why are there so few discrete transformerless mic preamps?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

ikiru

Well-known member
Joined
Jun 3, 2004
Messages
55
Location
vancouver, canada
I've often wondered, there seems to be a dearth of transformerless discrete mic preamps. Most transformerless designs seems to be IC based. Why is that? Is it more difficult to design without transformers with discrete components? fred forsell seems to make one...why aren't there more? Is it not worth it? I'll stop there...

-jay
 
I'm mixing a record right now, when I'm done, I'll scan my spectrasonics cards, and the schematics, Can someone help me find a place to host the images?
Thanks,
Peter
 
people around here seem to like place called photobucket. It is free! it does have some restriction on the hours that it accepts new members though. 10pm -2pm or 10am-2pm..100mb storage.

I would like see the spectrasonics peter, thanks!

are there any other notable designs?


-jay
 
The GML mic pres ( 83xx).. the 8900 compressor, and 8200 eq made by GML is discrete build without transformers..

I think that..that money, and quality are the to "main factors" why very few build discrete.

Kind regards

Peter
 
Isn't the Millennia HV-3B transformerless & discrete?

I think most of the transformerless discrete products are very expensive. It must have to do with the difficulty of removing DC from the circuit without using "sound-degrading" caps. Transformers make it easy, and usually impart a desireable color to the sound.

in fact, these days I'm wondering if it isn't so much tubes, but the transformers that they're coupled to that make some designs sound so good.

Designs I love are the ISA 110, API 312, Neve 1073, Great River....and all use transformers. In fact, the sweetness I hear in these pres is what I would have described as tube sound a few years ago, before I started really listening to nice gear.

Shane
 
[quote author="Category 5"]
I think most of the transformerless discrete products are very expensive. It must have to do with the difficulty of removing DC from the circuit without using "sound-degrading" caps. Transformers make it easy, and usually impart a desireable color to the sound.
Shane[/quote]

I highly doubt that...as most X-formers cost much more that the dc-servo circuits you talk about, and no its not that hard to do...*S*

Its "just" that its soo very time consuming to build/design/match etc with discrete designs..and yes on top of that you may have to have people (who wants a paycheck) to do this as well... transistor for transistor..cap/cap..resistor/resistor etc..not to mention "redesigning" every time parts manufactors deside to take a discrete component out of production..on top of that..most highend manufactor´s do not sell by the 1000 units a year..most by the 100, or less whitch adds to the cost pr units as well

you know..to buy a Lundahl x-former and a 5543 chip and put them together ( and call it a preamp) is not that hard to do...compared to the all discrete design ;-)

Kind regards

Peter
 
So i guess it is the individual attention needed to match up all the components...Seems like an ideal type of project for us diyers. As much as I like the simplicity of transformers, gml and little labs must be onto something!


-jay
 
hmmmm, that may be oversimplifying things a little bit, there are PLENTY of transformer balanced designs that rely pretty intensely on transistor matching, etc. etc. One could easily spend as much time desiging a transformered circuit vs.a transformerless circuit. It may come down to simple personal preference from a designer as expensive transformer designs sell well so...

dave
 
Buzzaudio's MA2.2: http://www.buzz-audio.com/products/ma2.2.htm

The insides are here: http://www.buzz-audio.com/products/ma2.2mbtour.htm
 
[quote author="Category 5"]in fact, these days I'm wondering if it isn't so much tubes, but the transformers that they're coupled to that make some designs sound so good.[/quote]

My buddy Mark who designs and mods guitar amps; the person that got me into DIY, always says, "The tone is the sum of all the parts." The tubes are only part of the "vintage" tone. The tubes, transformers, caps, carbon comp resistors, even the pots, wire, and the layout affect the tone.

For example: My DIY LA-2a has a more transparent tone than the original, because I used MF resistors, and Jensen transformers. It does however have the same compression characteristics.
 
