6SN7 / 6CG7 preamp

[rafafredd]

The NYDave one 6SN7 audio circuit is done, in a very little box, with an old beyer input transformer that came out of a mic and a cheap chinese (sorry NYD) 20k:600 20dBm output trafo that measures very high inductance despite being chinese :green: . In fact both transformers were near zero cost for me, but they do not test bad on the bench at all. I thought this would be a nice chance of trying these ****** cheap transformers and and hear their sound as this is a proto-proto-type. I´m curious about this cheap iron anyway.

If it all works nice, off coarse I´ll make a higher grade version, with four channels, rack case and expensive and wonderful Lynx Audio brazilian iron. :grin:

So...

I was thinking about making the psu outboard, as the box I built the amp in is so small... I don´t know why I did that. :? Now comes the trouble. I would really prefer to use 4pin XLR connectors for power, as I think it´s much much safer than 5/6/7/8 pin ones. At least that´s the case with the types I can find localy. 5/6/7/8 types have the pins much closer to each other, and the 4 pin version has it totally separated by plastic. Near impossible to short.

So, I thought about using the 4pin XLR like this: 300v, 48v, 12v, 0v. Would this be ok? Can I keep all the grouds in just one pin, like this? I never did an out of box PSU for tube preamps, but I thought that if I star ground the amp box and star groud the PSU box and then connect the two stars together throught a single 0v conductor, all would probably come along well...

Or do I need to run separate ground conductors for each rail, like 0vHigh, 0vPhantom, 0vHeatrs, and ONLY connect those together at the PSU star ground point? Thta would be more "stared" in my view, but do you think it would be really needed?

I know there´s no magic ticket in grounding, and sometimes we must try and see, then try again, and again until we get it right for the project in question.

But If I could just safely go for the four pin connector, that would be awesome. Just let me know if you tried similar configurations, and if that worked allright or not.

I hate to fight against ground problems in tube pres... I spent days once.

 

[Guest]

Grounding issues are what usually kill me, so I have become shy. I have done something similar, but used seperate ground line and it came out OK. I think you may be alright with one ground on a 4-pin connector if you use decoupling caps inside the preamp box. At this stage of prototyping to see if it all works I think it is worth a gamble. Good luck.

 

[skipwave]

Wanted to run an idea by the board. A box with the 6SN7 pre that Dave posted and a Collins 356E varimu limiter, with a switching jack for line-in to the limiter. Available gain will be high, 62db from the pre, plus the small step-up of a 1:1.4 interstage transformer to 6BA6 grids (in place of the 6386 in the Collins), and the gain of the limiter post-amp stage.

Triad interstage transformer:

The rated impedances are perfect. 15k primary to 60k secondary with CT. Nice match from the 6SN7 plate to 6BA6 grids. The only issue I see is rated level only +10dbm. Question is, will this little trans be hurting on the output of the 6SN7 pre? I think the answer is Yes, but how bad?

I already have all of the parts to build this as described, so I can try it and see. I just wanted some opinions from trusted sources before I proceed.

 

[rafafredd]

Hey, mine is almost almost almost done. It´s just a little more hard enclosure-work and it´s running...

I kow I´m slow, but I´ve been working on many many projects simultaniously, as always.

Also, I had some minor heath issues, and I´m still recovering. By next week I´m ready to finish this box.

Thanks Dave again for the help!

 

[skipwave]

Glad to hear that Raf. Hope you feel better.

 

[CJ]

that triad will mush out into a nice, compressed, saturated nickel sound.
this will be an "effect" amp, not a "Masenburg" by any means.

 

[skipwave]

[quote author="CJ"]that triad will mush out into a nice, compressed, saturated nickel sound.
this will be an "effect" amp, not a "Masenburg" by any means.[/quote]

Just what I always wanted! For clean tubed and comp'd I'm making another box with a One-Bottle (AV7) and analag's tube opto comp. Mmmmmmmmm. Sorry, I started drooling while thinking about these.

Speaking of effect mike amps, I recorded some guitar tracks with my 6EM7 pre and ShinyBox 46, and it made a crappy solid-state marshall head (which the guitarist insisted on using) sound pretty good.

 

[analag]

Rafa I'm sure you have some new simulations of this project...let's see it.

analag

 

[rafafredd]

My old simulations with circuit maker for this circuit worked ok. Not lots of headroom, but enought for most tracking purpioses. The LTspice simulation works much worst. I can´t say why. There´s something going on with the secon stage grid, I guess. Anyone, I´ll send you the file for you, if you wanna take a look. All other circuits I have been trying on LTspice works just fine and close to real world. Maybe something strange going on with that 6SN7 model???

