New Design, SE 6005 Directly Coupled Audio Amplifier

Status
Not open for further replies.
2-17-2021 REPORT

Installed and wired-in the new R-Core output transformers. These are wonderful !!

View attachment 54805

View attachment 54806

View attachment 54807

Military Spec wire is m22759/11. Multi stranded copper, silver plated, teflon jacket.

A second time, for your information : R-Core output transformers are different, and uniquely superior to all other types, as follows :

https://www.custommag.com/r-core-transformers

I have decided to name this new design " the REVELATION ", and am looking into a brass name tag design.

Thank you, for looking .
Jeff, Post your listening impressions once they have completed break-in. Would be interesting to note comparison with EI Vs R-cores.
 
Jeff, Post your listening impressions once they have completed break-in. Would be interesting to note comparison with EI Vs R-cores.
It takes a minimum of 100 hours, and typically 400 hours to break- in fully.

My plan is to use a CD player, on " Repeat ", and put load resistors on the speaker jacks ( disconnect the speakers ) and let the CD player and amplifier run for 17-18 days.

I have not heard every transformer in the world, but my initial reaction is very very positive.

I am pretty sure this will end up outperforming all existing output transformers, ( EI core, double C core, toroidal ) It has much lower distortion than all the others, due to it's intrinsic R-Core design. It " lures" the listener into the music, and plays melodically in a way that I think is impossible - for a non R-core.

These comments are after only about 11 hours of break in. I am sure of this initial opinion. I actually expect full break-in to reinforce the first impression.

Meantime, I am experimenting with extra filtering of the power supply B+ voltages, in addition to the existing double series shunting to the front end tube and to the 6005's G2 supply.

My recent design work has been productive. I will be using at least one more filter stage, in both of these two critical circuit areas. The added filter is an L/C ( choke - cap ) and the choke will be a small hash choke. The cap will be perhaps six film caps of different values in parallel, in other words, fully bypassed at all important audio frequencies, for full musical expression.

If I rebuild this amp into a larger stereo chassis, I might possibly decide to go to a double series L/C configuration, for both these B+ circuits. The filter to the front end tube and to the 6005's G2, may thus possibly become :

L/C, L/C, Shunt 1, Shunt 2 , .................... a quadruple B+ filter.

I need to simply listen to my reference music selections, and decide. Much is presently pending in this build and design. It is progressing quite well and " normal ".

Thank you for your interest.
 
Last edited:
Thanks for your comments... informative.

When you say
" It has much lower distortion than all the others, due to it's intrinsic R-Core design. It " lures" the listener into the music, and plays melodically in a way that I think is impossible - for a non R-core."

Is the lower distortion comment purely subjective ?

Or have you made measurements ?
 
It takes a minimum of 100 hours, and typically 400 hours to break- in fully.

My plan is to use a CD player, on " Repeat ", and put load resistors on the speaker jacks ( disconnect the speakers ) and let the CD player and amplifier run for 17-18 days.

I have not heard every transformer in the world, but my initial reaction is very very positive.

I am pretty sure this will end up outperforming all existing output transformers, ( EI core, double C core, toroidal ) It has much lower distortion than all the others, due to it's intrinsic R-Core design. It " lures" the listener into the music, and plays melodically in a way that I think is impossible - for a non R-core.

These comments are after only about 11 hours of break in. I am sure of this initial opinion. I actually expect full break-in to reinforce the first impression.

Meantime, I am experimenting with extra filtering of the power supply B+ voltages, in addition to the existing double series shunting to the front end tube and to the 6005's G2 supply.

My recent design work has been productive. I will be using at least one more filter stage, in both of these two critical circuit areas. The added filter is an L/C ( choke - cap ) and the choke will be a small hash choke. The cap will be perhaps six film caps of different values in parallel, in other words, fully bypassed at all important audio frequencies, for full musical expression.

If I rebuild this amp into a larger stereo chassis, I might possibly decide to go to a double series L/C configuration, for both these B+ circuits. The filter to the front end tube and to the 6005's G2, may thus possibly become :

L/C, L/C, Shunt 1, Shunt 2 , .................... a quadruple B+ filter.

