Printed Circuit Board Assembly & PCB Design SMT Electronics Assembly Manufacturing Forum

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Waves soldering: not enough solder remaining on pads

Views: 9959

Hi all, We run an old Seho 1135C wave solder with a 63/37... - May 21, 2017 by Zac  


Zac

#77610

Waves soldering: not enough solder remaining on pads | 21 May, 2017

Hi all,

We run an old Seho 1135C wave solder with a 63/37 pot at 250°C. It has been running well for a couple of years but we have noticed lately that all of the larger pads (terminal blocks, transformers, etc.) require additional solder to be applied by hand. I was not made aware of the extent of the issue until recently and upon closer inspection I feel that there is not enough solder applied to any of the pads.

I did some observation of the wave soldering today and tried a few adjustments (i.e. conveyor speed, preheater temps) but to no avail. We are using the delta wave only and are spraying flux by hand before loading the pallet into the machine.

Please see the attached photos (and please excuse my poor photography). The solder fillet does envelop the entire component pin and pad, but it tends to not want to go all the way to the edge of the pad. These are single sided PCBs with solder mask and HASL finish.

Next time we run the machine (it only gets used around once a fortnight) I will run a profiler through, as I suspect that the preheaters are not reaching the required temperature (despite what it says on the controller).

Does anybody have any other suggestions as to the cause? Thanks in advance!

Attachments:

reply »

#77621

Waves soldering: not enough solder remaining on pads | 22 May, 2017

There is one lead that is a total miss. I think you are too low with the wave there - you are barely touching the board. I would say first look at your wave height. If you have the possibility to change the angle of the wave(if there is a backplane there that could be controlled usually by a screw in front of the pot). If you don't have this option, try more RPMs. Also some of these leads look very long - you might try to pre form these in advance.

reply »

#77624

Waves soldering: not enough solder remaining on pads | 22 May, 2017

I agree. Trim the leads before insertion, then raise the pot height. With the information given, this is most likely your solution.

Our machine has a dam wall on the rear of the solder pot that raises the height on the solder pool. I would make sure that that dam wall is up to a height close to the solder nozzle.

Your conveyor speed may be too fast as well. If you have a profiler, use it to make sure that the whole assembly is getting up to temp. If the conveyor speed is too fast it would affect the solder joints.

If you just cannot find anything else that is the problem, you can send out a solder sample for analyses. The manufacturer of your solder should offer offer it as a service. They may even do it for free.

reply »


Zac

#77625

Waves soldering: not enough solder remaining on pads | 22 May, 2017

Thanks for the replies.

The single missed lead is a bit odd - we usually have an entire area missed and that's due to the board not sitting properly in the pallet fingers. I figured in this case it was likely due to poor fluxing?

We have always struggled with wave height: the pump is running at 100% and we only barely get the required 20mm of contact patch. All of the other manually adjusted settings for the nozzle mounting are as per the manual and any higher would give us problems as we're only getting about 4mm between the top of the nozzle and the PCB. I tried slower conveyor speeds and they made no difference.

We are concerned about trimming the lead lengths any shorter (we do use a forming/trimming machine) that they might jump out of the board during handling or when contact with the solder wave is made.

The wave angle and conveyor angle seem very fixed so there's not a lot that can be done there. I'll go through and check all of the heights again. Maybe add some more solder to the pot as well (even though it seems to be within spec).

reply »

#77626

Waves soldering: not enough solder remaining on pads | 23 May, 2017

hello Zac,

if you are running pump at 100% and there is not enough wave then you will probably need to empty solder bath and clean everything. also check pump parts for tear and wear. eached solder inlet can make a lot of problems when it comes to wave level.

reply »


Zac

#77628

Waves soldering: not enough solder remaining on pads | 23 May, 2017

Thanks Bukas. I've actually gone to a lot of trouble over the past 24 months to completely clean, rebuild and adjust the pump and nozzle assemblies and even through all of that we've still never been able to get theoretical maximal wave height from the delta wave.

One thing that just occurred to me to check is the oil around the pump shafts - I suppose some excessive friction there may have caused the wave height to drop lower than it had been and cause our recent issues?

reply »

#77630

Waves soldering: not enough solder remaining on pads | 23 May, 2017

the problem I had with pump was that turbine intake was eached by alloy. when I repaired intake, it was a irregular hole about 32mm in diameter, mechanics welded piece of plate with a 24mm hole (that was a factory size) we had to reduce pump speed from 95% down to 51% and wave became more stabil.

reply »


Zac

#77631

Waves soldering: not enough solder remaining on pads | 23 May, 2017

Oh wow that is very useful insight! I will remove and check next time we fire up the machine!

