New Design, SE 6005 Directly Coupled Audio Amplifier

Visit HiFiMART.com

drlowmu

Well-Known Member
Joined
Oct 28, 2020
Messages
205
Points
63
Location
Warrensburg, Missouri
The power supply design topology outlined on 2-22-21 in this thread :

( https://www.hifivision.com/threads/...tly-coupled-audio-amplifier.82963/post-929069 )

has been changed ( about 8-2021) to a ( hopefully ) better-performing design.


- - - - - - - - - - - - - - - - - - - - -

The latest design takes extra power supply parts. The two recently-developed build Modules, ( designed for this crowded stereo amplifier- VAM and RkM ) , will free-up chassis space for the new power supply topology - to be installed and accessed for service.

The two most sensitive areas of this amp is the B+ feeding the plate resistor of the Input / Driver tube ( Ra ) , and (G2 ) the internal second grid of the Output Tetrode,. Both sensitive areas get the same design : a quadruple sectioned passive filter. Let me try to show it with a typewriter !!

Note, both of these filter chains get fed from the last capacitor that feeds the Final's ( Output ) tubes ( "C2" ). First of all, allow me to show you that ( initial ) output stage topology :

Main B+ Power Transformer > dual tube rectifiers > Main L1-6 Ohms / Main C1 / Main L2-6 Ohms / Main "C2" ( to both output transformers ) and to BOTH of the new quadruple filters, ......................... all fed from Main " C2 ".



( each "quad" B+ filter is - effectively : SHUNT 1, SHUNT 2, L1- one Ohm / C1 and L2- one Ohm / C2 ) . ( The L1 and L2 are those two tiny one Ohm hash chokes, seen on top of each "VAM". )

Either will be like this : Here we are :


............................................................................> Second SHUNT Left Channel >( "VAM" ) 15uF + L1 / C1 / L2 / C2 > ( feeds Ra Left ( or G2 Left )
Main "C2" > Pre SHUNT One, shared L-R ( split to each channel, see above and below )

..........................................................................> Second SHUNT Right Channel > ( "VAM" ) 15uF + L1 / C1 / L2 / C2 > ( feeds Ra Right ( or G2 Right )



This above filter, fed by MAIN
" C2" is built twice in the stereo amp, once for the Driver tubes' Ra, and once for the Finals' tubes' G2.

How confusing is my explanation? Please, only those F.M.s who can follow it, kindly leave a " thumbs up " or " like" at the bottom of this post. There may not be adequate room for all of the thumbs down . :)

I am anxious to hear this. For the record, a usual tube amplifier would use one resistor and one cap, called a R/C decoupling section.

" In the end, we all listen to a modulated power supply....how good is it ? "

Jeff
 
Last edited:

Hari Iyer

Well-Known Member
Joined
Feb 8, 2010
Messages
3,175
Points
113
Location
Mumbai
The power supply design topology outlined on 2-22-21 in this thread :

( https://www.hifivision.com/threads/...tly-coupled-audio-amplifier.82963/post-929069 )

has been changed ( about 8-2021) to a ( hopefully ) better-performing design.


- - - - - - - - - - - - - - - - - - - - -

The latest design takes extra power supply parts. The two recently-developed build Modules, ( designed for this crowded stereo amplifier- VAM and RkM ) , will free-up chassis space for the new power supply topology - to be installed and accessed for service.

The two most sensitive areas of this amp is the B+ feeding the plate resistor of the Input / Driver tube ( Ra ) , and (G2 ) the internal second grid of the Output Tetrode,. Both sensitive areas get the same design : a quadruple sectioned passive filter. Let me try to show it with a typewriter !!

Note, both of these filter chains get fed from the last capacitor that feeds the Final's ( Output ) tubes ( "C2" ). First of all, allow me to show you that ( initial ) output stage topology :

Main B+ Power Transformer > dual tube rectifiers > Main L1-6 Ohms / Main C1 / Main L2-6 Ohms / Main "C2" ( to both output transformers ) and to BOTH of the new quadruple filters, ......................... all fed from Main " C2 ".



