[tangent]

Just to make sure everyone's clear on this: this trick only works if you use a multiloop configuration, which means you must use the buffers on the outputs, pointless as they may be in the face of the 100mA current drive of the LM6172. Without the multiloop, you get the full 80-100x gain out, which is way too much for headphones. The multiloop configuration lets you kind of cheat -- it bumps the op-amp's gain up for added stability and to get rid of some excessive bandwidth, and yet still lets you keep a reasonable 2-10x total system gain.

I suspect that if you configured the META42 as an über CMoy with a reasonable gain, you'd need to use the feedback path cap, and all the phase shift nastiness that entails.

And you're more than welcome -- I was curious about this, too.

 

[ppl]

Tangent;
thanks for your input regarding using opamps that are not normaly stable at low gains like the Lm Chip you tried. yes multi-Loop dose make Chips like this and Others stable at useable gains. I brought this up along time ago about using the OPA-637 that is not the most stable opamp in the world, however some head-Fi users Did not beleive me on this. So once again thanks for somewhat varrifying the ability of the Multi-Loop configuation to make these normaly unstable chips useable. true These Chips are not the best way to go since good stable opamps are available that are stable have low DC offset, might i sugest My new favorite opamp the AD-8610, Wounderfull smooth Clean and Articulate. it also is one of only 3 Opamps that handel Aplause with out adding strange artifacts to it.

Slindeman:
Regarding using high Input Bias current Opamps like the LM one mentioned The resistors in the feedback loop and the Non inverting input need to be Low value this would require a pot of about 2-3K with resistors scaled to match. this may result in an Input impedance too low for some sources like portable CD players line out that require 10k or better. just matching the resistors is not enought because of the Input bias current offset. this spec is often overlooked and becomes importent with Bipolar input stages. since you can match resistors all day and this Current imbalance allready is working aginst you. so i still maintain that if you wish to use this LM series opamp that an FET source follower between the Pot and the opamp is required.

 

[Audio-Me]

How about some sound characteristics/comparisons of:
AD825
AD8610
OPA637

EL2001
BUF634

 

[tangent]

If you were asking me A&M, I can't help you -- I don't have any of those opamps. As for comparing the BUF634 to the EL2001, I know ppl has some experience in that. You can't plug the BUF634 into the META42 board, if that's what you're asking -- the pinouts are totally incompatible. I guess I could air-wire it....

News: the second prototype is here and the physical layout problems are fixed. There's now some talk amongst ppl, Morsel and I about changing one electrical connection, but it won't require another prototype. It will just require deciding whether to do it, and that may involve hacking up one of these boards to try it before we make more boards. But gentlemen and ladies, the pro boards are all but ordered. I will announce details in the Marketplace forum when the boards are ready.

Here's a pic of the second prototype, populated somewhat differently than the first one:

A few things to notice:

1. I'm now using some large (10uF/16V) Wima film caps for my fast reservoirs instead of the 1uF/63V ones that are easy to find. These Wimas are what we had in mind when we did the layout -- I would have used them in the first prototype but they just came in today. For higher-voltage applications, there's a 6.8uF/50V part in the same series that's the same size.

2. I've populated this one with an ALPS Blue pot. Notice that it's in the same position as the Panasonic pot -- the two parts' footprints are superimposed so they collectively take up only as much board space as the ALPS pot.

3. The crowbar diode is a big 10 amp TO-220 package unit this time. (10TQ030)

4. Since the electrolytic capacitor placement issues have been solved in this iteration, I can use 10mm diameter caps as was intended. In this amp, I'm using 2x1000uF/16V on each side.

Now that I've got a better amp, I can finally get to my listening tests. I'll let you all know how it's going.

 

[ppl]

Tangent:
Cool board, i now see where the extra space that was showen on your first photo is for. great job. About those wimia caps I was unaware that thay made sutch a large value film cap, i will have to check into this since the larger the Bypass caps the deeper into the Audio range bypassing will happen and this is a good thing. I must say i am impressed by your result of making an Amp For all people. i would never have the energy to do that, but i make My Amps for optimum performance using what ever components mate together to do that, i gess i have a My way or the highway aproach to Circuit design or so i have been told.

Audio&Me:
The AD-8610 and the AD-825 sound alot better to me as well as some Other critical listeners that have also listened to these OpAmps. If you like the OPA-627 you will love the AD-8610 The Mid's on the 8610 are natural sounding with the correct Amplitude relitive to the bass and treble, where as the OPA-627 has a foward Midrange that dominates the bass and treble giving the sensation of reduces response in the Extreme ends of the Audio range. the clarity that the 627 is noted for is evean better in the 8610 and last but not least the Current drain is half of what the OPA-627 is. The AD-825 is extreamly bright compared to eather of these OPA's however the Articulation and microdynamics are superb on the 825. However the 825 should be used only with ultra cleen source components and recordings because sutch things as siblence and tape his are brought frount and center in comparrison to the 8610 and 627. The 825 used to be my faviorite opAmp for portable Amps untill the 8610 came along. With the 8610 i get the best of Both smooth and clear with Tolerence for less than ideal recordings, any flaws are heard however not Thrown in your face like the 825. The EL-2001 sounds faster and alot more articulate than the BUF-634. the BUF-634 sounds slow and slightly restricted in dynamics compared to the 2001. Also the 634 has alot higher DC offset voltage than the 2001 by about 3 - 4 times typicaly 35-60 Mv operated open-loop. this makes it's use with opamps that have high DC offset Voltage like the AD-825, Difficult to use with a Multi-loop configuation. The 634 is however a warm and musical sounding part and it has a slightly resessed midrange that mates great with the foward mid's of the OPA-627. if the Elantec's become unavailable i would then use the 634 with out reservation but use some of the trics i have used in the past to coax The best possible sound out of the 634. The AD-744 when the Comp pin is used to drive the 634's input rather than the normal output of the 744 IMHO gives the best sound out of the 634. the AD-8610 also sound real smooth with this Buffer but i feal some of the articulation the 8610 is capable of is somewhat lost with the 634, However i havent optimized this pair for max sonic performance yet but may do so if needed. lots of subtile details are of great importence with any circuit and combonation of components. Hope this help's.

