Sockets. Some good, some bad?

[AntKnee]

So I've been socketing a lot of diodes lately, building a lot of overdrives and distortions lately. I've found that the break-away single sockets I've been using are useless. Once I place a component in the socket and remove it, the next component I put in that socket barely stays put. This obviously completely defeats the purpose of the socket. Is it bad sockets, or do they all do this? I can't remember where I got them, but if it was Tayda, I suppose I shouldn't be surprised.

[pickdropper]

There are higher quality ones, but standard SIPs often expand when a component lead is installed.  If the lead is big enough, the metal in the socket is then permanently deformed.

The really nice individual sockets, like the ones from Mill Max, are designed to take a set when you insert a pin that's within their specified range.  At that point, they have a rated number of insertion cycles.  I can't remember exactly how many, but 1000 is the number that pops in my mind.

The downside of these is that, while designed to take a set, once the metal is deformed you can never use a pin (or component lead) smaller than the first one.  Nor can you use anything larger.  The other downside is that they are somewhat pricey.  You might spend 0.60-1.00 for each individual socket, not for a strip of them.

Sockets are a necessary evil, but not ideal for long term reliability.  If you are worried about quality, stick with name brands like Mill Max and skip the Tayda ones.  Those should at least fare better.

[Luke51411]

I've done ok with the tayda ones but typically I might change something out 2 or 3 times and the lead size definitely makes a difference. I also often remove the sockets once done testing and solder the component to the board.

[alanp]

You can also solder directly into the socket.

[mandrewbot3k]

I've often toyed with the idea of getting 6-12 position rotary switches and having them stacked with different diodes that I can test out in different circuits. This way there is only one insertion and removal before the next one, and you could use a small diameter lead to get them in there.

The unfortunate reality is that no two diodes are the same and you really need to test the ones you'll be using in the circuit if you truly care. Or at least have different FVs measured out on the different diodes.

In the end I don't have an answer for you. haha. I kind of hate the singles because they always break wrong for me. I started using IC sockets and cutting them in half. I don't socket diodes usually so I haven't really had an issue with them.

[mremic01]

I've often toyed with the idea of getting 6-12 position rotary switches and having them stacked with different diodes that I can test out in different circuits. This way there is only one insertion and removal before the next one, and you could use a small diameter lead to get them in there.

The unfortunate reality is that no two diodes are the same and you really need to test the ones you'll be using in the circuit if you truly care. Or at least have different FVs measured out on the different diodes.

This. I have several baggies of 1N34As and 1N270s, along with other random germaniums. I can nail the Klon tone with them, but only some of them, even from the same lot. A 1N34A can do the magic diode thing if you pick out the right one. All the '1N34As are fuzzier then D9Es' crap makes me facepalm.

The rotary switch box is a good idea. You can have some positions socketed to demo individual diodes against each other. Socket quality wouldn't even be that much of an issue for that since the thing doesn't need to be gig worthy.

[mandrewbot3k]

Thats true. Hmmm... I may have work on a diode socket PCB.... for testing...

[playpunk]

There is a shared oshpark project for this. Rullywow shared it


Sent from my iPad using Tapatalk