Finally starting on this convection oven -> reflow oven conversion project

[mindcrime]

I have an external SSR in the middle of a short extension cord that the oven is plugged into.  The SSR, and my controller are all powered via the USB interface to the controller. 

It's funny you mention that. I am actually using an SSR wired into an extension cord as well, but I consider that to be "just for the prototyping phase." I plan to eventually mount the SSR and the step-down transformer inside the oven case, just so it will be all nice and self-contained.

Just for fun, I've been trying to model the thermal behavior of the oven, and I've come pretty close with an electrical equivalent in LTSpice.  I ran some tests with the heater at 10% and 30% duty-cycle to see what the steady-state chamber temperature was for a given heater power, and logged the door-closed cool-down cycle to find the chamber thermal mass (capacitance) and the chamber-to-ambient thermal leakage (resistance).  I also empirically added a second R-C delay to the circuit to model the heating/cooling lag in the heater element.  This gives me my overshoot.  The component values look strange because I've been playing with time-scaling and other normalizations.  Now the scale is 1 second = 1 second, and a heater voltage of 10,000V = 100% duty-cycle.  It comes pretty close to matching my oven, but if I ever want to use the model for anything practical I will need to dial it in a bit more.

That's awesome! It would never have occurred to me to model thermal mass as capacitance, and so on. Very clever. I wonder if that's been done before?

[fourfathom]

I am actually using an SSR wired into an extension cord as well, but I consider that to be "just for the prototyping phase."
[...]
It would never have occurred to me to model thermal mass as capacitance, and so on. Very clever. I wonder if that's been done before?

I usually keep my "prototyping phase" lash-ups as long as I don't have to look at them.  Mine is behind and underneath the oven (the oven is standing off the bench, sitting on a couple of red bricks.)

Re: thermal mass as capacitance, I'm sure it's been done before.  We model thermal paths such as "theta J-A" (junction to ambient thermal resistance) as resistance, so thermal mass as capacitance just makes sense.  Hell, I used to model tidal flows in a river as  resistors and capacitors (lumped approximation) , and we sometimes use water tanks as an analogy for capacitors when explaining electricity (voltage = pressure, current = gallons/minute flow rate, etc.)  Sometimes analogy gives you a better feel for something than would a pure mathematical analysis -- at least it does for me.

[mindcrime]

Project Update, 05-28-2021

More (minor) progress. I have the transformer for the low-voltage rails mounted inside the case now.





Up until now, that transformer had its own independent AC mains connection:



So my next step now is to finally start hacking on the AC line coming into the oven, and tee off of that to supply the transformer. I'll also be adding a switch to the mains feed so I can turn everything off easily.

The other "next step" is to take these two boards:



And combine them.

I originally built them separately as I was prototyping select bits of this in a very piecemeal fashion. One board is just the low voltage power supply, which features a bridge rectifier and a couple of linear regulators. The other board is literally just a transistor, a resistor, and some headers. It's just there for level conversion so the  3.3V arduino pin can control the SSR which needs a minimum of 4V for the control side. 

Then I need to mount that combined board, the SSR, the Arduino, and the thermocouple amplifier board, add the switch, and wire everything up.

All of the stuff shown above with the keypad and the LCD will, I think, go in a separate enclosure, which I will mount on top of the oven, or on the side, or something. Once that's done and the software parts are all fixed up, I will have a more or less usable reflow oven (if I am standing there ready to open the door at the right time).  Adding the actuator for the door will come pretty close to last, I think.

I also ordered a piezo buzzer to use for alerting me when a profile run ends, or if any kind of error state (high temperature, etc.) occurs.

[mindcrime]

Aaaand, more progress. I have the SSR mounting taken care of now. It nestles in there right below the transformer.









Getting closer and closer to having this thing built! Still plenty of work to do, but progress is evident. 

[coppercone2]

did you measure temperature with the housing on to see what temperature that stuff in there will operate at?

For my oven I left the side wall in tact save for a gland/strain relief (separate one for thermocouple) and put the electronics in a big steel electrical box mounted to the side, with a grate on it, to ensure the electronics will not get hot. I have a feeling that area will get really toasty. The failure rate of electronics is proportional to heat so when you have stuff in there tight, it will have an increased chance of (dangerous) failures, and shorter life. Maybe putting a grate on the side will help vent it but that area gets toasty. Before you do more work, I suggest putting the lid on and measuring the temperature in that cavity. And, don't forget fuses now that you have added more complex electronics that can fail short.