How did you know the factory size of the intake hole?

reply »

#77632

Waves soldering: not enough solder remaining on pads | 23 May, 2017

i had 2 pumps xD also, propeller size can give you a clue. do some research on turbines of any kind, its same principal.

you can have someone check it for you on same machine, i'm sure someone will help.

reply »

#77676

Waves soldering: not enough solder remaining on pads | 5 June, 2017

Hi Zac,

In looking at your photos, you may not be having a problem with your machine, but with the solderability of the PCB. I've zoomed in to an area that looks to me like the solder has de-wet from the pad. The leads appear like it almost has a fillet, probably because the leads are solderable, but there's something going on with the PCB.

Attachments:

reply »


Zac

#77677

Waves soldering: not enough solder remaining on pads | 5 June, 2017

Hi Steve, I agree with your assessment, it would certainly seem like the pads are de-wetting. But what could be the cause of this? Contaminated solder? The boards are new out of vacuum sealed bags (with desiccant sachets), so I can't see it being the HASL finish). Insufficient preheating, perhaps?

reply »

#77678

Waves soldering: not enough solder remaining on pads | 5 June, 2017

Hi Zac,

My 2 cents:

We had similar problems on a board in the same situation (tin-lead HASL, very long leads, low pot height, high pump RPM). Trimming the leads shorter, moving the pot closer to the board, and dropping the pump speed down to sane levels took care of the problem.

The operators at some point decided that since our pot height was adjustable, they could leave the leads as long as they wanted to eliminate the risk of leads popping out of the holes (these boards must go through mass lead trimming, so that wasn't a consideration). I drew the line and gave them a maximum lead length.

After making this change, our solder skips/poor wetting problems went away with no noticeable increase in the amount of leads popping out of the holes. Might be worth a try, especially if your solder pot height is easily adjustable.

I'm guessing that most wave machines have a little unevenness across the width which is exacerbated when you're running very high pump speeds.

reply »


Zac

#77679

Waves soldering: not enough solder remaining on pads | 5 June, 2017

Thanks for the suggestions, Caurbach.

We do have the occasional lead jumping out even at this length, hence why we've been hesitant to cut them shorter. I have the wave nozzle to PCB distance set as per the machine instructions (4mm I think?). I certainly don't get the solder wave "depth" that the manual indicates at full speed. Can you recommend what distance you use from nozzle to board?

I have in the past tried overdriving the solder pump motor with a higher voltage and that certainly created noticeable turbulence in the wave that I could imagine creating problems. At the standard full speed though it seems pretty consistent.

reply »

#77681

Waves soldering: not enough solder remaining on pads | 6 June, 2017

Hi Zac, was that board finished with tin/lead HASL or leadfree/HASL? Also what type of flux were you using? Just because a board is fresh out of sealed packaging doesn't necessarily mean that it's going to solder well. The picture that I've attached is from a brand new board. It was supposed to be finished with lead free HASL but the board house screwed up and it wound up to be just nickel plated. Unfortunately we didn't realize this until we ran one assembly. The photo is one of the other boards where we just printed solder paste and ran it through the reflow oven. The whole board dewet...

Attachments:

reply »


Zac

#77684

Waves soldering: not enough solder remaining on pads | 6 June, 2017

Hi Steve, I just confirmed with the manufacturer and the HASL is tin/lead (or it is supposed to be, at least). Any suggestions for how to test for nickel vs tin/lead?

We're using an alcohol based no-clean flux from Alpha.

reply »

#77688

Waves soldering: not enough solder remaining on pads | 7 June, 2017

Hi again Zac,

Nickel appears a little darker and won't have as much of a deposit as it is plated on. But if you want a quick test you can buy a little kit from Amazon that tests for nickel. Some people are allergic to Nickel https://www.amazon.com/Athena-Allergy-Nickel-Solution/dp/B00175TYXQ

reply »

#77697

Waves soldering: not enough solder remaining on pads | 8 June, 2017

This may not be a solution to your problem, but something from my tuner car days jumped into my head. If you are running at 100% and not getting the height you need, change out one (or both) of your pulleys. A bigger pulley at the motor side, or a smaller one at the pump side will give you more pump revolutions for the same motor revolution.

