SOLVED Who else gets gremlins when socketing transistors?

MBFX

Well-known member
Tonight, I built a Fuzz Face with some Germanium transistors. It sounded amazing on the breadboard, so I committed it to vero using a layout I know to be good. Suddenly, it won't bias right and it sounds weak and shitty. I ended up rebuilding it on a breadboard, getting a good result, then re-socketing the transistors in the veroboard build. Suddenly, it is perfect.

Anyone else get problems with socketed transistors?
 
Solution
When using sockets, I've found some components have slightly thinner leads than others and can lead to connectivity gremlins. To overcome intermittent issues with socketed parts, I've started the practice of gently wrinkling the ends of the leads. (see attached image)

Wrinkled_Leads.jpg
When using sockets, I've found some components have slightly thinner leads than others and can lead to connectivity gremlins. To overcome intermittent issues with socketed parts, I've started the practice of gently wrinkling the ends of the leads. (see attached image)

View attachment 65766
Marking this as the solution because I just tried it and it's perfect!
 
No sockets for me. They cause more trouble than they resolve
I think I'm with you Harry. I used to be a believer in socketing transistors, but I've had my share of issues. I think from now on, unless I have a compelling reason, like I want to try a few different transistors, I'm just going to soldier them in. ICs are a different story - I'll always socket ICs -- had too many times where a bad IC was easily rectified by just swapping it out.
 
I like sockets in case I fuck up the pinout of an old transistor I paid $8+ for. It's a protection of investment thing. It also feels nice that they are more easily reusable, I guess. I should probably get over that.
 
So I decided on my latest build (PPCB Waypoint Fuzz) to skip the sockets…it’s an Si fuzz. Then I watch Grey Bench do a teardown, and the transistors are color coded. Mine all tested the same: 192hfe. Color coded almost definitely means sorted by hfe. My pedal is below unity and is more clean that anything…

I think I’ll go back to sockets…
 
So I decided on my latest build (PPCB Waypoint Fuzz) to skip the sockets…it’s an Si fuzz. Then I watch Grey Bench do a teardown, and the transistors are color coded. Mine all tested the same: 192hfe. Color coded almost definitely means sorted by hfe. My pedal is below unity and is more clean that anything…

I think I’ll go back to sockets…
f I do sockets, I solder the transistor legs permanently after selection is complete.
 
I'm not a fan of the machined sockets in any configuration; singles, SIP or DIP. You get one, maybe two in-and-outs and then you're boned. I did it to verify a board (quick, dirty and melty, I admit), and even that was a disaster. Hot glue to the rescue 🪄 🤬

apologies for shaky cam :S

kRvjC8B.jpg

soldered leads on the salvaged PNPs weren't related to the socket failure

I'll swear by (leaf spring) socketing ICs, however. #twoPhasedHippocrip
 
I'm not a fan of the machined sockets in any configuration; singles, SIP or DIP. You get one, maybe two in-and-outs and then you're boned. I did it to verify a board (quick, dirty and melty, I admit), and even that was a disaster. Hot glue to the rescue 🪄 🤬

apologies for shaky cam :S

kRvjC8B.jpg

soldered leads on the salvaged PNPs weren't related to the socket failure

I'll swear by (leaf spring) socketing ICs, however. #twoPhasedHippocrip
I'm currently repairing a circuit where i socketed transistors in machine sockets, but due to bad transistors, I had to replace them several times. In one socket I pulled out the machined part, leaving me with a socket that won't conduct properly, and I snapped off a transistor leg in another socket, leaving me with a socket that I can't insert a new transistor. After this experience, I think I may be done socketing transistors. ICs are another story. I agree with you that the leaf spring sockets for ICs are the better choice.
 
No problems with sockets, except those crappy fly-leaf IC failures-waiting-to-happen.

So I'll continue to use sockets and experiment — thus probably have tons of problems in the near future.


Meanwhile...

socketoomeesocketoomeesocketoomeesocketoomeesocketoomeesocketoomeesocketoomeesocketoomeesocketoomee...
 
Back
Top