One-Bottle Preamp

[CJ]

well, as soon as your input ratio climbs above 1:1, tha amount of ring will climb with the ratio.you can make the freq curve say anything you want, there is no magic values of r and c.
it's a damage control thing.
you could probably beat the formulas by using your ears with different networks applied to the sec.

 

[Svart]

ok another simple question that I am not sure of and am willing to risk coolness points asking.. adding phantom to the inputs.. with high precision resistors 1% or even .1% can I safely add phantom to the inputs without blocking caps? I know that there should be no DC flowing through the primaries, but that is the reason for using highly matched resistors right?

 

[pstamler]

Right.

Going back to the post-transformer attenuator: The problem with putting in a pot here is that typically the transformer wants to be loaded with about 10x its secondary impedance. Let's say the secondary is 15k; the pot needs to be 150k. At about halfway down (actually 5.19dB attenuation) the output impedance is (Rsec + Rattenuator) / 4, or 41.25k. That will give you more noise than the naked transformer, about 4.4dB worse equivalent noise with a tube or a FET active circuit following the transformer. And it happens when you have very little attenuation dialed in, which corresponds to a not-hot input signal, precisely the time when you need every dB of quiet you can get.

Peace,
Paul

 

[PRR]

> I wonder if a tapped grid choke has ever been used for input attenuation?

Tap the secondary: it is already there.

Do NOT reduce transformer ratio until the plate circuit is strained: for low noise from 200Ω source you need 1:5 or 1:10 ratio into a tube grid.

> could not go to far down the winding as DCR would get small

So? Tube is happy with zero DCR.

The problem with adding a tapped choke to a transformer is that you are nearly doubling the stray capacitance in what is already a too-high-Z circuit. Top-resonance will fall to about 0.7 of the tranformer alone. A 20K tranformer winding has enough trouble already: they tend to ring at 15KHz to 40KHz, you don't want to drop that to 11KHz or 28KHz. It won't quite be double-C or 0.7*Fh because the tube input capacitance is significant, but still....

Likewise, if you really have killer level at the output (seems unlikely with one fat triode), tap-down on the output transformer. The load on the tube will rise, reducing distortion.

 

[Svart]

who said a single tube would be easy.. :?

 

[skipwave]

My two channel One-Bottle box is done. I tested it briefly and was surprised by how quiet it runs and how hot the output is at the lowest gain setting.

My input iron is wired 150:30k, and I ran the output unbalanced straight to my A/D converter. I'm going to use Triad A69J as output iron, wired 6250:500. We'll see how that works.

 

[Svart]

:shock: :shock: :shock: :shock: :shock: :shock: :shock: :shock: :shock:

STOP the PRESSES!

I got the first version of the CJ single bottle pre working.

beyerdynamic input trafos 200:15k, 5842 firebottles, edcor output trafos 6k:600.

when I first hooked it up, the signal was through the roof with some modified SP B1s, so I cobbled together a 20db pad for each channel and tracked some drums using this setup for room mics.

BADASS.. these sound awesome. I truly didn't know if they would sound as good as they do, but I am definately building a rack full of these.

ghostly quiet, very high gain, and very flat.

definately need padding though.

:green: :thumb:

 

[frezen]

Happy new year!

I got fascinated with CJ´s posts on this and have a WE 417A in shipping from the states. I am a newbie in electronics and this will be my first tube project. I am in on the beyer groupbuy for an input transformer.

It would be very useful to follow your further developments regarding padding, gain control and balanced out.

Please keep posting on this project


Best regards for 2006

Frode

 

[Kingston]

[quote author="skipwave"]Speaking of input attenuation, I'm planning on using the Hampt0ne after-the-input-tx-pad, as shown in the jfet pre article:
http://home.comcast.net/~markfuksman/hamptone.jpg

I'm using Sescom MI-70 input TX, which aren't going to saturate easily, so I really like the idea of putting the pad after them. Hit those babies hard.:twisted:

Thoughts?[/quote]

Hi,

great news you've got a working preamp already. Did you do this hamptone post-transformer pad? How much are you padding and with what R values?

I'm building the two bottle version and I expect it to output hot even at lowest gain. I thought of putting two pads in there for flexibility, pre and post the input transformer...

Or maybe I should do a pad at the output... any thoughts on the "right" method? I'll probably be running instruments thru it that are already hot just to get the tube/xformer sound, extra pads might be necessary.

Cheers

 

[NewYorkDave]

You're talking about my 5879/12AV7 preamp, right? In that case, this statement...

I'm building the two bottle version and I expect it to output hot even at lowest gain.

...doesn't really apply, since there's a volume control between stages and you can turn the output down all the way to zero. But, by all means, include a switchable input pad. With about 40dB of voltage gain before the signal hits the volume control, you'll need it with hot mics.

Skip, congrats on getting your preamps up and running. As for the output iron, 6.2K primary is a bit too low; 10K:600 is the recommended output. Can your transformer be configured for a higher primary Z?

For the transformer-coupled-output versions of either the One-Bottle or the CJ pre, a good way to control the output level would be to add a 600-ohm bridged-T attenuator. Here's an 11-step version. Although it's actually an unbalanced attenuator, it's rarely a problem in practice.

 

[rafafredd]

OK, about CJ 417A one bottle...

