Reflow oven: further testing

A bit of boredom has set in… The oven is ready to go, for testing with actual boards and these – with requisite solder paste and components – are also good to go. But the stencils ordered have not yet arrived! What is one to do?

Final build steps

So far I hadn’t fitted any insulation, waiting on a welding blanket to arrive. Upon its arrival a pattern was duly created from paper, two layers of the blanket cut, Kapton tape put around the edges and then fitted. Photos are shown below. While I don’t really have any evidence either way for it’s effectiveness, it seems likely that it wasn’t really worth the effort for two reasons:

  • The non oven chamber side of the blanket becomes quite hot during operation. Maybe not as hot as the internal temperature and certainly nothing beyond it’s rating – at 600 °C or 1000 °C, I’ve forgotten – but I question its insulation capabilities.
  • Upon inspection of the oven design – and wishing I’d done this earlier – the internal and external back wall are one and the same thing(!).

In short: any heat saved is just going to go out the back wall anyway. Oh well – or c’est la vie –  not much that can be done. And it certainly now doesn’t seem worth the effort to pull the back shell off to seal the seams given the gold tape inside the chamber is doing the job.

Further testing

Saying all this, many further tests have been run and this time with the thermocouple taped to a sacrificial board and the temperature delta filtered with a 0.01 Hz low pass filter. Figures 1 and 2 show the results.

The filter used is equivalent to a simple RC low pass filter, as if there was an actual RC circuit on the thermocouple – but just to be clear, the results shown are simply post processing of the data received from the ControLeo2 controller. As for the cut-off frequency of the filter, this was chosen with a little trial and error to remove the noise to a reasonable level. Now one has a better view of the heating and cooling ability for the oven. The difference in the cooling rate between Figure 1 and Figure 2 is related to how far the oven door was open once the reflow was complete – Figure 1 with it open a bit, then fully opened, and Figure 2 with it fully opened straight away. (The oven has a potentially nice feature in that the spring loaded door can be opened about an inch at the top and it will stay in this position.)

Figure 1 – Test run 1 with the thermocouple taped to the board and filtered temperature delta also shown

Figure 2 – Test run 2 using the same set-up

Would it be worth investigating to implement such filtering on the thermocouple in either hardware or software? Not sure. In an email conversation with Peter Easton – the designer of the ControLeo2 controller – he’s stated that there’s already a certain level of filtering of the thermocouple signal. While I’ve looked at the code for the controller, I haven’t looked that close and besides the oven temperature profile seems to be fairly repeatable. We shall see once I try things out for real.

BTW – a thank you to Peter for his help with various questions I’ve had. Always a prompt reply! If you’re interested in the controller then do check out his website at: whizoo.com

The stencils should arrive this week and I’m waiting in anticipation to try things out “properly”. We’ll see how it goes…

By | 2017-04-02T22:16:06+00:00 Saturday, 1 April 2017|Builds|0 Comments

About the Author:

Colin is the chief designer at lagrangianpoint. His interests are in (too) many areas - to name a few: the Arduino platform, LED lighting, Software Defined Radio, statistics, trying to learn French, cooking and wine. Unfortunately lagrangianpoint isn't his day job just yet - but it's the plan. In the meantime he is Clinical Operations Manager for Cochlear EMEA. (Cochlear specialises in implantable hearing solutions for adults and children who are affected by deafness or hearing loss - check out www.cochlear.com)

Leave A Comment