trafoless microphone preamp wish list

[rafafredd]

[quote author="CJ"]I meant these:[/quote]

What? Do you mean these:

http://hps.infolink.com.br/rafafredd/DSC03889.jpg

http://hps.infolink.com.br/rafafredd/DSC03890.jpg

http://hps.infolink.com.br/rafafredd/DSC03891.jpg

http://hps.infolink.com.br/rafafredd/DSC03892.jpg

http://hps.infolink.com.br/rafafredd/DSC03893.jpg

These things are actually starting to crowd my place :?

payed $150 for all the lot :shock:

 

[Learner]

[quote author="analag"]No man....it's 250V, I use this regulator, although I juiced it up a little. Very quiet. And you will find the discrete regulator generates less heat and you can set the current limit to your specific needs.

Another kick ass regulator, I have this one in one of my mic pres.
Analag[/quote]

Right, looks good to me! :grin:

BTW, is that general op amp where any will do? Also, what kind of MosFet is that??

I am going to build one of those, I think I have to rebias all my tubes again for 300V.....

 

[Learner]

[quote author="rafafredd"][quote author="CJ"]I meant these:[/quote]

What? Do you mean these:

http://hps.infolink.com.br/rafafredd/DSC03889.jpg

http://hps.infolink.com.br/rafafredd/DSC03890.jpg

http://hps.infolink.com.br/rafafredd/DSC03891.jpg

http://hps.infolink.com.br/rafafredd/DSC03892.jpg

http://hps.infolink.com.br/rafafredd/DSC03893.jpg

These things are actually starting to crowd my place :?

payed $150 for all the lot :shock:[/quote]

wanna sell half of that ***** for $100 :green:

 

[tk@halmi]

The design should utilize selenium diodes and paper in oil capacitors as well. Anything else would be less than ideal!

 

[CJ]

Excuse me, but I count 16 or 17 noise making, thermally corrupt, spike sensitive solid state components that have "Kill Me So I Can Short Your Power Transformer Out" written all over them.

Tube version would use an EL34 pass tube just loafing around 24/7 for 30 years without a single problem.

 

[bcarso]

[quote author="CJ"]Excuse me, but I count 16 or 17 noise making, spike sensitive solid state components that have "Kill Me So I Can Short Your Power Transformer Out" written all over them.[/quote]

I think I need that T-shirt! :razz:

There is something immensely reassuring about a tube pass element, however inefficient. And once you've done that why stop there. But, I wind up with sand state even though the bench is littered with dead soldiers while getting there. Now if they keep dying in the field that's another story.

I designed a 700V voltage compliance constant-current power supply, all sand-state, for hollow-cathode spectroscopic reference tubes early on at UCLA. I didn't really know much but I loved power supplies and feedback. I kept banging away at this thing trying to get it stable against oscillations, initially not realizing that the gas tube load was about as impossible to manage as they come. To make matters worse the control electronics were floating up at high volts relative to the chassis, so making any changes in C's etc. was perilous (of course you got impatient cycling the power each and every time), and I got slammed hard a few times.

Some years later a prof. wanted to haul the thing off for a week, which would have disabled the field station telescope for anyone wanting to do spectroscopy. So I built another supply in a few days with a better design, as by then I knew a little.

 

[chrissugar]

Back on topic,

Brad, how is the transformerless preamp developing? What do you find are the main problems/compromises? It will be completely discrete or you will use some chips too?

chrissugar

 

[bcarso]

Chris, I'm moving towards either a paralleled FET input with some bipolar assist (compound pair-triplet sort of thing) or a paralleled bipolar basically diamond quad kind of arrangement. The latter is not the most quiet, but I think it will be quiet enough.

The last sims I did show, with ideally matched devices, a way of tuning out all of the third on each side. Then when the two sides are used differentially the second goes away. 4th and 5th are already very low. I'm seeing numbers that will surely be worse in actual silicon because of mismatching and second-order effects, and thermal distortions, but very intriguing. The sim program wants to bomb a lot which is a good sign something is cooking ;-)

The hope is to do the gain variation without feedback, other than local. This should allow something approaching the Stamler criterion of no sonic signature change with gain change. Also, if the gain is changed by mostly developing the signal from a current output across a variable resistor, the noise should remain low at low gains.

I wish I could work on this 24/7 but there have been some urgent distractions lately.

 

[jrmintz]

This is absolutely fascinating. Thanks. :thumb: :thumb:

 

[Winston OBoogie]

...

 

[bcarso]

John Le Carre :razz:
Actually (as I quipped before) more like Robert Ludlum

 

[Learner]

[quote author="CJ"]Excuse me, but I count 16 or 17 noise making, thermally corrupt, spike sensitive solid state components that have "Kill Me So I Can Short Your Power Transformer Out" written all over them.

Tube version would use an EL34 pass tube just loafing around 24/7 for 30 years without a single problem.[/quote]

yea..ahm, you refering to the this CJ?

 

[Svart]

damn, I go away on business all week to find that this is moving on without me.. both great and sad at the same time... :roll:

So how's it going Bcarso?

 

[rafafredd]

learner, I think he is talking about this:

http://www.duncanamps.com/images/dds2250v/cct_mosfet.gif

 

[chrissugar]

Hi Brad

Just a little bump to ask if you have any news about this project.
What input configuration you choosed, paralleled FET or bipolar?

chrissugar

 

[bcarso]

Still divided although the bipolar looks extremely promising from a distortion standpoint, since I realized how 3rd could be nulled out.

It does require a lot of devices though to get down to really low noise levels.

Unfortunately for this project, I have been swamped with consulting work lately, and aside from the odd post in here it's been tough to think about much else. Thanks for keeping it alive.

 

[Svart]

yeah we are still interested and as always, if you need some help with it, we are here!

:thumb:

 

[tk@halmi]

Brad,

It would be lovely if you let us stare at a preliminary schematics. Most of us take days to years to figure out what is happeing in a circuit. Let me know if it needs hosting.

Thanks,
Tamas
PS.: If it requires a lot devices BJTs are preferable too.

 

[pstamler]

[quote author="tk@halmi"]PS.: If it requires a lot devices BJTs are preferable too.[/quote]

But FETs have the advantage of ignoring RFI more than BJTs.

We seem to have come up with two Stamler Criteria:

1) Behavior shouldn't change with changes in gain.

2) No high harmonics. (To which I would add the codicil "and the low harmonics that are there should decrease monotonically with number": 5th < 4th < 3rd < 2nd.)

If this be my legacy, I will go to my grave content. Well, no, actually I'll be really pissed off at the fact of going to my grave at all, but this will mollify me about 0.2%.

Peace,
Stamler

 

[PRR]

> the low harmonics that are there should decrease monotonically with number

"monotonically" is way too generous.

And not strictly necessary: a harmonic can null to zero without requiring all higher harmonics to be zero. The ear will notice a "stand-out", it rarely notices "missing" harmonics in low-distortion sound. (Yes, missing harmonics are key to the sound of certain musical instruments, but these are "high distortion sources" where the overtones dominate the sound.)

At the least, harmonic distortion should decrease with the order of the harmonic. The 10th harmonic 1/10th of the 2nd harmonic.

Even that is very "phatt". 5% 2nd harmonic is nothing, but 0.5% 10th harmonic is grating.

The square of the order is probably a minimal goal for "clean".

See Radiotron 4th, Chap 14, pg 603-611 (pg 1-9 in the PDF), especially pg 610 (pg 8 in the PDF). Although over 50 years old, this work is still valid. Many of the faults of solid-state amps trace to blind lowering of THD, without weighting the higher more-offensive harmonics.