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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


QFN welding problem by reflux oven 10 zones

Views: 1886

Hi everyone I have problems welding QFN components in a 10-z... - Feb 11, 2020 by Sebas  

#84402

QFN welding problem by reflux oven 10 zones | 11 February, 2020

Hi everyone I have problems welding QFN components in a 10-zone reflux oven, we cannot make a welding fillet that covers the entire QFN pin. Currently we use paste T4 - 63Sn / 37Pb. I enclose the curve of the oven profile. Any help or support on this topic is appreciated by improving the oven profile or better understanding this process thanks.

Attachments:

reply »

#84403

QFN welding problem by reflux oven 10 zones | 11 February, 2020

Values and range definitions on your profile attachment are for LF paste not for Tin Lead. Please recheck and redefine, so we get the correct values. Also, which location is the QFN?

reply »


Rob

#84404

QFN welding problem by reflux oven 10 zones | 11 February, 2020

We had this problem at my old job. My understanding (from googling this problem previously) is that the exposed leads on the side of the qfn are not tinned and are not required/designed to have a fillet. Someone correct me if I'm spouting lies.

reply »

#84407

QFN welding problem by reflux oven 10 zones | 11 February, 2020

I was going to say, technically it's BTC, so I don't believe it needs a fillet, only evidence of wetting.

reply »

#84408

QFN welding problem by reflux oven 10 zones | 11 February, 2020

hello Evtimov Thanks for your answer. I will analyze the curve with our pasta supplier to reconfirm if it is adequate. It would also be helpful if you could share the tin lead curve profile to analyze.

The location of the QFN is assembled by the top layer of the PCB, the pcb only has an assembly of its faces. Forgive my ignorance if I did not fully understand your final question to what you want to refer to the location of the QFN.

reply »

#84409

QFN welding problem by reflux oven 10 zones | 11 February, 2020

Sebas,

on your profile you have listed U2, U101, U102 and I was wondering if the questioned part is one of them. I attach a typical profile specs - you should follow the one for your paste. In the one you attached the slope is too fast, soak is probably insufficient, TAL is over two minutes and overall the profile is too hot for Lead solder. Also it is pretty long about 6 minutes total. If you can add datasheet for the part this might help people here to troubleshoot.

reply »

#84410

QFN welding problem by reflux oven 10 zones | 11 February, 2020

Yep....I just looked at a copy of IPC-610-G and am not 100% sure but I think it's Note 5, Table 8-16, pg. 8-96.

reply »

#84411

QFN welding problem by reflux oven 10 zones | 11 February, 2020

Hi Steve Thomas

Effectively by IPC-610 standard specifies in the "content 8.2.13 Quad Flat Pack Plastic Package - Without Terminals (PQFN), Table 8-13, page 8-84" which depending on the class of manufacture notes 2-5 of the table, which in theory does not It requires the steak at its maximum height. But in some complex circuits of consumption in practice we show that we need to acquire that fillet for the card to work.

To perfect that welding fillet, we will analyze in depth the profile of the oven according to the comments of Evtimov.

reply »

#84412

QFN welding problem by reflux oven 10 zones | 11 February, 2020

Evtimov

The areas analyzed are a standard oven curve according to the pasta we are using. The channels used are only for measuring the oven temperature profile, not in detail of the QFN component sector.

reply »

#84413

QFN welding problem by reflux oven 10 zones | 11 February, 2020

Sebas,

I suspect that your finding of needing the fillet for the card to work is a false finding. One that I dealt with for many years in assembly.

As a bottom terminated device, a fillet to the side terminals of the device is not required. Adequate wetting of solder paste to the bottom termination is all that is required.

However, in practice, getting this bottom side wetting can be difficult. I had numerous issues with reflowing these parts, and human QC is, generally, not able to verify the bottom side termination. X-Ray inspection will reveal the true nature of your solder joint. We did not have x-ray available to us, and, as a result of a number of field failures, we set the process to require rework of these components to ensure complete reflow and wetting. The reworking of the components usually left a fillet, resulting in people believing that the only way to insure that the bottom side is soldered is to insure that you have a fillet.

