Any electrical engineers here?

Status
Not open for further replies.
Originally Posted By: Reddy45
OH yeah the tips on these cheap soldering irons don't last at all. I keep the very tip of mine tinned enough to work for the next time, but the rest of the iron is just corroded from heat.

I may actually pick up a rechargeable Hakko since the battery tech now gets them pretty dang hot and the lack of cords means no tripping/fire hazard.

Mine was actually pretty high quality. This was early 80s and Rat Shack was sourcing these soldering irons and tips from a Japanese manufacturer. However, I was 12 and had no idea what I was doing. I didn't have a cleaning sponge or solder suction remover. I didn't use flux. I just left it there on the stand long enough that the tip just started wearing away.
 
Originally Posted By: Reddy45
Update - I used some spare bits to jump 2 pairs of contact points, and that seems to have gotten things a bit further.

Now, I can press each button to generate a number (1 through 8) on the display.. but that's about it.


That means your soldering is broken (cold?) on the 2 pairs you jumped. Try warming up your iron real warm and solder again.

I'm "sorta" an electrical engineer but you do not want me to solder anything. These days a huge amount of EE work has become "software defined", because the risk of a tiny mistake in hardware (in house design or from vendor) can easily sink a 10-20M project and bankrupt a small company.
 
Last edited:
Originally Posted By: PandaBear
Originally Posted By: Reddy45
Update - I used some spare bits to jump 2 pairs of contact points, and that seems to have gotten things a bit further.

Now, I can press each button to generate a number (1 through 8) on the display.. but that's about it.


That means your soldering is broken (cold?) on the 2 pairs you jumped. Try warming up your iron real warm and solder again.

I'm "sorta" an electrical engineer but you do not want me to solder anything. These days a huge amount of EE work has become "software defined", because the risk of a tiny mistake in hardware (in house design or from vendor) can easily sink a 10-20M project and bankrupt a small company.



It's hard to explain, but if you look at the 2nd post of this topic where I put pics, you'll see what looks like a bare wire that goes through 2 holes in the PCB (one in the middle, one on the left). On the back side, these are actually solder pads so the wire is literally jumping the 'tracks' in 2 spots..

I literally used the clipped 'legs' from the resistors to jump these and that got the buttons working.
 
Originally Posted By: PandaBear
Originally Posted By: Reddy45
Update - I used some spare bits to jump 2 pairs of contact points, and that seems to have gotten things a bit further.

Now, I can press each button to generate a number (1 through 8) on the display.. but that's about it.


That means your soldering is broken (cold?) on the 2 pairs you jumped. Try warming up your iron real warm and solder again.

I'm "sorta" an electrical engineer but you do not want me to solder anything. These days a huge amount of EE work has become "software defined", because the risk of a tiny mistake in hardware (in house design or from vendor) can easily sink a 10-20M project and bankrupt a small company.

I kind of know what you mean. It's just various levels of abstraction. In my field (chip design) it's RTL/netlist/layout/masks where the engineer never has to ever touch the physical chip. And at all those levels we're just desk jockeys staring at screens all day.

But with the modern electronics industry it's simply required that everything has to be done on a computer. Besides that - almost all parts assembly work is automated. It's possible to attach micro-BGA parts individually, but it's generally impractical at the production level. And the testing generally isn't to detect what to fix, but what to send to the recycling heap after some parts go for failure analysis.
 
Originally Posted By: y_p_w
I kind of know what you mean. It's just various levels of abstraction. In my field (chip design) it's RTL/netlist/layout/masks where the engineer never has to ever touch the physical chip. And at all those levels we're just desk jockeys staring at screens all day.

But with the modern electronics industry it's simply required that everything has to be done on a computer. Besides that - almost all parts assembly work is automated. It's possible to attach micro-BGA parts individually, but it's generally impractical at the production level. And the testing generally isn't to detect what to fix, but what to send to the recycling heap after some parts go for failure analysis.


Actually, on the manufacturing side, a lot of contract manufacturer do replace parts when they are bad instead of throwing the whole board to scrap. They have machine for that too.

Even RTL level mistakes can be costly, I think what I heard is a few million per tape out these days, and a few months lead time. Projects often get cancelled if serious problem is discovered late.
 
Originally Posted By: PandaBear
Originally Posted By: y_p_w
I kind of know what you mean. It's just various levels of abstraction. In my field (chip design) it's RTL/netlist/layout/masks where the engineer never has to ever touch the physical chip. And at all those levels we're just desk jockeys staring at screens all day.

But with the modern electronics industry it's simply required that everything has to be done on a computer. Besides that - almost all parts assembly work is automated. It's possible to attach micro-BGA parts individually, but it's generally impractical at the production level. And the testing generally isn't to detect what to fix, but what to send to the recycling heap after some parts go for failure analysis.


Actually, on the manufacturing side, a lot of contract manufacturer do replace parts when they are bad instead of throwing the whole board to scrap. They have machine for that too.

Even RTL level mistakes can be costly, I think what I heard is a few million per tape out these days, and a few months lead time. Projects often get cancelled if serious problem is discovered late.

I've seen rework done before, but that was for prototypes. But I think it depends on the total cost and whether it makes sense to track down the error and what to replace. But a lot of companies will scrap an entire board just because of a bad connection or a single bad part. Certainly they'll do it for a board assembly that costs maybe $10-20.

RTL can't really be fixed, but I've done my share of ECOs to modify a setlist. Nothing quite like going through a netlist and spare cells to figure out how to change it with only metal layers. I think this is probably going over most readers heads.
 
Originally Posted By: Reddy45
Any idea what the heck this thing is supposed to do?
Yep...give you the opportunity to practice soldering on a PCB without lifting any pads. Several decades ago I used to play around with 7-segment displays and the IC driver along with a bunch of other stuff on my Heathkit electronics experimenter. I still use the BB and the PS still works!

I learned to solder when I was about 7yrs old, on wire lugs. Still soldering but now using a magnifier hood. I was amazed having gone through the EE program at a major university, having never picked up an iron. They didn't even teach it. Considered that 'technicians' work. Ha....

Guess how many electronics faults I've discovered over several decades due to LOUSY solder joints? Cracked rings-around-the-post, cold joints, lifted pads, burned pads & traces that needed repair...and that's with 60/40. The lead-free stuff is even worse.
 
Originally Posted By: y_p_w
I've seen rework done before, but that was for prototypes. But I think it depends on the total cost and whether it makes sense to track down the error and what to replace. But a lot of companies will scrap an entire board just because of a bad connection or a single bad part. Certainly they'll do it for a board assembly that costs maybe $10-20.

RTL can't really be fixed, but I've done my share of ECOs to modify a setlist. Nothing quite like going through a netlist and spare cells to figure out how to change it with only metal layers. I think this is probably going over most readers heads.


I can only speak from my limited experience. One contract manufacturer said they repair assembly up to 3x, and the assembly is about $100-700 worth of parts on a $14 service cost, so it make sense to rework instead of scrap.

We do metal layers work around in combination with firmware tweaks, because, metal layer is cheap compare to a completely new tapeout to fix the lower layer (in 14nm at least).
 
Status
Not open for further replies.
Back
Top