DIY 300B DHT SE Amplifier Project

Hi Ashok,

I can see Sivat-ji already putting a question mark behind our deliverable # 9

How about Damper Diodes in the power supply?
If I remember right,Viren mentioned that he was planning to try these out.

10.Must keep the exotic stuff away. No silver wires, cryogenic tubes and snake oil

Do Black Gates come under exotic stuff?
I know your preference for oil caps but I could send you a couple of the 100+100/500 Volt BG's.

A couple of the BG's after the choke in the 2A3 amp delivers tight musical bass.

Regards
Rajiv

Hi Rajiv

I am glad that you showed up. You perspectives will take this endeavour to its greatest height. BTW I have located the additional pair of 8uf oil caps. Let me know if need that.

Damper Diodes:
You mean the SS variety? If yes then I am open to using them. But where are they available? I have never asked my usual places. That is one very good suggestion

Blackgates:
They are a borderline case but on this build I was planning to use either elcos or Motor Run caps. Oil caps are too large and the build will be too big. Your offer of the BG is most generous and solves many problems. In fact I am planning a choke input PSU and would like to know your opinion on that.

BTW do you also have access to smaller BG’s? For cathode bypassing and such?

Our last DIY amp had generous contribution is the form of BG caps from Rjaiv-ji and the Main power transformer very kindly provided by Viren-ji. Both completely free delivered at my home. I take the opportunity to again thank these fine gentlemen.

Attaching a pic of that amp showing the said power transformer and the pair of 100+100/500v Blackgate WKZ caps.

Hi aks and Rajiv,

Can I put in a few suggestions:

- Circuitwise this is going to be a very simple amp. Inspite of the hype of single-ended valve amps, they are very easy to get going. (To make them superior takes a gigantic effort!) This should be a good candidate for DIY.

- By all means make this the best amp you can. But also give alternates and substitutes for parts that are more easily available, explaining the compromise. That would encourage more people to try building the amp.

This should turn out to be a super amp, that will put many a commercial amp to shame!

Viren

aks07 schrieb:

This thread is an offshoot of another thread here And our most distinguished member Shri Viren gave his most simple but enthusiastic “Go for it” thubs up. I hope he and Shri Sivat will be bringing the most practicle ideas to this. I can see Sivat-ji already putting a question mark behind our deliverable # 9

This is only a matter of choice Akh-ji. Tube's in PSU seem to add that little "euphony", which makes the sound warmer. Infact this is the sound which many tube-fan die for. I have to agree i'm amoung the minority here...with respect to this specific choice.

However i do have one feedback on the technical aspect of this. I was talking to a good friend of mine the other day....and an idea stuck me....the science behind audio is pretty much like "Chaos theory". We can only "assume" the reason for a "cause"...but never be certain about it. The biggest challenge in my many years of DIY effort....."Fixing a problem is very easy...but identifying the real source of that problem...can take years".

On the same line, i do not think the solid-state PSU itself makes the sound edgy or bad. I think it brings out the limitation of components used in the amplifier (right from resistor, to wiring, to capacitor...to chassis). The other way to put it would be....the "euphonic" or "warm" nature of tube based PSU covers some of the limitations of the other components. So it is a choice for one to make

For Cap try the Panasonic. I'm using the TS-ED range...its absolutely fantastic....black background with very little noise or phase problems. Its expensive ... but worth it. I will not spend the extra money on BG.

Locally available electrolytic caps generate a lot of noise...and many of them screw up the sound very badly.

For design with 845 ....how about the one from A.Ciffuoli...using 300B and an Interstage.

I'm using a similar design, but in a PSE configuration...and very happy with the sound.

Thanks for the valuable guidelines. Three very good points till now.

1. BG in power supply. Other option being Panasonic and Oil caps. Genric elcos are obviously out as they lot of their own generated noise.
   
2. Keep the design simple but execute a very high class job, while keeping options for key components should someone desire an alternative
   
3. SS rectification. Sivat-ji how do you implement this? I am concerned with the switching noise. In a tube rectifier its relatively easy. I do have a bunch of fast rectifiers.
   

