Creation of Robotic Electronic Soldering/Rework Station

At each company I have consulted for, there is one or more recognized company manual electroinic rework technician(s). They solder stuff. And some of them are like AWESOME at it (and some are far less awesome, but I won’t mention that here). As an EE consultant seeking praise and accolade often resulting in extensions and recommendations for other internal gigs, I am happy to assume the credit for any successes. But as for soldering, why do the work when one can delegate? So I have had years go by since needing to do any actual soldering.

Well, I got blindsided the other day when I was in the electronics lab at DMS. The lab is awesome now, and so my presence in there alongside the fact that I mentioned being an EE had resulted in a request from a gentleman needing a small amount of rework. I’ve done plenty of soldering back in the day, and all he needed was to fix the connector on a USB memory stick that had been broken. The connector that we all recognize as USB prior to plugging into the side of the laptop had broken off the board. Simple enough, right? I told him it should take me just a few minutes.

Well, many, many minutes later, I was still wrestling with this problem. In fact, I spent the whole evening working on this, and never even fixed it. It was embarrassing. I am shamed forever…it should not be so hard to do a simple four solder joints.

But it was, and I want to explain why. But most importantly, I see opportunity, and I want to see what others think,

The problem was that the pads had been lifted where the connector usually attached. They appeared goldenish to the naked eye, so I spent a good half hour wondering why I could not get the the solder to flow onto them when heated. I also had not used the equipment in our lab for soldering, so I had to get acclimated to which of the tools available were the good ones for this, that. And there is no stereo ring light optics, so I tried to use my eyeballs only.

Finally, I realized that I needed a magnified look, so I grabbed one of the old school magnifying things on adjustable stand (I hate these, we need optics in there). This helped and I gathered the notion that I had no pads, but wanted to confirm. So I used the magnified camera with platform system recently donated,. This thing was cool for looking at stuff, and I could clearly see that there was no pads.

So now, my plan changed to grab power and ground from nearby vias, then rub soldermask off the USB pair and flow solder to the point at which the trace ordinarily contacts the now missing pad. After a long time trying to do this, I was not able to get the soldermask off well enough to make solder flow. And even if I had, the traces were too small for me to keep them from bridging.

I tried soldering direct to the USB transceiver. This was the point where I started to realize that things had shrunk to the point where my hands were not a good tool for this work at these sizes.

And so that is where I am right now and the reason I am writing here. You see, I think a robotic soldering system is in order now for not only DMS, but I think it could be useful in the industry generally. I have not seen one in the industry for board design (chip probing is different setup, picoprobe big name). And I think that this is something easy to make for my background, as LabVIEW, robotics and cameras are strengths in my resume. I could whip up a nice gui for this in a few hours.

So what do others think? Does this sound crazy to you? Have you guys seen such a thing in the industry?

Now I need to differentiate what I am talking about vs. automated soldering equipment used in manufacturing. Automated soldering stations are a thing in manufacturing, but I think something more flexible and user drivable (while watching magnified ringlight camera) with iron (or hot air, or both), camera, solder, and wire holder, all controllable through the gui or game controller.

Am I crazy or is this increasingly becoming a good idea? Suggestions about how such a thing should be made to operate? Any items anyone has seen that would work well for iron? Solder dispensing? magnified machine movable camers? Movable mechanical grasper?

Got any links to something close to what you have in mind?
I would think that if this is a viable idea, it would already be out there.

One would think. However, one day long ago, a dude made a soldering iron. And there had never been one before. Had he fixated on that often correct train of thought, I might instead say things to you like, art, I prefer metcal soldering torches…

Additionally, that dude would have had trouble doing his internet soldering iron searches due to the fact that the soldering iron had not yet been invented, and hence the internet did not yet exist…

This is interesting, for probing semiconductors:

http://www.ggb.com/34a.html

Here is more info:

All of this sfuff is applicable to larger pitch soldering situations. But you can get a feel for what is out there presently, and what can possibly be modified into methods for extremely fine pitch work.

