Improving a T-962A Reflow Oven Part4

[@martyjohnston700]

Thank you for taking the time to help us amateurs. Very clear, concise and steady flow of information, I've watched several of your T-962 videos and have remained focused, that is quite an accomplishment in itself.

[@chrismr3972]

Aha! I had no idea you were doing it "properly" (most of KZbin doesn't). Thank you - I think there will be lots of people interested in professional instrumentation, at least there should be! I've been toying with making a vapor phase reflow oven, even got some Galden, because I have some fine pitch BGAs to cope with. At least with vapor phase you guarantee the temperature, and it definitely gets to everywhere. The idea is to use a 200mm deep stainless steel gastronorm pan (which I have) and then use an infra red heater underneath and a fan for the cooling cycle. That way there isn't much thermal inertia in the pan. I have seen one where they use a heating element, but it takes ages to cool. The other idea for heating it is to use the stainless itself.. stainless has relatively high resistance so if you take a high power toroidal transfomer and put a turn of thick copper wire though it, then weld that across the pan, that'll provide enough current to heat it. You can then turn the mains off for half a cycle and measure the pan temperature directly from the outside. I have a few things to make so I designed a simple "building block" system that has displays, routers and interfaces so you can connect lots of things up in any order to make more or less what you want. So I could make it control a Galden oven, a transformer winding machine, an anodising line and so on. All with more or less the same parts. The network speed is 32Mb/s using RS485 over CAT5 and the power is included in the CAT5 so there is no need to "hack" local power. Cost is tiny, because I am on a budget - plus it makes it more interesting! Boards should be arriving early December :o)

[@JerryWalker001]

Inaccurate measurements are just useless numbers and there are plent of those to be found :). My company develops instrumentation for medical, pharma and research industries so I tend to follow fairly formal procedures but I am hoping to make my channel informative and useful to others without making it tedious. As you know professional instrumentation quickly gets complicated so I am trying to strike a balance between useful information while remaining interesting. Your project sounds very interesting. I will be interested to hear how you get on.

[@chrismr3972]

@@JerryWalker001 Its going to be interesting to see how well someone with access to the right equipment, and has the right method, can make these little ovens work. I for one am looking forward to hearing of progress. Talking of measurements / precision etc, I used to run R&D for a company (to do with radio) who transitioned from commercial to military due to something we made. We had very little quality equipment, things like valve RF generators that had to be turned on at 5am and left in a closed room of their own to stabilise for 6 hours before you could use it without it drifting. At the point where they went military a "security officer" came round to check we were ok, he had to go in every room, including the one with the RF generator in it. I explained, he insisted. At the end of the tour he said something which has etched itself into me "in three years time this place will be full of engineers with the best equipment money can buy.. and they will achieve nothing like you have with so little. I see it everywhere I go". I left shortly after that... and he was right! Keep up the good work.

[@JerryWalker001]

Yes the instrumentation systems my company develops include custom RF links and I have a lot of very expensive equipment but that is not enough. It has to be used properly or it is wasted expense. Most of the high accuracy equipment is never switched off. You will notice in many of my videos that I use the simplest equipment in my lab as I want to keep the processes in 'reach' of anyone that wants to do the same. I used to develop and build all manner of things with nothin more than a moving coil test meter. More advanced equipment generally makes progress faster and more precise as long as it is coupled to proper methology. I actually mention this in todays video which I will be posting shortly.

[@chrismr3972]

@@JerryWalker001.. and there is always Price's law "50% of the work is done by the square root of the number of people" which neatly explains why so many great companies are basically two people (rolls royce, marks & spencer etc) and might give some insight into why there are two parents!

[@JerryWalker001]

Absolutely correct but you also need to factor in the additional work caused by the 'chocolate teapots' that we are forced to employ these days.

[@jontscott]

Well three thoughts come to mind. 1) On the oven would putting a thin layer of something to insulate between the boards and the metal tray help even out the heating? I was thinking maybe the metal tray becomes a heat sink and the edges are pulling more heat away than the center. 2) Would adding a little heater under the tray around the edge help even out the relatively cool edges? 3) On temperature accuracy I cannot say much, as I just have uni-t branded temp probes and every other tool I have is likely even less accurate. But I would be curious if you stick a basic multimeter temp probe in your high precision test setup how close does it read to the others?

[@JerryWalker001]

This is really an infra red heater so as long as the board is not in direct contact with the metal tray then it does not have much impact (see my earlier videos on this machine) because it is reflective as far as IR is concerned and so it already acts as a 'heater'. The way this machine works is by putting energy into the chamber at a 'specific' rate so I am not sure what effect a heater under the try would have but I suspect it may cause problems. The run cycles are only 7 minutes long. The problem with cheap measuring systems is that they are not predictable. Thermocouple systems are especially problematic when using cheap units (I may make a video on this). In a thermocouple system there are at least 2 measurement points (usually many more). There is one at the probe tip where the two metal wires are bonded and another where the probe plugs into the meter. The displayed value is calculated by measuring both locations and in theory the probe tip temperature can be resolved. With cheap units the 'cold' junction (the one in the meter) is frequently not well designed and can result in very large and unpredictable errors. Even using my professional units I have the actual measurement unit inside a temp controlled chamber both during calibration and during operation to minimise CJ errors. This is just an example of a possible error and this may vary on a single meter. I will add a cheap thermocouple meter for a few runs to show how they compare. What you are concerned with is the 'degree of uncertainty' in the values you record and this gets fairly complicated to determine. Let me know if this is something you would find interesting and I will make a video specifically addressing that issue.

[@jontscott]

@@JerryWalker001 it may go over my head but I enjoy your videos and like to learn. Thank you!