( each "quad" B+ filter is - effectively : SHUNT 1, SHUNT 2, L1- one Ohm / C1 and L2- one Ohm / C2 ) . ( The L1 and L2 are those two tiny one Ohm hash chokes, seen on top of each "VAM". )

Either will be like this : Here we are :


............................................................................> Second SHUNT Left Channel >( "VAM" ) 15uF + L1 / C1 / L2 / C2 > ( feeds Ra Left ( or G2 Left )
Main "C2" > Pre SHUNT One, shared L-R ( split to each channel, see above and below )

..........................................................................> Second SHUNT Right Channel > ( "VAM" ) 15uF + L1 / C1 / L2 / C2 > ( feeds Ra Right ( or G2 Right )



This above filter, fed by MAIN
" C2" is built twice in the stereo amp, once for the Driver tubes' Ra, and once for the Finals' tubes' G2.

How confusing is my explanation? Please, only those F.M.s who can follow it, kindly leave a " thumbs up " or " like" at the bottom of this post. There may not be adequate room for all of the thumbs down . :)

I am anxious to hear this. For the record, a usual tube amplifier would use one resistor and one cap, called a R/C decoupling section.

" In the end, we all listen to a modulated power supply....how good is it ? "

Jeff
Because of this confusion, I preferred a monoblock.
 

drlowmu

Well-Known Member
Joined
Oct 28, 2020
Messages
205
Points
63
Location
Warrensburg, Missouri
10-01-2021 .............................. PHOTO UPDATE / PROGRESS REPORT



Test fitting of the newly developed multiple film cap modules, inside the new STEREO chassis. Very happily to me, it all fits inside neatly.

Now to lay out the main Power Supply components. One must thoughtfully consider lead lengths and fields.

Then, mounting holes for tube sockets, and off to the Powder Coating shop, prior to final wiring and assembly.

These new Modules make the stereo amplifier ( with all these not-really " extra " parts ) viable , both for amplifier construction and any sort of future below-chassis work needs.




016 edited.jpg

013 edited.jpg

021 edited.jpg

009 edited.jpg

019.JPG



On the extra-heavy-duty ( 12 gauge steel ) mounting bracket - for the 320 VA R-Core main power transformer, I have devised a new simple double vibration deterrent mounting scheme. The bracket itself, is double isolated with brass pyramids. Photos coming-up soon. Stay tuned.

Thanks for looking and following along.

Jeff
 
Last edited:

drlowmu

Well-Known Member
Joined
Oct 28, 2020
Messages
205
Points
63
Location
Warrensburg, Missouri
Historical Information :

Can we name some vintage USA tube amps, that employed dual 5U4G rectifier tubes - like this thread's new 2021 SE 6005 amp does ??



1) McIntosh M-60 monos, ...........their better-performing models of that era. ( Better than a Mc 275 . )


2) Fisher 50 AZs.......lovely, mono.

and


3) ALTEC 345A, stereo ............highly collectable, all Peerless iron, good sounding. Here it is below:



ALTEC 345A AMP SCHEMATIC, Save, USE, EXPLAINED 1.jpg

And a die hard Audiophile :

................................................ fisher005.jpg
 
Last edited:

drlowmu

Well-Known Member
Joined
Oct 28, 2020
Messages
205
Points
63
Location
Warrensburg, Missouri
10-03-2021 ........................................ PROGRESS PHOTOS

Developed a new simple one - bolt attachment module today.

These enable mounting the large 50 Watt B+ SHUNT Power resistors wherever it seems best to position them inside the chassis. All this is used for extra-B+ filtering to the Driver tube and to Grid #2 of the Output tube. Both these circuit positions are super-sensitive to the cleanliness of their B+. This is done strictly to obtain outstanding amplifier performance.

This was gratifying to conceptualize over the last 48 hours, and build from scratch today.

