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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Wave soldering profiling

Hi, I currently work on SMT department to uptimize all pr... - Jun 10, 2003 by

Frank

#24751

Wave soldering profiling | 10 June, 2003

Hi,

I currently work on SMT department to uptimize all process. I've just selected the solder paste after testing 15 different products. To be able to do that, I had to create a procedure to evaluate each paste and to be sure that my profil was perfect. After doing that, I solved the problems of tombstoning and solder balls also.

Now I have to attact the wave soldering process. We currently use all information that solder wave supplier give me but I want to check if every temperature are respected correctly. Some times, we have some soldering problem after this step and I think the temperatures are not efficiency.

I've got a profiler for the reflow process so I'll be able to check the profil in the wave. I was wondering if there is also a profil to respect for wave soldering? Is there a document I can refer to get this information?

Thanks for your help

Frank

reply »

#24758

Wave soldering profiling | 10 June, 2003

Develop your recipe wave soldering exactly as you developed your recipe for reflow soldering. The only difference is that instead of using your paste supplier's recommendations as the starting point, use your flux supplier's recommendations as a baseline.

Sldrng, Wave, Setup, Profile Preheating: The temperature of the board (component side) during preheating should go as quick as possible to 45 �C (with maximum slope of 2 �C/Sec) and just before soldering an temperature of 85 �C. Wetting time: The time between the moment of first contact (between the parts to be soldered) and the solder, and the moment that the solder in the joints starts solidifying. Dwell time: The time between the moment of first contact (between the parts to be soldered) and the solder, and the moment of last contact with the solder. (This time should be between 2 and 4 seconds.) Soldering time: The time between the board actually contacts the solder and the onset of solidification. (Dwell time + approximately. 5-10 Sec.) Solidification time: The time solder is getting solid on the bottom of the board. Cooling time: Forced cooling after soldering reduces the maximum temperature reached in the areas on the component side of the board. However, forced cooling directly after solder bath, may cause unequal contractions, which will initiate cracks and therefore not advised. If cooling down is needed because of the temperature from the carrier is rising to high, the best place will be before or in the lift itself.

Good luck

reply »

Frank

#24763

Wave soldering profiling | 11 June, 2003

Thanks again Dave for your information.

Merci beaucoup Frank

reply »

#24764

Wave soldering profiling | 11 June, 2003

Il ne pas de quois.

Two most important parameters are: * Solder temperature * Dwell time

Immersion depth [1/3 to 2/3 of board thickness] has the least impact of any variable in wave soldering.

reply »


CAL

#24768

Wave soldering profiling | 11 June, 2003

Frank-In addition (Dave gave some good scoop) we use a Pyrex Glass board to run through the wave. This is useful as you can see through the board .....you can watch the flux coverage, preheat, and Wave (be it Chip, lamda, or??) as it goes through the process. I like it for the Wave part as it allows you to see how parallel the board is to the wave.

Lastly, I do not know what type of flux application station you are using but if it is not a sealed pressurized unit you may need to also monitor the chemistry ratio. Sorry to add stuff but it is part of that process you want to monitor.

Some good sites:

http://www.bobwillis.co.uk

http://www.smtinfo.net/Db/_Wave%20Soldering.html

www.empf.org

I hope this helps

Cal Driscoll Communications Test Design, Inc. www.CTDI.com

reply »

#24776

Wave soldering profiling | 11 June, 2003

Frank,

ECD (WaveRIDER) and Datapaq (Optimiser) have wave solder machine process control devices. They will check the following:

1. Conveyor speed 2. Preheat Temps 3. Solder Temps 4. Wave Dwell/Contact Times 5. Wave Heights 6. Waves Parallelism

Go to http://www.ecd.com or www.datapaq.com

If your profiler is an ECD SMG, the investment for their wave machine gizmo is shorter money than DQ.

Happy waving!

reply »

MA/NY DDave

#24783

Wave soldering profiling | 11 June, 2003

Hi,

All seem good answers and food for thought. Additions you probably know.

Flux Supplier Application Notes/Data Sheet. Many of them have run experiments to tell you what works for their flux on various machines. They might even give you options you can use for an experimental matrix (DOE is the usa popular phrase)

The Solder Pot Supplier's Application Notes.

