The diyAudio.com preamp project!

hi everyone,

ok so maybe this is the dumbest idea ever, but...

for the last several months i've been trying to figure out how to build my own preamp. i think there are a lot of nice commercial power amps out there but a good preamp is hard to come by for some reason. i've thrown some questions out there and have gotten great ideas from everyone. however every time i check out the forums, i see even more great information out there. basically what it comes down to is there are a lot of smart people here who know a hell of a lot more than i do about circuit design. soooo...

let's make the official diyAudio.com preamp!

i want to collect everyone's latest and greatest ideas on everything - circuit topologies, parts selection, power supply design, chassis fabrication, whatever - and distill everything into what will hopefully be a very cool and very nice-sounding preamp. kind of like those engineering projects we all (or at least some of us) did back in school, it will be a group effort. like any good project though, we will need some requirements. here is where i'd like to start:

- must be solid state
- must be DC coupled (no single-ended circuits i guess sorry :()
- HF response at least -3dB @100kHz
- ideally implemented with all or mostly SMT devices
- each channel will be a separate module
- must have a digitally-controlled attenuator with at least 1dB resolution over a 70dB range
- a microprocessor control unit will handle all control functions

these requirements may sound a little quirky to some of you but the point of it is to make a very modern, fully customizable unit that will work as a flexible but sonically-transparent control center for multichannel formats (e.g. SACD).

we can collect ideas on this thread, and i will put up a website dedicated to the development of the preamp where we can store data, pictures, schematics, etc. i will also do the prototyping, and if we get to the point where we can start building i will get PCBs printed etc. and pick up the tab when necessary. if enough people are interested i can even build extra modules for other people to play with and send it to them (in that case donations would be appreciated to offset the cost). basically i will do most of the legwork, i just want you guys to throw all your best ideas out there!

if any of you elders out there think this whole thing is just silly, please feel free to flame me and and i will stop the thread and go crawl back into my parts bin...

what do y'all think?

cheers,
marc
 
I stiil want to be alone

Once again.....
 

Attachments

  • garbo01.gif
    garbo01.gif
    44.2 KB · Views: 6,653
harry: you spend a lot of time on these threads. do you sit there and neurotically refresh your diyaudio.com window every minute or two like me? :p

anyway, two votes of confidence is good enough for me! i'm sure this thread will get long and messy eventually (or at least hopefully it will) but the web site i'll put up will help us distill the signal from the noise, so to speak.

joe: i want this to be a thoroughly modern design, so i think SMT is the way to go. traditional through-hole is more DIY friendly i guess but after asking a few people, SMT does not seem so hard to work with (at least building is not - disassembling is a different matter). it also has a lot of benefits: the absolute shortest signal paths, minimal parasitics, better thermal coupling, etc.

steve: i have not heard the pass labs stuff but there is no doubt in my mind that it sounds as glorious as people say it does. but if i'm not mistaken, don't all single-ended circuits require capacitor coupling? while AC-coupled circuits can sound excellent, again i am aiming for the absolutely most modern approach possible and i'd like to avoid capacitors in the signal path. it is just the design philosophy i would like to take, and not a put-down of the more old-fashioned approaches which can and do sound excellent. if we can do single-ended w/o the capacitor, then great!

harry (again): that's a great quote, and unfortunately i'm sure it will apply to this thread. but hopefully we can keep the tone of discussion constructive at all times and make some headway with the design.

ren: sorry you don't like my idea. :( is it just the idea of a group project, or maybe my requirements that bother you? just curious...

keep in mind that if we have a lot of conflicting ideas as to what will work best, we don't have to offend/favor anyone by choosing one over the other. i will just build them all! then i can listen and measure to see how they perform. it may not sound practical but it will be the most constructive.

cheers,
marc
 
Marc -

I deleted my earlier post because I wasn't certain if the Pass Stuff was actually DC coupled... I was going to check...

Regardless, I've heard quite a few preamps and to my ears the Pass stuff is damn near the best if not the best I've heard...

No, I don't work for Nelson and have no vested interest...

Find a local dealer, bring the best front end you know of (borrow from friends etc..) and put it up against the current X stuff. You will be very surprised!!! It will set your target reference...

Another angle would be to find a local Adcom dealer and audition the GFP-750 (also Nelson's design). Very simple circuit and very good sound... It is my understanding the it's cross between the Zen and the Aleph circuits both of which are available...

Just a few reference ideas...

BTW - I'm not against DC coupled or any other criteria; it's just that I've heard DC coupled preamps that can't hold a candle to the above...

I'm sure the tube guys will love this post...but if you want tube warmth with SS slam then go listen to the above...

Steve
 
dorkus said:
hi everyone,

ok so maybe this is the dumbest idea ever, but...

for the last several months i've been trying to figure out how to build my own preamp. i think there are a lot of nice commercial power amps out there but a good preamp is hard to come by for some reason. i've thrown some questions out there and have gotten great ideas from everyone. however every time i check out the forums, i see even more great information out there. basically what it comes down to is there are a lot of smart people here who know a hell of a lot more than i do about circuit design. soooo...

let's make the official diyAudio.com preamp!

i want to collect everyone's latest and greatest ideas on everything - circuit topologies, parts selection, power supply design, chassis fabrication, whatever - and distill everything into what will hopefully be a very cool and very nice-sounding preamp. kind of like those engineering projects we all (or at least some of us) did back in school, it will be a group effort. like any good project though, we will need some requirements. here is where i'd like to start:

- must be solid state
- must be DC coupled (no single-ended circuits i guess sorry :()
- HF response at least -3dB @100kHz
- ideally implemented with all or mostly SMT devices
- each channel will be a separate module
- must have a digitally-controlled attenuator with at least 1dB resolution over a 70dB range
- a microprocessor control unit will handle all control functions

what do y'all think?

cheers,
marc

I am in!!!!!!!!, but you know what you are asking for!? ;) I mean what is the pricetag on this project? Should it use solidstate switches or relays?

Sonny
 
steve: sounds good. i know the pass stuff is highly acclaimed so definitely worth considering for the project, DC coupled or not.

sonny: this is "reasonable cost-no-object." :p ok ok i know that makes no sense but i think you have some idea what i mean. i am willing to put in as money as necessary to the design as long the return on investment is good. now of course "return on investment" for sound quality is purely subjective but we can judge that as we go along. :)

definitely relays, not solid state! at least for input switching... for the attenuator, i might be willing to make a compromise.

cheers,
marc
 
If there was an answer to the question that started this thread, we could take down this forum.

At many points, dorkus is going to have to decide who's opinion to listen to and who's opinion to disregard.

Method of switching of inputs, what type of gain and how much gain, outputs (balanced or unbalanced) brand of connectors, wires, what type of power supply and flavor of regulation - if any.

Just look at "cascodes stink" thread.

Reminds me of a joke.
Q. How do you get 7 answers to 1 question about the economy?
A. Ask the question to 4 economists.

Weeeeee!

Aud_Mot
 
I think this idea is worth a try. Even if not everybody would agree on certain Dorkus' decisions the information generated through the thread would be useful to anybody thinking about building a preamp. I was playing with the idea of cost no object preamp for the last few years. I think that a preamp is the most interesting component to build from the whole audio chain (as well as the most difficult to succed). It is also the most useless component (if not using phono). I'm building Aleph P1.7 right now on my old chassis but would be interested in any ideas coming from this thread. I'm not using preamp in my main system right now because I still think that the best preamp is no preamp, but will need one for SOZ. Count me in.
 
Sounds like a great idea.

For the smt device, I'd consider the AD8610. I've been very pleased using it in both active x/o and phono preamp. Using a precision +/- 13V supply for the AD8610 to get the maximum output from it based around a low noise ZR431 precision voltage regulator with ZTX653 and ZTX753 pass transistors, this opamp can supply up to 50 mA of current for essentially no noise and wonderfully musical sound. (Supplied voltage for this precision shut regulation can be almost any voltage more than +/-15V so it's very flexible.)

For a more exotic approach using smt devices, take a look at the proposed preamp by Rod Elliott at http://sound.westhost.com/highspeed.htm

As for relays, I rather like the Aromat compact relay (Digikey part 255-1231-ND): nice and small, 12V, DPDT. Combine both poles for lower resistance and better reliability.

The remote control and/or display seems like the trickiest part to design.
 
ah, here we go... :p

i knew it was only a matter of a few posts before someone suggested an IC part, and i have no doubt it will start a stink around here... discrete vs. IC, here we go... :cool:

i personally am neutral on the matter, but many people feel strongly one way or another. i actually asked this question in a separate thread months ago, and after some many informative comments the conclusion seemed to be discrete is the way to go, but only if you're prepared to tweak like crazy. ICs are simpler and in implementation of course and can sound excellent, but will probably be beaten by the best discrete designs. sound reasonable?

having said that, i'm still very interested in trying out some IC circuits, and pitting them against discrete design. so please feel free to suggest IC topologies for this project. (the walt jung buffered/compensated opamp layout comes to mind.)

i'm not too concerned with the complexity of the control and display circuitry. i've got a Atmel AVR development kit to play around with now, and i did this sorta stuff while i was an EE/CS major so i think i'll be able to figure it out. besides, digital stuff is easy... either it works or doesn't work. :p wish i could say the same about analog circuits. :rolleyes:

HPotter: glad to have you onboard. :) i too agree with you, the best preamp is no preamp, but of course there are flexibility reasons to have a preamp. also, i've come to the conclusion that all passive preamps are prone to having somewhat diminished dynamic impact and slightly washed-out tonal balance... the sound is pure and clean but a little bit of the "drive" and beef of the sound is lost. this of course is highly system-dependent. hopefully we can come up with an active design that will come very close to rivalling the transparency of a passive design, but with the dynamics, flexibility, and drive capability of an active.

Aud_Mot: no doubt we will have conflicting opinions, and as HPotter says i may have to pick some ideas over others. but the point of this is not to find a single answer, but to pool ideas to try out. if there is an idea out there that sounds like it has potential and is not too impractical to implement, i'm willing to try it.


marc
 
Tree huggers unite!

Tried to warn you Harry.

I am sitting on the fence till more of the criteria is defined. A major no to integrated circuits for signal processing but SMD might be a good idea.

Harry has raised an important observation in that most products designed by commitee Suck with a capital S. This is almost a universal constant. We need a lead engineer (or DIYer) here for it to work.

If a Pass Labs product was designed by commitee it would probably sound like a Pioneer.

Jam