033 edit.jpg 025edited.jpg 012 edited.jpg 038 edited.jpg

Thanks for looking. I am pleased with this arrangement. More to do ! All is designed for high reliability, and is super conservatively operated . eg : These 50 Watt ARCOL HS resistors have a 1250 VDC Manufacturer's rating, and the B+ is 460 VDC. Many of the WIMA film power supply caps, ( shown in above-posts on this page ), have a lifetime use rating of 300,000 hours. That is over 34 years, of continuous operation.

Jeff
 
Last edited:

drlowmu

Well-Known Member
Joined
Oct 28, 2020
Messages
205
Points
63
Location
Warrensburg, Missouri
10-05-2021 .............. UPDATE

After doing all that work two days ago, to get those ARCOL HS 50 Watt resistors mounted onto Keystone 1569 Board Brackets, I have now decided to scrap that entire idea. ( This is all "part" of any careful amplifier development process. )

I just didn't like the way they fit into the amplifier.

Instead, yesterday I spent time having the ARCOL 50 Watt resistors be direct-to-the-chassis mounted. It is an additional 12 Watts the chassis will have to " sink ". Resistor positions are spaced to avoid chassis hot spots and to minimize resistor- to -resistor field reactions. The new chassis is large enough to do this.

One other thing, the capacitors will last the longest if they are kept cool.

The four main WIMA power supply caps ( 30 uF, 80uF, 50 uF and 50 uF ) will be mounted individually on spaced sheet metal " shelves "or " platforms " , 1/4 of an inch above the newly " resistor-heated " chassis floor.

Besides cooling and thus enhancing long term ( decades ) reliability of the film caps, wiring can now be freely routed below any capacitor, in the 1/4 inch open space beneath it. Win - Win, ..... if one is willing to do the extra work. Four WIMA capacitor platforms need to next be made-up, in the next day or two, including drilling and tapping 16 more holes ( 4-40 ) into the 14 gauge steel.

Hands and fingers have already gotten " beaten up " and bloodied. Carefully : measure twice, then cut or drill once.

Jeff

002 (3)  EDITED OK.jpg


004 (2) EDITED EXCELLENT CLOSE.jpg

Note : C2 Bypass GTO is positioned right between both Output transformer access holes of the chassis, for minimal and equal length feed of instantaneous pulsed music information.
 
Last edited:

drlowmu

Well-Known Member
Joined
Oct 28, 2020
Messages
205
Points
63
Location
Warrensburg, Missouri
LAYOUT, LAYOUT, LAYOUT .......................... 10-05-2021

The two layout photos ( in the above post ) pleased me on many levels, and I also liked its looks and symmetry.

However, two slightly different layouts have since come to mind today. I have not decided on which of the three to finalize upon, design-wise.

Here pictured below is OPTION 2 : The L2 choke was moved from the one long vertical wall, and brought closer to the other power supply components.




019 EDITED.jpg

020 EDITED.jpg

After looking at OPTION 2 photos ( above ) for awhile, one might ask, why not remove the 5VCT R-Core 5U4G Filament transformer, from it's side wall mounting ( NOT PHOTOGRAPHED )?? Why not bring it in closer, tighter-in towards the other supply components in the center of the amp ??

SO, below is the very very compact OPTION 3.

The only thing I do not know is how much fields will interact in this third option. R-Cores have about one tenth the radiated field of conventional E-I transformers, and the main R-Core power transformer sits on the top of the chassis / vertical, and should likely pose no problem. vertical. Option 3, moving the R-Core 5U4G filament transformer closer-inwards -( off of the side wall ), may work well.

I will show this to my Audio Mentor, Mr. Dennis Fraker, and seek his opinion.



031 (2)  EDITED BEST OPTION THREE.jpg

If I could get away with the tight spacing, field wise, I believe it could be the best performing choice. How will this Option 3 " look and feel " in a day or two ?? No more chassis hole drilling and tapping, until it is decided between these three layout options. If anyone has advice, feel free to express yourself...here or in a P.M. , nicely.

Thanks for looking. Matters like this above are a BIG deal, when seeking improved performance. The choice will soon be resolved ! Personally speaking, just three days ago I was terrified - that I would perhaps not develop a reasonable layout, being that it was a stereo amp with many parts.