YiEngr, MA/NY DDave

This message was posted Add this forum to your site! Click to learn more. the Electronics Forum @

reply »

MikeF

#24817

Wave soldering profiling | 12 June, 2003

Hexacon sells the LevChek, it is a special tempered glass plate for checking a wave solder machine. They make different sizes, so you can get one that is close to the maximum width of your wave. They also have grid lines etched on them so you watch it go over the wave solder and get an idea how even the solder contact footprint is, and how wide it is. If you know your board will make contact with solder for a 3 inch length, and you want it to make contact for 3 seconds, you know your conveyor speed should be 5 ft per minute, plus or minus a small amount.

I had one machine where I had to adjust one rail lower to get an even width of solder contact across the board because the wave nozzle was not quite level. I knew I was having problems along one edge, and the glass plate showed me why. The plate can also be used to see how even your flux is applied.

The drawback that even though the glass is treated to withstand the heat, when it is hot it is fragile. If it is bumped on something it will break. I've cracked a corner off one, and cracked a second one in half, and both times it was a light tap that did it.

Mike

reply »

#24819

Wave soldering profiling | 12 June, 2003

Try using ECD Wave Rider. This system will helps you on taking the parameter of the machines. It will helps you. Let me have your eamil ID so that I can send some information to you with an attachment.

Regards Danny Hui

reply »

Dreamsniper

#24847

Wave soldering profiling | 17 June, 2003

Hi,

All these guys have provided you good solid infos already! All I can say is that check your wave solder machine's mechanical set-up too...and the Molten Solder Level in the bath must be maintained at an accurate level/height specially after dedrossing as this will affect everything you have achieved if it is not in a constant level during every operation.

regards,

reply »


CAL

#24870

Wave soldering profiling | 18 June, 2003

Frank- Dreamsniper touched on something I totally forgot about....DROSS.....This is a subtle way for $$$$ to be lost.

If it is silver and Shiney...it does not belong in the dross recycle bin. That is good Solder being tossed.

You can also add into the process reclaimation.

Cal

reply »

#24878

Wave soldering profiling | 18 June, 2003

Rules of thumb :-

1. Never over flux - not necessary and waste's money. Never top flux neither 2. Top Board temp. around 70C - 105C or there about's pending what laminate you use. Obviously the more flux you apply the higher heat required/denser PCB more heat etc. 3. As temperature is not really that super critical for most new fluxes it is extremely important the wave is level across and the correct dwell time. If unlevel wave, bridges will occur with a noticeable build up of flux down edge of low side of wave. If dwell time too low then flux build up on back edge of PCB and bridges across PCB. 4. Watch out for waxy, greasy residue's this is not flux related to new no residue fluxes and a red herring to a board issue with resist's extremely common. 5. Dont blame blow holes on plating - look at HASL flux contamination as this is commonly the cause but rarely blamed. The tools we use for wave riding is a company called Circuit Master Design Ltd ProWave, actually produces 2 D picture of both chip and main wave along with dwell times/length brilliant bit of kit well recomend it.

reply »

Vic Banuelos

#24906

Wave soldering profiling | 19 June, 2003

Hi Frank In adition I recomend you reading Chapter 12 from SMT principles and practice by Ray P. Prasad, Chapman & Hall. You'll find a good guide for profiling.

Good luck

reply »

david coombes

#27987

Wave soldering profiling | 11 April, 2004

hi please could you tell me where the glass you use came from

reply »

#27998

Wave soldering profiling | 12 April, 2004

Levchek [ http://www.hexaconelectric.com/levchek.html ] is sold by Hexacon and most of the electronic assembly supply houses [eg, Technitool, HMC, Jensen, etc]

Alternately, you can purchase heat tempered glass from your local glass store, but it will not have the nifty grid lines. It will be just as fragile, but less expensive than a LevChek.

reply »

Ron Herbert

#28002

Wave soldering profiling | 12 April, 2004

Correct preheating of the assembly is critical.

Check with your flux manufacturer regarding the recommended topside pcb temperature as it enters the first wave. It is important that this temperature be achieved without burning the flux on the bottom of the board. A topside tunnel, or perhaps even topside preheater, may be necessary in the case of large ground planes or multilayer boards.

Bear in mind that the amount of heat required to meet this temperature will increase as the flux gets thicker. Therefore be sure to maintain the flux within the recommended density as reommended by the manufacturer.

