[HD-5000]

*

 

[morsel]

No.

 

[andrzejpw]

no (listened to both), but why not use the 8610, I mean, just in case. . .

 

[HD-5000]

good enough for me

 

[tangent]

Quote:

why not use the 8610, I mean, just in case. . .

I think the difference between these chips comes down to usability, since two -10s cost the same as one -20 both for the chips and the appropriate Brown Dog adapter. I played with the idea of offering -10s with the META42 board because I could buy more chips for a given amount of money, which would make the overall price lower because I could get a better price break. But the -20 won out because you can solder it directly to the board.

Choices, choices...

 

[eric343]

Well, the difficulty with soldering it direct-to-board is that it makes it nearly impossible to swap opamps if we find a better one down the road... Have you ever tried desoldering one of those things from even a BrownDog adapter?

 

[tangent]

Quote:

Have you ever tried desoldering one of those things from even a BrownDog adapter?

Yes, and while I wouldn't want to do it repeatedly, it's not terribly difficult. It's certainly a lot easier than desoldering DIP chips. I've also desoldered SO-8 parts from the new META42 boards already, for what that's worth. A little desoldering braid and a sharp pry tool is all you need.

 

[eric343]

Without damaging it? True, easier than DIP. But it helps to have a soldering iron with a relatively long tip (or at least longer than mine)...

And I have yet to see a cheap/small SMD socket...

 

[tangent]

I guess you're thinking to heat all four pins on a side at once and then lift the whole thing up. That's one way, and there are special iron tips that make this feasable.

The way I do it is to remove all the solder I can with braid on both sides of the chip. You can get the pins pretty clean this way. Then I put the pry tool (a pick or a small, sharp screwdriver) under the edge of the chip and then start heating the pins in quick succession. Enough heat builds up that the remaining small amounts of solder give way under the pressure and the chip lifts up. Removing the other side is even easier. I usually switch to a pair of plyers for the force to get the second side off.

Now, I'm using words like "pry" and "force", but this is all pretty gentle. It's a matter of skill and speed. If you do it clumsily, you'll lift pads and traces and maybe damage the chip.

 

[aos]

I can testify that it is quite hard to do this without damaging the board and possibly the chip too. You must be extremely careful not to overheat the chip which is very easy to do when de-soldering. Tangent's technique is the same that I use, but the success factor varies on your skill, quality of PCB and your iron and so on.

 

[andrzejpw]

What's the best way to draw off heat when desoldering them? Hopefully, I won't have to attempt it, but just in case. . .

 

[aos]

Using desoldering braid is a good way to draw off the heat. The other is to let the chip cool often, and work piecewise instead of trying to pull off everything in one step (unless you're very skilled; I myself can't do that, and I certainly don't suck in soldering).

 

[andrzejpw]

I know about the desoldering braid. . . I found that out by myself!

I'm just worried about applying so much heat to all of the legs on a soic part. . .

Maybe it's just me, but I'm sure that if I blow on some of these parts too hard, I'm going to ruin it.

 

[aos]

Be worried. It's good to be worried, you'll be more careful. However, many parts do seem to be able to survive tons of heat. Their performance might suffer though in subtle ways afterwards.

 

[aos]

By the way, in regard to original question, it is very unlikely that AD8620 sounds any different than AD8610 - as long as only one opamp is used. If you use both - which is the only sane thing to do - you will get some crosstalk, getting higher as frequency goes up. Depending on opamp, application, board layout, decoupling and voltage regulation, this might or might not be audible. I think what Morsel said is that with META42, difference is inaudible. It is possible however that in a different environment such as dual mono configuration or even separate voltage regulation per channel this would be audible. I think ppl tried it in his amp and found no difference though.