MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

IC sockets -- machined pin vs leaf type
MUFF WIGGLER Forum Index -> Music Tech DIY Goto page 1, 2  Next [all]
Author IC sockets -- machined pin vs leaf type
cygmu
I've noticed several photos of through-hole builds recently in which IC sockets with machined holes were used. These look really weird to me because I can't see how the flat pins of most ICs will make a convincing connection with the circular holes. The leaf-type sockets which grip both sides of the IC pins seem more likely to make a good connection, but this is just based on instinct rather than anything informed. Evidently both types work or they wouldn't be in use.

So, I am wondering: which type of IC socket is preferable and why?
aabbcc
Machined, simply because they feel better when inserting and removing the ics lol
lasesentaysiete
aabbcc wrote:
Machined, simply because they feel better when inserting and removing the ics lol


I am of the opposite opinion. I prefer the leaf-type for ease and feel. But I think that machined sockets are more durable and therefore more professional.
aabbcc
lasesentaysiete wrote:
aabbcc wrote:
Machined, simply because they feel better when inserting and removing the ics lol


I am of the opposite opinion. I prefer the leaf-type for ease and feel. But I think that machined sockets are more durable and therefore more professional.


With those I always seem to get the ics just slightly misaligned when inserting and then I have to use a bit of force to push them further until they make this click sound, always feels like I'm about to ruin something with those hehe :(
Graham Hinton
cygmu wrote:
which type of IC socket is preferable and why?


Don't put a square peg in a round hole--or rectangular in this case. Machined sockets are intended for round pin headers.

Face wipe (as opposed to edge wipe), gas tight are better and most sockets are like that now, but they weren't always so. Texas Instruments used to make some particularly poor ones, both face and edge wipe, so beware of old stock "bargains" on eBay.
kashmir
Lots of people recommend the machined ones for everything because of reliability and so, but this makes much more sense! I prefer the leaf-type and I´ve seen it in many "professional" devices, unless the chips had a round pins of course.
Graham Hinton wrote:
cygmu wrote:
which type of IC socket is preferable and why?


Don't put a square peg in a round hole--or rectangular in this case. Machined sockets are intended for round pin headers.

Face wipe (as opposed to edge wipe), gas tight are better and most sockets are like that now, but they weren't always so. Texas Instruments used to make some particularly poor ones, both face and edge wipe, so beware of old stock "bargains" on eBay.
keninverse
I realize this thread is about sockets but at some point one reaches a point in their "career" that they're comfortable with just soldering ICs, right? Lately I just solder in op-amps because A) they're cheap and B) I rarely experience op-amp failures. I worry about CMOS chips and more expensive ICs so those get socketed into nicer AMP dual-wipes (which are markedly cheaper than machine pin). I just don't think that machine pin sockets are necessary unless it's being placed in a high vibration environment and I'd consider soldering those in rather than relying on a socket.
pre55ure
keninverse wrote:
I realize this thread is about sockets but at some point one reaches a point in their "career" that they're comfortable with just soldering ICs, right? Lately I just solder in op-amps because A) they're cheap and B) I rarely experience op-amp failures. I worry about CMOS chips and more expensive ICs so those get socketed into nicer AMP dual-wipes (which are markedly cheaper than machine pin). I just don't think that machine pin sockets are necessary unless it's being placed in a high vibration environment and I'd consider soldering those in rather than relying on a socket.


Interesting. I have always just never understood why someone would solder any IC directly rather than use a socket. They don't add any additional pins to solder, they're pretty cheap, and they make troubleshooting/future replacement so much easier.

Not saying your wrong, just different philosophies about it.
Graham Hinton
keninverse wrote:
Lately I just solder in op-amps because A) they're cheap and B) I rarely experience op-amp failures.


Unfortunately ICs don't fail according to cost. Either socket everything or nothing and if nothing the pcb and the components have to be tested before assembly.

I had an expensive HP Universal Counter fail recently. Fantastic build quality. All the larger or more expensive ICs were socketed, but the general glue wasn't, moreover most had custom HP numbers. It was a single board device with built in multiple regulators and came with HP's proprietary "signature analysis", but one of the cheap chips had failed and the power supply shut down. So much for signature analysis identifying the faulty chip in that situation. There were 30-40 unsocketed ICs and it just wasn't worth the effort of repairing it, which would have been easy if every IC could have been taken out and put back in again until the psu came up.


Quote:
I just don't think that machine pin sockets are necessary unless it's being placed in a high vibration environment and I'd consider soldering those in rather than relying on a socket.


Actually, if you vibrate a pcb at the right frequency ICs will just jump out of machined sockets.