 

[Audio-Me]

ppl - thank you very much for the compare and contrast. Do you know if the AD8610 can be used with the META42 board? I can't stand bright and forward, especially sibilance. Smooth and clear yet forgiving is just want I want.

 

[Budgie]

All I can say is..Very Cool ! I want this amp.

 

[tangent]

Yes, A&M, you can use AD8610s with the META42. These are single-channel SO-8 surface-mount chips, so you need to use an adapter from Brown Dog that takes two of these surface mount opamps and converts them to a dual DIP-8 pinout. Morsel convinced them to build this adapter for this very purpose -- more accurately, she designed it and then convinced Brown Dog to make them.

Because one of the op-amp chips goes on the bottom of the board and the other on top, the whole assembly takes little more space than a DIP-8 chip. That would matter in the META42 board if you chose to use the high-speed bypass caps that are placed right alongside the opamp chip position. I haven't had to use those yet, but it's nice to know you could if you had to.

ppl, these caps are the MKS 2 series. They apparently obey some kind of footprint standard called "PCM 5", so you can find caps of this type from other manufacturers as well. I've found some 3.3uF ones from Vishay Roderstein (via Mouser) and they will be coming out with a 4.7uF one soon, according to the datasheet.

 

[Audio-Me]

hehe, coolness, eric said he already took the initiative to order those for me.

 

[tangent]

More news on the LM6172 front:

I was able to figure out how to get rid of its DC offset in a multiloop configuration without a feedback cap, and was able to keep it reasonably stable. The board change we're about to do probably invalidates my method, so I won't tell you what I did, but I will be able to figure it out again in the context of the new board, I'm sure.

So, LM6172 fans, yes, you can use this chip without a band-aid. ppl's happy, I'm happy, slindeman's happy, and NatSemi is happy.

 

[ppl]

Tangent:
It is good that your board has now so mutch interest, i think you will sell all of them and the Buffers you have got so as to get back your investment. i would suspect this will be more popular than the Hanson board that headwizers/Head-fiers have used in abundence.

I am also quite sure that it will sound alot better than the hanson board. Real buffers rather than opamp followers are The best way to drive reactive loads. Hopefully the El-2001 will be available for some time since this buffer gives the best bang for the Buck in terms of cost and Performance per Ma of Current drain and i cant' see any crossover distortion at low output levels. however if this dose become unavailable the BUF-634 is and would be the best replacement for the 2001's if the 2001's become unavailable. I have used this series of buffers from elantec since thay were introduced in 1989. and thay have never failed to please.

Now while this Board is great at being all things to all people there are certin items missing i think are Verry importent like a current source on the opamp output. this makes alot of difference and really cleans up the mids and high's on the AD-823 as well as most all Opamps with some notable exceptions like the AD-845 whitch is class A up to about 3 Ma anyway however this Opam draws 10 Ma so not battery frendly. But I gess this is why i must do the PC board artwork for my amps myself. and since i do this by hand it is somthing I really hate next is the finishing of the case. i am used to just supplying a schematic and others do the Actual assembly and mechanical design. Oh well>>>>>

Anyway great job and i comend you for providing this for others that need somthing better than the hanson board. now crossfeed Not my cup of tea eather however some like it. Good luck with this

 

[aos]

Where can one get those small yet large value WIMA capacitors?

Nice project btw! Especially if boards will be available, it has the potential to become the next step/level after CMOY and we will see some more comparisons between higher end DIY amps and commerical amps, something I'd really like to hear about.

 

[tangent]

I got mine at CapacitorOnline. The real name of the company is Bertech-Kelex, so don't be surprised when you call and they don't answer "Capacitor Online, how can I help you?"

They're 'spensive little buggers, though: $6.44 apiece for the 10uF/16V ones, and you have to buy at least 20. The 6.8/50V ones are $4.65 each, and the minimum is also 20 pieces.

There's no online ordering: you have to call them up, get a quote, and then wait a day or so for them to get the things in because they don't stock them.

Despite my carping, that deal is better than the few alternatives I've found so far. The best thing I've found so far from the DigiKeys of the world are some 3.3uF Vishay Rodersteins from Mouser that are $5.53 apiece. On the plus side, you can buy those one at a time if you want.

 

[morsel]

PPL, perhaps after this project is done, and I have found time to research a source for high quality 9mm dual log taper pots, build my amps, my pocket signal generator, finish my test CD, and all the other zillion things going on, we will design another board that is all out, no compromises for small cases or alternate components, and we will have a current source for the opamp.

 

[tangent]

I was sitting there regretting not adding just one more resistor for driving the op-amps into class A ("poor man's current source") and then it hit me: you don't need it.

If you don't mind building the current source on a small chunk of protoboard, the PCB is set up to allow you to tie it in fairly easily: just use the processing stage pads (the ones that are there mainly for adding crossover circuits) and the extra power source pads we've provided. The only reason I can think of that this wouldn't work is if the long hookup wires required would be a problem.

Whaddaya think, is it practical, ppl?

Here's an image of the PCB: