Author Topic: Hot plate reflow soldering and solder beads  (Read 3123 times)

0 Members and 1 Guest are viewing this topic.

Offline tounhoTopic starter

  • Newbie
  • Posts: 3
  • Country: de
Hot plate reflow soldering and solder beads
« on: May 23, 2023, 07:29:23 pm »
Hello,

I have made my own hot plate PCB using the JLCPCB aluminum PCBs and a simple Arduino (for now) + MOSFET + Thermistor control circuitry.
I have been trying around a lot but I keep having problems with solder beads.
Initially I had been using many years old, never refrigerated Chipquick Sn42Bi58 paste. Since the ageing might be my problem, I recently bought new MG Chemicals 4902P Sn42Bi57Ag1 paste (from Amazon, maybe a mistake), but I am still having problems.
I am applying the paste directly onto the pads per hand via the syringe. I tried different amounts of paste and different temperatures, going up to 200C.
My latest attempt's temperature curve looks like this:
(a)
The target was 180s at 120C and then 200C for 60s. I know that might be excessive for 138C paste. 160C from the manufacturer's datasheet don't work either.
Using the new paste, with the temperature curve above, the result looks like this:
(b)
I tried using more or less paste at different reflow temperatures (180C, 200C) but nothing works.
I also tried using the old paste with the same temperature curve and it weirdly looks better (even so I used way too much paste)!? (only pay attention to C5)
(c)
A video of the attempt from (c) can be found here (d)
All above are 603 components.
At this point I don't know what to try next. I have seen other people using the hotplate method before, so in principle it should work.
The MG Chemicals 4902P paste has the batch number 21248Y. Is the MFG coded into that by any chance? Maybe it was lying in a warm Amazon warehouse since 2021?

I would be incredibly thankful if anyone here had some ideas what I am doing wrong or what I could try next.
 

Offline asmi

  • Super Contributor
  • ***
  • Posts: 2884
  • Country: ca
Re: Hot plate reflow soldering and solder beads
« Reply #1 on: May 23, 2023, 08:38:43 pm »
Floating solder balls usually means that there is too much paste, and can indicate that it's held in a molten state for too long for the viscosity of the alloy. I don't know physical properties of bismuth alloys, but I find that the less viscous is the solder in a molten state, the more often I tend to see these balls all over the place (especially nasty if they are under BGA because they can short something eventually if not washed away). In my experience, I get those balls a lot more often with leaded paste (Sn-Pb) than with lead-free SAC (Sn-Ag-Cu). I'm not a materials expert though, so take it for what it's worth.
 
The following users thanked this post: tounho

Offline thm_w

  • Super Contributor
  • ***
  • Posts: 7835
  • Country: ca
  • Non-expert
Profile -> Modify profile -> Look and Layout ->  Don't show users' signatures
 
The following users thanked this post: tounho

Online tooki

  • Super Contributor
  • ***
  • Posts: 13386
  • Country: ch
Re: Hot plate reflow soldering and solder beads
« Reply #3 on: May 24, 2023, 07:19:16 am »
I would be incredibly thankful if anyone here had some ideas what I am doing wrong or what I could try next.
Your paste is fine. The problem is that you are applying way, way, WAY too much solder paste. I’m guessing at least 5 times as much as you need.

It’s really hard to apply paste in controlled amounts by hand using a syringe (there’s literally special equipment just to help with that), so either accept that you will need to rework it afterward, or just make your life easier and use a stencil.

What will help with syringe use is very fine conical tips. They require much less force than needle tips. (I’d recommend getting some 25ga (0.26mm) conical tips.)

Take a look at some photos and videos of solder paste applied by stencil. You’ll see it’s a thin layer applied to less than the full area of the pad. (A typical stencil is 0.125mm thick or less!) That should give you an idea of how little solder paste you actually need.
 
