@esden What solder paste do you use? Do you have two with different melting points for double sided boards?


@marble There is no need for different solder paste. First of all flux is decreasing melting temperature, after reflow the solder has higher melting point because flux is not working any more. Second the temperature on the bottom of the board is lower than on the top as the heating elements are on the top of the board. So independent of the solder paste PB or no PB you can just assemble and reflow one side an then repeat the process for the other side. Unless you have very heavy components.

@esden @marble I had bad luck with the cheap ali infrared oven, although my PCBs are weird (non-uniform copper pour, thin substrate)
I wonder if the same applies with a convection oven? How about vapour phase?

@uint8_t @marble The problem with the unmodified T962a is that it likes to overshoot the set temperature. (or even measure the temperature too low and result in overheating) Also it is a good idea to FR4 between the bed of the oven and the board. That helps keep the temperature of the top side of the board closer to the ambien temperature and thus not require the curve to be adjusted too high. That all also requires the thermocouple interface upgrade.

@esden @uint8_t Would it make sense to install a small fan in the oven and run it at low speed to spread the heat?

@marble @esden We just bought a regular 1800-W convection oven, installed a few thermocouples and RTDs, TRIACs and an ESP

@uint8_t @marble That is a much bigger job than reflashing the existing controller and adding the TC interface to a chinese oven. They really work great with the mods. I use two of them for production. esden.net/2017/11/27/upgrading (Updated mod tutorial with photos and video is in the works.)

@esden @uint8_t @marble FWIW: The reason I ultimately decided against the T962 is that it likes to expel its heat through the bottom, and I don't have any surfaces to place one that could tolerate that in the long term.

I use an unmodified toaster oven lmao.

@cr1901 @uint8_t @marble Sure, I use them on a concrete table on the front porch. I really do not want to expel the fumes indoors...

@marble @uint8_t The problem is not circulation. It is the software on the controller, and sub par thermocouple interface. Reprogramming and adding proper cold junction compensated interfaces solves the problem beautifully. I use two of those ovens with those mods since several years for production. The rest of the oven hardware is sound. I do not trust consumer pizza ovens to hold up in production. esden.net/2017/11/27/upgrading (An improved mod tutorial with photos and video is in the works.)

@esden @marble Hm it doesn't really change the melting temperature, but removes the oxide layer (which shouldn't have much effect on how hard the components stick unless its really thick).
For double-sided boards, the components on the bottom side usually are held by surface tension. For heavier components there is epoxy glue which cures at like 150 °C and can be used to hold the components. (I'd suggest reflowing side 1, apply glue to concerning components and then reflow side 2, the glue turns solid before the solder melts.)

@x44203 @marble Ok maybe not melting temp, but definitely surface tention. Just make the experiment yourself. Solder a joint melt it with the soldering iron. Then add flux and melt it again. Observe the difference of how hard it is to melt the solder joint. At the end does not matter what the exact reason is. The fact is you don't need glue. Glue is only necessary if you use really massive parts.

@esden @marble Yes, surface tension is sufficient to hold most components in place.
The iron taking longer to melt without flux is because of the bad thermal conductivity through the oxide layer. In a forced convection reflow oven they should melt approximately at the same time.

Sign in to participate in the conversation

chaos.social – a Fediverse instance for & by the Chaos community