Like I said before, this may not be the solution to your problem, but it bothers me that you are already at 100% and not happy with the wave height, and I tend to trust my gut.

reply »


Zac

#77700

Waves soldering: not enough solder remaining on pads | 8 June, 2017

Thanks for the suggestion, Steve, I'll look into sourcing a test kit here in Oz! I did however talk to the guys doing the touch ups of this batch of boards and they are sceptical that it's nickel: they have had to solder to nickel in the past and reckon that it is noticeably harder.

reply »


Zac

#77701

Waves soldering: not enough solder remaining on pads | 8 June, 2017

SuMoTe, that is a fair suggestion, but I believe will be a non-starter. The solder pump motors are actually 40V motors being run from a PWM 24V supply. I did one time try hooking up to a separate power supply up to 40V but all this seemed to do was massively increase turbulence in the solder wave without having sufficient impact on the wave height.

I agree with your assessment regarding already being at 100% - as suggested earlier in the thread I'm going to pull the pump nozzles early next week and check the pump inlet hole for wear.

reply »


Zac

#78549

Waves soldering: not enough solder remaining on pads | 31 July, 2017

Hi all, it has been a bit of a delay but I've finally had a chance to do some further investigation on this.

First off, we spend some time to clean the solder bath using Kester Solder Saver and scooped out a lot of crap. We then added about 5kg of brand new 63/37 solder bars.

The solder nozzles and pumps were both removed so I could investigate the pump inlet. Please see the attached photo. The inlet holes were the same size on both pumps and perfectly round. There was no wear evident.

The second photo is of a spare impeller that came with the machine.

After reassembly we ran two different products through the machine. One was the same board that I shared photos of to start this thread and the results were the same. The second board was a similar product and I noticed that although the pads seemed to wet much better than the other board (this is referencing Steve Gregory's comments above), we still had insufficient solder.

Interestingly, as per the third photo, on many of the larger pads there is much more solder on the trailing edge.

So at this stage bar any issues with the solder pump impellers themselves I've ruled out every possible issue regarding the wave height: - Solder pot level correct (less than 10mm from top of bath when pumps are off). - Running motors at a higher speed than the designed maximum simply increases turbulence, not height. - The pump inlet hole seems correct. - Pump shafts covered in oil. - No slipping of pump drive belts. - Delta-wave nozzle rear baffle adjusted so that there is only a minimal amount of back flow (most of the solder flow is towards the PCB).

And we're still not getting enough solder deposited on pads - the solder fillet is clean but just very lean.

Any further suggestions would be very appreciated!

Attachments:

reply »

#78554

Waves soldering: not enough solder remaining on pads | 1 August, 2017

Zac,

how about the following:

1. Thermal profile for the flux 2. Is your wave leveled? - If you see better joints on one side, you should probably level the machine 3. Just for elimination, do you have another flux, that you could spray on a board and run it through?

reply »


Zac

#78567

Waves soldering: not enough solder remaining on pads | 1 August, 2017

Thanks for the suggestions.

1. By this are you referring to the flux specifications or the actual profile of the machine?

2. Yes, the entire machine, the solder pot and the wave nozzles have all been levelled. The better joints on one side are perpendicular to the direction of movement.

3. Yes, I think we have some other flux we can try. We manually spray flux on the PCBs though so I don't see how it can get contaminated? We do thin it down as necessary.

reply »

#78603

Waves soldering: not enough solder remaining on pads | 3 August, 2017

Looking at picture 20170801_095540.jpg, the leads look pretty long to me. This would cause you to have to lower your solder pot so that they don't get hung up. Have the leads cut to ~0.150" (0.200" max) long and then raise your pot up.

reply »


Zac

#78604

Waves soldering: not enough solder remaining on pads | 3 August, 2017

That may just be an optical illusion. I've just measured and they're cut to just over 4mm (0.157"). The pot is at the level specified by the manufacturer (5mm from the bottom of the PCB if I remember correctly).

reply »

#78609

Waves soldering: not enough solder remaining on pads | 4 August, 2017

Zac,

for the profile I mean both - if your machine temp profile is inside the process window of your flux. Trying other fluxes might give you some answers.

reply »


Zac

#78621

Waves soldering: not enough solder remaining on pads | 6 August, 2017

OK thanks. I had assumed that since the soldering was generally smooth and we were having no issues with bridges that the flux would be operating fine. But there really isn't much left to check so I'll get my profiler fixed (the thermocouple has been damaged) and get some useful data.

I'll report back with the results!

reply »

#78637

Waves soldering: not enough solder remaining on pads | 9 August, 2017

I recommend examining the problem pads and a representative bare board using SEM/EDS to determine if the problem is related to a board solderability issue. I can support this type of analysis (www.semlab.com).

reply »


Zac

#78648

Waves soldering: not enough solder remaining on pads | 9 August, 2017

Thanks Ed. I'll keep that in mind but at this stage I'll rule out the other options such as flux and heat profile first...

reply »

Manufacturing Software

Metcal soldering rework