Canuck Audio Mart Hifi and Audio ForumCanuck Audio Mart Hifi and Audio Forum
It is currently Wed Oct 17, 2018 2:03 am

All times are UTC - 8 hours




Post new topic Reply to topic  [ 58 posts ]  Go to page 1, 2, 3, 4  Next
Author Message
PostPosted: Sun Sep 09, 2018 11:38 am 
Offline

Joined: Wed Sep 14, 2005 11:47 am
Posts: 4057
Location: Ottawa, ON, CA
I've looked at the price of a number of tube preamps - and the cost Yikes !!!

So in the spring, I examined about 5 different tube preamp topologies
and decided that a gain of 1 was enough,
so I decided to develop a preamp / buffer based on a cathode follower.

The idea here is that its a circuit that could be shared with CAM members
and could be built at a reasonable cost.

After spending about 2 months developing it, I discovered John Broskie
had published the same circuit - AUGH !!!

The sonics are very transparent - but it can be "warmed up" by using an output coupling cap
like a Clarity Cap CSA.

So sticking my neck out here, I wonder if there is enough interest in a few CAM
members building this ?

Here is the basic concept .

The circle with the arrow indicates a CCS (constant current sink) and it sets the bias current.
The CCS can be implemented in many different ways, but the big surprise is how much it affects the sonics.
In the end, I decided its best to implement the CSS with a triode and a resistor.


Attachments:
img_2696_Cathode_Follwer_Schematic.jpg
img_2696_Cathode_Follwer_Schematic.jpg [ 118.86 KiB | Viewed 1688 times ]
Top
 Profile  
Reply with quote  
PostPosted: Sun Sep 09, 2018 11:57 am 
Offline
User avatar

Joined: Wed Feb 06, 2008 2:04 pm
Posts: 771
Location: London, ON, CA
Are you concerned with cost issues with the power supply section?

What is required DC. Voltage and Current?

What filament specs?

Good on you for tackling this project!


Top
 Profile  
Reply with quote  
PostPosted: Sun Sep 09, 2018 3:06 pm 
Offline

Joined: Wed Sep 14, 2005 11:47 am
Posts: 4057
Location: Ottawa, ON, CA
jst_canuck wrote:
Are you concerned with cost issues with the power supply section?

Not so much the cost of the power supply - but the fact that it will need to be a higher voltage.
PLus getting the components can be a pain.
I have the experience to guide someone through a basic supply
with all the grounding requirements (hydro grounding, shielding and circuit star grounding).

However, as CAM member Nakamichel has pointed out, the supply is 1/2 the amp.
So hopefully, he can push the level of a power supply beyond
a basic C L C L C C filter - which is already better than most tube amps.

jst_canuck wrote:
What is required DC. Voltage and Current?


For a 6dj8, I would suggest dual supply rails of +/- 100 Vdc.
I see John Broskie also used +/- 100 Vdc.

The bias current depends on the tube.
I would need to look at the plate curves for a particular tube.
But its important to have plenty of drive current for the power's amp front end.
For a 6dj8, I would set the bias current as high as 10 mA.

So that means a 6dj8 would be dissipating 1 Watt
which is well within its maximum plate dissipation of 1.8 watts.

jst_canuck wrote:
What filament specs?


The topic of heater voltage and current is an entire topic within itself.
Many tubes guys prefer to heat the heater filament with AC.
However, I prefer to use DC with quasi current limiting.
The data sheet for a 6dj8 says 6.3 V @ 365 mA per tube.

A 6dj8 is good choice for a cathode follower because it
has a mu (amplification factor) = 33
and a rp (plate resistance) = 2.65 K ohms
This gives the follower a Zout = rp / ( mu + 1) = 78 ohms
which is quite low for a tube preamp
Rc will increase Zout, but it can be by passed with a cap.

jst_canuck wrote:
Good on you for tackling this project!

Thanks - I hope there is enough interest
and also some of our tech savvy CAM members will join in.

.


Top
 Profile  
Reply with quote  
PostPosted: Mon Sep 10, 2018 6:22 am 
Offline
User avatar

Joined: Wed Feb 06, 2008 2:04 pm
Posts: 771
Location: London, ON, CA
Would you say that a low cost power supply is a target for your project :?:


Top
 Profile  
Reply with quote  
PostPosted: Mon Sep 10, 2018 9:03 am 
Offline

Joined: Wed Sep 14, 2005 11:47 am
Posts: 4057
Location: Ottawa, ON, CA
jst_canuck wrote:
Would you say that a low cost power supply is a target for your project :?:

In a word - No.

