mic pre concepts

[Svart]

if we do that then we have to be smart about it. it seems that for small problems we always pile on complex solutions.. So i would make it a goal to work through any problem for the SIMPLEST solution, this should also prove to have the least amount of parts in the audio path too.

opinions?

 

[tmbg]

I love that idea! Less parts means less chance of screwing it up somehow

 

[CJ]

a fixed gain micpre with no negative feedback would be very simple.
we could split this thread yup into tube vs solid state, then have a shootout between the two designs after they are built.
we would have to decide on what mic to optimize both for so the testing would be consitent.

 

[tmbg]

standardize on the U87! Everyone has one...

















.... don't they?

 

[CJ]

No, but I bet everybody and their one legged stepchild has a SM57 laying around?

If you want, we could trade, the solid state group goes for color and the tube heads go for clean?

 

[Svart]

I hAtE my sm57s.. :green: truly.

i love my m201, C1, B1s(studio projects) oktavas.. mostly condenser mics.. but i have never liked that sm57 sound. My drums are either sharp and high pitched or low and tight.. that 57 is neither.

:green:

i would rather vote for a SD mic to design around... i guess i gotta be difficult right? :guinness:

 

[CJ]

OK, maybe we should design around a vocal mic, not a guitar cab mike.
But which one? Medium capsule condenser? I guess most people own at least one caspule mic, rather it be expensive or not.

 

[Svart]

well lets look at what the GroupDIY target audience has in stock..

I would bet that most have a cheap mic with a 797 capsule and some sort of modded circuit (tube or FET based). now the original circuits are usually shrill and/or boomy with lack of definition or some spectral anomoly sitting there screwing the sound up and giving us that 99$ sound instead of that 9999$ sound..

most mods are attempting to fix those issues, a lot of those mods with the cheap but decent capsule are based on a tube. since we have the tube sound, do we want to keep that exactly as it comes out of the mic? I would think so. so now we must have a clean pre with no color. my solid state condenser mics output pretty hot too as I'm sure most do when phantom/powersupply is involved. do we REALLY need 60db of gain when 40 will do? my console pres have 60db of gain and i have NEVER turned them more than 5/8 of the way up. seems like a little overkill if you are like me and cannot afford a ribbon.

so is 40db of gain enough? how about a fixed gain stage(like mr CJenrick suggested) with no feedback and switched gain output (almost back to classic, pre gyrator designs eh?) I'm just asking questions here for the sake of trying to spawn ideas.. how about a sidechain circuit for plugging in a tube into the audio chain for a simple hybrid?(yes i know a lot of cheapo companies try this as a gimmick, but can we do it *right*? pres*nus, we are coming for you..) I think something like that would be fun and challenging, moreso than the one or the other type of design.

ah just my .02$

:guinness:

 

[CJ]

if the parts count is kept down, we could put a tube and solid state mic pre in the same box. never seen that done before. a simple V72 circuit modded to omit the feedback loop and have 40db fixed would be very easy.
an EF804/EF86 type driving a 6SN7 might sound pretty good.
or a 6N1P first stage. I have been hearing good things about this tube, if you find the right ones.

 

[Ian MacGregor]

So are we talking about a fixed 40dB (or so) gain stage with an attenuator on the input? Or output? If it is on the output, the gain stage must have a lot of headroom right? Are we talking about gain control at all? I kinda dig the idea so far. I would personally like to see the gain stage be BJT/FET based (only because that's what I'm doing in school at the moment).

Cool stuff so far...

Ian

 

[Svart]

ah now see, that's what I'm talking about! would that be something we could switch into and out of the circuit on the fly?

I would LOVE for it to be FET based. If you can't tell from my postings, I'm a FET nut.. :green:

 

[CJ]

well, i was talkin about a stereo micpre, one tube, one transistor. now on the transitor side, you could maybe have a switch between fets and bjt.
I think attenuators degrade stuff, I mean if you use mic placement as your attenuator, you save that much more noise. I think it was fairly common to have fixed gain mic pre's in consoles in the old days (Langevin AM16 comes to mind). Saves parts and wiring, thats for sure. Attenuators are a pain to source/build and they degrade the signal. you just use whatever comes after the micpre to control level, as long as you have the headroom.