Machined sockets are a good example of a product that amateurs assume are better and "more professional" because they have a higher price.
lasesentaysiete
Graham Hinton wrote:

Machined sockets are a good example of a product that amateurs assume are better and "more professional" because they have a higher price.


I do not think the assumption is based only on the higher price. I bet their look and name have contributed to the misconception, too. And they are very often found on BOMs provided by manufacturers of synth DIY products.
Rex Coil 7
Edited ... meh seriously, i just don't get it ... what do I know.

nodnod



I'm just yet one more talking ape with a computer.

lol

thumbs up
nurbivore
I was rolling my eyes a little at the talk of problems with the machined pin sockets. But then just today, one of my modules stopped working, and when I pulled it out, the main IC was hanging half out of its socket. So, you know, be careful what you think.
The Real MC
Back in 1995, I installed machined pin sockets for the ribbon cables in my Memorymoog.

It is still going strong.

My vote goes to machined pin.
AlanP
I could be wrong, but IMO, cheap machined pin sockets are crappier than cheapo dual-wipe leaf sockets.
Graham Hinton
The Real MC wrote:
Back in 1995, I installed machined pin sockets for the ribbon cables in my Memorymoog.


Ribbon cable DIL headers usually have round or, at least, formed pins not flat like ICs. That is what they are intended for.
The Real MC
Graham Hinton wrote:
The Real MC wrote:
Back in 1995, I installed machined pin sockets for the ribbon cables in my Memorymoog.


Ribbon cable DIL headers usually have round or, at least, formed pins not flat like ICs. That is what they are intended for.


Memorymoog DIL headers have flat not round pins. I neglected to mention that I prefer gold plated pins on the sockets.
The Real MC
AlanP wrote:
I could be wrong, but IMO, cheap machined pin sockets are crappier than cheapo dual-wipe leaf sockets.


You are not wrong. You get what you pay for. I used Augat machined pin sockets (with gold plating on the pins) with good results.
Graham Hinton
The Real MC wrote:

Memorymoog DIL headers have flat not round pins. I neglected to mention that I prefer gold plated pins on the sockets.


Some DIL headers are intended to be soldered in and are called a "pcb transition". Not all transitions are DIL, but they are not gold plated and hence cheaper.

That's the other thing, don't mix tin and gold platings. Not many ICs have gold plated pins now, but "side brazed" ceramic packages do and these need a gold plated socket. Face wipe is still preferable though.
The Real MC
Graham Hinton wrote:
That's the other thing, don't mix tin and gold platings. Not many ICs have gold plated pins now, but "side brazed" ceramic packages do and these need a gold plated socket. Face wipe is still preferable though.


I studied some physics in college and am aware that certain combinations of elements and alloys can cause adverse reactions.

But I don't think the tin/gold cautions are all encompassing though.

Before Al Pearlman started ARP, he designed products for the military and aerospace industry which require compliance to design conventions. Some of those conventions carried over into ARP products. I have an ARP ProSoloist with mixed tin/gold interconnects, original from the factory. I bought it in 1994 and those interconnects have been zero trouble.

I don't know how to explain why my Memorymoog experience was positive, but 25+ years of tin/gold with no trouble is a hell of a testament and was far better than the original tin/tin interconnects which barely lasted eight years.

Maybe I was lucky enough with the right composition of tin alloy with gold.
Rex Coil 7
Graham and I were discussing this very issue not long ago (the gold with tin thing). And it seems as though I recall him saying it has something to do with abrasion rather than chemical interaction.

I could be wrong about that, however. I often times am, indeed, wrong.


... but I'll be a sonovabitch if I don't always try my best! ... that's certain!


thumbs up

EDIT: Nerp ... it was a discussion about pins used with various types of contacts in sockets ... mechanically speaking, that is. Something about 3 wipers vs one ... or something. OH WAIT! ... I think it had something to do with the multi-connector he uses on the Euro power cables he makes. Um, I think.

Geez, I'm pretty useless here. hmmm..... I'll uh ... just get my coat.

Now ... where is that rock I was under?

meh.
Graham Hinton
The Real MC wrote:
But I don't think the tin/gold cautions are all encompassing though.


You don't understand the problem. Listen to what connector manufacturers tell you.

Quote:

Before Al Pearlman started ARP, he designed products for the military and aerospace industry which require compliance to design conventions. Some of those conventions carried over into ARP products. I have an ARP ProSoloist with mixed tin/gold interconnects, original from the factory. I bought it in 1994 and those interconnects have been zero trouble.

I don't know how to explain why my Memorymoog experience was positive, but 25+ years of tin/gold with no trouble is a hell of a testament and was far better than the original tin/tin interconnects which barely lasted eight years.

Maybe I was lucky enough with the right composition of tin alloy with gold.