As CAM member Nakamichel has pointed out - the power supply is 1/2 the sound of an amp.
So a pretty serious effort should be spent on the power supply.

I can easily provide the schematic for a basic dual power supply,
which would better most tube preamps.

Also note the circuit itself has a CCR (constant current source) - the circle with the arrow though it.
This will greatly improve the PSRR (power supply rejection ratio).

Add to this - Notice how the output voltage is set to 0 Vdc BEFORE the output coupling cap.
At first glance this seems unnecessary.
However, the dual B + and B - power supply allows the noise in each supply rail to cancel out
(providing the ripple and noise are equal and opposite in phase on each rail).

But even with all the advantages of this circuit - the power supply still affects the sonics.

Here is the circuit schematic from John Broskie.
The CCS is implemented with the lower triode and 2 supporting resistors.


Attachments:
Tube Buffer - Preamp.jpg
Tube Buffer - Preamp.jpg [ 280.01 KiB | Viewed 1434 times ]


Last edited by Uunderhill on Mon Sep 10, 2018 10:37 am, edited 1 time in total.
Top
 Profile  
Reply with quote  
PostPosted: Mon Sep 10, 2018 10:28 am 
Offline
User avatar

Joined: Wed Feb 06, 2008 2:04 pm
Posts: 771
Location: London, ON, CA
Uunderhill wrote:
jst_canuck wrote:
Would you say that a low cost power supply is a target for your project :?:

In a word - No.

As CAM member Nakamichel has pointed out - the power supply is 1/2 the sound of an amp.
So a pretty serious effort would be spent on the power supply.

I can easily provide the schematic for a basic dual power supply,
which would better most tube preamps.

Also note the circuit itself has a CCR (constant current source) - the circle with the arrow though it.
This will greatly improve the PSRR (power supply rejection ratio).

Add to this - Notice how the output voltage is set to 0 Vdc BEFORE the output coupling cap.
At first glance this seems unnecessary.
However, the dual B + and B - power supply allows the noise in each supply rail to cancel out
(providing the ripple and noise are equal and opposite in phase on each rail).

But even with all the advantages of this circuit - the power supply still affects the sonics.

Here is the circuit schematic from John Broskie.
The CCS is implemented with the lower triode and 2 supporting resistors.

OK, I misunderstood...... I was thinking that you were looking for more ideas and solutions from those who may have more experience with electronics :?:

Still great that you are going to do all the prototyping and testing and then share the schematics and parts list with CAMers 8)
All good :D


Top
 Profile  
Reply with quote  
PostPosted: Mon Sep 10, 2018 11:07 am 
Offline

Joined: Wed Sep 14, 2005 11:47 am
Posts: 4057
Location: Ottawa, ON, CA
jst_canuck wrote:
OK, I misunderstood...... I was thinking that you were looking for more ideas and solutions from those who may have more experience with electronics :?:


I'm absolutely interested in getting power supply ideas from other CAM members.

With + /- 100 Vdc rails this points to a passive power supply.
A simple C L C L C C is is quite straightforward, but if any one wants to improve on that
that would be great.

Also, the power supply for the heater is quite important too.
The topic of heater currents is seriously debated.

However, I am quite comfortable with topics such as Star grounding for the audio circuit
and lifting hydro earth.

jst_canuck wrote:
Still great that you are going to do all the prototyping and testing and then share the schematics and parts list with CAMers 8) All good :D


I've already built a low voltage version using a 6gm8 tube and this version is sitting in my big system.

6gm8 tube pre ===> Belles 350 A power amp ===> B&W N804's

Putting my ear up to the B&W tweeter, I can't tell the system is on.
However, please understand this preamp / buffer is very transparent.
Think of it as a tube version of Nelson Pass's B1 buffer.
Actually, it was Erno Borbley who should get credit for the B1

When I first started prototyping, I had no idea that the CCS could have such an impact on the sonics.
So my first version used back to back MPSA18 bi polar transistors to implement the CCS.
I tried all sorts of variations on this.
Then I switched to implementing the CCS with different jFET's
and in the end used a 2sk246 - definitely an improvement over a bi polar CCS.

However, a 6dj8 is an improvement on a 6gm8
and implementing the CCS with the same triode instead of a jFET is a better option.

After prototyping this circuit for about 2 months, I know each resistor on a personal basis.
However, I have not tried a version with a high voltage tube like a 6dj8.

