I'm not sure if this is practical with this part, but some people dip no lead parts into a leaded solder to re-tin the leads prior to using leaded solder for assembly.

Rohs PCBs are made of the different materials. It is specially design for rohs component soldering.

The reflow profile of rohs soldering and non rohs soldering are different.

Since it is BGA soldering so not recomend for non rohs component on rohs PCB.

There should be no reliability issue with the soldering application that you are suggesting. The use of Sn63/Pb37 solder paste for lead-free BGAs is not all that uncommon.

It is recommended to reach a higher peak temperature than you would normally consider with Sn63/Pb37 solder paste. The reason for this is that you want to insure that the Lead-free balls on the BGA melt so that proper collapse can occur.

This generally means achieving a reflow profile with a peak temperature in the neighborhood of ~225C on the BGA. (The melt point of the SAC305 balls is 217-219C.) This means that the peak temperature on the lower-mass components may reach 230C or above.

Mixing of lead free solder and tin/lead solder balls will result in an alloy somewhere in between and you should be aware that this will have different properties to both alloys.

Also, if the alloys do not fully mix then you may be left with a non-homogenous structure which could cause premature failure. They must be allowed sufficient time above liquidus to fully mix. At the end of the day the drivers are the end use of the product and whether it needs to be truly lead free.

If it is in a critical application with long life expectation or severe environmental conditions I would not recommend it. You could always have the solder balls changed to lead free by a specialised rework company if need be.

With a device of this size and complexity, the issue will be thermal consistency across all the I/O positions.

In a typical reflow process where the transfer medium is gas [air or N2] the outer leads of this device will receive heat quickly and effectively, the inner leads will be starved until the whole PCB is able to conduct heat into the inner areas of the array [ it is impossible for the gas mechanism to reach under the device.]

The result is differing thermal profiles from the outer edges to the centre of the device, this in turn may affect the joint structure, integrity and longer term reliability. The recommendation would be to use a Vapor Phase System which uses an inert fluid as the basis for thermal transfer.

As the fluid is boiled It creates vapor and as the vapor condenses onto the PCB it transfers energy [heat] to the PCB. The efficiency factor over gaseous systems is ~10x and given the natural capillary action of a fluid it is able to move in under the array device and come into contact with very I/O, all 1500, and conduct heat to each equally quickly and effectively, stabilizing the thermal profile across the device ensuring the integrity of every joint on the device.

For a Pb-free ball & Sn-Pb paste, from a reliability standpoint - it is better to completely melt the Pb-free ball and have a homogenous joint - see figure below.

• this means that the peak temperature would be around 230-235C which the Sn-Pb paste is also exposed to most Sn-Pb pastes are designed for peak temperatures of around 225 C

When the Sn-Pb paste sees 230-235C, some of the things that need to be considered are:

• there is more oxidation at higher temperatures; can the flux in the Sn-Pb paste remove oxides even at higher peaks?
• flux residue becomes harder to clean off with high peak temperatures and could actually char - is cleaning affected?
• depending on the flux chemistry, some no-clean fluxes outgass/voids at higher peak temperatures - if so, is the voiding level in the BGA acceptable?
• can the other components on the board withstand a higher peak temperature?