The only caviat is that you need good grounding if you add a box to it for safety reasons, you don't want the metal box you add to be left floating.

You might be really disappointed by the temperature you will get in that cavity now that you put work into it.

When you measure temperature, run it for like 30 min to see what it gets to, you want to keep options for use open because having a good temperature controlled oven like I said before is a good asset in the electronics lab and it would be a shame to limit its capabilities to save on a steel box (some where between free and $20). you put alot of work into it so its best to make sure you get your value without surprises on operating limits.

[mindcrime]

did you measure temperature with the housing on to see what temperature that stuff in there will operate at?

Not yet, but that's in the plans as soon as I get the SSR, the transformer, and the switch mounted, and the AC side wiring done. I'm already committed to having a box mounted on the side/top to hold the keypad and display, so yeah, I've considered that it might make sense for most of the other electronics to live there as well. I *think* I can get away with the SSR and the transformer where they are, but I do plan to do testing around that very point. And the good thing is, given the mounting holes I drilled, I can easily just flip both of those components to live outside the case as well.

Before you do more work, I suggest putting the lid on and measuring the temperature in that cavity.

You might be really disappointed by the temperature you will get in that cavity now that you put work into it.

Yep, definitely going to be doing that. I'm also planning to measure that both with and without additional insulation in the inner chamber to see how that affects things. I've also considered that I might add an exhaust fan to move air through that cavity if necessary.

[coppercone2]

if you add an exhaust fan without insulation its going to cool the oven and make a thermal gradient because its like putting a big flat heat exchanger on the side of  the oven wall, minor, but the wider the PCB the more an effect it will have

[mindcrime]

if you add an exhaust fan without insulation its going to cool the oven and make a thermal gradient because its like putting a big flat heat exchanger on the side of  the oven wall, minor, but the wider the PCB the more an effect it will have

Right now I'm leaning towards the idea that I will be adding insulation. I haven't done it yet, but I was chatting with my buddy who built a similar setup, and he made a good point about that. That being, even if the oven gets up to temp with no problem without extra insulation, having the extra insulation reflecting the IR around inside the chamber should help with creating a more even temperature distribution.

I doubt I'll ever be soldering boards big enough for that to matter, but I'm kinda leaning that way. It may also help reduce some heat transfer into that side chamber inside the case that we were just talking about. The downside, if there is any, may be some reduction in the cooling rate on the cool-down phase. But since I'll be opening the door anyway, I don't think that will be a big issue.

That said, "test all the things". I'm big on empiricism and actually measuring and seeing what happens. I just finished up a lot of the work that needs to happen inside the oven case itself, so very soon now I'll be ready to put the outer cover back on, and start doing more temperature measurements and testing things out. 

[mindcrime]

Project Update, 05-29-2021

I've finished up a big chunk of the work that has to happen inside the oven case. OK, I suppose that technically nothing has to happen inside the case, as you can do this kind of conversion with keeping everything outside. But in this case, I chose to package some of the bits inside the existing oven shell. How that works out remains to be seen...

Anyway, I put the rocker switch in, and re-did the AC mains feed so that it goes through the rocker switch, and then after the switch tees off with one branch going to the SSR (for the heating elements) and one branch going to the transformer that will power the low voltage stuff.  All that really remains to be done in here is connect the wire coming off the output side of the SSR back to the wire going to the heating elements. I left it disconnected to do some initial testing, just to make sure I had everything else wired up correctly, and that the transformer worked, etc. So far everything looks good.



Switch now mounted alongside the other components.



Another view.



The switch, seen from outside.



Hard to see in this picture, but it lights up when switched "on"



And, as expected, we get our 12VAC from the transformer when the box is powered up.



This connection right here is one of the last bits remaining in here. That will join the SSR output to the wire that runs through the existing oven controls, and ultimately to the heating elements and what-not. The SSR datasheet recommends a fast-blowing fuse in this circuit, so I have fuse-holder I need to wire in.

