Darling En Vogue Amplifier with Parallel 1626's SE

ask your general tube related questions here

Darling En Vogue Amplifier with Parallel 1626's SE

Postby PIANOLYDIA » Wed Jun 05, 2013 12:41 pm

Hi All
Sorry, But the size of the drawings and limited conversion abilities on this site, It came out a little unreadable to say the least, but;..If you click on the image then go to View image and zoom in you should be able to see the writing if you will. However, Due to the fact that My friend and I are using Old good working Wards Console Tube transformers. We have decided to use this amplifier in a Tri Amplifier Speaker setup, using this one as the Compression driver Amp Only, So about 2000hz, on up! If you see anything I missed let me know. You will notice the PSU II Simulation I ran and this is as close to reality as it will get!. What about biasing the 12AV7's with battery's instead of Cathode biasing? And, Is the bypass switch for the Power on/off and the Filament switch hooked up right? Any Advice would be appreciated, and Shannon Parks if your there, You already Know about this amplifier, and If you see anything that could improve this design let me know. Thanks and bless all of you! Lydia

P.s. CHANGES HAVE BEEN MADE ADDED TWO LED's TO BIAS 1/2 12AV7, PER CHANNEL
Attachments
1626 amp.GIF
Taking into consideration from Ed Brown, I placed (2) Red LED's in series to bias the 1/2 12av7 per channel
1626 amp.GIF (20.55 KiB) Viewed 4913 times
1626 En Vogue Darling.JPG
1626 En Vogue Darling.JPG (382.49 KiB) Viewed 5011 times
Last edited by PIANOLYDIA on Fri Jun 07, 2013 3:04 am, edited 9 times in total.
PIANOLYDIA
 
Posts: 19
Joined: Sat May 04, 2013 2:04 am

Re: Darling En Vouge Amplifier with Parallel 1626's SE

Postby EWBrown » Thu Jun 06, 2013 9:55 am

Looks good! If you can determine the 12AV7's cathode voltage, then try using an LED (or LEDs in series) in place of the resistor and its bypass capacitor.

LED cathode biasing works very well for voltage amplifier stages. They have a very low effective impedance, and will avoid the
pitfalls of a capacitor "coloring" the sonic quality, all the more important at the higher frequencies. Cathode LEDs should not be capacitor bypassed.

HTH

/ed B
Real Radios Glow in the Dark
User avatar
EWBrown
Insulator & Iron Magnate
 
Posts: 6389
Joined: Wed Mar 19, 2003 6:03 am
Location: Now located in Clay County, NC !

Re: Darling En Vouge Amplifier with Parallel 1626's SE

Postby 20to20 » Thu Jun 06, 2013 12:56 pm

EWBrown wrote:Looks good! If you can determine the 12AV7's cathode voltage, then try using an LED (or LEDs in series) in place of the resistor and its bypass capacitor.

LED cathode biasing works very well for voltage amplifier stages. They have a very low effective impedance, and will avoid the
pitfalls of a capacitor "coloring" the sonic quality, all the more important at the higher frequencies. Cathode LEDs should not be capacitor bypassed.

HTH

/ed B


Ed,

Do LED's at that location nullify the self-generated FB like a cap would? How would a diode's action there be different from a full frequency bypass cap?
Headed for Tishomingo to sing in a can...
User avatar
20to20
KT88
 
Posts: 566
Joined: Wed Apr 21, 2010 1:48 pm
Location: W-S, NC

Re: Darling En Vogue Amplifier with Parallel 1626's SE

Postby EWBrown » Fri Jun 07, 2013 7:16 pm

The bypass cap, especially if it is an electrolytic will have inductive properties, and can create weird undesirable resonances, and uneven frequency response.

The LED should not nullify the inherent degenerative feedback that an unbypassed cathode resistor would cause, though the self-generated negative feedback effect might be a less.

The use of an LED is dependent on the expected cathode bias, different colors will have different forward voltage drops, the "old style" cheap red, orange,
amber, yellow and "slime green" LEDs are best, the newer 525 nm "true green", blue, white, pink, UV, etc, might not perform as well - I have not tried these newer generation LEDs for cathode biasing.

Though I have read of using 3 larger (5050) SMD white LEDs, connected in series for EL84 cathode biasing, with a B+ supply between 300 and 325VDC, the LEDs will place the EL84 cathode at 10-11 VDC above ground, which is correct for this B+ supply voltage range, I have not yet tried this method either..

