tube line amp-does this look ok

[bigugly]

I threw this together after reading everything I could about tube amps that were posted on this board.

My previous post about the tube pwr supply will be running this line amp. I am building it for my passive mixer.

If anyone has sugestions for improvement they would be very welcome.

thanks,
James

 

[pstamler]

For some reason I'm having trouble viewing the schematic in any detail...be that as it may, from what I can see the second section of the second tube has no connection to B+, and is directly connected to the cathode of the first section.

Peace,
Paul

 

[alk509]

[quote author="pstamler"]from what I can see the second section of the second tube has no connection to B+, and is directly connected to the cathode of the first section.[/quote]

Hey Paul,

I think it's some kind of series amp (SRPP? WCF?) drawn in a really weird way, that's all. :wink:

Peace,
Al.

 

[moamps]

The both second triodes haven't a DC grid bias.
The first triode's grid bias in the second tube can't be 0V.

Regards,
Milan

 

[NewYorkDave]

That output stage looks like an attempt at a White cathode follower, but it's all wrong. Rather than do your debugging for you--which wouldn't help you to learn--I'll just say that you should study up some more on WCF circuits and try again.

Oh, and there's also a big problem at the grid of U1B. Can you spot it?

The parts values tell me that you were probably looking at some of my circuits. Compare mine to yours and that might help you to spot some of the errors.

 

[analag]

This is a pentode input Mic pre utilizing a NFB technique that NYD used in one of his pre amp schemo's. This is a standard pre featuring two gain stages and a buffer. The problem with using a pentode leaves the designer with the added consideration of what to do with the second stage of the 12AT7, one answer is to make it a CCS which also effectively splits the B+ rail almost down the middle, and thus bias the WCF optimally. No cathode bypass is used as it is my personal preference to avoid them as much as possible, to my ears it is a tradeoff between incremental gain and sonic resolution.

Analag

 

[NewYorkDave]

"NFB technique that NYD used in one of his pre amps..."

But I didn't put a volume control inside a feedback loop :wink:

 

[Samuel Groner]

But I didn't put a volume control inside a feedback loop.

Now that's a clever way to adjust the global feedback factor!

Samuel

 

[PRR]

> ...trouble viewing the schematic in any detail...

If you keep clicking, the image ultimately opens to 3000x2000 pixels. It is prefectly clear... the one corner that fits on my big monitor.

It is mostly white-space. A quick crop brought it down to 2000x1100. Still too big to view, even on my 1280 pixel monitor, without some tool that lets me scroll around the image. And I hate that.

Further reduction is not possible, cleanly. The "wires" are 3 pixels wide, so a 33% resize should be possible, except much of the text is 1 pixel wide and vanishes or smears.

If he wants general comments, he needs to adjust the drawing program and his layout style to put bigger lines/text in less total area, and crop it to the bone.

Here is an almost readable 767-wide hack.

 

[skipwave]

[quote author="analag"]
This is a pentode input Mic pre utilizing a NFB technique that NYD used in one of his pre amp schemo's. This is a standard pre featuring two gain stages and a buffer. The problem with using a pentode leaves the designer with the added consideration of what to do with the second stage of the 12AT7, one answer is to make it a CCS which also effectively splits the B+ rail almost down the middle, and thus bias the WCF optimally. No cathode bypass is used as it is my personal preference to avoid them as much as possible, to my ears it is a tradeoff between incremental gain and sonic resolution.

Analag[/quote]

Nice design. I remember you saying that it wasn't absolutely necessary to hoist voltage of the filament supply to the CF.

What heater voltage did you use in this pre?

Thanks,

Brian

 

[pstamler]

[quote author="alk509"][quote author="pstamler"]from what I can see the second section of the second tube has no connection to B+, and is directly connected to the cathode of the first section.[/quote]

Hey Paul,

I think it's some kind of series amp (SRPP? WCF?) drawn in a really weird way, that's all. :wink:[/quote]

Oh -- duh. Need more coffee.

Peace,
Paul

 

[bcarso]

Agree with NYD---you will learn more with some reading and pondering. But, especially, read up on grid bias.

 

[analag]

Control within a feedback loop is done every day, in fact some circuits use it as a means of controlling the volume.

Floating 6.3V at 50V is good enough for this design.

 

[PRR]

> But I didn't put a volume control inside a feedback loop

But when you turn the knob down below about 25%, there isn't any NFB. So what's your objection?

> Control within a feedback loop is done every day, in fact some circuits use it as a means of controlling the volume.

Well, either you have effective feedback, or you don't. When you turn down enough, there is no feedback. When you turn up enough to have effective feedback, the gain hardly changes.

What we have is a system that, for strong signals, is just a zero-NFB passive volume control, and for weak signals is variable-NFB without control of gain.

If it works for you, kewl.

 

[bigugly]

Thanks to all for the feedback.


I'm sorry for the messy graphics. I'm redrawing the schematic using single triodes for better clarity.

NYD, you are correct. Your line amp was my inspiration for this. I will continue my reading of WCFs and post the revised schem when I'm done.

Thanks again to all.

James

 

[analag]

It's not a compressor, we don't have enough feedback action going on for that. I use both compression and feedback the same way "lightly".

 

[Kingston]

[quote author="analag"]This is a pentode input Mic pre utilizing a NFB technique that NYD used in one of his pre amp schemo's. This is a standard pre featuring two gain stages and a buffer. The problem with using a pentode leaves the designer with the added consideration of what to do with the second stage of the 12AT7, one answer is to make it a CCS which also effectively splits the B+ rail almost down the middle, and thus bias the WCF optimally. No cathode bypass is used as it is my personal preference to avoid them as much as possible, to my ears it is a tradeoff between incremental gain and sonic resolution.

Analag[/quote]

This looks like an interesting project. Did you build this one already, or is this just an example for this thread? How is it sonically?

I'm shopping for the next project as I just built newyorkdave's twobottle. I wouldn't mind another good tube pre, if it sounds different enough.

What transformers would you recommend for this? something like 1:6 stepup and 2:1 stepdown maybe (for 600ohm out)? or something more exotic?

Michael

 

[analag]

For input you can use 1:5 all the way up to 1:10, when I built it I used a Jensen JT 115 KE I had lying around. A 4:1 output was my choice, but 2:1 should work.

It can go from clean to vintage warm, unfortunately I scrapped it to feed my DIY craze, and I find myself missing it's sound when I listen to my old recordings with it. I think I'll use the LL1538 input trannie this time around.
and a not so exotic EDCOR 10k/600 for output. M6 saturation curve is all over Rock-n-Roll why not use it in the preamps as well.

analag

 

[Kingston]

Thanks,

looks like it's OEPs for me on this one.

The input zobel network looks weird, only termination and no cap. and is the 100ohm resistor going to grid a part of it?


or is it just room for experimentation or transformer specs?

Mike

 

[analag]

http://www.jensen-transformers.com/as/as020.pdf
http://www.jensen-transformers.com/as/as017.pdf
Take a look at that. Not all input transformers need a zobel network it depends on the transformer and what the manufacturer specify. Obviously the JT-115-KE don't. The grid stopper resistor usually come in flavors of 100-1k in most preamps.

analag