[coppercone2]

just, keep this in mind.
https://www.digikey.com/en/products/detail/bud-industries/JB-3954-KO/2674114?utm_adgroup=Boxes&utm_source=google&utm_medium=cpc&utm_campaign=Shopping_Product_Boxes%2C%20Enclosures%2C%20Racks&utm_term=&utm_content=Boxes&gclid=CjwKCAjwzMeFBhBwEiwAzwS8zA7IlaWvJAmPnVkRLjdH3Tx2dZPwq3WUBJkOaCig-Pf3czLVLQQybhoCxVYQAvD_BwE

it makes your life alot easier and you just need to cut or buy 1 piece of metal for the top, the only hard part is making the holes for the display etc. you put tape on the knockout and pour epoxy on it to seal it up if you want then take the tape off

[fourfathom]

I've seen these toaster-oven conversions with the electronics inside the case, so I guess it can be done, but I went with the external Bud Box solution. (I like that epoxy display window idea!)  The only modification to the toaster was to drill a hole for the thermocouple cable, everything else is external.

[coppercone2]

I think I read one mod where the control electronics had the wrong temp because of the heat, did dave jones not have this problem when he left his controller on top of the oven also?

[mindcrime]

Today has been a long day, but finally some significant progress.



The SSR line to the oven electronics is connected now. Yay!



Not the neatest looking job, and I still have some insulating sleeve material I may need to put around some of this wiring, but we'll see.



Two key things come outside - the low voltage (12VAC) rail, and the INPUT side of the SSR. With those two things and the thermocouple, I can have all the "real electronics" packaged up in a box that mounts to the top or side or whatever.



Yes that is ugly and hacky - and more to the point, TEMPORARY. I don't routinely leave wires rubbing on semi-sharp edges. This is just so I can do some testing with the oven basically assembled again. I'll do holes with grommets and wire clamps and/or whatever before calling this "done done".



Low voltage rail working - again. This cost me SO much time. I had the low voltage stuff working earlier (see above) but then it quit working for some mysterious reason after I finished wiring up the SSR and everything. In the end, I had out-smarted myself. I used some (shitty) pluggable connectors to connect the wires to the transformer, thinking "If I ever need to replace this thing, it's better if I can just unplug it."  Apparently the connectors were tarnished, or made out of crappy material, or both, as it quit connecting at some point when I was moving stuff around. Took forever buzzing around with the continuity tester and what-not to finally find the fault. Once I did, I ripped all that connector crap out, and just spliced the wires directly together, soldered the  joint, sealed it with heat-shrink tubing and called it done.





All buttoned up and back in one piece. Now for more thermal testing, evaluating use of additional insulation, checking the temperature in that side chamber, etc., etc, etc. Given that this is a "3 day weekend" here in the States, I may actually be able to do some test soldering before the "weekend" is over. That is, if I have any solder paste here that isn't too old to be usable.

[coppercone2]

put a few layers of heat shrink over the wire going through the vent hole, the teflon kind if you have it

also for deburring the vent, there is NO better tool then a abrasive rubber point for a dremel, they wear quick but your vents will be like, ultra smooth

https://www.amazon.com/Dremel-463-Rubber-Polishing-Point/dp/B00004UDHR

get the black ones if you can find them for that


also, this is the video where dave jones has problems with his reflow oven overheating the controller


if I am doing a wire route without a gland (you should actually use a gland, like the ones they sell for electrical boxes at home depot, you drill a hole , put the wire through a peice of silicone tubing and clamp it if you don't want to buy proper glands, they are nice and metal), I use these for deburring, they will leave an amazingly smooth surface for wire, beware punched vent holes, they are often sharp.

https://www.ebay.com/itm/322139737579?epid=2164364418&hash=item4b01064deb:g:yQsAAOSwxt5fA28y


For my oven I used a $50 panasonic SSR, you want to treat the cheap ones real well, and downrate to ~50% of their real rating, some people going down to 25% on the forum


But anyway, those black points won't get too much deburred unless you are real careful, but they are amazingly easy to use and fast and convenient, so its good for prototyping at least, unless you wanna get out 5 pieces of different grit sand paper lol, it stays professional

[mindcrime]

also for deburring the vent, there is NO better tool then a abrasive rubber point for a dremel, they wear quick but your vents will be like, ultra smooth

https://www.amazon.com/Dremel-463-Rubber-Polishing-Point/dp/B00004UDHR

Interesting. Never seen one of these before. Thanks for the pointer, I just ordered one. Funny that would come up in this moment, as I got a lot of use out of my Dremel tool earlier today, working on this thing (mostly to do with mounting that switch). I don't use the Dremel all that often, but when you need  it man oh man, is it handy.

also, this is the video where dave jones has problems with his reflow oven overheating the controller

Yep, I've seen that one. Probably wouldn't hurt to go back through it again though...

