I would have joined this thread much earlier if I had noticed it, but I just discovered it after Ian
posted a link to
Alan's blog post on Twitter. I just don't spend as much time in the EEVBlog forums as I should.
I have both a T-962 and a QS-5100 which have been gathering dust for quite some time. By a funny coincidence, I am also finally getting around to trying to tame the wild temperature profile of my QS-5100, so I was very interested in Alan's recent work. I may not bother to do the same with the T-962, which has dramatically inferior construction to the QS-5100. The T-962 is wrapped with masking tape as many of you know, and its oven drawer window is held in place with hot melt glue!
The QS-5100 has no such dangerously brain-dead construction features (but it sadly has no oven drawer window at all). I have made tear-down videos of both ovens, which I'll link to at the end of this post. While the two ovens are similar in concept and they can easily be confused with each other at first glance, they're clearly different designs once you take a close look at their innards.
I have a
Pax Instruments 4-channel temperature logger, and I'm waiting for some thermocouples to arrive from Digi-Key which are rated for the temperatures they will encounter in the oven. The probes which came with the logger were having trouble in the oven heat. Their cloth jackets char, and one of them became intermittent at high temperature. I'm using multiple probes so I can get a better look at the temperature distribution at different points in the oven. I expect that there are hot and cold spots that I'll need to work around.
I don't have a plot handy, but my first attempt to plot the temperature profile showed that my actual board temperatures were MUCH higher than the oven's profile settings. I was surprised to see Alan's plot showing temperatures much closer to a proper soldering profile. I'll gather more data and share my results once I get my new probes.
I suppose that the easiest fix would be to monkey around with the profile settings until my measured profile is reasonably close to what I want, but I have doubts about the repeatability of such an open-loop approach. I think that I will probably want to replace the oven's thermocouple with one in contact with the board being soldered, rather than hanging in the air right between the two heater tubes. I had an idea about placing a contact thermocouple in the drawer to contact the bottom of the board, but a quick experiment with thermocouples on both sides of a test board showed that would not work much better than the oven's dangling air probe due to the temperature differential between the two sides of the board. That's a shame, because it would have been so convenient to just set boards on top of a sensor mounted in the drawer. I have the beginnings of an idea of how to mechanically mount a thermocouple to make it easy to have it contact the top of the board while making it easy to swap out boards without knocking components off.
Here are my T-962 and QS-5100 teardown videos: