[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: [N8VEM-S100:372] What DIP sockets to use in building boards
Would like to pass on some hard earned lessons here!
Beyond doubt the Achilles heel of these circuits boards is the sockets. If a board seems to be "acting up" the first thing you should do is press in all the IC chips. Particular the 14 pin ones. (While you are at it clean the S100 connector with either a dry lint free towel or better "TV tuner" ). Cheap sockets are simply not worth it for anything other than a quick and dirty "one time" prototype board.
Over the years I find the "Single Dual Wipe Solder Tail Low-profile IC sockets" the best. They are quite common. For example the Jameco 14 pin "JVP" ones are #112214. For a "final board" if you can afford it, use the "Machine Tooled Solder Tail Low-Profile" IC sockets. Again common. The Jameco equivalent would be "JVP" #37197.
Unless you have a completely reliable burned in working board and transfer all the chips across to socket-less board I honestly would not recommend this approach to most people. If you do it with untested chips and the board does not work it is far more difficult to debug. Worse still, for many getting out the IC chip without raising the solder pads from the board is a serious hazard. If you do find yourself in this situation and you don’t have a multi-pin desoldering tool or vacuum desoldering gun its simply best to cut all the IC pins on the (suspected) faulty IC and remove them one at a time. Then use a needle/pointer to enlarge the hole after melting the residual solder.
Also if you find a board "flakey", for example it works/stops when you flex the board, this is almost always due to a pin being bent under the IC and only making a partial connection with the socket. Always check for this on a finished board.
I have to say that I have probably hand done over 200 S-100 board over the years, to this day, I still find solder connections I have to "touch up" when all components are on the board. I use a magnifying glass and go over the back of each IC socket one at a time. I don't know how many times I could have sworn the board was perfect only to find a marginal connection. So if you are a beginner, save yourself a lot of time later by doing this upfront. Make sure traces are not bridged by too much solder. In particular make sure Ground and +5V pads are not over or under soldered.
Finally when you find yourself soldering is all those 0.1 buffer caps, resistors etc., insert then all into the board bending the wires so they don’t fall out on the back. Then when soldering, do only one side of each of them first then come around again and do the second side. This way there is less danger of frying them.
From: n8vem...@googlegroups.com [mailto:n8vem...@googlegroups.com] On Behalf Of Lawrence Lewis
Sent: Saturday, July 30, 2011 7:48 AM
Cc: 'Lawrence Lewis'
Subject: RE: [N8VEM-S100:372] What DIP sockets to use in building boards
This kind of information should be on the forum.
I think a lot of the original S-100 board producers soldered the ICs in to save money and avoid problems.
From: n8vem...@googlegroups.com [mailto:n8vem...@googlegroups.com] On Behalf Of Bill Lewis
Sent: Saturday, July 30, 2011 5:55 AM
Subject: Re: [N8VEM-S100:370] What DIP sockets to use in building boards
Connectors are generally the most failure prone components in electronics.
The most reliable way to mount ICs is to solder them in.
I know people in this group are constantly playing musical chairs with their ICs, between 74LS, ALS, 74F, HCT, etc. trying things out. So sockets are still called for.
But don't skimp on socket quality. It is a false economy.
On 7/30/2011 4:33 AM, Douglas Goodall wrote:
> Now I am wondering if my desire to be conservative and socket
> everything, coupled with being new to the socket sourcing business has
> combined to contribute to the general flaky behavior some (many) of my
> boards have exhibited.
> During debugging I have experienced a situation where the pin on the
> other side of the board doesn't have continuity with the pin of the
> chip in the socket, and I assumed it was iffy soldering, but it could have been lousy sockets.
> I guess I will try Jack's suggestion (Phoenix Ent) in hopes that these
> problems will go away. If the sockets are the source of my troubles,
> the money saved isn't worth the trouble.
> On Jul 29, 2011, at 11:04 PM, Nick Papadonis wrote:
>> I ordered some inexpensive non-machined sockets from China on Ebay.
>> This occurred before I knew anything about socket quality. They were
>> pretty much junk and their price said it all. The IC barely held
>> into the socket. They can't be used for socket in socket adapters
>> because the pins pop out. I prefer the machined sockets even though
>> they are more expensive. They are better constructed.