IR LEDs have forward voltage drops between 1 and 1.2V, but will not emit visible light, except for some of the 880 NM IR LEDs which may appear as a "dark cherry red" to some people - my vision can detect these in a darkened room. 880 nm is the longest wavelength for most human vision.


The LEDs I use most often are the small HP (now Avago) HLMP6000 which have a forward voltage drop of 1.56 to 1.6VDC, these are fairly inexpensive, and are halfway between being a "leaded" LED and a surface mount, the leads are very thin and flat, and can break off rather easily, if excessive tension is placed upon them..
I bought a "lifetime supply" of these from Jameco several years ago, and the voltage drops have been very close tolerance, they were from the same manufacturing batch. I've also used the cheap 3mm red LEDs from Radio shack, they fall between 1.6 and 1.8 VFD. Amber and yellow LEDs fall between 1.8 and 2 VDC, and the lime green ones, 2 to 2.2 VDC. If higher voltages are necessary, just connect two in series (LED cathode towards 0 volts / signal ground, anode to the tube's cathode.

I would suggest, start with the cathode resistor, then try the LED, and determine which approach sounds better.

The "ultimate" application would be to use the LED for cathode bias, and an appropriate constant current source in place of the B+ resistor, this gets the most gain,
and generally lower noise, out of a voltage gain stage, than the standard plate and cathode resistor approach.

I have used this method in a 6DJ8-based phono stage, which is otherwise quite similar to Shannon's "Budgie" design, the main difference being that I operated the triodes at around 3.5 mA, with a 150VDC B+ voltage. Otherwise, 18K plate resistors, and 430 ohm cathode resistors would be used.


I had some confusion with the R/C coupling to the 1626s' grids, until I re-read your explanation about using it as an HF compression horn driver, with an LF roll-off
of 2 KHz.

HTH

/edB
Real Radios Glow in the Dark
User avatar
EWBrown
Insulator & Iron Magnate
 
Posts: 6389
Joined: Wed Mar 19, 2003 6:03 am
Location: Now located in Clay County, NC !

Re: Darling En Vogue Amplifier with Parallel 1626's SE

Postby DeathRex » Fri Jun 07, 2013 9:07 pm

I like to use a small resistor in series with the LED to give a little cathode FB. Like instead of a 480 ohm resistor and 200uf cap bypassing it, use a LED or 2 to give the bias + a 68 ohm resistor.
At first I wanted to be a ET. Now I are one.
User avatar
DeathRex
KT88
 
Posts: 1014
Joined: Wed Nov 18, 2009 8:13 pm
Location: Cortez, CO

Re: Darling En Vogue Amplifier with Parallel 1626's SE

Postby 20to20 » Sat Jun 08, 2013 7:16 am

DeathRex wrote:I like to use a small resistor in series with the LED to give a little cathode FB. Like instead of a 480 ohm resistor and 200uf cap bypassing it, use a LED or 2 to give the bias + a 68 ohm resistor.



I guess the thought or (question still to ponder) in the back of my mind about using LED's on the cathode was whether or not a diode would act as a precision voltage regulator there and so either filter the cathode like a cap would of all the degenerative FB, or if it wasn't as good at filtering or regulating as a cap then there could be unwanted frequency response tailoring. I guess it boils down to how well an LED regulates at mV levels and would it block (regulate) just half of the AF there by diode action.

And what would an LED let you do on the cathode if you wanted to bring GNFB in there? Would it mess that up?

This sounds like a question for "Super (scope) Man"!
Headed for Tishomingo to sing in a can...
User avatar
20to20
KT88
 
Posts: 566
Joined: Wed Apr 21, 2010 1:48 pm
Location: W-S, NC

Re: Darling En Vogue Amplifier with Parallel 1626's SE

Postby jonnyeye » Sat Jun 08, 2013 4:22 pm

LEDs look like a voltage source in series with a resistor of varying linearity:
Image
(source: http://www.telatomic.com/electricity/planck.html)
From the graph you can see that above a certain current (roughly 8-10mA for red and green, maybe 15mA for yellow) the graph is largely linear and can be approximated by a voltage source in series with a small-valued resistor - from these graphs perhaps 2 ohms for the IR/red LEDs, maybe 20 ohms for yellow. While these graphs show the DC impedance I can't think of any reason why the AC impedance of the diode would be significantly different, at least for audio frequencies. Beneath 10mA the graphs are all quite non-linear and the series resistance increases significantly, which would decrease gain and probably affect distortion.
You could apply feedback to the cathode, but because of the low impedance it would have to be current feedback, not voltage feedback.
Vacuum tubes: now (mostly) chemical free!
User avatar
jonnyeye
 
Posts: 109
Joined: Thu Feb 12, 2009 3:04 pm
Location: Sunderland, ON

Re: Darling En Vogue Amplifier with Parallel 1626's SE

Postby 20to20 » Sat Jun 08, 2013 9:56 pm

jonnyeye wrote:LEDs look like a voltage source in series with a resistor of varying linearity:
Beneath 10mA the graphs are all quite non-linear and the series resistance increases significantly, which would decrease gain and probably affect distortion.
You could apply feedback to the cathode, but because of the low impedance it would have to be current feedback, not voltage feedback.


....And the sign said, "12ax and 12at, needn't ought to apply..."

And anything that won't handle at least 15-20ma.

Will a common LED even light off with only 10ma?
Headed for Tishomingo to sing in a can...
User avatar
20to20
KT88
 
Posts: 566
Joined: Wed Apr 21, 2010 1:48 pm
Location: W-S, NC

Re: Darling En Vogue Amplifier with Parallel 1626's SE

Postby PIANOLYDIA » Sat Jun 08, 2013 11:11 pm

Ed, do you want to comment on this? especially since we are using a 12AV7 Black plate Sylvania's
PIANOLYDIA
 
Posts: 19
Joined: Sat May 04, 2013 2:04 am

Re: Darling En Vogue Amplifier with Parallel 1626's SE

Postby PIANOLYDIA » Sun Jun 09, 2013 4:14 am

Let's take this a step forward especially for linearity purposes!

Transistor constant current sink: There are many possible CCS circuits using transistors, many of which become quite extensive. The one provided here is the simplest, requiring one NPN transistor and a minimum of components.
The red LED sets the base voltage at about 1.6V so the transistor is always 'on'. Rb is simply the current limiting resistor for the LED. Since it has to drop most of the HT at a few milliamps, it will usually need to be rated at 2W or more, unless there is a convenient, small positive supply voltage available. Alternatively, the base voltage could be acquired from the cathode of an LED biased gain stage.
The base-emitter drop is about 0.6V, therefore the voltage across the emitter resistor (Re) must be at 1.6 - 0.6 = 1V. The emitter resistor sets the current since there must be 1V across it, use Ohm's law to find its value. For a constant current of 1mA:
1 / 0.001 = 1k
Note, this circuit may not work if Re is less than about 400 ohms because Vbe increases with current. If you need to sink more than about 2mA it is better to raise the base voltage by using an LED with a higher forward voltage, or zener diode.

The AC impedance obtained will be roughly equal to:
r(ccs) = Re * (2 * hfe) + 1 / hoe
Where:
Re = the emitter resistor
hfe = rated current gain (the "2" comes from the fact that hfe tends to be about double its normal value when Ibe is less than about 10mA)
hoe = ac impedance of the transistor
But hoe is typically very small indeed, then to a close approximation;
r(ccs) = Re * (2 * hfe)
Below" IMAGE"
Below that, other versions of biasing images
Attachments
ccs4.jpg
ccs4.jpg (21.65 KiB) Viewed 4778 times
etf2004bias.jpg
etf2004bias.jpg (11.22 KiB) Viewed 4778 times
BB Section.jpg
BB Section.jpg (80.94 KiB) Viewed 4783 times
BB Section (cont).jpg
BB Section (cont).jpg (40.19 KiB) Viewed 4783 times
PIANOLYDIA
 
Posts: 19
Joined: Sat May 04, 2013 2:04 am

Re: Darling En Vogue Amplifier with Parallel 1626's SE

Postby EWBrown » Sun Jun 09, 2013 9:53 pm

I've used LED biasing with currents as low as 2 to 3 mA, with good results, the actual light intensity has no relevancy to the tube's operation.

Another approach, mentioned elsewhere on this forum, is to connect a 100 to 200K resistor from the anode of the LED to the "hot" end of the plate resistor -
what this does is to make the LED draw an extra mA or two of current, and may make it more stable, in low-current (1 to 2 mA) applications.
This should not significantly raise the cathode voltage, or effect the stage's actual plate / cathode current.

/ed B
Real Radios Glow in the Dark
User avatar
EWBrown
Insulator & Iron Magnate
 
Posts: 6389
Joined: Wed Mar 19, 2003 6:03 am
Location: Now located in Clay County, NC !


Return to tube 101

Who is online

Users browsing this forum: No registered users and 10 guests