[da_burl]

Ok, everyone. Since I am not going to use the prefab board, I will not have a META42 (as you can see, I've been doing a lot of searches).
My Jung Multi-Loop Buffered Output Headphone Amp , is going to be used exclusively at home, even more specifically, in my studio/office, with a high quality PC Sound Card (Delta Dio 24/96 analog outs), which might become a Delta Revo someday.
I own Alessandro MS-2's, and Grado SR80's, and some off the wall Senn HD-595's, that I got carried away one day on Ubid, lets' just say they're not bad, but I'm a Grado man I guess.
I will use my version of a High Quality +/- DC Supply, its more or less like the SDS PS, although someday, I might tweak it to be more like the Gilmore PS. Its sitting at +/- 14.6V right now, but if I need to go up or down a little, its just a matter of picking different resistors on the 317/337 control pins. I use either 4700uF or 8200 uF as the preregulator caps, and I also have a choke just before the bridge (adapted from a very old post by good ole' Apheared).
I have at least two prototypes of these made up already, I have been experimenting with Hansens, CHA47's, and even Szerkes Class A's (AC and DC versions).
I do not think the DC PS is extremely critical (to a point) on this design, but I will emulate the filtering and reservoir caps, and obviously skip the voltage splitter and buffer, i.e. 1000uF Electrolytics and some kind of Polyester box caps (I don't have 6.8uF, but I have some nice 3.3's, and some Wima 1uF's, which I might play with paralleling.
I've read about alternatives to the EL200x buffers, and just to be a guinea pig, and because I am already ordering the AD843's from Newark, I threw in 4 of the HA3-5002's, I will try them single and stacked, single at first. Since I am prototyping, I can probably pretty easily "stack" them on the protoboard, they are the DIP package.
Morsel, I will be glad to let you know how these sound, there doesn't seem to be anybody definitively saying "Yes, I have HA5002's in my META42, and I love them"!!
So anyway, I do have a few questions:

What gain would I shoot for for Grados? The set of resistors I have dug up right now would be:

R3 - 332
R4 - 4.75K
R5 - 2.21K
R6 - 220K

These are all Vishay/Dale 1% from Mouser, from my collection that has resulted from my previous experiments.
I show that to be a gain of a little over 13 on the calculator, and that, along with a +/- 15V PS, I think I am going to have plenty of voltage gain. Is this too much for 32ohm Grado's? If so, what should I change?
I have already ordered the AD843's, but what I am shooting for here is detailed, refined sound. I like to hear the breath of singers, space around musicians, soundstage, all that stuff. If I just want to ROCK, the CHA47 is already pretty good for that Did I make the right choice of Opamps? Is the 8620 more "detailed" ?. If so I could try some of those too eventually. I have ruled out most of the others, the 8610's are more for battery people, and the OPA627 doesn't seem to have the detail I am looking for. I only have personal experience with the 2132/2134's in the Cmoy/Hansen and CHA47's.
I am looking forward to any input and experiences from Grado/META42 people.
I will follow up on my first listening tests when I get it finished. I realize that the fun of this design is in the tweaking, these resistor values are just to get started, I don't seem to have the exact values in the standard schematic, and have seen lots of people using all kinds of different values in theirs, just not much on tweaking it for lowZ cans.

 

[tangent]

Quote:

R3 - 332 R4 - 4.75K

Way too high for Grados. Full line level (i.e. from your sound card) is going to be 1-2V RMS, but your Grados only want like 0.5V RMS. So in fact, what you want is more like a gain of 1 or 2. Coupled with the attenuation of the volume control, you will get a reasonable volume level.

I only use gains of around 5 for Senns, and only crest 10 when I'm also compensating for the ~9 dB drop of a modified Linkwitz crossfeed in addition to running from portables and driving high-impedance headphones all a the same time. (whew!)

Quote:

Did I make the right choice of Opamps?

According to me and all other right-thinking people, yes. According to others, no.

You'll have to make up your own mind, I'm afraid.