Have you developed tube circuits, or worked with tubes before ?


Top
 Profile  
Reply with quote  
PostPosted: Mon Sep 10, 2018 12:39 pm 
Offline
User avatar

Joined: Wed Feb 06, 2008 2:04 pm
Posts: 771
Location: London, ON, CA
Uunderhill wrote:
jst_canuck wrote:
OK, I misunderstood...... I was thinking that you were looking for more ideas and solutions from those who may have more experience with electronics :?:


I'm absolutely interested in getting power supply ideas from other CAM members.

With + /- 100 Vdc rails this points to a passive power supply.
A simple C L C L C C is is quite straightforward, but if any one wants to improve on that
that would be great.

Also, the power supply for the heater is quite important too.
The topic of heater currents is seriously debated.

However, I am quite comfortable with topics such as Star grounding for the audio circuit
and lifting hydro earth.

jst_canuck wrote:
Still great that you are going to do all the prototyping and testing and then share the schematics and parts list with CAMers 8) All good :D


I've already built a low voltage version using a 6gm8 tube and this version is sitting in my big system.

6gm8 tube pre ===> Belles 350 A power amp ===> B&W N804's

Putting my ear up to the B&W tweeter, I can't tell the system is on.
However, please understand this preamp / buffer is very transparent.
Think of it as a tube version of Nelson Pass's B1 buffer.
Actually, it was Erno Borbley who should get credit for the B1

When I first started prototyping, I had no idea that the CCS could have such an impact on the sonics.
So my first version used back to back MPSA18 bi polar transistors to implement the CCS.
I tried all sorts of variations on this.
Then I switched to implementing the CCS with different jFET's
and in the end used a 2sk246 - definitely an improvement over a bi polar CCS.

However, a 6dj8 is an improvement on a 6gm8
and implementing the CCS with the same triode instead of a jFET is a better option.

After prototyping this circuit for about 2 months, I know each resistor on a personal basis.
However, I have not tried a version with a high voltage tube like a 6dj8.

Have you developed tube circuits, or worked with tubes before ?

I have worked with tubes and most other electronic components in my career and even my hobbies with vintage radio and audio equipment.
Very familiar with point-to-point wiring but would never build any active electronics using this technique.
Printed circuit boards are so much more consistent and stable :idea:

Sadly, I'm not one of those rare individuals still living who can claim to have "developed tube circuits"
as very intensive research long ago discovered and applied nearly all imaginable arrangements of hollow state circuitry.

In John Broskie's Guide to Tube Circuit Analysis & Design https://www.tubecad.com/2016/05/blog0346.htm
he refers to this patent from long before I was born:

https://patents.google.com/patent/US2358428 titled Thermionic valve amplifier circuit arrangements assigned to EMI Limited in 1943 :!:

Would be very humbling to sift through all of that state-of-the-art IP from the period from 1935-55 when the vacuum tube was King :D

EMI, RCA, Telefunken and Bell Labs are just a few that hired the best and brightest to improve tube applications around the world.
The RCA Tube Manuals, Bell Labs Journals and all US Patents are free online to learn about the "high tech" of those times.

Always possible that some new hybrid designs can still be developed by combining solid state, hollow state or even active magnetic devices.

The end goal being to use circuits that "sound good", regardless of when they were originally developed, right :?:


Top
 Profile  
Reply with quote  
PostPosted: Mon Sep 10, 2018 12:41 pm 
Offline

Joined: Wed Sep 14, 2005 11:47 am
Posts: 4057
Location: Ottawa, ON, CA
Here is the schematic for the actual version I'm using.
It uses a 6gm8 for the follower - this is a low voltage VHF tube designed for car radios.
These tubes are difficult to find.
The CSS is implemented with a 2sk246 jFET and two resistors.

Here is the completed version in my system right now - listening to Mozart Piano Concerto No. 9.
The only to way tell if its on is that the music is playing.
It dead silent - no hiss - no hum.

The wiring was done P to P - but the bread board allowed me to make changes to the CCS.

However - like a true audiophile - "It sounds fabulous - but I want to improve it."

A 6gm8 has a gm ~ 2.6 mA/V where as a 6dj8 has a gm ~ 12.5 mA/V
so a 6dj8 would give a cathode follower a much lower Zout.


