DIY Tube Preamp

[Brett]

Howdy FBLEAGH,

Things seem to have gotten a little off track, so I want to back up a bit and clarify some things. You said you wanted to make a line stage and use the project for the experience. No matter what project you choose you will gain valuable experience, but it is the part around the line stage spec that has me curious.

What do you plan to use the linestage with, ie, what sources do you have? What is the input impedance and sensitivity of the power amp you want to use?

My guess for the answers are, between a CDP (or other high level source) and a poweramp, >20kohm and about 1V. So you need a line stage with an output Z of less than 2k (lower the better) and a mA or two of drive at least. But what you DONT need is any gain. The Red Book standard for CDP outputs is 2Vrms, so you already have twice the voltage drive you need to send most power amps into clipping. So adding a stage with 20x gain just means you're going to have to waste all that gain away in the volume control. If you put a pot before the tube, you'll lose your signal to noise ratio, and one ofter the stage will give you a variable, but highish output impedance. A two stage line stage is idiocy. Even with a stage gain as low as 4 and a CDP as source, you will be able to overdrive EVERY amp I've ever seen, and some by over 20dB. I drive my poweramp with my CDP,DVD, VCR and tuner through a passive TVC, and am still 20dB or so from max even when it's exceptionally loud. At full gain I cannot bear to be in the room > 115dB SPL.

Almost every system has too much gain and too little headroom. Doing it DIY means you can optimise it for what you need, not trying to have a universal stage (like most manufacturers try to do) that does nothing at it's best in your system.

Any of the circuits like the basic anode follwers in the link you first posted, or the SRPP Frank suggested will work, but far from optimally. My suggestion would be to breadboard one of these on a simple chassis and listen for a while (the offer of the 6N1Ps still is there). Most of the passives can be reused as can the iron, so the experimental cost will only be a few dollars. Later try something like a CCS or choke loaded 12B4A or even a 6C19pi, both of which are cheap tubes and will give gains below 6, low output impedance and a huge amount of grunt to drive capacitive cables (if needed). Another option would be to use an output transformer. The CCS is the cheaper option, maybe $A150, and the choke loaded or transformer stage for about double that. All will easily exceed the performance of anything you can buy here for a lot more money.

Cheers

[Joel]

My mistake guys - I was in such a hurry to finish the drawing, I attached the ouput to the wrong point.
Here is the correct schematic. Any other mistakes??

[Bas Horneman]

Joel...

What do you use for drawing schematic..

I use the free program switchcad..serves me well...but i like the look of your schematics...

Cheers,
Bas

By the way..For the guy asking about the transformer...a decent price for something you saw would be around 50 US$.

Why don't you guys try to find a "ordinary" transformer shop..anyone can wind something like that in a jiffy really.

It's the output transformers that needs some practice.....

[Joel]

Originally posted by Brett
But what you DONT need is any gain. The Red Book standard for CDP outputs is 2Vrms, so you already have twice the voltage drive you need to send most power amps into clipping.

I actually agree with you for once. I don't think anyone who uses a cd player as source needs a preamp at all. But the guy asked for a preamp design - who am I to argue?

So adding a stage with 20x gain just means you're going to have to waste all that gain away in the volume control.

So? Better to throw away gain than to not have enough. And I found that when being used as a headphone amp, you DO need gain afterall. Even with a CD source.

A two stage line stage is idiocy.

Nice comment.

Any of the circuits like the basic anode follwers in the link you first posted, or the SRPP Frank suggested will work, but far from optimally.

"Far from optimally" in what sense? They are all proven designs.

[Joel]

Originally posted by Bas Horneman
What do you use for drawing schematic..

Hi Bas,

It's just a bunch of clipart files basically. The suite is called "TubePad", and I think you can find it here http://www.mines.uidaho.edu/~glowbugs/

You need to use microsoft Paint, in Windows, to put the drawings together though.

[Bas Horneman]

Thanks Joel..

[Brett]

<b>But what you DONT need is any gain. The Red Book standard for CDP outputs is 2Vrms, so you already have twice the voltage drive you need to send most power amps into clipping.</b>

<i>I actually agree with you for once. I don't think anyone who uses a cd player as source needs a preamp at all. But the guy asked for a preamp design - who am I to argue?</i>

Even a passive pre has to compromise between the impedances seen and driven adequately by the CDP output, the output impedance of the passive stage, the input Z of the power anp, and the assosciated capacitances in the betwork. Most CDP output stages harden considerably in sound as the loads get lower, often 20kohms is the minimum for most opamp output stages to drive (irrespective of what the datasheets might indicate), esp the 4558 and the 5534, the most common devices seen in commercial players.

<b>So adding a stage with 20x gain just means you're going to have to waste all that gain away in the volume control.</b>

<i>So? Better to throw away gain than to not have enough. And I found that when being used as a headphone amp, you DO need gain afterall. Even with a CD source.</i>

Why have all the gain thrown away? Why not have something better optimised to the gain requirements of the circuit? Most systems IME have way too much gain to start with, so few systems ever see full gain, even for those CDs etc that are mastered low. So, already with an excess, you throw it away in a pot before the stage, so your SNR is affected by maybe as much as 30dB.

What relevance does a headphone stage have to the original poster's question? You might also note you need the gain because you are losing some in the CF (< unity gain) and through the losses in the stage's output Z, esp driving a low impedance load.

<b>Any of the circuits like the basic anode follwers in the link you first posted, or the SRPP Frank suggested will work, but far from optimally.</b>

<i>"Far from optimally" in what sense? They are all proven designs.</i>

In all the circuits with a fair bit of gain, mainly noise as I've explained. Many of the anode followers have quite a high output Z.
The SRPP will vary it's distortion and sonic characteristics into different loads. A lowish output Z does not neccessarily equate to decent drive abilities.

Sure the circuits are proven, in the sense that they work. But a 1948 sidevalve V8 "works" and is reliable. But it hardly makes it a high performance design.

[Joel]

Brett,

I'm not quite sure why you are so opposed to the designs that Frank and I have offered him. This is supposed to be a simple, affordable project to get the guy started in DIY. You have now successfully clouded the entire discussion to the point where he and many others will now doubt that these are "good designs". Congratualtions.

Also,
1) I did not suggest that anyone use an input impedance of 20k. My schematic shows an input impedance of 100k - which is just peachy for a cd player to work into.
2) A "headphone amp" IS a preamp. I said you could use it as one.
3) I didn't mention that the CF loses gain because I assumed everyone knew that.
4) I hardly think the 20 or so gain of the 12AY7 is going to add 30db of noise! This is ludicrous.
4) If you have a better alternative for a newcomer to build, please offer it.

[fdegrove]

Hi Joel,

Here is the correct schematic. Any other mistakes??

Not a mistake but if you want to use this circuit as a linestage preamp wouldn't it be wise to add a second smaller value coupling cap and say a 100K shunt resistor.

You can then make both outputs switchable: headphone/preamp.

Loading down a lineoutput with a small value resistor is going to take down the input impedance of the amplifier offering a severe load.

Just my 2 eurothingies,

[FBLEAGH]

Thanks for the intersting discussion guys.

are hammond power xformers supposed to be really good of something ??

302AX 300-0-300 200mA 2.5v 2A, 5v 3A, 6.3v 6A
$250.50AUD + GST

why such a big difference between that and my custom one ?