But first of all lets firm up if it is indeed a 300B project. We shall then come to a specific design and what is the best possible way to implement it.

I too am inclined on an IT coupled amp similar to what Andreas has.

Hi,

This thread on AA about how Cary and Quicksilver HT power supplies is interesting.

http://db.audioasylu...oshiba+gz34&session=

I do not know anything about the sonic or technical merits of this type of implimentation,but it is a novel way for power supply design.

Regards
Rajiv

hi aks,

a 300b tube amp sounds like a great project. one request though do go step by step and S-L-O-W-L-Y for electrically challeged guys like me

would you mind expanding on why you want to avoid capacitors? point 8. is it cos of reducing components in signal path or do caps do something bad to the sound?

regards

Rajiv schrieb:

Hi, This thread on AA about how Cary and Quicksilver HT power supplies is interesting. http://db.audioasylu...oshiba+gz34&session= I do not know anything about the sonic or technical merits of this type of implimentation,but it is a novel way for power supply design. Regards Rajiv

Funny that you bring that up. I have seen many implementations of the hybrid rectification, but your linked post very clearly explains the advantage. I guess being an unconventional approach its yet to find wide spread application apart from a selected few gurus. Viren-ji and Sivat-ji, what is your take on this?

Steve-ji
Don’t worry. Don’t worry . We are all sort of newbe when it comes to the ancient technology of tubes

Elimination of the coupling cap as such doesn’t reduce the part count. You need to replace it with something. Often a more expensive interstage transformer. Even the best of the caps does more than what its asked to do and in the process degrade the SQ. We shall come to that when we freeze our design. Perhaps Sivat-ji will like throw more light on this.

Ok. I spent my morning to round these lovely ladies for a photoshoot so that you yahoos can salivate a bit

First a pop quiz. Can you guys ID the tubes? 8 tubes from left are pentodes/tetrodes, and the rest are power triodes.



The pentodes/tetrodes



The DHTs

DROOOOOOOOOOOOOOOOOOOOOOOOOOL

the DHTs - 300b, 45, 2A3, 300b, 845?

I do not think you can eliminate the switching noise completely in SS...regardless of how fast this is. I did not say SS is a "perfect" solution Life is all about compromises. Using fast diode & good caps in PSU does help a lot.

However, i do not like the non-linear behaviour of a tube rectifier. "current" dynamics are still in important in a tube amp (even if they are voltage based). I assume this is what causes the "euphonic" nature of a tube based PSU.

stevieboy schrieb:

the DHTs - 300b, 45, 2A3, 300b, 845?

Great. You got two correct!
Lets wait for what others say.
BTW exept two, all of them all mint fresh. Never used.

aks07 schrieb:

Elimination of the coupling cap as such doesn’t reduce the part count. You need to replace it with something. Often a more expensive interstage transformer. Even the best of the caps does more than what its asked to do and in the process degrade the SQ. We shall come to that when we freeze our design. Perhaps Sivat-ji will like throw more light on this.

I' have no knowledge of Hybrid PSU designs...sorry.

On Caps...you are perfectly correct. The Capacitors can really screw your happiness...if you are not careful with them. You need to use the right value (more is not always good) and more importantly a good quality product. No compromises on Caps...which remains one of the biggest expense in all my DIY project

Interstage is infact a very good alternative. But then again, the difference in price can be so high....and that investment, if made else-where might make more sense. This would be a subjective decision....no easy answer here. It depends on design...and overall cost of other components.

Sivat-ji

If you recall, during the Santa Claus build period we had posted some technical reasons why tube rectifier is inferior to SS. Purely on technical basis. Just to summarise, SS has a lot less internal resistance than tube. It can deliver much more current. Its cheap, easy, doest produce all that heat and does away with an additional wiring on the PT. It can take greater abuse. Its effect on final sound quality will be consistent.

Its got so many things going for it. But as I read you really need a very well filtered, preferably a regulated, design to get rid off all noise. Specially crucial in a SET amp which have inherently poor PSRR as compared to push pull designs.