The following users thanked this post: tounho

Offline tounhoTopic starter

  • Newbie
  • Posts: 3
  • Country: de
Re: Hot plate reflow soldering and solder beads
« Reply #4 on: May 25, 2023, 08:31:40 pm »
Floating solder balls usually means that there is too much paste, and can indicate that it's held in a molten state for too long for the viscosity of the alloy. I don't know physical properties of bismuth alloys, but I find that the less viscous is the solder in a molten state, the more often I tend to see these balls all over the place (especially nasty if they are under BGA because they can short something eventually if not washed away). In my experience, I get those balls a lot more often with leaded paste (Sn-Pb) than with lead-free SAC (Sn-Ag-Cu). I'm not a materials expert though, so take it for what it's worth.

My hotplate has a slightly too high resistance because I forgot to untick the plating thickness box in the Saturn PCB Toolkit. Now it is ~1.4Ohms instead of the 1Ohms I wanted. Because it heats up very slowly above 180-200°C, I haven't tried "normal" solder paste yet and stuck to the 138°C one. I don't want to cause damage to components because it is spending too much time in that 180°C+ region.
I will order a new set with <1Ohms the next time I order stuff at JLCPCB and then I can try "normal" temperature paste like the SAC you recommend.
I will also try better application of the paste on the pads too.
Thank you very much!

https://www.eevblog.com/forum/manufacture/are-these-balls-because-of-soak-or-did-i-squirt-too-much-paste-(reflow-help)/

Thank you for the link. So it is just too much paste or paste not properly on the pads.

Your paste is fine. The problem is that you are applying way, way, WAY too much solder paste. I’m guessing at least 5 times as much as you need.

It’s really hard to apply paste in controlled amounts by hand using a syringe (there’s literally special equipment just to help with that), so either accept that you will need to rework it afterward, or just make your life easier and use a stencil.

What will help with syringe use is very fine conical tips. They require much less force than needle tips. (I’d recommend getting some 25ga (0.26mm) conical tips.)

Take a look at some photos and videos of solder paste applied by stencil. You’ll see it’s a thin layer applied to less than the full area of the pad. (A typical stencil is 0.125mm thick or less!) That should give you an idea of how little solder paste you actually need.

Okay, it's good to know my paste is fine. The syringe is for T3, so not very fine. I will try the conical tips with a smaller opening. My hands are very shaky so I have troubles applying paste perfectly without tools. I have avoided stencils since they are more expensive than a full set of PCBs. I will try them next time I order a batch of PCBs.
Thank you very much!

I have also thought maybe I can use some type of stand for the syringe. I have seen there are expensive, fully automatic paste application machines like this.
Would anything speak against putting the syringe in a stand like this https://www.amazon.de/-/en/Universal-Adjustable-Professional-Microscope-Magnifying/dp/B0797PPX8D? I could move the PCB under it to the right position using my microscope, lower the syringe a bit, apply paste and move on.
The syringe came with an adapter to put in a compressed-air hose. I don't know what to google yet to find what I am looking for, but I am assuming if the stand was too unstable and pushing on the piston moved it too much, I could put in some cheap, hand powered air pressure device there. Something like a hose going into a bellows/food-pump/hand-pump to push the piston without touching the syringe.

Thanks again to everyone replying, it was all very helpful.
 

Offline mcconkeyb

  • Contributor
  • Posts: 28
  • Country: ca
Re: Hot plate reflow soldering and solder beads
« Reply #5 on: May 25, 2023, 09:03:20 pm »
Repairs or new production?
Repairs might work best if you use a hot air station.
New production would work best if you get the solder paste mask with the bare PCB's, then you will get the right amount of paste on all of the pads.
 
The following users thanked this post: tounho

Offline tounhoTopic starter

  • Newbie
  • Posts: 3
  • Country: de
Re: Hot plate reflow soldering and solder beads
« Reply #6 on: May 25, 2023, 09:30:31 pm »
Repairs or new production?
Repairs might work best if you use a hot air station.
New production would work best if you get the solder paste mask with the bare PCB's, then you will get the right amount of paste on all of the pads.