[quote author="ikiru"]I've often wondered, there seems to be a dearth of transformerless discrete mic preamps. Most transformerless designs seems to be IC based. Why is that? -jay[/quote]

Because we want symetrical input (from asymetrical transducer).
Transformer in mic pres do two thinks,
Make symetry and do noise match.
If we want transformelless amp, we must do it symetric
and noise matching must be done by paralelling transistor elements.
Theoretically it must be TR*TR , where TR is transformer-ratio.
It is not exact, because we must compensate tax paid from symetry transformerless
circuit. This tax can be 4*TR*TR.
It is very much transistors and it is good to integrated it.
In modern IC I do n t know, why noise parameters
of it is worse than discretes.
There is advantages. Small dimensions implies less spurious signals,
There is disadvantage: lower power voltage (it is limited by ? somethink
as parasitic thyristor) and then lower dynamic range.

If you want to build discrete (mic) preamp without limitations
of ICs, you must use power bipolars or highly paralel fets (i.e. 2sk170),
and must use relative high power voltage (+-60 V).
Some modern low noise bipolars are high voltage.

xvlk
 
Don't forget Valley People. It does have an opamp output, but the main sound is from the descrete transistor input with descrete constant current source.

www.vacuumbrain.com
click on the green thingy.
cj
 
right, Allison Research predated the valley and Im pretty sure that all that stuff was discrete transformerless. I have an AD&R compex that was unbalanced discrete transistor, I wound up adding transformers to it and its pretty cool either way.

dave
 
Define discrete. The definition of transformerless is obvious but what do you all consider discrete?

The Spectra Sonics 110 card is (by my definition) is a discrete amplifer because it is built using individual transistors, resistors, caps, and diodes. But it requires an input transformer (it also has a fixed input impedance of 600 ohms because it uses a grounded base input stage) for use as a mike preamplifer, so it's not a transformerless design. The Valley People module did use individual transistors in front of IC opamps as does the HV-3. Should we call them discrete designs because if a couple of transistors? I wouldn't. You might.

So how do you define discrete when it comes to a mike preamp circuit? I wouldn't call it discrete if it uses IC opamps in the signal path, period. But that's me. If there are opamps used in the signal path made from individual components( like the SS 110 or my JMP-1 and JMP-2), then to me that is discrete. My dictionary defines the word as meaning (in part) "made up of distinct parts". To me monolythic (IC) opamps do not fit that description. They are build from a single substrate (hence the word monolythic) So, to me, a mike preamp made up of individual transistors AND IC opamps is a hybrid circuit, not a discrete circuit.

Your mileage and opinion may vary.
 
Id agree with fred's definition.

I thought the spectrasonics 110 could be used completely transformerless, I guess just the output transformer is optional, I must have read the data sheet incorrectly. Sorry for any confusion.

dave
 
I both agree and disagree to some degree with Fred's definition.
What we should consider is manufacturing technology used in design and production of an IC opamp. In most technologies used today the manufacturing process by it's virtue does not result in discrete - by which i mean physically and electrically isolated from each other - transistors. There are parasitic PN junctions between the transistors and the substrate, these are normally reverse biased and should not affect the functioning of the cuircuit unless the junction 's reverse breakdown voltage is exceeded. This voltage is also what is meant when manufacturers talk about "30V or 45V linear CMOS or bipolar process". These parasitic PN junctions also contribute distributed parasitic capacitanaces.

However, in DI - Dielectrically Isolated process each transistor resides in its own well physically and electrically isolated from the substrate by a a thin layer of silicon dioxide. The only connections between the transistors are those that form the actual circuit. So, by Fred's own definition "made up of distinct parts" an IC opamp made in Di process should qualify.

But all this may just be splitting hairs in two.

THAT uses this process as well as Intersil who still manufacture some classic Harris products.
 
[quote author="buttachunk"]It's been used to record vocals of Sheryl Crow, Bruce Springsteen, Don Henley, Suzanne Vega, Soul Coughing, Peter Gabriel, etc...[/quote]

Despite that, I'm sure it's still a fine preamp.

(Sorry... couldn't resist!).

Actually, I was a big fan of Gabriel's until he entered his Armani suit period.

I saw a really interesting preamp at AES last year, transformerless, built with shitloads of discrete transistors. Two channels completely filled a 1RU case, and the build quality looked top-notch. I wish I could remember the manufacturer's name now.
 
That's Grant Carpenter's "Gordon" preamp.

The discretest of all, I dare say.

I agree with Fred that the gain elements should be individual transistors throughout to be called discrete. While an LM394 is built using IC technology, it is still just a super transistor pair. No biasing circuits, no current sources, just NPN junctions.
 

Latest posts

Back
Top