Also, as the solder builds up in contaminates its melting temperature, and therefore its flow characteristics, change. Have your solder analyzed on a regular schedule.

reply »

Vince Whipple

#28070

Wave soldering profiling | 14 April, 2004

Frank, Your initial question about temperature profiling is a critical point to a good wave solder process. Especially if you have a No Clean flux. As recommended by many in the thread above, get the Process Data Sheet for your flux (this is not the MSDS). There will be a recommendation range for topside temperature on the sheet. ( eg. 190f to 220f, depending on the flux) Attach the thermocouples at different areas on the topside (pick areas of low, average and high thermal mass). Adjust temperatures of the preheats such that the average mass areas fall within the middle of the range and the low and high mass areas are still within the range limits. If you have smt on the board, make sure you follow the same recommendations for your smt components for temp rise. Most places settle somewhere between 2-3 deg c/sec. check your component specs. (you may have to play with the preheats along with the conveyor speed ). What flux are you using? Are your boards HASL or OSP? How are you applying the flux? The above 3 questions will have an impact on further recommendations to hone in the process. I have a wave solder process sheet that assumes spray fluxing. It lists solder problems with possible causes by likelihood. It may be in the SMTnet archives or you can contact me to email it. best of luck. vwhipple@sono-tek.com Vince

reply »

Vince Whipple

#28072

Wave soldering profiling | 14 April, 2004

Is this thread still live? The starting post date is from last year. Frank, you still there?? Vince W

reply »

Ken

#28079

Wave soldering profiling | 14 April, 2004

Just a thought....

Some Fluxes require temperatures upto 260F (127C).

Are you using water based flux or solvent (alcohol). I have noticed that many wave machines are NOT capable of preheating adequately when using water based materials. This can be compounded with selective pallets.

Wave dwell can be limited by flux selection. NC flux tend to have shorter dewll times (2-3 sec) where as OA's can survive out to 4+ seconds.

Is your flux dual wave capable. Low/Ultra-low solids NC's tend to NOT be dual wave capable. You'll know it by the icicle's and solder flags.

What alloy are you using? I think everyone is assuming Tin/Lead. All the above will be different with LF.

Oh, and finally you will find that wave flux can have a very large process window (especially high solids OA). NC's tend to be smaller requiring tighter process control.

I have one product that uses a high solids OA. Spec sheet calls out 190-220F target preheat temp. This 7lb, .200 thick backplane reaches a preheat temp of 300F. Solders like a dream! Keep in mind that your start point is the specification, but ultimately you will need to determine you process window in YOUR process.

reply »

#28080

Wave soldering profiling | 14 April, 2004

Vince

Frank is probably not still here. Management at SMTnet chooses to keep all threads alive, turning a deaf ear to forum user requests to close threads after a reasonable period of time.

The person that reopened this thread did so in all good conscience. Unfortunately you, Ron, and Ken are responding to the air. Good information, just not very useful, at this moment.

Thanks for responding.

reply »

Frank

#28126

Wave soldering profiling | 19 April, 2004

WOW...

Hi Dave and everybody. I'm very surprise that this thread is still alive. From all your comments and your good tips, I was able to confirm that my current profil is good. Concerning the small problem of soldering I still have, I've decided to get more information concerning the direct temperature of the component pins to solder. Using an High tempersture solder, I place thermocouple directly on the pin juste above the top side of the PCB. I use DataPack to get the exact temperature of the pin and the maximum temperature ( directly in the wave ) is 181.5degC. So what I think is ... if you pin do not reach 183degC at least, is it possible to get a good solder quality.

Thanks again for all you comments Frank

reply »

Bruce

#28449

Wave soldering profiling | 6 May, 2004

I am glad this thread is still alive as I have a concern. I have just entered the wave soldering business after many years in smt. After wave I am getting icicles or flags on my component leeds. My temps seem to be all with in spec as does my NC flux. Any one have an idea on how to get rid of these icicles?

Thank

reply »

Ken

#28451

Wave soldering profiling | 6 May, 2004

tell us a your setup parmeters.