IMHO the greatest practicable advantage that a tube rectifier offers is the negligible noise characteristic and a slow start-up of main HT.. This allows keeping things very simple. To achieve theses objectives in SS will require for several additional circuits.

As far as sound is concerned it’s a subject issues. Tube rectifier is my first choice reverting to SS where its not practical.. I have built both varieties and difference in SQ is ultimately minimal all things remaining same. But obviously SS can never reach the coolness factor of tube rectifier.

Do you think we should look to regulate the driver rail by using a gas discharge tube?

aks07 schrieb:

Sivat-ji If you recall, during the Santa Claus build period we had posted some technical reasons why tube rectifier is inferior to SS. Purely on technical basis. Just to summarise, SS has a lot less internal resistance than tube. It can deliver much more current. Its cheap, easy, doest produce all that heat and does away with an additional wiring on the PT. It can take greater abuse. Its effect on final sound quality will be consistent. Its got so many things going for it. But as I read you really need a very well filtered, preferably a regulated, design to get rid off all noise. Specially crucial in a SET amp which have inherently poor PSRR as compared to push pull designs. IMHO the greatest practicable advantage that a tube rectifier offers is the negligible noise characteristic and a slow start-up of main HT.. This allows keeping things very simple. To achieve theses objectives in SS will require for several additional circuits. As far as sound is concerned it’s a subject issues. Tube rectifier is my first choice reverting to SS where its not practical.. I have built both varieties and difference in SQ is ultimately minimal all things remaining same. But obviously SS can never reach the coolness factor of tube rectifier. Do you think we should look to regulate the driver rail by using a gas discharge tube?

I'll agree on this technical aspect. But the linearity is still missing in tubes...so it affects the SQ...and it becomes a matter of taste.

Regarding regulation...is it really required. I have couple of OA2, which i can send it across to you..

Thanks Sivat-ji for the kind offer on 0A2 regulators. I too have a pair, so thanx again for your most generous offer.

What is your opinion on a Sakuma type topology?

I collected some info on damper diodes as I
considered them for my amp.
Damper diode tubes (like the SS variety) were used in
TV to control the ringing across the deflection coils.
Here's a typical psu using damper diodes -

http://nullspace.us/45/schematic.jpg

They are praised for their smooth sound, slow
warmup, higher current capacity and lower voltage
drop than GZ34 types.

Some damper diodes suitable for rectifier use -
6AU4GT, 6AX4 (easy-to-use octal socket), 12AX4, EY500,
6CJ3 (Lynn Olson used 6CJ3.
6CG3 can even go up to 350mA.

See info here on 6D22S -

http://www.tube-town.net/info/sed-techbulletins/technoteNo52.html

Here is a hybrid Graetz bridge supply -
http://www.audiodesignguide.com/my/se.html

6AU4 type damper tubes are half wave requiring a pair. If we are building a 845 type SET calling for more than 1,000volts then these damper tubes make eminent sense

If we decide to build 300B SET (as seems most likely) we are perfectly well off using the simple 5U4G or my personal favourite the GZ32.

I am proposing a choke input power supply. What do you guys think about that?

Aks

Methinks, Unless you get anyone else with a better tube idea, let's freeze on the 300B's.

Besides the gz32, What would be your close second choice ?

Manek

That’s 2 votes for 300B !

Next to GZ32 , I like 5U4G, 5R4GY and GZ37. Each have very individual sonic signature.

A driver tube to be considered, esp with 300B is the ECC99.

Hi aks,

Please go with the 300B DHT. There are so many flavours of it available, that DIY guys can have a field day!

Also, a vote for a choke input power supply. Eminently suitable for a class A design. There are many ways of regulation of the supply. I favour a more benign way - high capacity power transformers, all choke filters.

On the subjective nature of valve rectification, it all depends on what your listening reference is. I try to emulate the sound of acoustic intruments playing in a natural acoustic space. Very important is to reproduce the harmonic nature of instruments, and the rise and decay of notes. If you achieve that, all types of music will sound more real. Valves just come much closer to that - I consider them more accurate!

Viren

300B is okay...but 845 makes the sound complete

ECC99 is a good choice for driving 300B. I tried with in amp.

