[tangent]

I come down between these two rework points of view. Yes, blindly reheating all the joints can cause more problems than it solves, because you can actually make more cold joints this way. If you have some liquid flux, adding a bit of flux to each joint before you reheat it will remove this risk. Even so, this is the hardware version of shotgun debugging -- like waving a shotgun around and pulling the trigger at random, hoping to 'shoot the trouble' if you pull the trigger enough times.

I'm also not a fan of completely desoldering and resoldering everything just to find one bad joint. You can almost always find a cold joint just by looking at it, but this requires experience. If you don't yet have that experience, the reheating method is probably best. To get the experience to be able to identify cold joints, retest the amp between joint reheats so you can remember what the joint looked like before you reheated it. Eventually you will be able to identify cold joints just by looking at them, which means you won't leave cold joints for the testing phase to begin with -- you'll see them right after you make them, and fix them immediately.



Getting back to the problem at hand: finleyville, now that you have a working amp, remove your temporary modifications one by one, getting back to the way you had it before, testing each step of the way. I would first put C1 back in, then replace the pot, and finally swap out your small 'lytics with the big ELNAs you want to use. If bad assembly technique was your only problem, then you can just resolder all of the parts to the board and it will work this time, but you don't know that yet. You need to re-test at each step so that you know what really went wrong. You may have bad parts, or a cracked trace or something else entirely.

On your RCA jacks, are you using a metal case? If so, your chassis grounding might be incorrect.

 

[finleyville]

Tangent,

I did reheat all of my solder joints and have begun to systematically replace the removed components. I have replaced a panasonic pot and reconnected my off-board Cerafines. Either the battery or the wallwart does power it, but there is a slight hum to the wallwart. Probably just the AC power. I'll hook up the AD8620 and C1 next one by one. --- With the C1's touching each other and the ground buffer, will that give me problems? ---

The RCA inputs still do not work. My amp still isn't cased so a bad ground to the RCA's may not be the problem. (But I wouldn't be surprised) I did borrow a digital camera and took pictures but am unable to post them because they are too big.

Thanks for the help. Stay tuned for the continuing adventures of Menaceman and his META...

 

[tangent]

Since you're talking about replacing the op-amp next, does that mean you're turning the amp on with no op-amp in the socket? If so, then what you hear through the headphones doesn't mean anything, hum or otherwise. If you're getting hum with an op-amp in the socket, that power supply might not be quite so well regulated as you think. The META42 isn't the best amp when it comes to rejecting ugliness on the power lines.

The only for-sure way I've ever been able to test a power supply for noise is with an oscilloscope. You'd think you could get meaningful results from a VAC meter, but what I see on the scope is often very different from what the meter tells me.

The C1s touching each other shouldn't matter. I also don't see why it would matter if they touch 2001G, unless of course it's their leads touching the pins on the 2001G.

 

[finleyville]

I was test the amp with the OPA2134 installed instead of the AD8620. That way if I did a bonehead move, it would only fry the BB and not my nice AD.

 

[Gariver]

Jon Risch says: "Solder only has about 10-15% conductivity compared to copper, so a good solder joint is a bad connection anyway, a bad solder joint is a REALLY bad connection!"

Knowing this, I always make sure of a good mechanical connection before I apply solder. I mean, I try to "power squeeze" the wires with my small needle-nose pliers. Why? Well, I want the best possible conductivity at every joint. Solder alone will not give me the best possible connection. Thus, to get the maximum possible conductivity you must squeeeeeze!

To learn more about all this, read Howard H. Manko's 485 page book: "Solder and Soldering." Oops...4th Edition is now 576 pages. That book keeps growing with each new edition.

 

[finleyville]

Gariver,

Do you bend a components leads back on itself and squeeze them in even on premade boards?

Talking about good solder joints... When I use that RS silver solder all my joints appear dull. The only time that they are really shiny is in the liquid state which cools very fast. Have I done something wrong with this and my last three amps? Does anyone know what I am talking about?

 

[finleyville]

I don't know if this pic shows my soldering job well enough....

 

[tangent]

Silver oxidizes quickly. Hot silver oxidizes quicker. I don't recommend solders with high silver content to newbies. If you get a solder with a little bit of silver in it for the hell of it, that's fine, but most "silver solder" is like 4% or more, and often has other undesirable properties like a high melting point.

I use Kester 44 with 62/36/2 Sn/Pb/Ag blend. It gives shiny joints.

