This Works

[PRR]

> my suggestion of joining R14 and R16? 20 dB more CMRR is always handy, especially if it even saves one part, improves DC offset and shouldn't have any sonic side effects.

Agree. With that change, I would not worry about output offset. I would however like to see 100Ω output resistors for long-line stability, and perhaps a place for Phantom input caps.

R17 is pointless. Change R1 R2 R18 R19 to 90Ω and the max gain is the same, noise the same, DC match triflingly better, and we save a part.

Distortion on large outputs may not be point-oh-oh-oh stuff but should be small and inoffensive.

Noise on passive (dynamic and ribbon) mikes is not theoretically low, but close enough that it is a non-issue unless you record harpsichord in Iceland.

I would prefer BIG parts in the input. TO92 transistors and 1/4W-1/2W LEG resistors. There is some component of resistor noise that is inversely related to size. When you have 10 zillion carbon atoms, you don't hear each one vibrate; when you have only a zillion atoms you start to hear their individual random agitation. This was not an issue until these teeny-weeny SMD parts came along. If you don't like drilling boards, put down 6mm square copper pads and butt-solder the resistors. I wouldn't go on truck-tour with the Rollin Stones that way, but I do slam butt-soldered PCB amps in/out of bags on my gigs.

 

[Svart]

I have the schematic transcribed to my flavor of circuit drawing thingy and will apply the suggested changes tommorrow before i start to layout.

I know you guys hate SMD but it is the future. :green:

I'm going to do a simple layout with headers for input, output, circuit power and phantom power. Sorry guys, i'm going to pack this as tight as possible because I have a need for some preamps like this in a very small package for some possible field use. As a concession I will use 1206 size resistors and have large pads for caps in either through-hole, SMD or through-hole mounted like SMD(which is fine for us DIYers..)

let me know what else you want for now. If this turns out good i might be persuaded to do a through hole layout.

:thumb:

 

[bcarso]

In my brief but intense crisis with excess low-frequency noise from resistors in a client's sub amp front end, detailed in an earlier post, I did find the bigger the better to hold true for the thick film parts running d.c. 1210's were the largest I tried and they helped a lot.

I never did get fancy thin-film parts but the Matsushita garbled-text reference does show them as being much better than the thick films. Then there are the MELF parts which are just like spiral-cut metal films without leads---but manufacturing will hate you since they tend to roll off their pads.

The d.c. operation simply reveals the excess noise capability, so even if there is no d.c. the lower noise parts should be used in critical places---otherwise you will have the excess noise appearing as a signal-modulated source.

 

[Samuel Groner]

With that change, I would not worry about output offset.

Why not? because it is essentially CM?

Samuel

 

[Svart]

:evil:

someone in my department decided they needed my boxes of BC550 and bc560 more than i did.. when i find them they will be owned.

 

[analag]

Samuel I don't go into "audiophooleric" sound desriptive terms...sorry.
As far as transformerless input SS preamps go, this is very good, with speed, presence and clarity. Very clean!!! I refer to this as my "transparent" pre, and we all need one of those.
I suppose I could join the resistors for improved CMRR but it works and my attention has already shifted to my next challenge, but I will try it.

I also built this pre specifically for my tube mics, so no Phantom Pwr provisions were made. For those intrested in doing so...a pair of 47uF caps should work fine, a tightly matched pair of 6.81K one percent or better preferably metal film resistors...

 

[smallbutfine]

Hi Analag,
i wonder what is C4 supposed to be? a bipolar electro?
How many power the whole circuit drains?
Would a +-15V supply severely degrade the sound quality?

This one seems so easy to build that i'm already doing it on a spare veroboard right now immediately after reading this thread.
Using hfe around 430 matched BC560 and 0.4W metal film resistors... :grin:
I will try to go with OPA2132 or OPA2134 as TI suggests me and i have them...or would MC33078 be a better drop in for the moment (bjfet)?
What do you think?

Kind regards
and thank you for bringing this one up (at a time when everybody wants a good pre for mic comparing :grin: )

:thumb: :thumb: :thumb:

Martin

 

[Svart]

Dunno about C4, i think either would work here. +-15 should work fine, around 1-2 db of difference i believe.

I have most of the circuit layed out. should take up around 2 square inches including phantom caps and connectors.