For my oven I used a $50 panasonic SSR, you want to treat the cheap ones real well, and downrate to ~50% of their real rating, some people going down to 25% on the forum

I went with an Autonics SR1  SSR. Not the cheapest ,not the most expensive. It's rated at about twice the expected current consumption of the device, so I think I'll be OK there.

[coppercone2]

crap, get the black one off ebay if you want to debur, that one will also do it but its finer and wears more, their supposed to be sequential, I think just the black one is enough, when you follow up the black one with the other ones it feels like smooth wire

if you have really good tool control they last a longer time, but they are somewhat expensive and wear quick compared to other abraisve types.

oh yeah, beware, the ebay one needs a smaller collet, so only do it if you got the dremel collets that are smaller size.

a worthy investment because that is a stupid thing that can really hurt you, a wire shorting to a chassis

ALso, I only used mine for stainless steel, they might be alot longer life on other materials. just don't press too hard. its a recent acquisition (last week, which I had them 10 years ago).

with the blue one I expect if you hit it with some rough sand paper haphazardly before you debur it will wear less. the problem is you get sharpish stuff thats painted, the paint makes it feel smooth, but there is potentially a knife edge under neath.

I also suspect their going to be very good for stripping thicker magnet wire, cleaning corrosion around ground points in old equipment, the ultra fine ones might be good for cleaning contacts too, but I prefer radial nylon abrasive brushes for those, since their more delicate, but anyway its a damn good tool so long you don't save your budget too much thinking about when to use it.


Right, they are also the PACE soldering recommended way of cleaning charred materials, silkscreen, corrosion, etc off circuit boards you are repairing, all the pace stuff is either ruberized or leatherized abrasive (leather dust mixed with abrasive and hydraulic glue or rubber dust mixed with abrasive and heated).  and you can clean up a ring terminal

[beanflying]

I think I read one mod where the control electronics had the wrong temp because of the heat, did dave jones not have this problem when he left his controller on top of the oven also?

If you drop the controller on top then sure bad idea on the side and in particular lower down on the side of the oven it won't be much of an issue in particular if you thermally insulate it from the oven either by material or forced airflow over the board. Part of the reason for this is if your temperature sensing uses cold junction compensation and the ambient where the board sits ramps along with the actual oven it will throw your sensing off. Before you get to the issues of Electronics impacting the electronics from general heat issues.

[mindcrime]

oh yeah, beware, the ebay one needs a smaller collet, so only do it if you got the dremel collets that are smaller size.

Hmm... not sure what size collet mine has. It's not an actual Dremel brand tool, it's the Craftsman version. Not sure if I even have the manual or anything around here. Maybe if I can find the model number on it I can Google it up.

Anyway, no big deal. I have the one from Amazon coming, so that'll be handy enough in the short-term.

a worthy investment because that is a stupid thing that can really hurt you, a wire shorting to a chassis

True. Although in any meaningful context I would never leave a wire running unprotected through a panel without using a rubber grommet at a minimum, and more likely a purpose built cable gland of some sort.

[coppercone2]

if you do it for long enough you eventually forget something, imo

[mindcrime]

Project Update 05-31-2021

Did some more ramp-up / ramp-down thermal testing just now, to see how things look with the outer cover back on. At first blush, it actually doesn't appear to have made a huge difference, but I need to spend some more time crunching the data to be sure. Note that this plot is misleading if compared to the previous ones, because the scale of the x-axis is different.

Briefly, though, it took 182 seconds to get to 250° C with the cover on, as opposed to 217 seconds without the cover. Note that this isn't a very scientific test as I am not controlling for the ambient temperature in the room, etc. But it does suggest that the outer cover helps hold heat in a little more effectively, which one would expect. Also, with the outer cover on, there was a little bit more overshoot past the target 250° point.  With the cover, the temp went up to 261°, whereas it peaked at 253° without the cover. I would say that is also totally as expected given better trapping of heat inside the inner chamber.