Attachments:
Finished Preamp in System.jpg
Finished Preamp in System.jpg [ 296.59 KiB | Viewed 1393 times ]
Preamp Schematic- jFET version.jpg
Preamp Schematic- jFET version.jpg [ 98.66 KiB | Viewed 1393 times ]


Last edited by Uunderhill on Mon Sep 10, 2018 1:13 pm, edited 1 time in total.
Top
 Profile  
Reply with quote  
PostPosted: Mon Sep 10, 2018 1:12 pm 
Offline

Joined: Wed Sep 14, 2005 11:47 am
Posts: 4057
Location: Ottawa, ON, CA
Uunderhill wrote:
. Have you developed tube circuits, or worked with tubes before ?
jst_canuck wrote:
I have worked with tubes and most other electronic components in my career and even my hobbies with vintage radio and audio equipment.

Very familiar with point-to-point wiring but would never build any active electronics using this technique.
Printed circuit boards are so much more consistent and stable :idea:

Sadly, I'm not one of those rare individuals still living who can claim to have "developed tube circuits"
as very intensive research long ago discovered and applied nearly all imaginable arrangements of hollow state circuitry.


Excellent - so you are OK with working with +/- 100 Vdc supply rails ?

If you have any suggestions concerning the power supplies - I'm absolutely listening.

I spent 2 months "developing this circuit" - only to discover John Broskie
had already published it.

This first version is not up to Frank's standard of building - however, its a prototype.
Plus I made the case myself.

An all triode version could be built using P to P
but having circuit boards would make life so much easier.
That's a hint directed towards Frank.

Notice I used a shunt volume control.
Some DIY types frown on this - but it allowed the input to be connected straight from the RCA into the Grid
via 1 resistor - no carbon pots - or switch contacts in between.
In order to do this, an extender rod was needed for the volume control.
Plus a plastic extender rod with a wood base isolates the person from the voltages.
A circuit board would make it much easier to set up an extender rod.
.


Top
 Profile  
Reply with quote  
PostPosted: Tue Sep 11, 2018 1:25 am 
Offline

Joined: Sun Dec 28, 2008 7:49 am
Posts: 1937
Location: Fairport Beach (in Pickering), ON, CA
Here's an excellent free program for designing power supplies:

http://www.duncanamps.com/psud2/

_________________
Tom

Collecting vintage tube Pilot (Pilotone) HiFi gear.


Top
 Profile  
Reply with quote  
PostPosted: Tue Sep 11, 2018 3:58 am 
Offline

Joined: Wed Sep 14, 2005 11:47 am
Posts: 4057
Location: Ottawa, ON, CA
Tom_r wrote:
Here's an excellent free program for designing power supplies:
http://www.duncanamps.com/psud2/

Tom,

Looking at John Broksie's schematic, notice how Vout = 0 Vdc BEFORE the output coupling cap.

So if the noise and ripple on the B + and B - rail are equal, but opposite in phase,
then much of the noise is canceled at the output.
Its a neat trick.

So its critical that any power supply filtering on the B + rail, is identical to the filtering on the B - rail.

Even though this is a tube circuit, I would suggest solid state rectification.
Its simpler than getting into a matched pair of tube rectifiers.
They are WAY in excess - but those silicon rectifier bridges I used make life so much simpler.

However, CAM member Nakamichel pointed these diodes out
https://www.mouser.ca/ProductDetail/Inf ... FydCXMQ%3d
These soft recovery diodes may be total excess - however, if they remove the need for a snubber circuit,
then they are well worth it.
.
In terms of power supply current, looking at the plate curves for a 6dj8,
they can safely handle 10 mA at 100 Vdc.
So the total power supply current would be 20 mA for both channels.


Attachments:
Tube Buffer - Preamp.jpg
Tube Buffer - Preamp.jpg [ 280.01 KiB | Viewed 1348 times ]
Top
 Profile  
Reply with quote  
PostPosted: Tue Sep 11, 2018 8:22 am 
Offline
User avatar

Joined: Wed Feb 06, 2008 2:04 pm
Posts: 771
Location: London, ON, CA
Uunderhill wrote:
Uunderhill wrote:
. Have you developed tube circuits, or worked with tubes before ?
jst_canuck wrote:
I have worked with tubes and most other electronic components in my career and even my hobbies with vintage radio and audio equipment.

Very familiar with point-to-point wiring but would never build any active electronics using this technique.
Printed circuit boards are so much more consistent and stable :idea:

Sadly, I'm not one of those rare individuals still living who can claim to have "developed tube circuits"
as very intensive research long ago discovered and applied nearly all imaginable arrangements of hollow state circuitry.