New production. I'm just a hobbyist and this is for my home workbench. That's why I prefer it to be as cheap and low space requirement as possible. I got the hot plate because it is low space too, otherwise I would have tried modifying a pizza oven.
I understand that the stencil is the best solution. I will try that the next time I will order PCBs. I was avoiding it and trying doing it manually since stencils are as expensive as a set of PCBs and I would only use them once or maybe a few times.
Thanks a lot for the input!
 

Online tooki

  • Super Contributor
  • ***
  • Posts: 13386
  • Country: ch
Re: Hot plate reflow soldering and solder beads
« Reply #7 on: May 26, 2023, 06:18:18 am »
Your paste is fine. The problem is that you are applying way, way, WAY too much solder paste. I’m guessing at least 5 times as much as you need.

It’s really hard to apply paste in controlled amounts by hand using a syringe (there’s literally special equipment just to help with that), so either accept that you will need to rework it afterward, or just make your life easier and use a stencil.

What will help with syringe use is very fine conical tips. They require much less force than needle tips. (I’d recommend getting some 25ga (0.26mm) conical tips.)

Take a look at some photos and videos of solder paste applied by stencil. You’ll see it’s a thin layer applied to less than the full area of the pad. (A typical stencil is 0.125mm thick or less!) That should give you an idea of how little solder paste you actually need.

Okay, it's good to know my paste is fine. The syringe is for T3, so not very fine. I will try the conical tips with a smaller opening. My hands are very shaky so I have troubles applying paste perfectly without tools. I have avoided stencils since they are more expensive than a full set of PCBs. I will try them next time I order a batch of PCBs.
I do recommend storing your paste in the fridge. It'll last longer.

FYI, your T3 paste is ideal. Finer paste, while necessary for extremely small stencil apertures (for pads far smaller than anything you'd ever consider doing by hand), is actually more susceptible to bead creation, and its shelf life is shorter.

Shakiness: if you can get a decent optical (not digital) microscope to work under, you will find that you shake far less.

Stencil cost: I understand. For a one-off board that can be soldered by hand, I often skip them too. But for anything where you either a) need to make a bunch of them, b) have components that cannot be hand-soldered easily, or c) where you want to minimize the risk of assembly problems, a stencil is a big help.

Something I have done in the past to save cost on the stencil (especially for larger boards, where the sheer size of the stencil ends up substantially increasing the shipping cost from China) is to create a partial stencil only. For example, I had a board with three identical channels, so I just ordered a stencil for one channel and used it three times.

Note also that the openings in the stencils need to be smaller than the pads. 15% reduction is typical, but for stencils used by hand -- that is, not in a stencil printer -- a bit of paste can often squeeze under it, so I've had the best results with 20-25% reduction. Or you can order a thinner stencil if your manufacturer offers it. Large hidden pads (like ground/thermal pads under QFNs) need even less paste, often around 40% reduction.

Most PCB manufacturers will reduce the openings by 10-15% by default. If you want them reduced more, either tell them to do that, or tell them to not reduce them at all, and reduce them yourself in your PCB layout.

Another approach I've used with decent results is to apply paste by hand, but to clean it up carefully before reflowing. I use a curved scalpel blade (with no force applied) as a scraper to pick up and remove excess paste. What is quite important, possibly more important than the absolute amount of paste, is that the amount of paste be even on the pads of a part, so that the part doesn't tombstone or get pulled away once the solder melts.

For example, on chips with entire rows of pins (SO, SOIC, QFP, QFN, etc), I will lay down a "noodle" of paste across a whole row of pins. I start squeezing on the board a cm or two before the first pin, then lift the syringe tip so the noodle is actually squeezed out in the air, which keeps it nice and straight, and then lay it down across the entire row of pins, ending a few mm after the last pin of the row. Starting and ending outside of the row of pins keeps the amount much more even, since the ends tend to have blobs. Then I use the scalpel to "trim" the noodle to one pin's width longer on each end, and then scrape the excess paste away.