 

[smallbutfine]

I fished up a 1000u audio electro, this should go - to R17, + to Pot, am i right?
(ehem confused right now )...

Thanks Svart.
(hmmm...verry small, what you are working out... )

Martin

 

[PRR]

> Why not? because it is essentially CM?

With R14 R16 joined, the output DC should be just a couple milliVolts. Any Line-In that can't eat a few mV does not belong in a studio.

 

[Svart]

Smallbutfine, go ahead and leave out r17. not needed. change the other values as suggested also.

 

[analag]

The pre will work without R17, it is there to prevent full gain so it can be avoided. C4 can be any type of electro it doesn't matter in this application. +/-15V will work just reduced headroom. Right, the cap goes from R17(or from transistor) to pot. The Pot should be 10K connected in reverse taper, in other words when the resistance across the pot is lowered the gain goes up, so if you wire the pot up as we normally do...turning it anti-clockwise would increase the gain, just tie the wiper to the side that normally goes to ground.

Analag

 

[Svart]

Analag, didn't you have a tube/ JFET hybrid circuit? I can't find it through the search..

 

[smallbutfine]

Thanks alot for the fast replies!
Yes i leave R17 out and change the R1/R2/R18/R19 to 90 Ohm as PRR suggested.

I also stuffed 2x 100u electro paralleled with 2x 0.1u film input caps on the veroboard, like i saw it in the green mkI schem, for phantom power issues.
I will solder the phantom supply, as seen in the green I schem, directly point to point *later*, (green R36:300R at the switch and green C15:10u 63V from both 6k8 to ground at the in XLR).

Overall, the transistors and caps are stuffed, i left some space for the resistors and now its going on to the dual opamp section.

Here i will substitute R14/R16 with a 300 Ohm resistor between the opamps like Samuel suggested (and PRR explained in detail :thumb: ) and add the 2x 100 Ohm for improved long line capability (as seen in the green mkI scheme, too!).

Since i have a small bulk of rotary switches spare from my green pre diy-ing, i will make another gain switch for the reverse log 10k instead of a pot (yeah, maybe i will make some more of them afterwards!).
Thanks Analag, for explaining the gain pot (i would have asked about the type of pot used a few minutes after your post for sure, it came exactly when i began to think about that :razz: , perfect timing :thumb: )

As it seems this one will take place side by side to my Green MKI in the same 1HE 19", sharing the psu! Very nice for comparing, a 'clean mic pre'-box!

I will post some pictures when i'm done with the veroboard, if my wife lends me the camera and doesn't kill me for doing electronics all night instead of doing all that things women want us to do... :wink:

Thanks to all of you for your kind help, it is great DIY fun for me tonight!!!

:grin: :thumb:

Martin

 

[PRR]

> electro, this should go - to R17, + to Pot

Strictly, it should be a non-polarized cap, in which case there is no + or -.

In fact, the AC voltage is so very low that everybody uses polar electros. Put it in either way: both are equally "wrong" and work just fine.

> The pre will work without R17, it is there to prevent full gain

Max gain is limited by the emitter resistors. If you want the same gain as before, without R17, they should be 90 ohms. Unless we need to match the prototype, they can be 100 or 68 ohms. They could even stay 43 ohms, for 6dB higher maximum gain, but with typical 10K taper pots there will be a big jump between 9.9 and 10. The original values (43||43+47+43||43) are good but not critical.

 

[analag]

didn't you have a tube/ JFET hybrid circuit?

Yes but the schemos I put up were the initial idea, not the final thing.
It has a powerful and aggresive sound to it. It can cut through any mix.

 

[smallbutfine]

Yes, i saw the theoretical need of a non-polar cap as there had to be some ac voltage at this place, but i wondered where to find a 1000u non-polar cap tonite :grin: :wink:
Thanks, PRR for explanation.

Nearly ready with this one (two resistors left, taper, wiring and *testing* :grin: ), but now some sleep is needed, it begins to look very ugly on the board, totally forgot that i *hate* veroboarding... :shock:
More later today...

Have a good night...or day or whatever :wink: .

Martin

 

[Svart]

any chance of seeing the final schematic of the JFET/valve combo?

 

[bovox]

Hey ,maybe I´m totally out of date..but where can I find this scheme from analag? Cheers Bo

 

[flaheu]

I also would like to se this schematic, is anybidy kind enough to re-post it ?

Thanks