mic pre concepts

[tmbg]

I think I'm strong enough in my Transistor-Fu to do something similar, I can envision a multistage discrete transistor amplifier which gives me three main gain stages of 20dB each, plus an output follower if necessary, and an input buffer if necessary... I don't actually OWN any audio xformers, and I'm not fully comfortable with the hows and whys of them, but I'm getting there...

as for feedback, the emitter resistor in a common emitter transistor stage is a sneaky sort of feedback

 

[Svart]

excellent, back to the interesting stuff..

i vote for:

FET i/o
multiple gain stages and/or some way of controlling gain without feedback.
something like a tube being switchable into the audio path.


now.. for phantom power.. is there a way to float the input stage so that we could directly apply phantom power without blocking caps?

I'm still wondering about changing gain through varying voltages on FET sources instead of a steady voltage.. no one has even discussed why or why not.. :? I assume current noise and non linearity but does that necessarily mean "bad" sound?

 

[CJ]

That 20-40-60 db idea sounds great. I also vote for external pwr supply.
Ghostly quiet.
I was thinking transformers for the tube amp and maybe transformeless input for the solid state amp with transformer output for less coloring.
Get one of those sexy Burr Brown opamps to do the innput balancing.

I wonder if a mic could drive a transformeless tube input stage?
It would have to be push-pull to cancel noise. Never seen that done before either, probably because the idea sucks!

 

[tmbg]

if it's an input trafo, I'd guess we'd need one which could handle DC without saturating if we had no blocking caps, and from what I've read, I think that means gapped. Not sure what else you trade off if you have that 48v DC on the input trafo...

 

[rafafredd]

what about the parallel k389´s for the transformerless inputs?

 

[tmbg]

ok so explain to me how an output transformer provides less coloring? I'm not being snotty, I just don't know!

Also, what are the benefits of some boutique op-amp for input balancing over just using a 5532 or OPAwtfever, etc?

also, what exactly is "DC servo balancing"? Lots of marketing hype these days!

 

[Kev]

I think CJ was referring to the transformer-less solid state input section as being less colour ...

Added to that a 1:1 output trafo can be made to be more clean than people give credit to. Those API type trafos that many of us got recently really are very clean.

There are many other places that bring colour in front of a 1:1 output trafo. Gapped core and the circuits that require it and low frequency, high dB and low ohms loads do bring colour from 1:1 trafos.

I hope that made sense :? now you can explain it to me.

 

[tmbg]

well ok, but what is the BENEFIT of an output transformer over just having a nice low impedance transistor driver output?

just that the output transformer gives you simple, clean balancing? or is there more to it than that?

The way I'm looking at it, we have these parts that cost anywheres from $60 to $300 apiece, and aren't especially accessible, it may be my ignorance showing, but it seems like we should be avoiding them in solid state circuits.

 

[Kev]

Well this is all part of why I wanted to direct the thread in a slightly different direction and get back to the original question.

Most people want to jump to a design ... and a design that is most likely a staple of the past. Not that, that is a bad thing.

Transformers do provide a differential feed, AC couple ... err :roll: ... impedance/level/loading matching perhaps.

 

[tmbg]

kev, you are certainly helping my cause here!

ok I'm just gonna spout off about things the way I sees them...

impedance matching, in my experience, is used when you want to get as much power as possible somewhere or other, or if you want to turn a particular voltage into some other particular voltage. That's ATTENUATION!

Having a given Zout that's greater than 0 ohms, being loaded by some Zin that's less than infinity ohms, causes attenuation. Our goal is to pull a weak microphone signal up into a usable line level through amplification, then deliver that line level in such a manner that any particular load will not cause it to drop below the specified line level.

NOW, if we're talking about gear that's all transformer outs into transformer ins, I can understand where we want to match impedances in such a manner that the input of the next piece of gear loads the output of the previous to exactly +4dBu or whatever... BUT WHY?! We have the capability to design circuits that get way close to infinite Zin and 0 Zout, so why don't we?

 

[tmbg]

perhaps it has something to do with the fact that a bunch of people already have all these fancy transformers, and they really really want to use them!

:guinness:

 

[Ian MacGregor]

Haha.. very true. I suggested using transformers because I see it as being the easiest and simplest way to do the output balancing. I'm pretty damn green when it comes to audio circuit desiging and I've had more exposure to Neve Class A circuits than anything else, so I tend to head there. I am pretty interested in going transformerless, could make a design much cheaper/easier to source parts. What would anyone suggest as a transformerless input? What about using some form of JFET 990-ish opamp to do the unbalancing?

