[hell0.com]

As many noobs here, I've decided to try to make a CMoy as my first DIY audio project, but I have a few questions before I start out. I've ordered the standard parts that tangent specified from digikey, but once I polish up my soldering skills, I would hopefully like to make a few bucks on the side from selling a few pocket amps. For this, I have a few questions about higher quality components and solutions.

The Virtual Ground:
I've read tangent's article on virtual grounds and the standard resistor divider solution, but would like to eventually create a VGnd using the TLE2426 Rail VGnd solution. I whipped up a schematic of how I would like it to function. Basically, all I've done to tangent's design is add a LED, resistor, and switch, but would like any input, such as on the placement of the capacitor. I've read many posts concerning this, but it seems the jury's still out on where the capacitor should be placed (ie before or after the TLE). Also, is 470uF too high, too low, or ok? Tangent's original design used a 220uF.

Also, is it correct to assume that the Amp circuit is still the same, despite the change to the power supply circuit (from resistor divider to TLE)? I plan on using OPA2277PA or OPA2134PA. Will this supply be sufficient for those opamps? Lastly, I plan on using the TO-92 version of TLE2426 for their compactness. Is there any reason I shouldn't?

Power Caps:
So, i've read about the importance of good caps until my eyes've bled, but still have yet to find a good cap to put into my cmoy. From what I've gathered 470 uF is a good value for the power caps, and so, after looking through digikey's catalog, have decided on the "P5552-ND". These are electrolytic, 470 uF, +-20% tolerance, 25V, and radial. Any input? Are these satisfactory? Any better recommendations? Is a 1000uF too high? Should I stick with 470uF? Digikey's catalog makes my head spin.

Input Caps:
Now, this is where I've really been having trouble finding quality large caps. I've gathered that polyproplene caps are the best (for the price point), but can't find any from digikey that have a 1uF capacitance, except these (digikey prod #):
P10761-ND - CAP 1.0UF 450V FILM BOX TYPE 10%
P12207-ND - CAP METAL POLYPRO 1.0UF 400V 3%
Both of these are expensive and have a higher volt rating than needed. Also, the former has a high tolerance, which I'm not sure if that's good or bad. I didn't see anything about "tolerance" in the tangent's cmoy cap guide. I've been drooling over the excellent work of "biosciencegeek" at ebay, and noticed the compactness of his Caps.

Any idea what those blue "box" caps are? He doesn't specifiy, in his auction, any of the specifics of the input caps.

Well, that's about it for now. Thanks for reading my seething mass of questions. Any input would be very much appreciated!

EDIT: Whoop! I guess that's not it

. I forgot one thing. Would it be possible to put in parrallel 2 0.47uF polyproplene caps to get nearly .94uF for the input caps? How would this affect the audio?

 

[ATAT]

TLE - put capacitors after the TLE. The TLE sources / sinks a pretty pathetic amount of current, so the capacitors are pretty much necessary unless you want to put a buffer after that TLE.

Basically, the tradeoff for TLE capacitor placeement is as follows.

Before TLE - you can get more capacitance since you have one capacitor for +/-

After TLE - you get less rail capacitance since you have 2 caps / rail (+ -> vgnd and vgnd -> -) but it helps the current drive capability of the TLE.

470 is just fine, be sure to put a fair bit of capacitance post TLE though.

You can use whatever version of the TLE you'd like, and it'll work with any opamp.

IMHO good power caps are pointless in a CMoy. id probably still use Nichicon UPWs though. they're fit for the job, any more "high quality" and you're probably wasting your money (or panasonics, if you're using digikey)

75-MKT1826522055 would be hte cap you're looking for.

close Tolerance isn't really needed, since the caps dont really "do much" they just block DC.. their -3 db point is high enough that a slight shift wont matter..

Yes you can parallel.. but why? just use a 1uF film cap.

Also, you could just skip the input capacitor..

 

[Andrea]

Good electrolytics are never, ever a waste of money, on the contrary, they're compulsory in whichever "hi-fi" application in order for the op-amp to perform up to its potential.

A cmoy with a good & well suited op-amp & good power supply can do really great things, particularly with high impedance headphones.


As always, care for details is a key to audio goodness.

 

[JahJahBinks]

Blue boxes are polypropylene capacitors.