I need to simply listen to my reference music selections, and decide. Much is presently pending in this build and design. It is progressing quite well and " normal ".

Thank you for your interest.

Jeff, Thanks for the detail update.
If I remember the shunt 1 and shunt 2 are for the driver tube and shunt 1 only for the G2 Finals. The finals are connected directly to B+ via the OPT. So how does these Shunt impact B+ filtering? Are they for regulating only the driver B+ supply?

Indiaears,
Imo distortion is a function of noise floor. If your ambient noise floor is low then you can subjectively hear distortions else not. Even during objective measurements if your noise floor is higher than your distortion then the distortion gets masked. This is indicated as grey in the distortion graph as measured by me in REW.
 
Thanks Hari.
In well designed modern day amplifiers ... even tube amplification, noise is atleast 6dB to 12 dB below distortion components.

Noise masks detail but in a generally unobtrusive way.

Distortion on the other hand is generally far more audible and often in an obtrusive manner.
 
Thanks for your comments... informative.

When you say
" It has much lower distortion than all the others, due to it's intrinsic R-Core design. It " lures" the listener into the music, and plays melodically in a way that I think is impossible - for a non R-core."

Is the lower distortion comment purely subjective ?

Or have you made measurements ?

Totally subjective. But when you read the URLs I gave on the R-Core, you can deduce what is happening easily !

Think about this : Unlike all other outputs, in a R-Core, the primary is wound on one leg of the R-Core, and the secondary is would on the other lag. And this Softone's R-Core winding is done in a "balanced" fashion.

In all other outputs, the primary and secondary are wound on top of each other !!! Compared to Softone's R-Core, that is highly undesirable engineering.

To gain efficiency, the better conventional output transformers use a technique called "interleaving". The " more interleaving the better ",..... so it is said. But there is an almost never-discussed and realized trade-off to interleaving, VS: the R-Cores' needing no interleaving.

Every interleave introduces opposite phase-current anomalies to the winding, and it confuses the original signal, such that all the non R-Cores only play gross information well, but in a confused and muddled way, VS a Softone output.

I am not so sure that some of these things can be easily measured. However, from the very first time five of us audiophiles heard an RW-20 , ( with only about 80 hours of break-in time on it, about 20 months ago ), it was very easy for all of us present, to hear how special sounding, and low in distortion, it played.

Since then, that amp's owner tells me the RW-20 has fully broken in and displays a lovely extended top end. He and I both agree, the RW-20's bass transient delivery ( I can already hear ) is lightning fast !!

I hope what I explained is more meaningful, understood, and useful to people who listen than " distortion dropped 25% from 0.04 % to 0.03%. !!!!

Please do note this, in both cases above, the RW-20's stock leads ( which are very decent ) have been wire augmented, as shown in this thread. These outputs are used in similar Directly Coupled two stage tube amps with 6 Ohm DCR chokes.

I hope this post helps people !!! Thank you for viewing this.
 
Last edited:
Very Interesting.

R Core transformers have been used for decades as mains transformers....

Any thoughts on why R Core transformers are not so commonly used in High End Valve Audio Power Amps ?

I am no expert in this, so if they are used, could you name a few brands & specific models ?
 
IndianEars, I have done an extensive search for 3+ months on Google and not found a single manufacturer other than Softone who does a commercial built of an R-core Audio OPT. If you are able to find then do let us all know. I waited for 6+ months for the Softone but even today they are not available after end of lockdown. Hoping it to be available probably next year.

What I could understand is - it's difficult to airgap in an R-core as you need to slice the core precisely and then join them to prevent saturation. I think this technique is unique only to Softone.
 
Very Interesting.

R Core transformers have been used for decades as mains transformers....

Any thoughts on why R Core transformers are not so commonly used in High End Valve Audio Power Amps ?

I am no expert in this, so if they are used, could you name a few brands & specific models ?

They are not used because people do not yet fully understand them, nor have they heard them in really good circuits.

Here is what happens - I can tell you this : a typical E.E. or technical audiophile type, will read " Inductance 12 HY ." on an RW-20 data sheet, and immediately totally "dismiss" and totally miss the point of the entire transformer. This is their highly unfortunate loss !!!!