Maybe you were lucky with a low number of mating cycles. Two instances are not a valid statistical sample.

Your post is an outstanding example of multiple fallacious reasoning. While I have always admired Alan Pearlman's circuit designs, the production engineering at ARP left a lot to be desired and the people responsible caused the company's demise.
PrimateSynthesis
Graham Hinton wrote:

Unfortunately ICs don't fail according to cost. Either socket everything or nothing and if nothing the pcb and the components have to be tested before assembly.


It's not that IC's fail according to cost. It's that when more expensive IC's get damaged during assembly or disassembly they cost more to replace.

Putting chips in sockets makes repairs easier. But it also costs more money and can be less reliable. It depends on the quality of the socket, chassis, and how the gear is used. I've seen plenty of equipment that didn't work only because of dirty sockets, that were fixed with a can of spray cleaner. I've also seen guitar pedals that stopped working after chips were knocked loose because pedals often get dropped and kicked around.
Rex Coil 7
PrimateSynthesis wrote:
Graham Hinton wrote:

Unfortunately ICs don't fail according to cost. Either socket everything or nothing and if nothing the pcb and the components have to be tested before assembly.


It's not that IC's fail according to cost. It's that when more expensive IC's get damaged during assembly or disassembly they cost more to replace.

Putting chips in sockets makes repairs easier. But it also costs more money and can be less reliable. It depends on the quality of the socket, chassis, and how the gear is used. I've seen plenty of equipment that didn't work only because of dirty sockets, that were fixed with a can of spray cleaner. I've also seen guitar pedals that stopped working after chips were knocked loose because pedals often get dropped and kicked around.
I suppose one could design the PCB with holes on each side of an IC socket for using a zip tie placed all the way around the IC to secure it in the socket. Or a pair of short standoffs on each side of the socketed opamp could be used to hold down some sort of securing "lid" or "cap". All depends on how dedicated one is to addressing some need of the end/user.

If that's not an option perhaps using a "breadbag tie" (one of those plastic covered tie wires used to close up a bag of bread) ... or even just a length of insulated hook up wire ... could be threaded long-wise under the IC socket and twisted/tied down on the IC to keep it in place.

Blob a small bit of silicone goop on one or more spots on the IC/socket to secure the IC.

Solder just one pin into the socket to keep it in place.

Sometimes an easily changed IC is desired so that the end/user may install different opamps to produce different sound. This is a commonly done thing on certain overdrives/distortion units, it's also done with certain compressor circuits. In some compressors/overdrives changing opamps can be as effective and personal as changing tubes in guitar amps. That said, using IC sockets is pretty much a necessity in those situations.

I've even used I/C sockets as multi-connectors ... in one custom unit I used an IC socket to accept another IC socket (used as a sort of ~plug~) which had wires soldered into it ... those wires ran to an area within the enclosure that housed a PCB of my own design that had four different opamps in it. Using a toggle switch array the user could switch between four opamps from the front panel, opamp combos were also possible depending on how the switches were set. It worked.

I/C sockets do have their uses, that's for sure.



The Real MC
Graham Hinton wrote:
Maybe you were lucky with a low number of mating cycles. Two instances are not a valid statistical sample.


You assumed I was successful with two examples. What you do not know is that I have successfully applied this practice to many vintage analog synths.

You assumed that low mating cycles was the key to my success. But with tin/tin interconnects there was also low mating cycles, and the oxidation wreaked havoc on sensitive control voltages.

Do not be so quick to assume.

Quote:
Your post is an outstanding example of multiple fallacious reasoning. While I have always admired Alan Pearlman's circuit designs, the production engineering at ARP left a lot to be desired and the people responsible caused the company's demise.


What part of 40+ years of failure free history of my ProSoloist built in 1973 falls under the definition of "fallacious"?

Yes ARP had their moments of questionable engineering. This was an exception. It was early in ARP history before intransigent management practices set in.

Do not be so quick to apply knee-jerk over-generalizing labels to experiences that contradict your own. You wrongly assumed low mating cycles and low statistical sample set did not warrant a conclusion. You harbor a deep suspicion of ARP design conventions which illustrates that you may be incapable of accepting that there were decent ARP products. Not a fine example of statistical integrity.

Lighten up. It's OK to question, but it is not professional to blindly dismiss without knowing all the facts.
The Real MC
Graham Hinton wrote:
The Real MC wrote:
But I don't think the tin/gold cautions are all encompassing though.


You don't understand the problem. Listen to what connector manufacturers tell you.


You don't understand the exceptions. Listen to what my experiences since 1994 have told you.
MUFF WIGGLER Forum Index -> Music Tech DIY Goto page 1, 2  Next [all]
Page 1 of 2
Powered by phpBB © phpBB Group