 

[hell0.com]

Quote:

Originally Posted by ATAT 470 is just fine, be sure to put a fair bit of capacitance post TLE though.

So, putting them both before and after the TLE, as follows, is best?

Furthermore, if I want to choose just one (either after or before the TLE), the best choice would be after the TLE. Did i understand you correctly?
Quote:

Originally Posted by ATAT 75-MKT1826522055 would be hte cap you're looking for.

How will 2.2uF affect the audio (I assume this is for the input caps)? The highest value i've heard of someone using for the input caps is 1uF. Also, I'd prefer to stick to prolyproplene, if possible.
Quote:

Originally Posted by ATAT Yes you can parallel.. but why? just use a 1uF film cap.

All I could find for the input caps were .47uF panasonic polyproplene caps at digikey (P3474-ND), and so I thought maybe putting them in parrallel to reach the desired capacitance of 1 uF (.94uF) would be a solution. Upon further consideration, though, I think I'll just stick with a single .47uF per channel, as I want to make it as compact as possible, unless someone can recommend a 1uF polyproplene cap.
Quote:

Originally Posted by ATAT Also, you could just skip the input capacitor...

Hmmmm... I've read on tangent's guide that this could possibly lead to a damaged headphone driver.

 

[cetoole]

Quote:

Originally Posted by hell0.com So, putting them both before and after the TLE, as follows, is best? Furthermore, if I want to choose just one (either after or before the TLE), the best choice would be after the TLE. Did i understand you correctly?

With just a TLE providing the split, if you only want caps rail-rail (before TLE), or rail-gnd (after TLE), stick with the caps rail-gnd, as they will assist the TLE in sinking current.
Quote:

How will 2.2uF affect the audio (I assume this is for the input caps)? The highest value i've heard of someone using for the input caps is 1uF. Also, I'd prefer to stick to prolyproplene, if possible.

Larger values will lower the -3db point of the high pass filter, so bass will increase with larger values (up to a certain point).
Quote:

All I could find for the input caps were .47uF panasonic polyproplene caps at digikey (P3474-ND), and so I thought maybe putting them in parrallel to reach the desired capacitance of 1 uF (.94uF) would be a solution. Upon further consideration, though, I think I'll just stick with a single .47uF per channel, as I want to make it as compact as possible, unless someone can recommend a 1uF polyproplene cap.

Hopefully this link works, just select the values you desire in the filter list. Plenty of polypropylene caps >1uf.
http://www.mouser.com/search/Refine....user_Wildcards
Quote:

Hmmmm... I've read on tangent's guide that this could possibly lead to a damaged headphone driver.

I would stick with input caps if you are selling the amps, because there is no telling what kind of offset the sources used with the amp will have.

 

[hell0.com]

Quote:

Originally Posted by cetoole Hopefully this link works, just select the values you desire in the filter list. Plenty of polypropylene caps >1uf. http://www.mouser.com/search/Refine....user_Wildcards

Ahh. Very useful link. I may have to start ordering from mouser rather than digikey. Unfortunately, all of the 1uF polyproplene capacitors I could find were huge (size wise). I think I may just break down and buy a metalized polyester cap for the input cap. Would I regret it? I've been eyeing the MKT1817510064. At $.77 and 5.5x11.5x7.5mm, the size and price are certainly right! Would I be making a mistake buying these for input caps? Quote:

Originally Posted by JahJahBinks Blue boxes are polypropylene capacitors.

Polyproylene caps that small? Even .47uF polyproplene caps don't get that small, do they? Of course, "biosciencegeek" only calls them "high value box capacitors" so maybe that's just advertising talk for "bigger than tangent's standard input caps". As you can see, I'm no expert on caps, but from browsing several electronics catalogs, they look closer to polyester film than polyproplene.

 

[Dreamslacker]

The blue film caps are most probably Metalized Polyester caps. You can find Wima and ERO 3.3uF/ 50v caps in that size. That colour looks like Evox Rifa caps but I might be wrong.

 

[peterpan188]

Several things you should consider, first a larger value cap means larger physical size, so if you are going to put you CMoy into a mint tin, you have to watch out for that. My first Cmoy used big stuff, and I end up with only 1 9V battery in the tin. IMHO, two 9V batteries helps more than a bigger cap.

You should stick with Panasonic ECQ-P for input cap. 0.47uF is big enough, giving you a cut-off frequency at 3.386Hz with a 100K resistor. IMHO, low enough for a CMoy.