I´ve got some 10k/600 for the outputs and some 150:15k for the inputs. That makes +20dB at the input and -12dB at the output... So, I got a total of 8dB passive gain. Not much... I´ve noted CJ used a 1:30 transformer for the input for more passive gain.

So, the simple question is: When loaded with a 10k transformer, how much gain this single bottle active part of the circuit will give me?

 

[skipwave]

[quote author="NewYorkDave"] But, by all means, include a switchable input pad. With about 40dB of voltage gain before the signal hits the volume control, you'll need it with hot mics. [/quote]

I'm using the values shown in the hampt0ne schemo, thru 162k and a 40k to ground.

[quote author="NewYorkDave"] Skip, congrats on getting your preamps up and running. As for the output iron, 6.2K primary is a bit too low; 10K:600 is the recommended output. Can your transformer be configured for a higher primary Z?[/quote]

Right, I forgot about the 10k minimum. The Triad A69J can only be configured as 6.2k or 25k primary, but the secondary can be 500 or 12.5 ohms. What if I tried it as 6.2k:12.5?

 

[kiira]

[quote author="NewYorkDave"]Skip, congrats on getting your preamps up and running. As for the output iron, 6.2K primary is a bit too low; 10K:600 is the recommended output. Can your transformer be configured for a higher primary Z?.[/quote]

mmmmm.... I think that pretty much might mean I cannot use the 5:1 Haufe 5223 I was planning on using for the output as its primary ohms is far below 10k.

I'm still working on the two bottle pre.... I only have to finish soldering everything together but it's going slow.

Congratulations Skip and Svart! It's always amazing what a tube and a handfull of parts (and a good designer can do. :razz:

Kiira

 

[rafafredd]

I think it would work just fine as 6.2k to 500... Triad transformers have very high inductance for their rated impedance, so they can actually work at higher impedances han the rated impedance, like all the very good old stuff, like UTC, peerless, etc... If ou don´t really need a true 500 ohms output just think about it as a 10k:800 transformer... Got it?

Haufe 5223 is another problem. It probably has very high inductance windings. Haven´t measured it, but just because HAUFE makes the very best in the transformer world it´s probably over engeneerred in the impedance\inductance aspect. But it´s a too small transformer, for this application. Won´t handle the levels. Use it at the input of another tube pre.

 

[kiira]

[quote author="rafafredd"]

Haufe 5223 is another problem. It probably has very high inductance windings. Haven´t measured it, but just because HAUFE makes the very best in the transformer world it´s probably over engeneerred in the impedance\inductance aspect. But it´s a too small transformer, for this application. Won´t handle the levels. Use it at the input of another tube pre.[/quote]

OK, I see. Hmmm... yes it is rated at 23dB which is low I guess for such a high gain pre. I think I'll finish it up with just the unbalanced out and then buy a Cinemag (or something on ebay that's right) when I get some money.

thanks Rafa (Raphael?)

Kiira

 

[NewYorkDave]

Right, I forgot about the 10k minimum. The Triad A69J can only be configured as 6.2k or 25k primary, but the secondary can be 500 or 12.5 ohms. What if I tried it as 6.2k:12.5?

Why would you want to do that? Just strap the primary for 25K, the secondary for 500.

The primary inductance, when strapped for 6.25K, will likely not be high enough to avoid low-frequency rolloff. Remember, the tube plate is a fairly high impedance source in any circumstance. The plate resistance of the 12AV7 under the conditions shown is about 7-8K. Put this in parallel with the plate and feedback resistors and you have a good approximation of the source impedance driving the transformer.

Don't assume the primary inductance is high enough to use at above-rated impedance (as suggested above) unless you're equipped to measure it and confirm it for yourself. Besides, there's more to it than just primary inductance; using a transformer outside its recommended range of impedances can easily shift some nasty but normally out-of-band resonances down into the audible region. You'd really have to load it up with the source and load impedances you plan to use and test it with a signal generator and scope.

Besides, even if the primary L would be high enough, and there were no problems with resonances, doing as you describe would gain you nothing. Let's do the math:

25000/500=50, Sqrt 50=7.07, Insertion loss=17dB
6250/12.5=500, Sqrt 500=22.3. Insertion loss=27dB

Another thing: with those pad values, you're loading down the 30K secondary of your input transformer. Simply raising the value of the resistors is not a good solution because the additional series resistance looking into the grid (with the pad engaged) would cause high frequency rolloff. I know you're into the idea of putting the pad after the secondary, for whatever reason, but I still recommend the type of pad shown on the original schematic.

Kiira, why not just use the Edcor 10K:600 you bought? It would be perfect. I was under the impression that that's what you had planned to do all along.

 

[rafafredd]

Yeah, Dave is right. If you use a 10k:600 one you will have a 12dB loss.

If you use a 25k:600 you have a 17dB loss.

Only 5dB difference in gain and max output.

But Probably the 12aV7 will swing a little more with a 25k load, so you will have even a lower maximum output difference.

 

[NewYorkDave]

Exactly! :thumb:

 

[skipwave]

Thanks!!!!

It's times like these that I'm reminded how much I love this board. Almost brings a tear to my eye. :wink: :green: :sam: :sam: :sam:

 

[Svart]

I used 6K:600 for the outputs and it's pretty hot. with a SPB1 these clip an MX-2424 pretty badly. with a 20db pad at the mic, it's just right.