Darius Loftin White explained

[oldeurope]

On topic please

On topic please, thanks

 

[oldeurope]

waiting ...

Hello,
I am still waiting ...

Originally #42 posted by SY
I'm sorry, I just don't have an English word for it. I'll look in a couple of texts and see if there actually is one.

I am the only who didn't know this well known factor, sorry.
It is the most important mathematical tool in triode circuit design.
Thus I am sure the experts here are able to tell me the name of it.

Thanks,

Darius

 

[oldeurope]

Loftin White ... "Ultrapath" ... Broskie

Hello,
anyone noticed this blog by Broskie? read me

quoted from Radio News Feb.1930: "A condenser, C2, of about 1 microfarad is needed to form a local signal circuit in the output circuit including association of any suitable loud speaker as indicated." link

This is the so called "Ultrapath" and the green line shows this
local output circuit in my fundamental schematics.
http://rl12t15.blogspot.com/2008/03/ecc83-300b-darius-loftin-white.html

Schematic

http://rl12t15.blogspot.com/2007/09/pc86-rl12t15-darius-loftin-white.html

Schematic

I am sure Broskies text will help you understanding the "Ultrapath
aspect" of a true Loftin White design. Thanks to JRB.

Kind regards,
Darius

 

[Sheldon]

Go ahead and say it Darius. All that work and he almost got back to 1930, where the entire amp is floating with the output cathode - and no capacitor adjustment, no big supply caps, and no coupling cap.

Sheldon

 

[korneluk]

Darius,

In your schematic, where is the 1uF cap from the ground input to the cathode resistor of the input tube as shown on the original schematic? Is it necessary?

An externally hosted image should be here but it was not working when we last tested it.

Happy new 2009,

 

[Sheldon]

Darius exited the forum. A little to loose here for him, maybe. In any case, C2 on his schematic is the 24u cap, and Chb is the 2.2u cap. C2 on his schematic is just shown on the right side of the tube, instead of the left. But if you look closely, you will see that the connection is exactly the same (top of cathode resistor to B+). Yes, it is necessary.


Sheldon

 

[korneluk]

Darius exited the forum. A little to loose here for him, maybe. In any case, C2 on his schematic is the 24u cap, and Chb is the 2.2u cap. C2 on his schematic is just shown on the right side of the tube, instead of the left. But if you look closely, you will see that the connection is exactly the same (top of cathode resistor to B+). Yes, it is necessary.


Sheldon

C2 and Chb are clear. I do not see the C1 cap at the input and the 6SN7 grid leak resistor goes straight to ground.

here is a clearer picture:

An externally hosted image should be here but it was not working when we last tested it.

 

[Sheldon]

Let's do this in one or the other thread, not both.

Sheldon

 

[JLH]

I want to bounce something off the group here. Which example would be the better solution in your opinion and why? Thanks in advance for your constructive criticism.

Example #1

Or
Example #2

Rgs, JLH

 

[Sheldon]

In either case, I'm not sure that I see the benefit. Darius' triode LW version references the signal circuit to the top of the cathode output resistor, to eliminate PS ripple from the output. The original does the same, but uses a pentode input tube.

In version 1, the screen grid already isolates the output tube from the effects of ripple in B1. The CCS on the input tube should isolate it from B2 ripple. The ripple compensation circuit will eliminate ripple from B2 on the screen grid, but it should be easy to get that low current supply quiet anyway.

In version 2, more or less the same things apply, but with one less cap. I'd do that one.

Sheldon

 

[oldeurope]

Originally #67 posted by korneluk


C2 and Chb are clear. I do not see the C1 cap at the input and the 6SN7 grid leak resistor goes straight to ground.

here is a clearer picture:

An externally hosted image should be here but it was not working when we last tested it.

Hello korneluc

thanks for your question. I am using an input triode. The 'well known factor' for this triode is > 4 therefore I don't need this cap.

Kind regards,
Darius

 

[oldeurope]

Pentode output

@ #70
Thanks Sheldon
@ #69
Hello JLH
Example #1 brakes the 'ultrapath'. #2 is the better solution, you can take the g2 voltage from +B1.

Kind regards,
Darius

 

[korneluk]

Folks,

The Cunningham CX-350 data sheet specifies that "the '50 is not recommended for use in resistance-coupled circuits having a grid resistor". However, in the original LW circuit the grid of the 50 is directly connected to the plate of the 224. See schematic posted above.

Does this arrangement violate the spec?

-- josé k.

 

[kmaier]

Very true.... but later when the type 50 in ST glass became available, RCA changed this (newer manual)... stating that any conventional type of coupling may be used provided the resistance added to the grid circuit by this device does not exceed 10,000 ohms. That's a pretty low resistance, implying that the 50 can get into a runaway condition.

In any case, you bring up a valid concern. In a traditional self-bias circuit, you could get runaway with a higher grid resistance and things would get ugly... I have a couple 2A3 tubes which do this, hence they may be a bit gassy. As you have a DC coupling from the 224 with a current flow of 1ma (thru the 224), it's unlikely to have the runaway problem as the 1ma of current should be far greater than the grid leakage current of the 50 triode. I'd probably be inclined to increase the driver idle current to be on the safer side which would improve the slew rate for the driver.

Regards, KM

 

[oldeurope]

@ #73 and #74

Please read the radio news articles, thanks.

http://rl12t15.blogspot.com/2007/09/pc86-rl12t15-darius-loftin-white.html

 

[korneluk]

Re: @ #73 and #74

Please read the radio news articles, thanks.

http://rl12t15.blogspot.com/2007/09/pc86-rl12t15-darius-loftin-white.html

Hallo Darius,

I have part 3 of the Radio News article series (not in your blog) where it shows the actual part values using a 45 tube. Please email me offline and I'll be glad to send them to you.

Grüße

-- josé k.
jose_korneluk AT bellsouth.net

 

[oldeurope]

@ #76

Thanks José,
I sent you an email.

In my old circuit I had some trouble with drifting bias current for the RL12T15.
The resistor 220KΩ+390KΩ was much too high and little voltage drops made
big changes in bias because of the relatively low cathode resistor 1KΩ.
The Loftin White topology solved this problem and reduced the AC load impedance
for the driver stage. schematic
comes from

Kind regards,
Darius

 

[oldeurope]

Re: @ #76

correction:

The Loftin White topology solved this problem and reduced the AC load impedance
for the driver stage.

The Loftin White topology increased the AC load impedance for the driver stage.
Means less load for the driver stage.

Have a look at #69 The triode works in no load operation (current source). Loftin White makes this possible.
See this one bootstrap

comes from

 

[kenpeter]

Does your 220K resistor present higher dynamic resistance
to the ECC83's plate when so bootstrapped? More constant
current than pure resistance?

If so, was that your main intention for this bootstrapping?
Or you had some other important reasons in mind?

 

[Sheldon]

It takes some work to get all the operating values set for a true LW amp, as they are all interdependent. However, once set, the operating points are very stable and self correcting.

On question Darius, on the clever bootstrap. How is the output impedance calculated? The feedback from the bootstrap should increase the impedance. From this article: http://www.tubecad.com/articles_2002/SRPP_Deconstructed/page5.html

It suggests that the amount of the output fed back to the input is calculated according to the ratio between the resistor from plate to feedback cap, and the rp of the input tube (in this case, about 225K). So for this example, about 50% of the output is fed back to the input, in phase. I assume that the output impedance of the follower would be increased by approximately double?

Sheldon