Code: [Select]

0,32.07 # Turning OVEN on
2,32.02
4,32.00
6,32.14
8,32.19
10,32.27
12,32.58
14,32.73
16,33.04
18,33.50
20,34.02
22,34.76
24,35.52
26,36.57
28,37.66
30,38.92
32,40.30
34,41.87
36,43.73
38,45.73
40,47.59
42,49.76
44,52.20
46,54.50
48,57.03
50,59.53
52,62.30
54,65.22
56,68.00
58,70.88
60,73.99
62,77.14
64,80.24
66,83.70
68,86.94
70,90.18
72,93.50
74,96.86
76,100.28
78,103.61
80,107.03
82,110.55
84,113.97
86,117.37
88,120.83
90,124.18
92,127.52
94,130.96
96,134.51
98,137.98
100,141.34
102,144.68
104,147.96
106,151.04
108,154.31
110,157.45
112,160.57
114,163.62
116,166.90
118,170.00
120,173.01
122,175.91
124,178.99
126,181.87
128,184.95
130,187.66
132,190.41
134,193.12
136,195.82
138,198.71
140,201.43
142,203.96
144,206.62
146,209.16
148,211.86
150,214.37
152,216.88
154,219.20
156,221.53
158,224.07
160,226.59
162,229.07
164,231.50
166,233.75
168,236.12
170,238.33
172,240.43
174,242.69
176,244.88
178,246.95
180,249.17 # Turning OVEN off
182,251.67
184,253.84
186,255.20
188,256.70
190,257.71
192,258.92
194,260.00
196,260.85
198,261.16
200,261.45
202,261.39
204,260.24
206,257.87
208,255.59
210,252.52
213,249.80
215,247.70
217,245.39
219,243.34
221,241.15
223,239.01
225,237.31
227,235.59
229,234.06
231,232.01
233,230.41
235,228.35
237,226.08
239,224.27
[snip]
823,39.70
825,39.62
827,39.50
829,39.41



Next up, do some measurements of the temperature in the side chamber. And I'll probably try adding a layer of insulation inside the core chamber and then re-test everything to see just how much of an impact that has.

[mindcrime]

Just for the sake of comparison, here's a standard Kester reflow profile for leaded solder. Not sure if I'll be using Kester solder paste primarily, but this is probably fairly representative. Looking at this, I don't think that matching a profile like this relatively closely is going to be too much of a challenge.

[coppercone2]

Don't worry too much about room temperature with such a high temperature instrument, a livable area won't see anything past 15 above room temp and thats really bad, and you clearly have a very big difference because of the cover. the insulation will change it tons more

[mindcrime]

Up next, take all of this stuff, along with the keypad and the LCD display, collapse/combine things as necessary, and package it all into a proper enclosure. And route those wires properly that are stuck in those ventilation slots right now.

[coppercone2]

you can mount the heatsink on a bracket made out of bent metal and put the circuit boards up on hex standoffs if you have the room

for wiring if you want it nicer you can bent brackets and put zip tie holders on them.

[mindcrime]

OK, more testing results.

Before adding any insulation or doing anything, I did a 6 minute "full on" test to see how warm that side chamber was getting. As expected, it was a bit toasty. It made it up to a whopping 155° C. Now I would never do 6 minutes of "full on" for soldering anyway, but still, that's clearly not a good position.

Soo... went in and made a bunch of heat shields out of pieces of aluminum flashing, wrapped in the same insulation I bought to use inside the main oven chamber. Formed and twisted those around to make an isolated pocket near that back wall, where the SSR and the transformer are mounted.  Also added insulated sheathing around some of the new wiring I added, and the fuse holder. Added a layer of insulation to the inside wall of the outer cover, on top and on the left hand side (*not* the right hand side, since I very explicitly don't want to trap extra heat over there!).

Put the cover back on and did another 6 minute "full on" test. Now the temperature in the pocket where the SSR and transformer live stays down to a much more manageable 76° C. It's still a little bit toasty, but that's also an unrealistic scenario for soldering. And I still haven't added a layer of insulation on the inside walls of the inner cooking chamber, which should help even more.  So right now I'm pretty content with this situation in terms of using the oven for soldering. If / when the need to use it for some other use cases arises, I can look at adding a fan to pull air through the SSR/transformer chamber. With all the insulation that's either already there, or will be added, and assuming it's a small fan, I don't expect that will mess with the temperature gradient in the inner chamber enough to matter. 

No pictures this time, as I forgot to take any before putting the cover back on. Shame, as I was quite pleased with how those heat shields came out.