Excellent - so you are OK with working with +/- 100 Vdc supply rails ?

If you have any suggestions concerning the power supplies - I'm absolutely listening.

.....
.


I'm comfortable working with supplies up to about 400 VDC :)

Thanks to Tom for the link to Duncan's very useful software for a foundation for a linear power supply http://www.duncanamps.com/psud2/

Wondering how you plan to evaluate use of components such as soft-recovery diodes, snubber circuits and toroidal transformers?

Is component cost a consideration for the project:?:


Top
 Profile  
Reply with quote  
PostPosted: Tue Sep 11, 2018 8:28 am 
Offline

Joined: Sat Sep 01, 2012 5:08 am
Posts: 699
Location: Montréal, QC, CA
Hi Uunderhill

I guess you already know that I will never build or even design a PS at +/-100v.
It's too dangerous.

But if you want to build a good PS yourself, I can briefly tell you what I think is important to know when designing a PS.

The ripple and noise are not the only important things.
A constant Z factor at all frequencies is what makes an amp transparent. More than anything else in the amp.
It's why the PS is so important, it plays a big role in the Z factor.

A PS makes some small ripple at idle. But as soon you take current, the ripple grows. Higher the current demand, higher the ripple.
Why ? It's because the energy is just not there. Never we will have a flat DC at any current demand, impossible.
There's no energy (instantaneously) available higher than the RMS voltage. Capacitors are just a storage. But if we use 'em, we have to refill 'em.
So, capacitance filtering is not enough. The solution is to lower the sinewave and by consequent, using the real energy a bit longer.


Does my clipper PS would be good for your preamp ?
Absolutely. The clipper is good not only because of the lower ripple, but also because the Z becomes flatter and stronger at audio frequencies.
This is the trick (not my clipper, but having a flat Z) that makes a good power supply. But since my clipper help to equalize the impedance...

Refer to my thread on power supply.
About the clipper itself, the shunted Zeners are only to get the reference point voltage. The clipped part is not shunted to the ground
as some guys thought it was the way it works. The clipped part is dissipated through the output. So, depending on the current demand,
the clipped part can actually provide instant current very quickly to withstand that current demand.

Here a very basic clipper, which nevertheless will improve (in transparency) any power supply.
In this version, the clipping is done after the rectifier. It's an easier version for DIYers.
Only 3 parts; Z R Q.

Attachment:
basic-clipper.jpg
basic-clipper.jpg [ 51.81 KiB | Viewed 1300 times ]
Choose the zener the same as the RMS transformer sec. voltage (not .707x DC).
So if you need +100v DC, you choose a transformer 100v AC (and not 70v AC). Once rectified and clipped, it will provide 100v DC.
The resistor will set the right current to bias the transistor's base according to the current wanted.

And do not forget, the zener diodes are not the only way to clip a sine, there are better things...

After the clipper, you can choose any filtering method you want, hopefully better than just a regulator as showed above.

Mik

_________________
Keep things simple and even simpler
Nakamichi BX-300/Balance BW-1010 DAC, Balance MR-50R, Infinity RS 8 Kappa
Nakamichi Dragon/Onix Dac25b, Audio Research SP-9, Spectral DMA-100s/Nakamichi PA-7II, Martin-Logan CLSII/Balance Helicia


Last edited by Nakamichel on Wed Sep 12, 2018 11:23 pm, edited 1 time in total.

Top
 Profile  
Reply with quote  
PostPosted: Tue Sep 11, 2018 4:01 pm 
Offline

Joined: Wed Sep 14, 2005 11:47 am
Posts: 4057
Location: Ottawa, ON, CA
Its interesting watching a preamp design, evolve into a discussion about a good power supply design.
As good as the PSRR (power supply rejection ratio) of this tube preamp circuit is,
having a good power supply is still 1/2 the sound.

I thought the clipping stage needed to be implemented BEFORE the diode bridge,
however, if a simpler version can be installed after the bridge that's great.
Simple is good.

For those who haven't read it, here is Nakamichel's forum describing clipping power supply theory in simple terms.
http://www.canuckaudiomart.com/forum/vi ... 21&t=49583

Here is the data sheet for Zener diodes - basic bread and butter Zeners are available up to 100 V
https://www.vishay.com/docs/85816/1n4728a.pdf
.


Top
 Profile  
Reply with quote  
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 58 posts ]  Go to page 1, 2, 3, 4  Next

All times are UTC - 8 hours


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group