Besides being shiny, a good solder joint will also be slightly concave. Also, with boards like the META42 that have plated-through holes, the solder should wick up around the component leg and show through at the top of the board. If you're not getting good wicking, you're not heating the component leg enough, probably because you're melting the solder with the iron top and not with the heated component leg. Or, it's possible that your solder has poor flux or no flux.

 

[finleyville]

I guess that is why I have dull joints. My solder is 4% silver. Oh well.

 

[Gariver]

I don't like your description of your dull-looking, RS solder. Maybe that's the source of ALL your META42 problems. If I were you, I would buy a better solder. You can get any of the best available solders for under $16.00. That's the price of Wonder Ultra Clear Solder. In my case, I prefer Cardas Quad Eutectic Solder. You can get a 1/4 pound roll of this superb solder for $11.00 at PercyAudio.com. That's not too much money, right? I mean, you can spend more on a pair of opamps!

Here's a good analogy: If you own a 2002 Porsche 911, don't use Vokswagen beetle parts in your Porsche. It just doesn't work! The same goes with your META42: If you use expensive OPA627 or OPA637 opamps, don't use a cheap solder and expect great results.

Electric Desoldering Tool from Parts Express...
If you do a lot of DIY work, this desoldering iron is great. It combines the heating and solder removal in just one tool. It costs $26.95. Take a look...

http://www.partsexpress.com/pe/showd...ID=10654&DID=7

Proto Boards...
I have been using prototype boards for my amps. Reasons? I've got a bunch of Riken 1/2 watt resistors. These Rikens are longer and wider than Vishay Dale RN55 or RN60. Thus, I have to use proto boards to fit the Rikens in. The other reason why I use proto boards is that I have a good supply of 22 gauge silver wire and Teflon tubing (from Michael Percy).

When I can, I do bend back and twist the wires before I apply Cardas Quad Eutectic Solder. It's almost a knot! I also like to give these twisted wires a power squeeze with small needle-nose pliers. In this process, I am going way beyond avoiding cold solder joints. My aim is to get the best possible conductivity in each and every joint. With the squeeze, I strive towards good metal-to-metal contact. I believe this is the only way to get the best possible sound out of my opamps, buffers and amp circuit. So far, so good! It works very, very well!

Another thing: I'm not building a META42 with a PCB. Instead, I am building a Sijosae style META42 on a small proto board (RS 276-149). Once I am finished, I will be able to quickly change resistors if the need arises. I mean, it's just like a breadboard. Take a look...

http://headwize.powerpill.org/ubb/sh...20020817203644

On RS solder...
Jon Risch simply hates Radio Shack solder (all of them). He says they don't sound good to him. Hey, I took his advice. It's good advice. Elsewhere on the net, I found out that RS sources its solder from different manufacturers. Thus, you might get lucky and get great solder. Or you might get unlucky and get lousy solder. I simply don't like this type of lotto. I want consistent quality! So I use Cardas Quad Eutectic Solder. It's an excellent solder! You will get shiny solder joints that work perfectly every time!

BTW, Jennifer Crock of Jena Labs did listening tests using solder as wires. In these tests, Cardas Quad Eutectic Solder came out on top. That's one of the reasons I picked Cardas as my solder. Another reason I like Cardas is that it solders clean. I mean, there's hardly anything to clean up after you solder. Nice!

Why should you use eutectic solder? Eutectic solder goes from liquid to solid instantly, without a plastic state. This is very important: Most cold solder joints are caused by tiny movements of the hand during the plastic state. So you should avoid plastic state movements by all means possible. The best way, of course, is to use eutectic solders: You can use Sn63/Pb37 or Sn62/Pb36/Ag2 (2% silver). However, stay away from regular Sn60/Pb40. This solder has 15 degrees of plastic state.

Now, if you want to "overdo" your soldering, use heat-sink clips to dissipate heat and minimize movement. Or you can tape your joints before you solder to minimize unwanted movements. This type of perfectionism will help you build amps that work exceptionally well.

Here's some brands of eutectic solders...
--Cardas Quad Eutectic Solder (Tin, Lead, Silver, Copper).
--Wonder Ultra-Clear Solder (contains Au so its expensive).
--Kester "44" (or No-Clean) in an eutectic mix: Sn63/Pb37 or Sn62/Pb36/Ag2 (a silver solder).

Michael Percy sells Cardas and Wonder solders....
http://www.percyaudio.com

OK, finleyville! Good luck to you! I hope this info helps you solve your META42 problems.