is very different from what most people have and use presently in their tube amps.  Conventional EE techniques and 100 year old formulas are not particularly adhered to by me, for very good reasons ( most importantly to me, with superb results !! ).  

 

Some general comments on this  direct coupled KT88 build.

 

(1) Besides the LSES power supply, I have

 

(2)  a separate supply for the G2 B+, and

 

(3) the screen's B+ will be SHUNT regulated. with a simple 100 Watt wire wound draw-down resistor,  KISS rules, just one part !! 

 

(4) Unlike most EEs and their Formulas, I will have NO NEGATIVE FEEDBACK in my Pentode amplifier, which in itself, is highly unusual to the average bloke " into tubes and Pentode Finals".  I hate what negative feedback in amps does..... to the music, what NFB sounds like.

 

This amp will get built and listened to, long term, as I describe it.   

 

(5) My KT88 G2 Pentode B+ voltages will be much different than everyone else's, and I can only tell you " it will be more triode-like in operation." 

 

(6)  My concern is to design this amp to power my loudspeaker  - for the first 1/4 to 3/4s of a Watt,  which is all the power my 8AWG Mil Spec rewired VOTT A7-8 speaker's will need.  This KT88 tube amp does more power, yes, sure, but it is not important to me.

 

(7) At a later time, after living with the amp long term, I plan to convert it to what you would call pseudo triode, connect G2 to the plate, and listen to that. 

 

Since no one I know of has ever designed and run a KT88 in Pentode, precisely as I plan to, Daniel, there is NO way to predict in advance, what operating mode will sound best !!   If I wasn't curious to design, build and hear this,....... I DO own nine brand new JJ 2A3-40s on my shelf, and two blank welded steel chassis.  I could have done monoblock DC JJ 2A3 - 40 amps.  This KT88 tube design, however, even as a stereo amp possibly may beat that ..... I have to find out !!!!

 

Ohh, for outputs, I would like to have used 3.5K, but I have Magnequest DS-025s, a great sounding output, and with my LSES supply, and general execution, and triode-like Pentode operation, I can get away with the DS-025, whereas the typical tube amplifier build perhaps can't.

 

 

Share this post


Link to post
Share on other sites
On 4/5/2019 at 3:40 PM, tube fanatic said:

When you have time, please clarify this issue (and check my calculations as I did them in a hurry) and I'll try to post other questions as my extremely limited time permits.

 

Best regards---

 

 

Maynard

 

it'll never happen Maynard...  all the pontificating of design superiority and other "innovative yet unexplained or immeasurable design" is pure crap...

sooner or later people on the Klipsch site will recognized this as total BS as others have where Medwin has been banned for propagating the same foolishness

Share this post


Link to post
Share on other sites

A single ended pentode final stage without feedback is going to get you a high output impedance.

This is of no concern if your speaker impedance curve is flat, which is not the case for older Klipsch speakers.

Since you are interested in the first watt, operation in pseudo-triode mode will result in a much lower output impedance, as I now use in my amp.

This is just my humble opinion, to each his way in this fun hobby.

 

Dan

Share this post


Link to post
Share on other sites
21 hours ago, Schu said:

P1010028.JPG

 

Okay... now you're just trying to pull our leg

The proper way to create a good solder joint is to make a solid mechanical connection first.  At that point you can solder the pieces together.  This looks as if the wires were just laid next to each other and a blob is holding everything in place.  The same is true when wiring a house.  Make a good twisted mechanical connection and then install the wire nut.  Don't lay two pieces of wire next to each other and rely on the metal coil in the wire nut to pass current.

  • Like 3

Share this post


Link to post
Share on other sites
15 hours ago, Danieln said:

A single ended pentode final stage without feedback is going to get you a high output impedance.

This is of no concern if your speaker impedance curve is flat, which is not the case for older Klipsch speakers.

Since you are interested in the first watt, operation in pseudo-triode mode will result in a much lower output impedance, as I now use in my amp.

This is just my humble opinion, to each his way in this fun hobby.

 

Dan

 

 

Hello Dan, 

 

Yes, what you say is technically correct.  Your humble opinion is correct, up till now !!

 

The real question is, can I design and build an amp with the sonic advantages of a KT88 Pentode, and not have to revert to hearing a Negative Feedback Loop, which I detest sonically, as it is a music-destroyer to my ear !!  Will it play my speaker??   This will be the essence of this new amp build!!

 

I am well versed in the sound of real triodes Dan.  I built my first-ever amp in 1978, a Push-Pull 2A3 DHT amp, from an AudioNote kit.  'Built many excellent all triode amps, with zero feedback, since then.  My audio nick-name for decades has been " drlowmu triode " ....as is my email address.