Larger power cap does give you a more impactful bass, since it stores more energy. Again, 470uF is a good value, Nichicon UPWs are great.

Regard,
Peter

 

[hell0.com]

Quote:

Originally Posted by peterpan188 You should stick with Panasonic ECQ-P for input cap. 0.47uF is big enough, giving you a cut-off frequency at 3.386Hz with a 100K resistor. IMHO, low enough for a CMoy.

So, the 0.47uF panasonics will fit into a tin using tangent's protoboard layout? They're much bigger than the metalized polyesters I was looking at. I'm trying to decide if the smaller size would be worth the (slight?) loss in quality.
Quote:

Again, 470uF is a good value, Nichicon UPWs are great.

So, something like the UPW1E471MPD6?

 

[peterpan188]

Quote:

Originally Posted by hell0.com So, the 0.47uF panasonics will fit into a tin using tangent's protoboard layout? They're much bigger than the metalized polyesters I was looking at. I'm trying to decide if the smaller size would be worth the (slight?) loss in quality.

Well, the layout that Tangent has can be expanded, as long as the input is connected to the lead of the input cap. If anything I would scarfice value over type, so Panasonic is my choice.
Quote:

So, something like the UPW1E471MPD6?

That looks good.

Regard,
Peter

 

[hell0.com]

So, while looking at the M3 tutorial I came upon some Vishay/BC MKP416 (digikey #: BC2076-ND) metalized polyprolene caps. Judging from the picture it looks like they might even be the one biosciencegeek was using in his amp, albeit maybe a smaller version. That's a pic of the 1uF version (10x16.5x17.5mm). Anyway, anyone have any experience with those? I'd assume the M3 wouldn't recommend caps with bad sonic properties, but, on the other hand, I'm not exactly sure what the M3 uses those for. I was trying to look at the schematic of the M3 and couldn't tell if they were used in the signal path.

On a completly different note, mouser is out of the OPA2277 and TLE2426

EDIT: Oops, looks like the OPA2227 and OPA2277 are two different things.

Sorry for the false alarm.

 

[peterpan188]

Well, apart from sonic advantage, Pany ECQ-P is also pretty flexiable in terms of mounting them on the board. You can easily bend its leads if you think they don't fit. Also, box caps takes more space, since it's squareish and thicker than some ECQ-P, you might want to consider that when you picking caps out.

I usually get all my ICs from Digikey. They got the fastest shipping, and best packaging, IMHO

I mean, really, just go with Tangent's guide, and build one first. Change some values if you want, but don't bother going with different types. Then you will really see what's happening. There are so many caps, resistor, ICs out there, and you will just get stuck with a CMoy if you keep reading without taking action.

My $0.02,
Peter

 

[threepointone]

using a .47uF ecqp in an altoids tin will NOT be fun. I've used .33uF ones, and it's a pretty tight fit. The .47uFs are too tall, so you'll have to do some pretty crazy lead bending. If you're using socketed resistors like I did, this will be pretty much impossible. A better idea might be to increase the value of r2--at 150k w/ a .33uF cap, it forms essentially the same high pass filter as a 100k resistor w/ a .47uF cap, and it doesn't add much space. The only tradeoff is slighly more noise, but you probably can't notice it. To use smaller caps, you could maybe try 200k or more for r2.

tangent's noise calculator: http://www.tangentsoft.net/audio/calc.html

 

[hell0.com]

Quote:

Originally Posted by peterpan188 I mean, really, just go with Tangent's guide, and build one first.

Aye, very true. I actually do have the parts for a basic CMoy a la tangent. They just arrived today from digikey, and I've built the circuit on a ratshack plugboard (universal breadboard) to play around with a bit before I finalize it. Unfortunately, I can't actually test it out until I get some aligator clips tomorrow. <rant>Also, fedex lost my package from radioshack with the protoboard...</rant>
Quote:

A better idea might be to increase the value of r2--at 150k w/ a .33uF cap, it forms essentially the same high pass filter as a 100k resistor w/ a .47uF cap, and it doesn't add much space. The only tradeoff is slighly more noise, but you probably can't notice it. To use smaller caps, you could maybe try 200k or more for r2.

Ahhh very interesting. I'll definitely give this a try on my plugboard.