What is your dwell time? What flux are you using?

reply »

Greg York

#28456

Wave soldering profiling | 6 May, 2004

Watch your dwell time especially if larger components. Problem is normally associated with lower temp as well as insufficient flux coverage with larger massed boards or ground planes. Try to hand spray a board to check that fluxer operation is OK. If long Leads ensure flux choice will last longer in the wave. Really need more information to try to give some idea's What machine? What flux? Is it dual wave? Type of PCB? Lead Lengths? Method of fluxing?

reply »

#28457

Wave soldering profiling | 6 May, 2004

Here is some more info with regarding my Wave.

1. PCB .065" thick 2. leed length .20" (.135" protrution) I know this is high but we use many thousands and cutting each is not practical. 3. no clean flux by way of foam fluxer 4. wave machine is a HOLLIS eng. +/-20y old (but does work well)

This machine has been sitting for a while and did not have a well performing wave, as in hieght and flow. I took the wave and pump apart, cleaned and once back together performed like new.

My problem is before the clean up all my test boards were great, after the wave restoration my test boards all have icicles.

I did notice reduced icicles with greater speed on the conveyor but then the board temp had become too low. I'm stuck for an answer....

reply »

#28458

Wave soldering profiling | 6 May, 2004

It does sound like your leads are to long, speeding it up will reduce volume on lead and most will wick up the joint but when slow excess solder is left as long leads will have little flux left to aid drainage when it leaves the wave, thus giving a spike. Only real options you have are

1. Increase solder bath temp(Very unusual) 2. Use low solids Rosin flux, rosin is not volatile so will stay around the solder longer in the wave a will aid drainage. Normal No Cleans/Residue's will volatise off with long leads and cause a spike if left in the wave too long.

reply »

#28459

Wave soldering profiling | 6 May, 2004

sorry one other thing are you sure your getting 2 -3 seconds dwell after you reset the wave? Or is the dwell much longer?

reply »

Ron Herbert

#28470

Wave soldering profiling | 7 May, 2004

Bruce,

Using a foam fluxer with no-clean flux is a very touchy situation. It is imperative that the specific gravity of any flux be maintained within the manufacturers recommendations. If it is not, you will get icicles and bridging. Maintaining the specific gravity is difficult with no-clean fluxes because the specific gravity is so low. Spray fluxing is generally recommended when using a no-clean flux. If you can't afford a spray fluxer you should check the specific gravity hourly.

Also, the Hollis machines normally used calrod preheat elements. This can be tricky when using a no-clean flux. Be careful not to overheat the bottom of the board when trying to get the topside to the proper temperature. The flux on the bottom of the board should be slightly tacky as it enters the wave.

You mentioned the machine had been sitting around for awhile. Did you have the solder analyzed to see if it was within specs? Are you sure all the preheat elements are working?

reply »

bruce @ knoll

#28472

Wave soldering profiling | 7 May, 2004

I did have my solder analyzed and it is clean.

The Hollis preheat coils are a bit hard to control but I have measured the board top side and I am getting a temp of 105c just before wave contact.

I would like to move to a spray fluxer but I am going to have to stay with the foam fluxer for the time being.

My dwell in the wave did increase once the wave pump was reinstalled but I have it back to 2-3 sec contact.

I have since run some other boards with smaller leed extention and I am getting icicles or flags on the smaller leed lengths as well. So this problem it is not limited to just long leeds.

I am thinking I may have to flush my flux tank and start all over.??

reply »

#28474

Wave soldering profiling | 7 May, 2004

Manage the speed of the board and the speed of the wave to be nearly same at the point where the leads leave the wave.

reply »

bruce @ knoll

#28476

Wave soldering profiling | 7 May, 2004

I will try that next. I did speed up the wave pump motor with a larger pully(20% more speed) This was done at the time to try to increase the size of the wave because I had component leeds touching the wave pot as they passed by. Due to the wave hight being so small. After I cleaned up the solder pot and wave pump impellers though the wave hieght increased dramaticly. And now perhaps the wave is flowing too fast. I will go back to the original pully and that should slow down the wave speed.