For now, and for the third time in this thread, I will post this URL :

https://www.custommag.com/r-core-transformers

Here below are your reputable / proven, high quality multiple transformer design choices :


Remember, do not early judge the R-Core, we like 400 hours to fully break in. After 100 hours, you can get a good taste, of what it is. Mine has only 11 hours so far !! But the one I heard 20 months ago had 80 hours on it then.

IndianEars, I have done an extensive search for 3+ months on Google and not found a single manufacturer other than Softone who does a commercial built of an R-core Audio OPT. If you are able to find then do let us all know. I waited for 6+ months for the Softone but even today they are not available after end of lockdown. Hoping it to be available probably next year.

What I could understand is - it's difficult to airgap in an R-core as you need to slice the core precisely and then join them to prevent saturation. I think this technique is unique only to Softone.

They are available to India, you told me this today. But, with Covid shipping restrictions, it is a long non-air shipping time, did you tell me 4 to 6 months !!!

Well worth the wait ............... I say !!
 
Last edited by a moderator:
They are available to India, you told me this today. But, with Covid shipping restrictions, it is a long non-air shipping time, did you tell me 4 to 6 months !!!

Well worth the wait ............... I say !!
Non- air shipping is risky here as it can easily get lost in transit. So it's as good as not available for small consignment like an OPT.
 
Jeff, Thanks for the detail update.
If I remember the shunt 1 and shunt 2 are for the driver tube and shunt 1 only for the G2 Finals. The finals are connected directly to B+ via the OPT. So how does these Shunt impact B+ filtering? Are they for regulating only the driver B+ supply?
The shunt lowers the impedance of the supply, and it makes the amp sound better. It can be viewed in a way, as a passive regulator, stabilizing the B+ . Two shunts in series are what I prefer. It is essentially " double B+ regulation " of a brute-force nature, passive. It gives the tube amp more detail retrieval of the original incoming signal, and it keeps the music signal intact.

Harri, your amps' G2 supply uses a single high powered ( 50 or 100 Watt ) resistor, B+ to ground, to develop a shunt current. It is further filtered at the point of use ( the tube socket ) with a 10 Ohm DCR 320 mHY filter choke, feeding a small bank of film caps.

For this 2021 amp design, " the REVELATION " I am not only double shunting the driver tube's B+, but also double shunting the other ultra sensitive spot,.... the output tetrode's G2 B+. In addition, as of the last 24 hours, I have decided to configure both B+es with four-way filters, using a double L/C filter in front of the double series shunting .

So, "the REVELATION" 6005 2021 amplifier will use quadruple filtered B+es, to the Driver stage and to G2 of the 6005, using a quadruple B+ filter, like so :

L1/C1, ................. L2/C2, ................... Shunt 1, .......................... Shunt 2.

This audio amplifier is destined for my personal long-term use, and I will do what I can, circuit wise, power supply wise, to make it perform.

A little more on R-Cores :

The amps we heard 20 months ago, were possibly / probably the world's very best sounding 2A3 amplifiers ever made. ( Serious Stereo 2A3 amps, heard right at the Manufacturer's facility ). Very expensive, highest sound quality amplifiers. It was the Softone RW-20 amplifiers ( monoblocks ) that stood out from all the rest.

I just came across a comparison done in lower level tube audio gear, with Edcor E-I transformers, and similarly, the R-Core Softones ( push pull types ) were preferred.

Read about it here :

 
Last edited:
drlowmu

What are your views on Active vs Passive regulation, for the B+ supply ?
 
drlowmu

What are your views on Active vs Passive regulation, for the B+ supply ?
My opinion only, I much prefer passive filtering. Less of a thumbprint, less parts, and less noise.

A voltage reference tube, as part of any regulator can oscillate whereas a shunt resistor will return the B+ hash to ground.

Also, there is no substitute for thoughtful use of hash chokes in tube amps, to remove noise, stabilize an amplifier, and make it sound better. Hash chokes can have a very positive effect on performance !! They often could be used to clean up the output of solid state regulators, when found in tube amps. ( eg : SP3A-1 preamps benefit greatly. ) There are more subtle ways to use hash chokes to good effect. The new 6005 prototype already has six hash chokes in it, and I now plan top add four more in a upcoming rebuild !!