 

[da_burl]

Thanks Tangent! I will keep plugging in values to the calculator! Great site you have, I might actually learn what I'm doing. I will probably have to order some resistors anyway eventually.

A couple more things I left out:

I wasn't planning on using C1 or the POT, just control the volume with the PC Volume Control, can save a lot of money and headaches that way. That's how I am using the 47. But now that you mention it, I may go ahead and put in a pot, I have some of the panasonics, those connections sure are flimsy if you don't board mount it!! But with a pot, it would be more universal if I ever wanted to take it somewhere, like to one of the Dallas Headphone meets, or to use with a regular CD unit. I measured the DC offset and my sound card is cool, no DC offset.

 

[Voodoochile]

I'm anxious to see how you like those 5002s... I just picked up several of them from Newark myself.

 

[tangent]

If you're going to use a cheap pot, another thing you have to consider is that the further down you turn it, the more likely it is that you'll get audible mistracking. (I.e. one channel quieter than the other given an equal input signal.) So, you need to tune the gain so that the pot stays mostly at 50% or greater of its rotation for full volume. That way, when you have to turn it down, the mistracking comes in only when the music's pretty quiet already.

As for a better pot for your use, ask Michael Percy to get some Bourns 51s for you. Morsel and I were evaluating these for small META42s, and had Percy lined up to distribute them if we liked them. Unfortunately, they weren't small enough, the electrical performance isn't better than the Panasonics, and they're a lot more expensive than the Panasonics. However, they're built better and they use standard 100 mil pin spacing, so you can solder them directly into your perfboard. These things may make up for the 3-4x price difference between them and the Panasonics.

Anyway, we rejected them, but Percy was talking about carrying them anyway for those who needed a smaller pot than the large Noble he carries (the same size as the ALPS Blue). You'll need to email Percy about it, as they aren't in the current catalog. I hope this just means that he hasn't yet decided one way or the other. If that's so, your request could tip the scales. It'd be good to have an alternative available that's more friendly to perfboard users.

 

[morsel]

Da Burl, good luck with your new project! The AD843s should be fine. While leaving out C1 is better for sound quality, leaving out the pot is not a good idea, because you want the hottest signal you can get from your sound card to keep it above the noise level of the computer. Doing the volume reduction at the amp is the better plan. Let us know what you think of the 5002s, and be sure to try some CDs, not just compressed MP3s.

 

[da_burl]

Guess I should have read this first!!

AD843 vs. AD8620 Review

This sounds like what I am looking for.

I do realize my source might end up being the limiting factor. I don't have any hi-end gear like Kelly or anything to compare, just a midrange Denon DCM380, and an old Parasound DAC. After a few comparisons, I just listen to the PC Sound card in the office, which is also my chill-out and listen to music, (and get away from the other half, just for a little while room.
As far as the music goes, I do have lots of MP3's, some of the older downloaded ones are pretty poor. As this is another of my hobbies, I have lots of self ripped wav's, lots of downloaded ape (lossless) files, and lots of self ripped and compressed mp3's and mpc's (aka mp+). Just recently I started giving mp3 another chance, after leaving it for a while for ape and mpc. But with the new lame, and the --alt-preset insane, I am having a very hard time distinguishing much difference between these and the original. Maybe my new amp will "reveal" the difference, and I will have to re-compress everything again!
Don't know if anybody else visits avsforum (the htpc forum), but there are a LOT of things happening in PC sound right now, for example the much anticipated M-Audio Revolution, which "soon" will have the ability to play DVD-Audio and (I believe) DTS Audio CD's on the PC, and a 24/192 DAC to boot. Sorry, getting off-topic. Can't wait to start building when the parts come in!

 

[erix]

Dig this...

$1.95 for a detented Alps dual 100K Audio w/50% tap.

They also have ones with (2) 50K sections and (1) 2K section.