My concern with matching the conveyor speed and wave speed is that I will have to speed up the conveyor (already at 4.5' per min)and in turn loose board temp. There for I will have to increase the preheat, which is already hot.

reply »

Ron Herbert

#28479

Wave soldering profiling | 7 May, 2004

Bruce,

You didn't mention whether you had checked the specific gravity of the flux. Also, most flux manufacturers recommend getting rid of the flux after 40 hours of use.

reply »

#28555

Wave soldering profiling | 12 May, 2004

woow, looks like I need to go back on the road... you all have great info. but what happen to asking what type of machine do they currently use, what is the configuration of the machine, what type of flux are we using, are running single sided or double... man ol man....some of our helpers need to slow down, take a few more years to understand and then give advise... if you email me direct I will send you a powerpoint presentation that will answer all your needs free, not to all of you, just this poor soul... have a good one, oh and by the way, been doing wave as a sr. process engineer and consultant for 25 yrs. and supported one of the largest manufacturers of wave and reflow in the world... just a little food for thought, but not that I am beyond learning from all... have a good one..hope to help

reply »

#28560

Wave soldering profiling | 12 May, 2004

May'be after all these years your eye sight could be failing,other postings including mine asked board type and machine type and set up etc etc. Oh well

reply »

#28562

Wave soldering profiling | 12 May, 2004

having a bad morning there Gregory.... take it easy, where you from? who's company? Do you rep this firm or is this yours? I surely did not mean to disrepect your input, but I am sure there are things you need to learn that are not in the operators manual... be nice, we all have something to share and I was just adding some humor.. by the way, eyes are still good, not that old, just started back with the real engineers that hold most pat. in the developement of Hollis and Electrovert.... got lucky early thats all....

reply »

#28564

Wave soldering profiling | 12 May, 2004

Hi Solderdoctor! Days been great especially as it's nearly over. Don't see many of those machines over here anymore due to healthy competition from local supplier. Only been in industry myself nineteen years so spring chicken I suppose. Mind you sold tens of thousands of litres of flux and tonnes of solder and developed our new lead free wave solder alloy so counts for a little I suppose.Travelled extensively commissioning flux and solder in various machines seen many problems and surprised sometimes at the cure as we learn new things daily. I would be only to pleased to check out your powerpoint presentation and give it the once over if you wish to send it to me! You have a nice day.

reply »

#28565

Wave soldering profiling | 12 May, 2004

Now we are finally breaking the ice, glad to meet you greg, once again did not mean to offend any one and glad to have guys like you on the chemistry end of things... have no problem with the pp file, its mainly set up to help those with issues like this one and to help trouble shoot defects, really if boils down to: when in doubt, go back to the basics, lot of things have changed in our time, but the physics are still the same, just chemicals and configuration have changed... hopefully we will have more time to know each other and help some of these younger ones...you know on my end, I find it sad that alot of experienced engineers don't help any more and they always leave out important training facts such as exit flow rate, sequence of operation, physics and applied science, really simple rule of thumb, can't trouble shoot if you don't know what takes place.... hope to chat again soon...where you located?

reply »

#28566

Wave soldering profiling | 12 May, 2004

Bruce, regards to all who are trying to help and have great input, with those type of lead length, I would be very impressed to see you get rid of the problem, and as the others said, no clean in a foam fluxer, not good but workable.... the real issue is lead length and the design and dynamics of the old hollis wave former, you can't really control the exit flow rate and minimize solder force like the newer ones, therefore the peel off will comply due to time/travel/temp. at exit.. you really need to look at the cost and try to fight to shorten those leads to .060 max on older machines... good luck...

reply »

#28567

Wave soldering profiling | 12 May, 2004

Here in good old Great Britain - UK - small island across the pond. speak again shortly By the way I consider myself still young at 36 yrs and willing to help anybody who will take my thoughts and try to interpret them - except my competitors that is they can go and get ................... All the best, got to go and have a cup of tea as we say over here - wishing it was a BUD - TRUE

reply »

#28572

Wave soldering profiling | 12 May, 2004

Hello Solder Doctor;

I would be very interested in your file regarding wave process improvements. I have sent an email to you and I hope it gets to you.

I have found some other tips from Speedline Technologies. They have a trouble shooting chart that lists problems and possible solutions. As well I received an exe file from Electrovert tech support to help with the same issues. You type in your flux, type of board and so on and then click on the problem you are having and the program gives you some issues to look at.

I will say though that I have yet to get rid of the icicle problem I am experiencing.

Any help is very much appreciated Thank you Bruce

reply »

reflow oven profiler

SMT Machines