AKS07 said:

If we decide to build 300B SET (as seems most likely) we are perfectly well off using the simple 5U4G or my personal favourite the GZ32.

Just for my info and learning... I haver read that Audio Research in their Top-Of-The-Line Ref-3 Preamp have used the 6550 Pentode as a rectifier ! ?

True or have I mis-understood ?

What befit ?

Hhi Guys,

For my learning, can some one please explain the advantages / disadvantages of :

Directly Heated Cathode vs an Indirectly Heated cathode ?

Any sonic qualities associated with Directly heated Cathodes ? ( not the sonics of different tubes, but the sonics / technical implications of Directly Heated Cathodes.. )

Thanks

Amp_Nut-ji

Every vacuum tube after all is a valve. I am sure someone must be rich enough to triode strap a 6550 and use it as a diode. On net I saw one amp which only had 300B’s ! Starting from rectifier to voltage regulator to driver and so on. It had 13 qty of 300B in a stereo build.

As for superiority of DHT versus IDHT, I have seen many references on why it is. From my personal experience DHT has a much better midrange (maybe as a result of its greater linearity?) and the aural quality is very relaxed. Otherwise the 6550 is my fav tube.

Good to see that we have finalised the tube (300B) and using a Choke input power supply. Driver suggestion is ECC99. These choices will certainly help in meeting our key deliverables.

Now let’s discuss the topology.

The simplest is using a driver tube and coupling cap and one of the simplest of these types is the JE labs or Angela design. A 6SN7 driving a 300B. Its simple, cheap and easy to get going. However, a lot will depend on the type of coupling capacitor and as Shri Sivat pointed out a good one can cost a little fortune. Moreover, these types of design need at least 3 stages. Each stage adds its own signature and the cumulatively adds up to a high source impedance. These two are issues are its biggest drawbacks.

One way to avoid that is to use a Mu follower or SRPP design. My 6550 SE amp uses this design. 6SL7 with a 1K2 resistor and no NFB provides phenomenal amount of low distortion drive. Sound pretty expressive and open. Its one of the simplest design to follow but as usual there are many who simply hate this claim the audio looses its life if put thro a SRPP.

Then we have the cathode follower type design. Many designers in Far East favour this. I think Sun Audio designs are CF too. My good friend Mr Liang (of Mr Liang Audio fame) vouches by this and his reference 845 SET’s have a CF driver using a 300B or 6L6. From my personal hearing of Mr Liang amps I can attest to this designs efficacy.

We also see many designs whereas a pentode is driving a triode. This will only need two stages and seem to meet one of our goal. Thorsten has a very famous design of EL84 driving a 300B. It his personal system. In Radiotron manual one can see many pentode > triode designs. I too have had good experience driving a 811 SET with a variety of pentodes like EL34/6L6/KT66. Separately I have tried using pentodes like EF86 and 6SJ7 to drive tubes like EL34 / EL84 and 6L6 in SE design. It does away with that dreaded Miller capacitance issues that bog triodes. Very nice sounding. But some people have problems with the driver pentode screen noise.

And then we have the really ancient designs that employ an Interstage Transformer. Many famous designers follow the same western electric idea for their amps. It’s a simple plot with just 2 triodes. IT coupling enables the output stage to really play throu its operating range particularly on a highly linear triode like the 300B. This results in a stunning life like quality for reproduction of acoustic music. Japanese tube guru Sakuma san has made an art of designing transformer coupled audio amplifier. He gets his custom wound iron from Tamura. Back home Shri Viren has always been using an IT design and I think he will do more justice in explain this philosophy.

Considering all objectives I am voting for a transformer coupled 300B SET. Comment and suggestion highly solicited.

Here we have basic block of a typical IT coupled amp. You can get going with only this many components and get an extraordinarily revealing and pleasant sound.



Subsequently, we shall get down to a discussions on op points etcs

Aks

Since we do have a good source of the transformers(delta) I would think a transformer coupled desisn would be the way to go keeping into account the simplicity of it all.

"Less is more"

Btw I do remember you saying you wanted to use the 5687 tube sometime so would this be the right time and build for it ?