A shunt resistor must be chosen for its sonic characteristic. We do have quiet and great sounding shunt resistors in 2021. My 6005 amplifier uses modern Caddock " power tabs " as the resistor to ground. ( Types MP-820, MP-930 and MP-925. ) Other acceptable shunt power resistors I use are vintage Ohmite Brown Devils, Mills MRA-12s, and ARCOL HS series chassis mount power resistors - which are engineered for pulse duty.

Back to the subject of R-Cores : Kitamura Kiden, who invented the R-Core in 1978, has McIntosh ( USA ) as a customer. We can learn about this Japanese company, who also supplies Softone their products, here :

http://kitamura-kiden.co.jp/product/rcore/

I hope this post is of some assistance to FMs.
 
Last edited:
My opinion only, I much prefer passive filtering. Less of a thumbprint, less parts, and less noise.

A voltage reference tube, as part of any regulator can oscillate whereas a shunt resistor will return the B+ hash to ground.

Also, there is no substitute for thoughtful use of hash chokes in tube amps, to remove noise, stabilize an amplifier, and make it sound better. Hash chokes can have a very positive effect on performance !! They often could be used to clean up the output of solid state regulators, when found in tube amps. ( eg : SP3A-1 preamps benefit greatly. ) There are more subtle ways to use hash chokes to good effect. The new 6005 prototype already has six hash chokes in it, and I now plan top add four more in a upcoming rebuild !!

A shunt resistor must be chosen for its sonic characteristic. We do have quiet and great sounding shunt resistors in 2021. My 6005 amplifier uses modern Caddock " power tabs " as the resistor to ground. ( Types MP-820, MP-930 and MP-925. ) Other acceptable shunt power resistors I use are vintage Ohmite Brown Devils, Mills MRA-12s, and ARCOL HS series chassis mount power resistors - which are engineered for pulse duty.

Back to the subject of R-Cores : Kitamura Kiden, who invented the R-Core in 1978, has McIntosh ( USA ) as a customer. We can learn about this Japanese company, who also supplies Softone their products, here :

http://kitamura-kiden.co.jp/product/rcore/

I hope this post is of some assistance to FMs.
SUPER Post on MANY counts, including the selection of Shunt Resistors. Many Thanks!

I am surprised about McIntoch... They have always made much that their Output transformers are all wound in house.

Edit:
I did a quick Google check on McIntosh using Kitamura Kiden R-Core Transformers.

I have reason to suspect that this is not true .... :(

This is the Kitamura Kiden page that says:
"Our customers include major domestic Japanese companies such as Shimadzu, Toshiba Schneider Inverter, Toshiba Medical Systems, Hitachi Medico, and Hitachi Machine System. We also have many major overseas clients such as Mackintosh, GE, and Varian Medical Systems."

None of these are Audio Equipment manufacturers, and the spelling of Mackintosh, is completely different from the Audio Company McIntosh Labs.

A search for Mackintosh put out this page on Linkdin :)
 
Last edited by a moderator:
I did a quick Google check on McIntosh using Kitamura Kiden R-Core Transformers.

I have reason to suspect that this is not true .... :(

This is the Kitamura Kiden page that says:
"Our customers include major domestic Japanese companies such as Shimadzu, Toshiba Schneider Inverter, Toshiba Medical Systems, Hitachi Medico, and Hitachi Machine System. We also have many major overseas clients such as Mackintosh, GE, and Varian Medical Systems."

None of these are Audio Equipment manufacturers, and the spelling of Mackintosh, is completely different from the Audio Company McIntosh Labs.

A search for Mackintosh put out this page on Linkdin :)

Of course it is true.

I have seen photos of R-Core XFRs inside McIntosh equipment, in recent years.

I have seen a published schematic of Nobu Shishido's Loftin White 2A3 DC amplifier, and it was mis - translated by the Japanese as Roftin- White, a R-W 2A3 amplifier !!! I think that - is funny !!

As I have already pointed out, it is early in audio's adaptation to R-Core output transformers. Only a few people in audio today actually realize what they do !!

I feel the R-Core audio output transformer easily out-modes, all that has come before, in terms of it's design and performance !! No contest overall, between a Softone R-Core and all the others, what-so-ever.