But I´m so close on ending this one to test the real thing. I have been put out of DIY for the last two weeks because of my teeth... I had to take my upper third molar out, because it was smashing the lower flesh. Argh. Now that I´m better I´ll probably have to take the other´s side out to. But yesterday I had a look at the PSU and after some metalwork it will be all done.

 

[analag]

http://www.birotechnology.com/articles/pspice/6SN7.cir

Copy and paste this in your tube.inc file or make a 6SN7.inc file.

analag

 

[rafafredd]

Thanks. The fact is that the output stage won´t do past 7dbM, or 2.5v peak. :sad: But I´m on it anyway. I´ll try to get my hands on this today, as I´m already recovered.

In the end, I think I´ll stick with a WCF output, and three bottles for two channels, if I ever mess with the idea of using only 6sn7/6cg7 in a tube mic pre again. It seems that one bottle 6sn7 driving 600ohms to good level is really not possible.

 

[rafafredd]

here´s a much more interesting configuration, with more gain, more output level, no feedback, and still only two tubes for two channels:

lesson of the day: 6sn7 sucks as output tube!

Make the grid resistor 220k a 100k to 330k pot.

 

[rafafredd]

mmm... I´ve made some more soldering tonight and now it´s really close.

I managed to get another model for the¨6SN7. Aperently it was made specially for LTspice. The circuit works a lot better with this model. Here it is:

*6SN7 LTSpice model
.subckt 6sn7 P G K
Bp P K I=(0.02003791851m)*uramp(V(P,K)*ln(1.0+(-0.07740549711)+exp((4.618036737)+(4.618036737)*((20.85288965)+(-110.4389272m)*V(G,K))*V(G,K)/sqrt((28.13407639)**2+(V(P,K)-(7.118597372))**2)))/(4.618036737))**(1.380047579)
Cgp G P 4.0pF
Cgk G K 2.6pF
Cpk P K 0.7pF
.ends 6sn7

here is where I´ve found it:

http://www.joeltunnah.com/JT_TUBELIBRARY.txt

Well, now the circuit can swing 15dBm at 600ohms load nicely. It makes me much more optimistic about the real box i´m building. I hope it behaves more like this...

 

[NewYorkDave]

"lesson of the day: 6sn7 sucks as output tube!"

Whether or not that's true (hint: it's not), I wouldn't make any such pronouncements based on a SPICE simulation. SPICE is a fun toy and sometimes a handy tool, but it only takes you so far. I don't believe in investing a lot of time in SPICE simulations because they usually end up being wrong to a greater or lesser extent, anyway. And if you've fallen in love with your simulated circuit it can be quite disappointing to have to modify it extensively during real-life experiments

 

[rafafredd]

Yeah, I know what you mean Dave. I´m working on the real life one, so we will see how it goes. I do think it will work nice and actually give me more output level than what spice predicts.

 

[analag]

[quote author="NewYorkDave"] SPICE is a fun toy and sometimes a handy tool, but it only takes you so far. I don't believe in investing a lot of time in SPICE simulations because they usually end up being wrong to a greater or lesser extent, anyway. And if you've fallen in love with your simulated circuit it can be quite disappointing to have to modify it extensively during real-life experiments[/quote]

I'm building the, (shall we call it the) RS660 based entirely upon a simulation. I believe it will be so close, no tweaks shall be necessary. It is an expensive way to test a point.

analag

 

[skipwave]

I'm running with the 6SN7 pre as shown. Of course, its only going to have to drive an interstage transformer to PP grids in my application. I bet it has too much drive capability and I end up reigning in the gain structure by applying feedback. We shall see.....

 

[rafafredd]

Mine is finally alive:

I know it looks funky, like all my latest DIY, but it sounds great, so... Who cares? Think about it more as a breadboard or something alike... Once the circuit is working, I like to keep the test versions, and make it a fully functional unit, so I do not waste anything.

I´m racking it now, so that all that mess stays inside, out of sight in the studio.

 

[analag]

Nice...

I was going back through the thread and I noticed NYD uses circuitmaker, no wonder he thinks SPICE is a toy. Circuitmaker is a toy.

analag

 

[soundguy]

dave just builds stuff. Stuff that sounds real good, too.

dave