tek Q's for geoffT.

[manning1]

dear geoff T. i understand you frequent this forum.
have a couple of questions for you of a tek nature.
although i'm a computer engimeer i'm still learning
audio circuits after building my own mic pre's and experimenting for a few years. i like the fun of it.
ive built many op amp and transistor circuits from easy to high esoteric end -
but i keep on returning to a simple one transistor stage
mic pre (basically straight wire with gain) with the output going
to a comp/limiter or line mixer to make up any added gain.
mainly because of the low component count and because it sounds sooo good.
the kiss concept you know.
my QUEST as it were is to do it all in one stage, rather than use multiple stages. because as i said if i need added gain i use gain in a compressor or a line mixer.
my questions are as follows,,,
1. what are the lowest noise highest gain transistors available ?
with low Rbb ? ive had good results with mpsa18, mat02 (low noise but not super high gain),
and was totally surprised at some varieties of 2n2222 (but once again
not high gain). any suggestions ?
2. ive been told often to come into the emitter for low impedence
but i find coming into the BASE of the npn transistor works better on some mics for me anyway.
3. on sm57 i'm aware that they are sensitive to loading.
ive tried this -> hot pin thru a cap to transistor base. is there
a better way ? some special resistor to ground perhaps in front of the transistor stage ?
or other mojo ?
sorry to trouble you. just trying to learn more.
thanks.
ps some songs done with my weird mic pre circuits
soundclick.com/bmanning
eg,,, hi tech man, up in the sky, limeyrap

[Geoff_T]

Originally posted by manning1
dear geoff T. i understand you frequent this forum.
have a couple of questions for you of a tek nature.
although i'm a computer engimeer i'm still learning
audio circuits after building my own mic pre's and experimenting for a few years. i like the fun of it. <snippety>

1. what are the lowest noise highest gain transistors available ?
with low Rbb ? ive had good results with mpsa18, mat02 (low noise but not super high gain),
and was totally surprised at some varieties of 2n2222 (but once again
not high gain). any suggestions ?

Hi

The first thing to remember is that it ain't what you use, it's the way that you use it. Using a low noise, high gain transistor (there's a battle of interests!) does not mean that's what you'll get if the circuit isn't optimised for that component. If you look at old Neve modules (the the original black faced ones) they used transistors like BC107's and still managed -128dBu noise ref 300 ohms (in the 1053). A very easy way of making a low noise amp (if you haven't anything against op amps!) is to put a differential transistor pair in front of an op amp. You'll certainly get gain and low noise that way!


2. ive been told often to come into the emitter for low impedence
but i find coming into the BASE of the npn transistor works better on some mics for me anyway.

If you browse some of the posts about the 283 amp in my Ask Geoff pages on my web site, I discuss the various merits of impedance and phase of a transistor in some of them. Basically, if you go in the base of a transistor the signal out of the emitter will be in phase and out of the collector 180 degrees out of phase.

If you go into the emitter and out of the collector you'll also get a reversal and a lot lower impedance options than via the base. If you look at any B238/338/438 Neve op amp, you see that the base is the non-inverting input and the emitter the inverting input.

To be honest, that's not the best way to wire an input and why I mention the differential input above, driving either a discrete or an IC output section.


3. on sm57 i'm aware that they are sensitive to loading.
ive tried this -> hot pin thru a cap to transistor base. is there
a better way ? some special resistor to ground perhaps in front of the transistor stage ?
or other mojo ?
sorry to trouble you. just trying to learn more.
thanks.

Hmmm.... I'm not sure that I follow you there. First, you have to bias the input stage to work with whatever power configuration you are using. It's going to have volts on it so you need an input dc blocking capacitor so that your source impedance doesn't change the bias of the transistor. You can certainly wire a resistor to 0v in front of that capacitor to produce a lower impedance but I wave a flag that this is an unbalanced circuit... not the best option to use for noise cancellation.

So, either stick a transformer in front of it or cobble together a differential input pair (that's two transistors joined together at the hip - their emitters) and that'll give you an electronically balanced input.

Honestly, you'd be surprised at what you can get out of two transistors and a TL071!

Leave the exotic components until you have a core working stage.



Geoff

[manning1]

dear geoffT.....thanks for being so kind in responding.
in reply....please feel free to critique.
i'm from UK btw. a physicist/computer engineer by training
with a lifelong love for this craziness. my wife thinks i'm nuts !
1. to clarify i DO use a blocking cap (low esr) on input
to transistor base. also of course on the way out of the circuit.
2. i have tried desperately to keep the use of resistors down to a bare minimum,
thus keeping the gremlins of thermal noise at bay,
in fact in one common emitter pre circuit i made i have only two resistor - from power
to the collector. the other from collector to base for gaining. it sounds very good as a pre with makeup gain by a line mixer etc.
i also like to experiment with hanging crazy combos of caps/resistors hanging off the resistor going from
collector to base. i know its crazy but sometimes i get some interesting things happening
adding weird topologies in different places.
3. i think ive tried every differential in circuit topology,
but one problem with op amps is the bipolar supply needs.
in fact ive been bugging op amp manufacturers for years to find some way of generating the -ve on chip thus eliminating the need.
so one just comes in +ve into op amp and smart circuitry on the op amp creates the -ve. thus instead of a 8 pin device. an op amp is reduced to just +/- inputs, supply pin, and output pin. and maybe a couple of pins for setting gain via a resistor.
4. ive tried tons of op amps but the ones that seem to give me this wierd goosbumpy
feeling ie: "the sound i want" is the op27 and varieties. i particularly like max427/37 not for any scope tek reasons but purely for the sound of my tracks.
what i do is ACTUALLY RECORD TRACKS and listen on playback while designing a circuit and trying different components.
in fact i got away from purely technical scope analysis awhile back
because i found that good scope results dont often translate to the sound i'm looking for.
any comments on the use of op27/37 in mic pre's.???
one prob is the big supply current needs. which has stymied me in portable apps where no mains power is available.
5. one of the things ive been trying to do is get something very portable going as there are some phenomenal little animal sounds i would like to record in this semi rural area i live which is peacefull. ie: using just 9volt batteries.
ive tried micropower op amps but the noise floor is too high.
any ideas ?? also this year ive tried a desperate search for any
low power vac tubes. to do the same thing. apparently they were around years ago. do you know of any ?
members of my family who used to be into tubes tell me there were many 50 years ago.
thanks for your comments geoff. i really appreciate any help/ideas from someone with
your great background. i'm still a neophyte.
ps - on the 57 i do use a blocking cap. i was just curious if anything else was needed because i'm trying to understand why 57's in particular are sooo sensitive to different topologies.
best regards to you for your help.