 

[Ian MacGregor]

[quote author="cjenrick"] Attenuators are a pain to source/build and they degrade the signal. you just use whatever comes after the micpre to control level, as long as you have the headroom.[/quote]

Ahh.. I agree. BUT, in today's Bro Tools/etc. world, there are many instances where the only thing after micpre is the AD. Varying mic placement may not be an option sometimes (close micing of any source with goal to minimise room sound for example).

This is just an idea, but what if our generic "gain block" was relatively low gain (~10-20dB) and the method of altering gain would be switching stages in/out? Sorta how Neve did it. Having the gain variable in 10dB (or so) steps would be better than none at all in my opinion. Would noise performance suffer as a result of this?

I like the brainstorming, it makes my brain work on something other than Laplace transforms and convolution.

Ian

 

[CJ]

Thats exactly what Grant Carpenter did on the Gordon micpre. Super low noise as you are not building the signal up just to cut it with neg feedback and attenuators. Might be a lot of parts, especially in the tube section. Maybe do that only for the solid state side and use voltage reduction to play with the gain on the tube side.

 

[Svart]

putting various gain elements into the path is a cool unorthodox idea. however, with higher gain we will have more noise to begin with and adding more gain stages, i.e. more stuff in the audio path, *should* introduce even more noise into the signal... however if the stage is designed well, shouldn't we get away with an acceptable noise floor?

I know current noise would be a problem with this idea but i'll throw it out anyway.. FETs as our gain elements, but with variable biasing voltages to increase or decrease gain instead of feedback/phase cancelling..? use some adjustable lm317/337s(with large resevoir caps..) to adjust the voltages to the FETs? we could use LARGE die fets to start. I wonder if something like IGBTs(my favorite) would work in this situation..?

wow that would be cool

 

[alk509]

[quote author="cjenrick"]we could put a tube and solid state mic pre in the same box. never seen that done before.[/quote]

There's a box that does just that. I think SPL Labs makes it.

For what it's worth, I'm in the process of designing a 6N1P-based mic pre (transformers in and out, two 6N1P grounded cathode stages with about 50dB of gain, a 12BH7 WCF and global feedback to control gain). I had it all nice and ready for prototyping, but PRR came and fucked with it... "This is too low here, that's too high there, that's not gonna work" :roll:

I just got some 12BH7's in the mail today (thanks Tom! :green and will start testing different WCF's soon. I'll keep you all posted.

Anyway, sorry for the hijacking. Carry on...

Peace,
Al.

 

[CJ]

Al, are you using the Svetlana stuff?
cj

 

[alk509]

Yes, sir.

Al

EDIT: Make that "No, Sir". They're actually Sovteks.

Peace,
Al.

 

[CJ]

Has anybody tried out any of the Cranesong stuff?
I think they take the approach of building different pres for different jobs.

Nice rack, eh? AES shot.

 

[Ian MacGregor]

Yes, the Cranesong stuff looks great and from what I hear sounds just as good. I remember putting a rack like that together for another hi-end audio company a couple years ago. It's fun to step back and think about having the rack available on a tracking session!

Back to the mic pre design brainstorming...

I've been mulling over the multiple amplifier stage idea and I 've thought of a couple different things:

-It could be say 3-5 amp stages in series, the last stage would probably want to be a low gain, high power (is this right?) output stage to drive the desired +26dBm (??) into 600 ohms or whatever. The first 2-4 stages would be fixed at some relatively small amount. I was thinking anywhere from 10-20dB or so.

-I would personally like xformer coupled in and out, while the amp stages are fully discrete Class A. No feedback, no attenuation. Possible??

-I forsee the amp switching arrangement and noise minimization the most difficult things to accomplish, but what do I know...

-Some have mentioned tube stuff, and although I know little about tubes, it might be cool to have some gain stages be tube, some solid state. Even possible to switch tube in/out, etc.

I dunno, bunch of ideas! What do you guys think??

Ian