Tilt it, baby.

ok,
erix

 

[da_burl]

I'm back. After further review, I think I kind of understand the resistor values, correct me if I'm wrong:

You want to set the inner loop gain pretty high like 100, to limit the bandwidth to approximately 100Khz (just a rule of thumb)
The overall circuit gain is set by the outer loop resistors

The AD843 had unity gain bandwidth of 34Mhz, and a full power gain bandwidth of 3.9 Mhz, did not see a value for GBP.
So which value is equal or similar to Gain Bandwidth Product?
Let's just guess its the UGB, at a gain of 100, my bandwidth would still be 340 Khz, should I go even higher on the inner loop gain? If its the FPB, I guess I would have to go lower , even though 39Khz is still almost twice the accepted audio bandwidth of 20-20Khz.
Just guessing I will be fine with 100

You want the inner loop "shunt" resistor (for lack of a better word) to be 5-10 times higher than the outer loop shunt resistor

So I came up with this

R3 - 1K
R4 - 1K
R5 - 10K
R6 - 1M

or
R3 - 470
R4 - 470
R5 - 2.21K
R6 - 220K

I like the 1st set, because I have all of them except the 1M, and it just seems more "standard" to me.

I am shooting on a gain of 2, based on Tangent's recommendations, for my Grado's. I would have to make R4 even smaller (470 or 220) to get closer to Gain of 1. You can't really get a gain of exactly 1, without using just wires in both places, I guess (foggy memories of opamps always at least having unity gain??).

Well, that's about it for now. I am definitely going to put the resistors on a DIP for easy experimenting.

Haven't had this much fun since Devry in '80

Thanks a lot

da_burl

 

[da_burl]

Erix:

By the way, I have some Allen-Bradley pots very similar to those you posted the link for, 50K Audio taper, just not Alps. There is a surplus place here that always has some interesting stuff, just usually not enough to complete an entire project So I can use those, or the Panasonics, or I also have some "full sized" Xicon/Alpha 50K Audio that caught my eye at Mouser, p/n 313-2420-50K. Guess I'll do a little testing and see which is the best sounding and feeling!! I am not going to spend $50 on an esoteric pot at this particular juncture, but someday, maybe
Then maybe a stepped attenuator!

 

[morsel]

Kurt's recent distortion measurements indicate that inner loop gain should be higher than 100, at least 200 and perhaps 400. See the end of this thread:

http://www4.head-fi.org/forums/showt...threadid=23610

 

[CaptBubba]

I run my amp (similar idea btw... thread here) at a gain of 1.47, and have plenty of power for uncomfortably loud volumes using V6s and HD580s.

As for the opamps, my initial impressions lead me to feel like you'll be perfectly happy with the 843s.

oh, btw....post pics!

 

[da_burl]

Morsel:

Thanks for that, I saw that thread, but didn't realize it had grown to include the factor of the gain. Good to know, because the HA-5002 wasn't lookin' too good at first. Maybe I will just put multiple resistors, and have selectable gain!
I'm going to need THIS! before this is all over

Capt. Bubba:

I'll do some pictures, if my soldering job comes out at least as presentable as yours, which so far they haven't . Or at least I will have to only show the top side.
What are you doing with those DIP Switches? That might be cool for selecting gain, but I couldn't leave them there forever, not very "audiophile", unless they're OFC Dip switches from Welbourne Labs or Michael Percy

 

[CaptBubba]

Quote:

Originally posted by da_burl at least I will have to only show the top side.

This is the dirty secret of the DIY trade: if it doesn't look good, hide it!

Quote:

What are you doing with those DIP Switches? That might be cool for selecting gain, but I couldn't leave them there forever, not very "audiophile", unless they're OFC Dip switches from Welbourne Labs or Michael Percy [/B]

They are a cheap stepped volume control. No, not audiophile like a goildpoint or something, but definatly better than most cheap pots you can find. Right now I'm actually using it in conjunction with a pot so I can "fine tune" the volume to my likeing.

 

[da_burl]

You know, I've been thinking about something like that too, with the soundcard volume control, all I really need is four or maybe eight levels, then do the fine tuning on the soundcard.