The industry is headed there. Your phone already is way too small to work on much by hand. And its only going to get worse. I know its not the easy question to answer, but nothing worth having in this life comes easy, cheap or free. Someone, somewhere is going to make a solid innovation, if it hasn’t happened already. If not, I’d elect for that somebody to be one of us and that somewhere be DMS…

Internet users are speaking highly of this:

https://www.amazon.com/WEP-858D-Soldering-Station-Suitable/dp/B0055B6NGE

Which we may already have similar. These also look good:

https://www.amazon.com/Baitaihem-Nozzles-Solder-Soldering-Station/dp/B00LUKMBN8/ref=pd_lpo_469_lp_t_4?_encoding=UTF8&psc=1&refRID=R657VTHFSF03KHMG3AC8

Do we have solder paste in our lab? The last video I included in the above shows guys using paste and hot air to do soldering. Note that in the comments, a fellow inquires as to who knows ‘the trick’. I think that video shows ‘the trick’, but we need more tricks because this still is not enough to get the fix that I wanted on that SD card.

I have placed a similar hot air station and nozzles on loan to Electronics, so it should be available. I don’t believe we have solder paste available, and not sure how we can. I keep a small tube in my fridge, since I have been given to understand that it can ‘go bad’ very quickly once opened if not refridgerated. Even refridgerated it is supposed to only have a 6-12 month shelf life. It is also pretty expensive stuff compared to regular solder.

Not saying I don’t think we should stock it, but I think we should think through some of those logistic issues before we do so.

There’s some in the e-lab, but it separates out the flux so you’ll need to stir it up.

It has a shelf life and is expired but you can at least play with it.

This project would be fun. Having recent experience with building some 3d printers, It’s not that much different. The same mechanics and electronics on a smaller scale (nema 14s maybe on some small profile extrusion).

Not only fun, but I think it might turn profitable. I have noticed that those that make it to that enviable financial spot are dabbling with all sorts of ideas right there on the bleeding edge of what is yet to exist.

If we can find appropriate hardware to set up a prototype, I determine the most appropriate financing channel to procure the parts, and get them. But I would be looking to be ultra cost effective (so as to leave maximum margin for profitability). In other words, I would want to check out Chinese gear as I have noticed that some of it is quite good, but still cheap. I’m cool with individual steppers or perhaps even something already in the form of multi axis articulating arm (probably exists and much cheaper). And we need a camera solution, and some other stuff. In fact…

ULTRA FINE PITCHED ROBOTIC SOLDERING STATION PROTOTYPE

  • Magnifying Camera On A Stick- First its on a mount that can sit flat on a table. It does not need robotic articulation, because that would be silly, you can move it with your fingers. But it would be nice for all the camera features to be controllable via software. This is a USB camera, no doubt, and it is even better if it has options for thermal sensing capability. B/ut I would like to be able to turn off the thermal aspect as it may confuse in some situations. Focus and magnification are best as computer controllable (through USB).

*Light On A Stick-and a stand, I’m sure you have likely seen one of these before. They resemble the ‘snake light’ in form, but much smaller and sit on a stand so you can bend its bright tip so that it illuminates whatever you are interested is checking out.

*Mounted PCB Holder-we will need the target board to be held in place by something, but why not have things set up where we can hold smaller PCBs in any orientation of our choice. I am documenting here my random thought that the weighting of the bases of all this equipment is going to be quite important. Heavy bases means stable soldering setup. I like the sound of that.

*Soldertip, I think, should be something that has its side sort of over to the side (heavy base) and the coarse adjustment is handled manually. But once you get it in the vicinity of operation, it has USB X and Y control, and then it also has and a possibly faster actuating ‘thrust’ z axis so we could quickly apply heat, then remove. It seems it may be a good idea for this particular piece to have the capability of holding soldertip and/or hot air nozzle. We also need to examine solder extruding solutions, and they may need to go on this piece as well. So it turns out that this will be the key piece. And while ;we are in the mood, why not have an option to put a camera and light on this one too. You can’t go wrong with too many lights or cameras. Its just not possible. Just make sure they have lots of feature control software registers I can hit over USB.

*Wire holder-a lot of work on the lab involves putting the green wires to fix oopses on Rev A. So we need systems for holding the wire in place. I think camera and robotic control is a nice option here as well since holding the wire still was such a huge problem for me the other day. But it may prove to be the case that the roach clip dohickeys with a nice heavy base is sufficient. We have got to work with this stuff more to know for sure.