If you're doing relatively low volume (as we were) this can be an acceptable means of processing these components (ignoring the possibility of overheating the components due to too much rework). If you're doing higher volumes, this can become over-burdensome.

Some things you can look at to help resolve the issue: 1. Solder paste planarity. Make sure that your paste deposition is as planar as possible...too much solder on, say, the ground pad, can cause the component to float too high to make a good wetting during reflow.

2. Amount of solder paste on the ground pad. Even if your paste deposition is planar, too much solder in the center of the component will cause floating. Thermal imbalances will cause a very slight tilting of the component, not allowing wetting on one side.

3. Soak time. Uneven reflow of the solder paste can cause tilting as above, due to thermal imbalances on the board (thicker traces for ground/power, etc). If you're constantly having the same pins not wetted, this is where you need to look.

4. TAL. Same as above in soak time.

cheers, ..rob

reply »

#84416

QFN welding problem by reflux oven 10 zones | 11 February, 2020

Your profile should be close to this.

Attachments:

reply »

#84418

QFN welding problem by reflux oven 10 zones | 11 February, 2020

hello robgd3 Thank you for your contribution based on your experience and I see that it is very useful for my case, since it is very similar, we will begin to analyze the points mentioned.

reply »

#84419

QFN welding problem by reflux oven 10 zones | 11 February, 2020

Evtimov thank you very much for sharing this profile, we will perform tests with this curve to see behaviors in the oven and pasta.

reply »


SMTA-Alex

#84431

QFN welding problem by reflux oven 10 zones | 12 February, 2020

There are several areas you can check. Your profile is very fast. You reach the peak at around 3 minute mark. That's not enough time for the fillet to form. Try to slow down your belt speed to around 90-95, to give it more time. Also, your temperature settings in the middle zones are way too high. You are burning off the flux too early, and there isn't enough to activate the oxidized component leads to wet. Zones 5-8 should be in the 170-180 range for Tin-Lead. Zone 9 should be around 220, zone 10 around 250. These are ballpark temperatures, without seeing the board. And, check your peak temperatures. They are too high on your profile. You should be around 225 on a bare board. A lot of old Tin-Lead components can't go over 220.

This message was posted via the Electronics Forum @ SMTASMTA

reply »

#84470

QFN welding problem by reflux oven 10 zones | 18 February, 2020

Evtimov SMTA-Alex

Hi After the comments, we have analyzed our oven curve according to our sheet of technical data of solder paste. We modify the curve taking the linearity to the maximum with what the file specifies, we attach the profile and the curve for your analysis comments, if the profile is ideal, thanks.

Attachments:

reply »


SMTA-Alex

#84477

QFN welding problem by reflux oven 10 zones | 18 February, 2020

Sebas, some things in your profile don't make sense. 3 of your thermocouples, #1 # 4 and #6 are consistent. The others show completely different data. The consistent thermocouples, also, show peak temp at 241 - 243 degrees. That is extermely high for lead paste. You should not be over 220. The highest zone setting, you have, is 230. That means you are reading 11 - 13 degrees higher, than the hottest zone. Are you sure those are the oven settings? I, also, see that you are over 183 degrees for at least 100 seconds. That is way tool long, unless you are running very large heavy boards.

This message was posted via the Electronics Forum @ SMTASMTA

reply »

#84484

QFN welding problem by reflux oven 10 zones | 19 February, 2020

Sebas,

Curve you attached is using different temp table. We need valid data to evaluate.

reply »


ben

#84494

QFN welding problem by reflux oven 10 zones | 20 February, 2020

Hello,

The oven profil is not the solution! The only solution for the solder mount on the side is to put gel flux on the side of qfn and make a second travel in reflow . The flux remove oxyde from the copper that is not tinned and the solder is able to mount…

reply »

#84516

QFN welding problem by reflux oven 10 zones | 23 February, 2020

Yes, this is for tin-lead profile. And for the QFN pins which has copper exposed that not required and not possible to have fillet on side surface as in IPC, see 8.3.13, in IPC-A-610-G

reply »

FPC* - Fluid Pressure Control - Dispensing Pump

Reflow Oven