Simple tube Mic preamp help

[quote author="NewYorkDave"]

60+ into a 10:1

Click to expand...

No, 60-plus dB with a 4:1 output transformer. You must have mixed that up with the 1:10 input transformer.

You'd better call Edcor back now before they start winding your 10:1![/quote]

ORK! I am so confused. So for the two bottle one botlle pre I need a 1:10 input and a 4:1 output. I ordered a 10k:10k and a 10k:600 from edcor. Ok... maybe I can use the 10:1 backwards and use my Haufe 5:1 23dB on the output. Yikes I always do this.

Posts are acting all wonky so I ended up answering wrong twice. yay me.

Kiira

:shock:

the forum is behaving oddly again. I can't edit my posts, slow/weird updates.

Hope this topic won't disappear like a similar simple tube pre thread did earlier. :sad: Best copy and paste the new info now.

[quote author="NewYorkDave"]

60+ into a 10:1

Click to expand...

No, 60-plus dB with a 4:1 output transformer. You must have mixed that up with the 1:10 input transformer.

You'd better call Edcor back now before they start winding your 10:1![/quote]

ok NOW I am really confused coz I re-read the one bottle pre thread looking for how to determine the zobel values and it seems like you did say use a 10:1 on the output? or you mean I'll just get more output than 60+ if I use a 10:1?

Kiira, I think you're thinking of one of my other circuits (the line amp with the 12BH7 WCF output). For the One-Bottle pre, or this new "Two-Bottle" idea, you'd use a 10K:600 output, wired in the normal fashion (using the whole primary and the whole secondary). A 5:1 (15K:600) output would work fine, too. Your max output level would be less, but the distortion near max output would be reduced, too (due to slightly lighter loading of the output stage). Just make sure those small 5:1 jobs are intended for use as output xfmrs, not inputs! They should be rated to handle +20dBM or more.

Click to expand...

so I can use either right? I'd rather use the 5:1 Haufe because the edcor won't fit in the space I have left. They are outputs... 23dBM.

The only thing I am wondering now is how do I figure out the values for the zobel accros the secondary? You used an ouncer but I am using an edcor (the 10:1 I ordered by mistake... I'll wire it backwards). Should I just put a square wave into it and fiddle until I don't get ringing? Yech.

thanks,

Kiira

Try it without a network first.

Networks are a minor tweak to the high end. No major damage if you don't use one, unless the transformer is a real piece. Might be enough capacitance in your wiring to mellow things out.

Where did I say to use a 10:1 on the output?

I'm not sure we're on the same page as far as how turns and impedance ratios work. When you state a ratio, such as 4:1, that refers to the ratio of primary turns to secondary turns (unless stated otherwise). But the impedance ratio is the square of the turns ratio. And the turns ratio is the square root of the impedance ratio. So, a 10K:600 transformer has a 4:1 turns ratio.

I wouldn't recommend using the Edcor as an input transformer. They're output transformers and the construction is all wrong for mic input use. On the other hand, it wouldn't hurt to wire it in just as an experiment, but don't expect great results.

I might have some small 1:10 inputs laying around. They're cheapies, but let me know if you're hard-up and maybe I can dig up something for you.

A 5:1 (15K:600) is fine for the output. The maximum output will be a dB or two less; but on the positive side, the distortion as you approach maximum output will be lower.

What CJ says about the input compensation network is correct. It's a tweak for flattest response. It usually won't make anything blow up if you omit it, although there are cases in which a preamp with an uncompensated input can oscillate if there's no mic connected.

To tweak the values, you need a scope and a square-wave generator. Do you have those?

[quote author="NewYorkDave"]Where did I say to use a 10:1 on the output?[/quote]

you didn't I am having a brain fart. I did the arithmatic in my head for the square root of 10K and 600 and got 10:1 ... xcalc is really much better though than whatever my head thinks arithmatic might be lol. :razz:

I wouldn't recommend using the Edcor as an input transformer. They're output transformers and the construction is all wrong for mic input use. On the other hand, it wouldn't hurt to wire it in just as an experiment, but don't expect great results.

I might have some small 1:10 inputs laying around. They're cheapies, but let me know if you're hard-up and maybe I can dig up something for you.

Click to expand...

Oh thanks... I prolly can dig something up tho. I have this Thordarson 1:10 thing with no can - not sure how good it is but I can put it in there just to see if I have things put together right. I have some other 1:10 too but they are too big. The THordarson is the same size as the edcors.

What CJ says about the input compensation network is correct. It's a tweak for flattest response. It usually won't make anything blow up if you omit it, although there are cases in which a preamp with an uncompensated input can oscillate if there's no mic connected.

To tweak the values, you need a scope and a square-wave generator. Do you have those?

Click to expand...

Yes. I'd like to eventually try and tweek this pre to use mainly with my Beyer ribbon too...

thanks!

Kiira

would something like this work better? Higher current from the 12at7 plus lower impedance so 1:1 transformer can be used on the output....thoughts?

anybody?

I don´t think that cathode follower can drive a 1:1 nicely. Use at least 2:1, but 3:1 or 4:1 should be better.

Many people thinks that cathode follower sounds bad. That´s because cathode followers should be loaded with as high a impedance as a plate wold be most of the time for better results (there was an article somewhere about this). On the Pultec MB1, two 12AU7 cathode followers in parallel are used to drive a 15k ohms 5:1 transformer, and that´s why it sounds great for that circuit.

For driving lower impedance transformers, a SRPP, or a White cathode follower or a MU follower would be better than a simple cathode follower.

If you want few bottles and needs to drive a 600:600 transformer, you could try a MOSFET/tube MU follower.

> would something like this work better?

Work better for what?

Define what you need to do. -10dBV, +18dBu, 10K load, 600Ω load, ?????

As I read that: the 12AT7 CF will pass maybe 10mA. At the extremes: 12AT7 is off, for small output swings the 15K resistor will pass 10mA; 12AT7 is driven full-on, passes 20mA of which 10mA flows to the 15K and 10mA is available to the load.

Now if the load is a good 10K transformer and the load is 10K, you have a ton of output. If the load or the transformer is 600Ω, you get at most 6V peak output, 4V RMS, about +14dBm, so safe nominal level is 14-16= -2dBm, which is much lower than the customary +4dBm.

Distortion for the 10K loading will be about 7% at 40V peak, divided by the CF internal feedback is maybe 0.2%, and much less at the levels we typically jam into 10K inputs. Say under 0.1%. Distortion for the 600Ω load at 5V peak will be more like 9%, divided by internal CF feedback which has almost vanished with the heavy load, say 2% at 5V peak which is only 3.5V RMS or +13dBm, maybe 1% at 1.8V or +7dBm. Nominal level should be 7-16= -9dBm which is awful lame.

With 10K transformer and load, it works. You could (and should) reduce the CF current: you are awful close to 12AT7 abuse.

With 600Ω load, I say it won't do.

Or the short answer:

> Higher current from the 12at7 plus lower impedance so 1:1 transformer can be used

Two changes that cancel each other, almost: even without analysis I would not expect much "better".

> Many people thinks that cathode follower sounds bad.

They are wrong, for the reason you cite: there are too many over-worked cathode followers out there, giving the trick a bad rap. They are perfectly transparent if not abused.

As long as you have a transformer anyway, load the plate and use step-down. The CF has only current gain, a plate-loaded stage has the same current gain and a lot of voltage gain, more total power gain. And ultimately, it is all about power.