Jeff
 
Last edited:

Hari Iyer

Well-Known Member
Joined
Feb 8, 2010
Messages
3,175
Points
113
Location
Mumbai
Looks nice!

Imo, the L1 is way to close to the resistors. The flux lekages from the L1 could interact with them. Try putting a shield inbetween them to avoid this. An aluminium wire- mesh could also be good enough. The mesh should be earthed. Or if possible encap the L1 in a metal case or something similar.
 

drlowmu

Well-Known Member
Joined
Oct 28, 2020
Messages
205
Points
63
Location
Warrensburg, Missouri
Looks nice!

Imo, the L1 is way to close to the resistors. The flux lekages from the L1 could interact with them. Try putting a shield inbetween them to avoid this. An aluminium wire- mesh could also be good enough. The mesh should be earthed. Or if possible encap the L1 in a metal case or something similar.


OK Hari, good point. Let us do this simpler than a shield.

By removing the L2 choke off of the left vertical chassis wall ( Option 2 ) , and the 5U4G filament XFR off the right vertical wall ( Option 3 ) , we can relocate many 50 Watt ARCOLS on those vertical chassis wall spaces. The mounting spaces become "freed up" - on either side of the chassis.

I will consider doing this on 10-6, but only when I am fresh and rested. We shall call relocating the resistors " Hari's Option 4 ".

Since the Option 3 photo was posted, I have moved the C1 Bypass GTO to be 1/2 inch closer to the chassis center, and the L2 choke 1/2 inch away from the center, to lessen possible field interactions between these two large power supply chokes.

In our filtering, the first choke, L1, has large current spike fields associated with it. This is because it is ( on purpose, to get low DCR ) less than textbook " Critical Inductance " rated....( thank goodness ). L1 is a Hammond 159ZA, with a generous 1 Amprere current rating, so it does a fine job with current spikes in the L1 position after a break-in of 75 hours. However, the 159ZA's use requires there always be a similar L2/C2 topology in series with L1/C1. L1 and L2 must be the same to avoid audible skews. Finally, we certainly want a choke input filter - that a 5U4-type rectifier "sees", not a capacitor.

We always use two sections of filtering " L1/C1/L2/C2 " to the Output stage. As you have found out Hari, these 6 Ohm 159ZA chokes perform wonderfully, especially when we combine them with our use of GTO bypasses and really good wiring to the Output stage.

Thank you very much for helping .

Jeff
 
Last edited:

Hari Iyer

Well-Known Member
Joined
Feb 8, 2010
Messages
3,175
Points
113
Location
Mumbai
OK Hari, good point. Let us do this simpler than a shield.

By removing the L2 choke off of the left vertical chassis wall ( Option 2 ) , and the 5U4G filament XFR off the right vertical wall ( Option 3 ) , we can relocate many 50 Watt ARCOLS on those vertical chassis wall spaces. The mounting spaces become "freed up" - on either side of the chassis.

I will consider doing this on 10-6, but only when I am fresh and rested. We shall call relocating the resistors " Hari's Option 4 ".

Since the Option 3 photo was posted, I have moved the C1 Bypass GTO to be 1/2 inch closer to the chassis center, and the L2 choke 1/2 inch away from the center, to lessen possible field interactions between these two large power supply chokes.

In our filtering, the first choke, L1, has large current spike fields associated with it. This is because it is ( on purpose, to get low DCR ) less than textbook " Critical Inductance " rated....( thank goodness ). L1 is a Hammond 159ZA, with a generous 1 Amprere current rating, so it does a fine job with current spikes in the L1 position after a break-in of 75 hours. However, the 159ZA's use requires there always be a similar L2/C2 topology in series with L1/C1. L1 and L2 must be the same to avoid audible skews. Finally, we certainly want a choke input filter - that a 5U4-type rectifier "sees", not a capacitor.