Ian

 

[tmbg]

ooh good idea! Replace the $60 input transformer with a $65 discrete opamp!

:green: :sam:

 

[CJ]

Yes, the tried and true 5532/34 dual opamp diff input thingy would be fine. No use in blowing up an expensive B/B opamp.
Doh, how would we apply phantom to a direct coupled tube input, oh well, sometimes lame mistakes lead to new ideas.
You could go tranfrmer in with a switch to go thru a dual opamp balance circuit, but now our parts count/price goes up.

 

[tk@halmi]

Getting a bit tired of transformers. I am gathering parts and planning to make a PCB for the Cohen AES paper preamp. It has been tested and many serious preamps have been based on his design since that paper was published. The cost for a complete channel is less than a cheap and crappy input transformer.

Tamas

 

[BradAvenson]

[quote author="cjenrick"]Doh, how would we apply phantom to a direct coupled tube input, oh well, sometimes lame mistakes lead to new ideas.
You could go tranfrmer in with a switch to go thru a dual opamp balance circuit, but now our parts count/price goes up.[/quote]

Have you seen this?
http://dogstar.dantimax.dk/tubestuf/shroom.htm
http://dogstar.dantimax.dk/tubestuf/graphics/shroom.gif

Maybe think along those lines.

 

[PRR]

> how would we apply phantom to a direct coupled tube input, oh well, sometimes lame mistakes lead to new ideas.

The circuit is do-able, in the sense that you can connect all the pins and get predictable results.

Problem is that the noise of a tube is around 2 microVolts, two tubes around 3uV, and the noise of a 150Ω resistor (or dynamic mike) is closer to 0.2uV. Direct-coupled transformerless tubes have horrible noise figure on 150Ω dynamic mikes.

But note that the high-output big-diaphragm condensers often have noise level around 1uV. So taking them through 3uV of tube-noise may not be unbearable.

Also tube noise reduces with cathode area and current. A dozen TV-tuner twin-triodes in parallel might give noise voltage below 1uV.

Note that while you can avoid an input capacitor, a cap is not a problem on a FET or tube because it can be very small and thus very extra-virgin gold-plate. It is only BJTs that need the fat caps which "have" to be aluminum electrolytic to be affordable.

Also note that this lack of noise current means a step-up transformer "is the ideal way" to couple a mike to a tube: it overcomes voltage noise without inducing current noise trouble.

Arrgggggh. I ripped the shit out of my back a few days ago. Sitting at the PC still hurts like hell. Over and out for now.

 

[Ian MacGregor]

[quote author="PRR"]>
Arrgggggh. I ripped the shit out of my back a few days ago. Sitting at the PC still hurts like hell. Over and out for now.[/quote]

Ouch! Take it easy man. I feel your pain. I broke my elbow a week ago skateboarding. It's no fun to type loads while injured, that's for sure.

Ian

 

[CJ]

It's always around the holidays when people get banged up. Get well soon you guys!

Thanks Brad! I nearly fell out if my chair when I saw that direct coupled tube circuit. Getting phantom from the cathods, never could have thought of that.

Radar tubes, hmmm, never did get around to trying those WE 408 types, supposed to be quitier than the best fets.

 

[Kev]

[quote author="PRR"]Arrgggggh. I ripped the shit out of my back a few days ago. Sitting at the PC still hurts like hell. Over and out for now.[/quote]
yes I do know how this is.
you young blokes need to look after your backs.
when things go wrong, whether it is a single event or years of a degenerative condition, it all amounts to the same thing.
Coughing, sneezing, shitting, sitting at the computer, sleeping ... it all hurts and apart from the sheer pain it just wears you out mentally as well.

I have just embarked on a Chiropractor for the first time and he has taken a series of x-rays on a new machine and the quality is sensational BUT the pictures tell a bad story. I'd love to post them on the site.

Anyway I'm on the Chiropractor and not the Physio this time and I won't do drugs and I don't want an operation. I don't do being sick at all well or go to doctors and end up just gettig grumpy and bad tempered so I paid up front for a series of consultations which is another thing I've never done before but if it forces me to go regularly and give it a chance to help me then it will be worth it. I'll let you know how it turns out.

get well soon to Paul.