Real talk, I think that is a reasonable probability of pot of gold at the end of this rainbow. Enough so that I am up for devoting some time to it. But I always prefer co-conspirators to pin blame on… lol…

I have created so very many such similar gui software systems, and I believe this one would come out quite nice, but its not as much fun for me as it used to be. So I’ll do it if thats just the way it plays out, although I am also not above stepping to the side some to let another ambitious soul have some fun. This one is about video/motion controlling guis with coolness oozing out of it all over… but who would want to work on something like that??

Additional suggestions from SMD soldering thread, very pertinent IMO.

*Consider the use of laser soldering system.
*Measuring the temperature of soldering target can be quite helpful. Doing so via camera could be good. There was also a suggestion to measure radiance at a few wavelengths, with peak at 300C being -5um and 20C is -10um.
*programming parameters and saving them to files for various part shapes could be quite helpful for soldering complete chips.

So if I had my wishes at this point, I am now soldering with computer movable laser, I may be using solder paste, but I also like having option to use regular solder, I am still brainstorming ways to hold a wire there as necessary, and I am watching things up close with flir type thermal camera (infared camera? can this be done with magnifying lens?)

Any more bright ideas (total absence of my usual sarcasm, there are nothing but bright ideas I am hearing from DMS folk concerning my solder station concept). Give me more, please!..lol…

I would suggest a 40watt 808nm diode bar with fiber lead. It will simplify your optics setup. The standoff distance needs to be > 3 inches to keep crud off of your lens.

I’ll check that out. Thanks, Ken!

Ken,

You are not allowed to say a bunch of technically correct answers, then just disappear off back into the ether leaving us mere mortals to wonder how you knew this!.. lol…

Why 40 watts? Sure, I want to save a lot of time evaluating, but I want maybe the user to be able to control the power delivery from laser to heated surface. Is this possible with lasers or am I stuck with just a given power level and can only control power delivered into target via duty cycle and duration of pulsed use?

Why 808 nm? I think this is invisible wavelength. Wouldn’t it be helpful for us to be able to see the beam? I shot a laser through a sleeve of a shirt I was wearing in a lab (former life). I couldn’t see it. I think its better if we can see it. Is this just out of the question? Why? Come to thinik, I have a million questions about 808. I don’t want to waste endless dollars on arriving at 808 myself. But I must know why, or my head will explode. You don’t want my head to explode, do you?

I can live with >3 " to eliminate crud, because I have no use for crud. I generally don’t want crud. In this case I am electing to specifically avoid crud. So it will be >3.

Thanks for the advice, and yes I am quite certain we are going to wind up with an invisible 808 nm bar type laser with about 40 W at > 3 inches. But I am hoping for an excess of two sentences of interface with my higher power when it comes to this topic…lol…

Visible is a no go unless you have proper glasses or enclosure. 808nm is the most available infrared wavelength in high powers. Can go higher like 975nm. These are mostly used to pump ND:Yag solid state lasers but make great focusable heat sources.

Looks like this but needs to be 10watts or higher:

http://www.ebay.com/itm/808nm-3W-Fiber-Coupled-Laser-Semiconductor-Diode-Fiber-Coupled-laser-Engrave-/141388631068?hash=item20eb6b4c1c:g:C9IAAMXQWlFRwBXD

40 watts lets you have a bigger heat zone.

The power supplies are pretty basic but can control power via PWM or constant current PSUs.

Why do I know this? I have lots of lasers, basically everything but a Copper Vapor Laser. Prettiest lime and emerald green lines. Pink Floyd used to use them a lot outdoors. Beams as big a a broom handle and visible for 10 miles.

Thank you very much for the input. You are allowed to again walk among the mortals.

I am hoping for power delivery ‘knob’, software or otherwise, this way we can use the laser to heat a surface that may be, and often is, attached to a reference or ground layer through multiple vias. Or perhaps we are soldering a coaxial cable to something, and the shield is the target so heat keeps draining down the wire. Whatever the case, we would want control so that we don’t overdo it and burn up components.

But I very much agree that laser is where its at for what I am proposing here. Maybe even two lasers, one low power visible for aiming, and the other for heat delivery.

Ouch… 40w… that sounds really expensive :frowning:

Kinda depends on what you classify as expensive and how long you are willing to look.

http://www.ebay.com/itm/Fiber-Coupled-Laser-Diode-58W-915nm-Checked-Verified-/122284938070?hash=item1c78c01356:g:meUAAOSw44BYMgGp

1 Like