I also suspect that Softone and the Japanese do not yet fully realize this. Their tube amplifier build capability lacks what some of us in the USA do, to make vivid aural determinations.

RW 2A3.JPG

I would suggest not to copy-build this as it shows.

I did a quick Google check on McIntosh using Kitamura Kiden R-Core Transformers.

I have reason to suspect that this is not true .... :(

This is the Kitamura Kiden page that says:
"Our customers include major domestic Japanese companies such as Shimadzu, Toshiba Schneider Inverter, Toshiba Medical Systems, Hitachi Medico, and Hitachi Machine System. We also have many major overseas clients such as Mackintosh, GE, and Varian Medical Systems."

None of these are Audio Equipment manufacturers, and the spelling of Mackintosh, is completely different from the Audio Company McIntosh Labs.

A search for Mackintosh put out this page on Linkdin :)

Notice this : the Linkdin report on John Mackintosh ( an E.E. ) has nothing to do with any audio involvement - going back 41 years to 1980. However, I am fairly sure I have seen R-Cores in modern McIntosh Labs audio equipment pictures. Were they for the output transformer use, or for power transformer use ? I do not know! If I recall properly, I seem to think seeing a total of three R-Cores in a single McIntosh chassis.

I know for a fact - that a small high end audio Manufacturer , Dennis Fraker of Serious Stereo in the USA, uses Softone R-Cores. He told me, this week, that " this is all he plans to use, going forward." In my experience and personal opinion, he consistently hand-builds the world's best sounding 2A3 and KT150 SE amps, directly coupled, since 1989.
 
I spent Quite sometime Googling for McIntosh using R Core Output transformers... Did not find any such reference of picture on the net.

I ofcourse do not rule it out. If you do however fid a specific reference, do share.

Incidentally, Shilchar Electronics, India is one of the Handful of R Core Transformer (5VA to 1500VA ) manufacturing companies worldwide.


View attachment 54854

I would suggest not to copy-build this as it shows.
Any particular reason that you posted this circuit diagram ?

Its a fairly typical typical & straightforward implementation of a 2A3 SET ...
 
Any particular reason that you posted this circuit diagram ?

Its a fairly typical typical & straightforward implementation of a 2A3 SET ...


Yes, I posted it......... because it is called a " RW 2A3." Did any one catch that, from someone from Japan?? A "Roftin"-White 2A3. Not a Loftin-White . It's related, ( a first cousin ), to Mackintosh.

That Shishido 2A3 circuit, as I pointed out, needs many improvements, and should not be copied by any DIYer as it is.

Here is the latest update on the 6005 Stereo SE amplifier :

As of this morning, I am listening to the two most sensitive B+ spots in the entire amp ( the B+ for the Driver tube's plate resistor ( Ra ) and now, the B+ for the output tube's G2 ), with each sensitive area having it's own separate quadruple filtered B+ supply.

The playback precision of musical instruments today, is breathtaking to hear. I am, as-of-now, very "sure" of this newest power supply design and configuration. It is 100% desirable to have extra-ordinary B+ filtering. I am hearing all of this, on playback.

B+ feeds need to be super clean, especially when both channels are operated from a single supply. Great care has been applied, to execute this new 6005 stereo amp this way. No one ever has done this before in audio, for the Ra and G2 both. The quadruple filter to each location has this configuration :

L1/C1...........L2/C2.............Shunt 1...............Shunt 2.

The Ls ( chokes ) referred to above are small - "hash" chokes, the size and shape of a 2 Watt resistor. These hash chokes are maybe 1 Ohm in DCR , and 500 mA. capable, 100 uHY or less in inductance, single layer wound. Very useful, and seldom seen.
 
Jeff, Will you use shunt 2 on a monoblock too? This being a stereo amplifier the power supplies are shared between 2 channels.
 
For decades, the importance of carefully regulating the B+ supplies has been ventilated.

Alan Wright (actively) regulated the B+ with ripple down to the milli-volt level.

I believe VTL also uses stiff, actively regulated B+ supplies.
 
Status
Not open for further replies.
A beautiful, well-constructed speaker with class-leading soundstage, imaging and bass that is fast, deep, and precise.
Back
Top