We always use two sections of filtering " L1/C1/L2/C2 " to the Output stage. As you have found out Hari, these 6 Ohm 159ZA chokes perform wonderfully, especially when we combine them with our use of GTO bypasses and really good wiring to the Output stage.

Thank you very much for helping .

Jeff
I am planning to add one 0.15uF, 630vdc Vishay MKP axial capacitor across my C2 to handle HF pulse to prevent them from reaching the KT88 grid2 or the driver plates. I already have a 5uF GTO and a 0.033uF, 800v EPCOS but the Vishay are very good at handling transient pulse.
 

drlowmu

Well-Known Member
Joined
Oct 28, 2020
Messages
205
Points
63
Location
Warrensburg, Missouri
10-07-2021 .............................. Mini REPORT, COMMENT

Designing and building your own DIY amplifier can become an amazing process, when it comes to laying out the internal parts.

Two days ago I posted three possibilities or " Options ". F.M. Hari Iyer suggested what became " Option 4 ".

I was dreading constructing individual "shelves or platforms", to raise each of four WIMA power supply capacitors off of a hot chassis floor. A lot of work, and time, for a minor embellishment.

Today, upon awakening, I realized now only two hot ARCOL 50 Watt resistors would be near the power supply's caps ( the others are now mostly on the side walls, as per me and Hari's Option 4 . )

A new idea instantly occurred.

Since space is tight, ( and these two ARCOLS will each be dissipating 2.9 and 3.2 Watts of power ), these can be mounted OFF the chassis floor, in free air with tiny / compact 10 Watt heat sinks added, on long 4-40 stainless steel threaded rods.

The two ARCOLS would " hover " vertically spaced apart and "over" each other - in the same ideal footprint-position ................which has already been determined.

( Recall, " build in three dimensions, not two ? " )

With six Watts less heat, ( and the two resistors no longer being chassis - cooled ) the power supply capacitors can now be attached directly to a cooler-running steel chassis floor. ............ No "platforms or shelves" required !!

- - - - - - - - - - - - - - - - - - - - - - -

What I wanted to share with you is, that sometimes ( often ! ) it takes several days, to come up with decent layout solutions. Usually, at the very beginning of one's day, when fresh and rested. new ideas come into our minds...." like lightning ".

Mounting the stacked / spaced apart two resistors on stainless steel threaded rods, above the chassis floor, was a truly very "happy " solution today.

Minimalist hardware and complexity........ just two long threaded rods, and several nuts for securing and positioning resistor heights. Elegant.

ARCOL HS Manufacturer's datasheet supplied below. ( FYI, a large ARCOL HS is a decent-sounding, OK to use power resistor IMHO ).

Pictures forthcoming, once executed.

Jeff
 

Attachments

  • ARCOL HS - Datasheet Rs.pdf
    297.6 KB · Views: 2
Last edited:

drlowmu

Well-Known Member
Joined
Oct 28, 2020
Messages
205
Points
63
Location
Warrensburg, Missouri
Hi Jeff, build coming along nicely.
Will you be installing heatsinks as well?
Susnick


Undecided until running and very scientifically finger tested........how many seconds can one hold their own finger on the part ?????

In datasheet posted above, a HS-50 Watt resistor shows " up to 14 Watts, with no heatsink, but ... at very elevated temperatures. "

Each HS-50 will be dissipating about 3 Watts.

These little sinks in the photo require a chassis backing surface. A slightly larger sink can employ the nuts on the threaded rods, to hold resistor to the mini sink, and that is what we need to find. Need a perfect size and shaped one...something ideal.