Manek

Yes DELTA is a very good source. I have used their power transformers and chokes last several years. Their standard catalogue of output transformer ofcource isn’t designed for hifi duty. But they do have a small transformer of 5K impendence that can be used for a EL84 SEP amp and it sounds very pleasant.


For driver tubes I have narrowed down to 5687, 5842 and ECC99. I have the 5842 with me. I had some 5687 and ECC99 that I gifted away just a few days ago to a friend ! I guess I should have held on to them a few days more. I know for sure that ECC99 is not available locally. I do hope that at least the 5687 is available. Lets see.

Aks

Viren now uses 5687 in his preamps so I assume there is a local source

Ecc99, that's a tube that I haven't heard much about...

Manek

I've compared the ECC99 with E182CC (Mullard) and 7119 (Amprex). The ECC99 is the smoothest one with the least distortion....

I prefer..by a very small marging...the Amprex 7119. The Amprex is more linear than ECC99. I think the ECC99 are very smooth...with a slightly forward mids.

Not OT I hope, as this is an integrated amp.
Anyone have more info on the Miracle transformer
volume control?

http://www.toroidal.com/controller.html

sivat schrieb:

I've compared the ECC99 with E182CC (Mullard) and 7119 (Amprex). The ECC99 is the smoothest one with the least distortion.... I prefer..by a very small marging...the Amprex 7119. The Amprex is more linear than ECC99. I think the ECC99 are very smooth...with a slightly forward mids.

What was your application?

quadtech schrieb:

Not OT I hope, as this is an integrated amp. Anyone have more info on the Miracle transformer volume control? http://www.toroidal.com/controller.html

Thanks for that link. Indeed we need a good pot
But the Miracle volume control is for line output speaker as found in PA application.

What type of pot are you using in your KT77 amp?

The 7119/E182CC/ECC99...is the input stage in my 845PSE...which is coupled to the 300B using a PIO cap..which in turn drives the two 845's through an Interstage.

sivat schrieb:

.. the two 845's .

Must be needing one man size OPT.

For our build I am suggesting 5842 > Lundahl LL1660 > 300B > 3K OPT

PSU:

440 - ct - 440 > 10H choke > 36u Oil > 2 x 20H 100u BG (split for left & right)> R > Driver section.

Comments.

What type of pot are you using in your KT77 amp?

100K stereo audio (log) taper Taiwan Alpha pot.

Hi aks,

Your topology looks good. However, since this is an integrated amp, you may find that you run out of drive in the input stage. A quick look at 5842 curves - a bias of -1.5V will give you a voltage swing of +/- 50V. May or may not be enough. You could get a bit of gain by hooking up the interstage to a 1:1.5 turns ratio.

The major advantage to an interstage in such an amp, driving a DHT, is the capability of allowing positive current through the grid, A2 operation. Helps the amp handle peaks in music, and makes for good dynamics.

Also, it's a natural for applying fixed bias to the DHT.

Good power supply topology too!

Viren

Thanks Viren-ji
I am looking to run the 5842 as per datasheet at 150~180 VDC, 20mA with a grid voltage of -2.0VDC for an mU of about 50 into 1:2.25 IT.
In this case the RP is only constrained by the primary DCR of the IT. We are well within the 4.5w Pd of the tube while still operating in deep class A. The secondary of the IT is almost unloaded and the reflected primary Z is high. The plate load line is at its most linear region.
As a person who has been using IT commercially in all his amps, I want your valued opinion if it seems ok to you.

Tomorrow I will run some simulations to check out. I am using 5842 simply because that is available at hand. ECC99 is not available in India and 5687 is currently not available in the usual places I buy tubes.

I am attaching Version 1 of the Amp and the PSU. To start with I am thinking of AC filament supply and hoping that it will be hum free. We are obviously going for fixed biasing.

Comments pls.

Quadtech-ji
Did you get those pots in India?

Ver 1 Amp


PSU

Aks

If you are unsure of ac heater hum then why use it at all ?

Dc ?

Sorry but that's a noobie question.

