Is a Convection Oven a Better Choice for a DIY SMD Reflow Oven?

[TheRuler8510]

Q:

Is a Convection Oven a Better Choice for a DIY SMD Reflow Oven, rather than a toaster oven? The fan is already built in.

I have not seen much real discussion on this--unless I missed it.

I'm only worried about the wind blowing around the parts--is that a risk?

Thanks,

[EEVblog]

It's an advantage. The convection toaster ovens are supposed to be better because of more even heat distribution. Toaster ovens and the low cost ebay SMD oven have a big issue with hot spots.

[bob808]

I am wondering this myself.
I just found an old microwave oven in the basement. I decided to gut it and found that besides microwaves it also has a convection system with a 1400W heater in the back, and two IR elements on the top (650W each). I ripped the magnetron/transformer/cap and the rest of the electronics, only kept the two IR elements, the convection system and the light.
I did a fast test and the oven gets to 230 degrees C in 3 and a half minutes.
IR only heats the surface, not the air around it.

[Kjelt]

It all depends if the convection oven can cope with the reflow soldering profile and I highly doubt that.
So first try to see how fast it can heat up from roomtemperature to 160^oC  and than again from 160^oC  to 230^oC if one of  these take a lot longer than a minute it is pretty useless IMO.

That is where the IR ovens shine, they heat up directly so very very fast and also cool down again fast since there is no thermal mass.
Normal ovens have a lot of thermal mass that needs to be heated up and then buffers the temperature which is ideal for cooking long times but useless for reflowing where you want very fast rampup times.

[EEVblog]

I have read, here and elsewhere, that it is the Infrared radiated heat that does the heating. Someone here recently said that is why you get cool spots behind rack bars. eg. on a pizza base or bread it is evident as lighter stripes.

I've shown that in an SMD reflow video somewhere. Although not a all analytically and repeatable.

Wouldn't that effect be less significant with pure convection? Everything would heat at a similar rate.

Thermal mass is thermal mass, and unless you are talking about very slow ramp times, it's still likely to be an effect.
Even the best professional industrial machines often must be tweaked to minimise this to get an even heat pattern. This is of particular concern on large thermal mass boards.

[IconicPCB]

Thermal mass in an oven is a good thing. It helps maintain set point as various thermal loads enter the chamber.

Thermal mass is augmented by the use of convection. This means a small mechanical structure can be made to approximate larger thermal structure with the view to improving process window.

The question how ever is : Is a Convection Oven a Better Choice for a DIY SMD Reflow Oven?

The answer is YES.  But better still, a vapour phase oven is significantly better than reflow ovens relying on air  and IR for transfer oh heat.

[Kjelt]

I did an experiment yesterday with my 1500W €600 kitchen combi microwave/oven and it took 5 minutes to reach 150^oC.  Fail for reflow

Why do manufacturers use convection ovens with high thermal mass for reflow, because they have multiple of them after each other set at the different temperature setpoints of the reflow soldering profile and a transportation belt transporting the pcb's between the different oven stages. Now if you want to build something like that  two or three ovens with the right setpoint and intelligent belt control (time delay for wetting stage etc.) and your set to go

[bob808]

I will test my oven once again tonight. I placed the temp sensor beneath the IR lamps. I guess a combo of the two should yeld good results.
I've seen a heating profile for some soldering paste, and it was pretty much linear to 230 degrees C, and it had 210 seconds to get there.
Also I didn't have a good seal on the front door as I had the probe wire going in. With seal should ramp up faster.

[Kjelt]

I've seen a heating profile for some soldering paste, and it was pretty much linear to 230 degrees C, and it had 210 seconds to get there.

Depends on the paste but I would check again, almost all pastes esp. no clean paste have a soaking stage of at least a minute at around 160^oC then ramp up again to 230 and cool off. If you keep on ramping up you might be stuck with flux residues or worse badly placed parts (tombstoning).
Just Google with "reflow soldering profile" and click on pictures. You see more like these ones below but I would suggest checking the profile with the manufacturer of your paste.

[JimmyM]

My reflow oven started off as a 1500W (1300W measured) convection oven. It has the standard iron heating elements. Not IR. The ramp speed was maybe 1C/sec. I took the oven apart and added insulation between the inner and outer walls. That improved the ramp speed, but only by a little. So I bought another oven and used the heating elements from it to add to the original oven (Series connected). Now I run it on 230V and get ~2600W of power. The convection fan evens out the air temperature in the oven and shortens the response time to changes in heat input. The ramp is now faster than it needs to be at full power, but that's where the controller comes in.