Heatsinks?? Well hello " Nebraska Surplus Sales " !! https://www.surplussales.com/sitemap2/alphalisting.html Click on " H ". :)

Jeff
 
Last edited:

Hari Iyer

Well-Known Member
Joined
Feb 8, 2010
Messages
3,175
Points
113
Location
Mumbai
Jeff,
These Arcol resistors get pretty hot imo. I have mounted 4 of them on my bottom aluminium plate and can actually cook bread on them literally. Efficient heat dissipation / distribution is most important for any amplifier. Do not worry much about how they will look internally as it won't matter. Thick about how you would couple them with the chassis for distributing heat.
 

keith_correa

Well-Known Member
Joined
Jan 17, 2009
Messages
3,302
Points
113
Location
Mumbai
Two small "U" shaped sinks. One stacked over the other with contact between them via bolts. The resistors mounted to the bottom valley area of each of the "U". The lower heatsink bolted to chassis.
You get the benefit of heat dissipation from the sinks and from the chassis too. Doable?
 

drlowmu

Well-Known Member
Joined
Oct 28, 2020
Messages
205
Points
63
Location
Warrensburg, Missouri
Jeff,
These Arcol resistors get pretty hot imo. I have mounted 4 of them on my bottom aluminium plate and can actually cook bread on them literally. Efficient heat dissipation / distribution is most important for any amplifier. Do not worry much about how they will look internally as it won't matter. Thick about how you would couple them with the chassis for distributing heat.
Dear Hari,

Each of your own KT88 monoblock amps is dissipating a total of about 20.56 Watts into that aluminum bottom plate. It is using only four power resistors. Yes, it will get very hot. Each ARCOL 50 Watt resistor in your amp averages 5.14 Watts of dissipation.

My all-steel chassis has at least twice the surface area of your KT88's single bottom plate.

This new stereo amp, ( with double shunting to all stages ) dissipates 39.5 Watts total.

The new amp's dissipation however, is spread over 13 different power resistors. This averages 3.03 Watts per device.

Of this new stereo amp's 39.5 Watts total, 14 Watts is dissipated by two paralleled "large-surface-area" black anodized heatsinks, ( seen in the middle of the chassis.) The chassis has been laser cut purposely - 14 ventilation holes below it, and seven above the black heatsink.

The stereo amp's remaining is 25.5 Watts ( 39.5 minus 14 ), is dissipated by the steel chassis' side walls.

All I am trying to do is relieve the chassis' dissipation on the floor of the chassis - near the supply caps, by re-positioning the two ARCOLS positioned close to the film caps, ( each resistor dissipating 2.9 Watts and 3.1 Watts. )

I think it is a reasonable idea, for the long term longevity of the power supply's film capacitors, to have these two resistors totally air cooled. As a precaution, I will add compact 10 Watt heatsinks to each. This will assist the resistors - from operating at needlessly high internal temperatures.

Fortunately, I have had adequate time to contemplate this execution.

Jeff
 
Last edited:

drlowmu

Well-Known Member
Joined
Oct 28, 2020
Messages
205
Points
63
Location
Warrensburg, Missouri
Two small "U" shaped sinks. One stacked over the other with contact between them via bolts. The resistors mounted to the bottom valley area of each of the "U". The lower heatsink bolted to chassis.
You get the benefit of heat dissipation from the sinks and from the chassis too. Doable?


But the fields of the two resistors will interact with each other. I want space between the power resistors, as per the photo. Thanks !!!

Jeff
 

keith_correa

Well-Known Member
Joined
Jan 17, 2009
Messages
3,302
Points
113
Location
Mumbai
But the fields of the two resistors will interact with each other. I want space between the power resistors, as per the photo. Thanks !!!

Jeff
I suggested that the 2 sinks be stacked over each other and not the resistors. There will be sufficient space between them.
 

drlowmu

Well-Known Member
Joined
Oct 28, 2020
Messages
205
Points
63
Location
Warrensburg, Missouri
I suggested that the 2 sinks be stacked over each other and not the resistors. There will be sufficient space between them.
OK, I misinterpreted what you meant. Understood now.

I actually will be using the two small heatsinks, partially shown in the right top corner of the posted photo. I will not have either device or it's heatsink touch the chassis, for reasons described in the response to F.M. Hari Iyer, above.

Your thoughts and kind contribution are appreciated. Thank you !

Jeff
 
For excellent sound that won't break the bank, the 5 Star Award Winning Wharfedale Diamond 12.1 Bookshelf Speakers is the one to consider!
Top