 

Most KT88 tubed amps sound good to me, even with their less than optimal internal builds.  The KT88 tube sounds good, and actually, subjectively to me, in some areas, better than even the best 2A3s ever made.  How and why?  Let me just say the extra grid in the tube, puts " more meat on de bone", than the best 2A3s.  The KT88 can play big and bold with authority in different ranges.  Yes, the pentode KT88 tube sounds different than a 2A3  triode !! 

 

Treating the KT88 tube in the unique way that I intend to do in this build,  (  outlined in my " just one post above ", points 1 through 7 ) ,  I will have an opportunity to compare it against the world's best sounding 2A3 mono block amps.   I have access to these amps, thankfully, once a year in Denver at the RMAF show.  The Manufacturer who builds these 2A3 amps I have referred to, and myself, do not know "for sure" how this KT88 amp will do.  However,  he tells me, this KT88 prototype certainly  " has the possibility " to out-play his own 2A3 tubed amplifiers.

 

All he and I want, is the best possible result in a tube amp, with zero feedback - to play back music on our speaker loads.

 

I am TOTALLY aware of the high impedance factor, that is " common knowledge", but everyone's cure, adding feedback, is unacceptable as it creates far more sonic problems than it solves !!   I would't care if it was a banana peel that "did it", as the Finals device!!  Few prejudices.  We shall see...or in this case, ......hear !!

 

Thank you - for posting so nicely Daniel.

 

Jeffrey

Share this post


Link to post
Share on other sites
26 minutes ago, CECAA850 said:

The proper way to create a good solder joint is to make a solid mechanical connection first.  At that point you can solder the pieces together.  This looks as if the wires were just laid next to each other and a blob is holding everything in place.  The same is true when wiring a house.  Make a good twisted mechanical connection and then install the wire nut.  Don't lay two pieces of wire next to each other and rely on the metal coil in the wire nut to pass current.

 

 

You are 100% correct sir !!  Most people know this....I even do.

 

But that can't be done easily once the choke is already made, and one seeks to minimally invade the unit. 

 

Lets get off soldering, and address the fact that 98% of the people who read this thread, are seeing SIX Ohm chokes with 13 AWG equivalent mil spec wiring used in an audio amp, probably for the first time in their lives. 

 

Ponder that.............................why was it done ? 

 

That, interestingly, importantly,  gets discussed up here,  .....this week.

 

Thank you for posting.

 

Jeffrey

Share this post


Link to post
Share on other sites
On 4/6/2019 at 10:09 PM, Jeffrey D. Medwin said:

Lately, I even set my car's cruise control at 62 to 65 MPH, a Golden Ratio of my MiniVans' maximum speed.  Gets better mileage than cruising at 70 MPH plus !!

Fuel mileage has nothing to do with golden ratios rather the coefficient of friction or "drag" at different speeds.  You'd get even better fuel mileage at 55 mph.

  • Like 1

Share this post


Link to post
Share on other sites
On 4/5/2019 at 9:50 PM, Jeffrey D. Medwin said:

Just always remember this fellas: " The bigger the glob, the better the job ." 

 

 

"Tune for maximum smoke"

Share this post


Link to post
Share on other sites

Jeffrey,

 

I kind of agree with you on the negative feedback. The better the open loop, the less correction required.

So, the solution regarding the high output impedance effect Vs Speaker load is to have a flat impedance curve speaker.

Problem solved without feedback :-)

 

dan

Share this post


Link to post
Share on other sites

 

The issue of using nfb with pentodes/beam power tubes has raged for a very long time!  Jeff, have you tried using a small amount of voltage fb at the output tube only?  I find it to be extremely effective without any deleterious audible effects.  Of course, at only 0.25 to 0.75 watts out, it may be a moot point as the distortion at such a small percentage of maximum power output is going to be low.

 

Another option which may be worth considering is a filter across the opt primary to reduce some of the high frequency issues created by the varying speaker impedance.  

 

In any event, I look forward to seeing what the amp looks like on the 'scope when it's all done.  It's always fun to correlate what we hear with what we see when looking at square waves.  In fact, once familiarity is gained with that method, it's possible to design an amp simply by looking at its appearance on the 'scope as you will know what the resultant sound will be.

 

Maynard

'scope pic.JPG

  • Like 2

Share this post


Link to post
Share on other sites
Guest
This topic is now closed to further replies.
Sign in to follow this  

×
×
  • Create New...