Manek

Hi,

AKS,Viren,Siva, what are your opinions on this post.

http://www.audioasylum.com/forums/set/messages/5/58985.html

I have interacted with Jeff Medwin the poster to rebuild my RCA field coil drivers.(it is still a work in progress,after I damaged one of the drivers voice coils due to my ham handedness.)He is an old timer and one of the few people who have the skill and knowledge to repair and restore vintage drivers.

Would this type of power supply increase the cost substantialy?

Regards
Rajiv

Mod-sahib

I am sure of the hum but unsure just how much of it will be there.
If tolerable then I will prefer AC heating for the simplicity of it. DC is always an option for later.

Rajiv-ji

Low impendence is always a good thing in a PSU. I completely agree within reasonable limits.

I always aim for the lowest impendence possible in the PSU but in the end its always a study in compromises. The PTX donated by Shri Viren for the Santa Claus amp had a secondary DCR of 22ohms. I had deliberately overspeced the current rating so that DCR is low.

For this build I will need to order custom wound PTX and I am hoping that Primary DCR will be less than 7R and secondary DCR about 20R. In PTX such ratings are feasible.

The DELTA 10H/250mA Choke has a DCR of 45ohms. Quite acceptable. But I am a little worried with the DELTA 20H/180mA Choke. It has a DCR of about 120R.

Chokes with DCR of 10ohms are frightfully big and expensive.

There was one big and couple of small errors in the PSU (VER 1.0) schema posted yesterday. A corrected version (VER1.1) is posted below. I regret the errors.

This is a birds-eye-view of the amp layout.



I prefer the Japanese way of laying out things. Function over form. The Japanese gurus put a great emphasis on how signal and power lines should move within the amp, maintaining isolation from each other and follow the shortest path. The result is often not a very beautiful or symmetrical amp layout, but a very sensible and industrial like approach. They seem to apply Kaizan practices in every sphere of their activity.

Our amp somewhat tries to fit this “Gemba” philosophy. The Yellow Lines are power. The Red is signal. Both lines follow a very logical step-by-step travel path and stay away from each other. Yellow enters from back, makes a turn at filter condensers before ending with the OPTs. The Red line (audio signal) enters from right side, travel just a few inches to the DACT attenuator; travel a couple of inches to the 5842, another inch from there to the interstage transformer and then a short length (thro a ferrite bead) to the gird of 300B. The ground bus will be 2.5dia copper and segregated by section before converging at one central point. The earth will be “lifted” from the main chassis safety ground.

This is how it will look from side 3 quarter view. Please note that the Interstage transformer will reside inside the amp. Also, the Power transformer as show is actually the Output transformer from my 845SE DHT amp. Placed here just for reference. I believe the actual power transformer will be somewhat similar in size.

The driver tube we are using is 5842 super tube. It’s a little power house designed for military/industrial use. An audio application was never intended before the tube gurus discovered its virtues.

Don’t go by its diminutive size. It can pull in almost 40mA at its max. Try doing that with a 6SN7 and you have a quickly melting tube.

Just compare its size with a typical small signal noval tube like the ECC82


Anybody wants to read what, why and how we are deciding the driver operating parameters? We shall try in our usual haphazard way to present that.

aks07 schrieb:

Ok. I spent my morning to round these lovely ladies for a photoshoot so that you yahoos can salivate a bit First a pop quiz. Can you guys ID the tubes? 8 tubes from left are pentodes/tetrodes, and the rest are power triodes. The pentodes/tetrodes The DHTs

Hello Screamgigi, gang:
Looks like this is a very interesting discussion happening w.r.t. DHT amps & I have been missing it for quite some time. I have to admit that I have been busy @ work & home & have not spent much time on this forum except for an occasion post here/there.
Looks like I have much catching up to do on this thread before I can make any suggestion (if any).

I'm not very good at tube identification given my limited experience but to show everybody my ignorance here are my answers:
the pentodes/tetrodes:
L->R: 6550, EL34, looks like a TungSol 6550, ?, ?, KT88, ?, ?

the DHTs:
L->R: 805, 6AS7, 300B, ?, 845.