You can similarly use a scalpel blade to remove excess from the pads of resistors, caps, etc.

With that said -- because boards reflowed with syringe-applied solder paste often need lots of rework with the soldering iron anyway -- if possible, I just solder them with the soldering iron to begin with.

Invest in some soldering iron tips for rework. For this type of work, I really like PLCC blades (often called a "knife" tip), since their edge can fit into nooks and crannies to pull away excess solder. I've attached a picture of the PLCC blade my soldering iron uses, so you can see the tip shape I mean.

Also make sure you have a good gel/tacky flux, which makes a world of difference during rework. For example, MG Chemicals 8341 or Chip-Quik SMD291NL (my favorite so far). A liquid flux pen is also handy on occasion.
 
The following users thanked this post: ebastler

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2869
  • Country: us
Re: Hot plate reflow soldering and solder beads
« Reply #8 on: May 26, 2023, 03:00:57 pm »
I would avoid the Bismuth pastes.  Components are designed to do 230C reflow with lead free pastes.  But, the genius is all in the flux!  After using GC10 paste (available in the US under the Loctite brand, elsewhere under Henkel) it has solved all these issues.  The stuff is totally amazing.
Jon
 

Offline Solder_Junkie

  • Frequent Contributor
  • **
  • Posts: 509
  • Country: gb
Re: Hot plate reflow soldering and solder beads
« Reply #9 on: May 26, 2023, 07:17:25 pm »
As another “kitchen table” home brewer, I too have suffered from solder balls, tombstoning and parts getting moved when using a hot air blower.

I have returned to hand soldering, which doesn’t suffer from the above issues. This video shows the technique:
https://youtu.be/eZtPR8_x0nc

Incidentally, why use 138C alloy? From what details I can source, nowhere is it recommended, or used, for PCB assembly work.

SJ
 

Online tooki

  • Super Contributor
  • ***
  • Posts: 13386
  • Country: ch
Re: Hot plate reflow soldering and solder beads
« Reply #10 on: May 27, 2023, 02:19:13 pm »
I would avoid the Bismuth pastes.  Components are designed to do 230C reflow with lead free pastes.  But, the genius is all in the flux!  After using GC10 paste (available in the US under the Loctite brand, elsewhere under Henkel) it has solved all these issues.  The stuff is totally amazing.
How is the availability now? With Henkel having sold the solder paste business to Harima, the whole supply chain for it got shaken up. See https://www.eevblog.com/forum/manufacture/henkelloctite-gc10-solder-paste-source/

(Plus, I don’t think anyone ever carried it in hobbyist-friendly sizes. A 500g container would be something like 200 years worth of paste for my home use…)
 

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2869
  • Country: us
Re: Hot plate reflow soldering and solder beads
« Reply #11 on: May 27, 2023, 03:47:09 pm »
How is the availability now? With Henkel having sold the solder paste business to Harima, the whole supply chain for it got shaken up. See https://www.eevblog.com/forum/manufacture/henkelloctite-gc10-solder-paste-source/

(Plus, I don’t think anyone ever carried it in hobbyist-friendly sizes. A 500g container would be something like 200 years worth of paste for my home use…)
Newark seems to be the only distributor that has some in stock right now.  Yes, they only sell it in 500g minimum jars or syringes.  But, that is only a little more than $100 and it lasts a LONG time.
Jon
 

Online tooki

  • Super Contributor
  • ***
  • Posts: 13386
  • Country: ch
Re: Hot plate reflow soldering and solder beads
« Reply #12 on: May 27, 2023, 05:19:50 pm »
I need to find some people who want to split a tube. Cost aside, I can’t imagine I’d use 500g before it went bad. It may be long lasting, but it won’t last forever. I’m still only halfway through my first 35g syringe of paste I bought in 2015 or so, so when I said 500g is a 200 year supply for me, I’m not exaggerating! 😂
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf