EEVblog #558 - Beta Layout DIY SMD Thermal Reflow Oven

[EEVblog]

Dave plays around with converting a toaster oven into a surface mount SMD reflow oven with a beginner kit and controller from Beta Layout.
Includes basic calibration and thermal profiling of the oven.
Also a look at an example PCB from PCB-Pool, and reflow of an example board with lead-free solder paste.
And using an Agilent bluetooth multimeter to remote log the temperature.

http://www.beta-estore.com/rkuk/order_product_details.html?p=13
http://www.beta-estore.com/rkuk/order_product_details.html?p=242
http://www.maximintegrated.com/datasheet/index.mvp/id/3149

[marshallh]

Convection can make all the difference. The first run failed being undertemp, but notice the parts that didn't reflow were directly above the metal of the tray.

[peter.mitchell]

Those boards look absolutely beautiful, aside from the silkscreen, such a letdown. however.... if you dont use silkscreen on your products, 10/10!

[mixt]

I noticed that as well. The parts that didn't reflow were sitting directly over the metal grill. Perhaps for non-convection ovens, it would be good to float/support the board(s) from the corners in some way so that the grill isn't adding any thermal mass to the PCB. Even if the reflow is successful, the grill would definitely create some thermal stress due to uneven heating.. which is what we're trying to avoid, right?

[EEVblog]

Convection can make all the difference. The first run failed being undertemp, but notice the parts that didn't reflow were directly above the metal of the tray.

Ah, didn't notice that.

[EEVblog]

Those boards look absolutely beautiful, aside from the silkscreen, such a letdown. however.... if you dont use silkscreen on your products, 10/10!

Yeah, just the poor silkscreen, bummer.

[bookaboo]

Yeah that metal grill def caused an issue on the first attempt. Maybe some FR4 or similar as a baseplate to give more even thermals.
Was thinking the E4 would be a great job here but it wouldn't see through the glass, could still get a snapshot on opening the door at the end though.

[JuKu]

I have that too, and just like Dave, ended up makign manual adjustments to the profile. After that, I too like ot and consider it reliable. 

[Wytnucls]

If you liked the controller enclosure, they are available from Conrad in different colors and sizes:
http://www.conrad.de/ce/de/product/523105/Universal-Gehaeuse-Polystyrol-Rot-135-x-95-x-45-1-St

[EEVblog]

Yeah that metal grill def caused an issue on the first attempt. Maybe some FR4 or similar as a baseplate to give more even thermals.

I was curious to see what a big piece of copper clad board would do acting as a heat shield and heat sink...

[jeroent]

Hi Dave.
I got a Severin sliced bread toaster. Wondering if I can use it for soldering as well.... :-)

[nitro2k01]

As noted, the grill absorbed heat locally. Maybe it would be a better idea to actually use the tray to get a more even thermal distribution. And maybe even tape down the probe to the tray without hooking it on to the PCB. Yeah, the tray will have thermal mass, but that's what the overshoot calibration is there for, right?

I'm also wondering about the temperature offset. You changed it at room temperature because it gave a too high reading and when you logged the temperature, then the oven controller returned a too low reading (compared to the Agilent, 51:30-52:30 ish.) Maybe the probe, or amplifier circuit in the controller, is crusty and doesn't give an accurate reading at room temperature. Maybe the probe offset should be calibrated around a references temperature of 100 or more degrees C?

[GeoffS]

Hi Dave.
I got a Severin sliced bread toaster. Wondering if I can use it for soldering as well.... :-)

Of course you can! Courtesy of mikeselectricstuff 

http://www.electricstuff.co.uk/smtoast.html

[eV1Te]

That temp controller is not very good at all, 7 Degree offset and no decimal points in the reading or offset calibration value. I have previoualy designed thermocouple controllers myself, and you would need to really try hard to mess it up by that much!

Could you please make a follow up video where you trace out the analog part of the controller and check what voltage and temp reference they are using. Are they using the built in 10 or 12 bit ADC? In that case I would expect some white noise generator so that they can oversample and dither the reading.

If anyone is interested, almost every PID controller you buy today, can 'learn' by them selves and any serious company would have a specification of maximum +-.5 degrees in offset (then the thermocouple would be an additional +-1 degree). Also common PID controllers can handle a variety of thermocouples and also PT100/1000 for better acurracy since you rarely reflow your boards at 1300 degrees Celsius. 

[EEVblog]

That temp controller is not very good at all, 7 Degree offset and no decimal points in the reading or offset calibration value.

That's likely just the display output rounding.

Could you please make a follow up video where you trace out the analog part of the controller and check what voltage and temp reference they are using. Are they using the built in 10 or 12 bit ADC? In that case I would expect some white noise generator so that they can oversample and dither the reading.

I already posted the datasheet in the video description.

[AndersAnd]

Most importantly, does it come with a heating profile for pizzas too? 

[JoannaK]

Nice video of the home (small business) reflow owen, especially showing the process and the traps there are (calibration, metal grill etc)..

I wonder though, if you should mentioned a bit more clear that you didn't actually purchase the owen and controller (was gifted), and that the entire set (including materials, delivery etc) would have cost close or over 300 Euros.

[mikeselectricstuff]

A coarse metal grille can cause cold spots - I tend to use a fairly fin mesh sheet which works fairly well.
Another option is to make up some long triangular (think Toblerone) sections of PCB strip (e.g. old tooling bars), about 10mm on  a side, by taping 3 strips together with kapton tape, and using these as 'trestles' to seperate the PCB from the shelf. This is also a useful way to do double-sided reflow. 
 

[JoannaK]

Most importantly, does it come with a heating profile for pizzas too? 

For  ols school coders.. There's always C1541 .. Disk drive with pizza warming feature pre-installed.

[eV1Te]

That temp controller is not very good at all, 7 Degree offset and no decimal points in the reading or offset calibration value.

That's likely just the display output rounding.

You are correct since the ADC has 0.25 degrees resolution after reading the datasheet.

After looking carefully on the video, I think they might have made the mistake of not using a proper thermocouple-wire from the connector to the circuit board inside the box. The wires from the connector is soldered to the board, but I have never managed to solder the Nickel based alloy used for Type-K thermocouples, the solder does not wet this alloy. If this is the case, you can just replace that wire and possibly gain several degrees of better accuracy, and you can even leave the box on top of the oven without any problems 

I already posted the datasheet in the video description.

I completely missed that, thanks! 

But for the same amount of money (123 Euro) you can get much better performance, for example this industrial miniature thermocouple controller (which I have used as an oven controller previously).
http://www.omega.com/pptst/CN7500.html
Cost 70 Euro (and then you buy a solid state relay and a fancy box for the remaining 50 Eur you have saved)

This can handle all thermocouples types (even other signals such as 10 mV per Celsius and PT100) from -273 to +1800 Celsius with 0.25% accuracy out of the box without calibration (better with calibration of course). It has dual temperature display and up to 64 different ramp/soak steps with build in extra relays for customizable alarms. Omega also have free software for changing parameters and proper logging/graphing capability for more than 200 devices on the same RS232 or RS485 cable if you ever would like to expand your toaster oven reflow capacity. 

[peter.mitchell]

Just wondering, whats the go with using SSRs in these things, compared to using TRIACS? Is it primarily just heat dissipation.

[mikeselectricstuff]

Just wondering, whats the go with using SSRs in these things, compared to using TRIACS? Is it primarily just heat dissipation.

A SSR is typically just either a triac or a pair of back-to-back MOSFETS, plus isolation etc. in the same package.
I think some use photovoltaic isolators to drive MOSFETS, which probably reduces losses- a triac drops typically 0.8V, so produces a fair amount of heat when switching heavy loads.
SSRs also have stuff like isolation & approvals taken care of.

BTW I didn't like how it looked like that low-voltage wire could touch the fuseholder.
 

[Kohanbash]

Hi
Has anyone here tried a small infrared reflow oven such as  http://www.ebay.com/itm/T962-INFRARED-IC-HEATER-REFLOW-OVEN-SMD-BGA-T-962-d-/260799459468?pt=LH_DefaultDomain_0&hash=item3cb8dbf08c

For $245 USD it is about the same cost as this thermal controller+toaster.

[eV1Te]

A SSR is typically just either a triac or a pair of back-to-back MOSFETS, plus isolation etc. in the same package.
I think some use photovoltaic isolators to drive MOSFETS, which probably reduces losses- a triac drops typically 0.8V, so produces a fair amount of heat when switching heavy loads.
SSRs also have stuff like isolation & approvals taken care of.

Is there some other reason to use MOSFETS, I mean at 10A you need a MOSFET with 80 mOhm on-resistance (and 300+ volt capability) which is quite expensive, just to compete with the 0.8 V drop in the triac?

BTW I didn't like how it looked like that low-voltage wire could touch the fuseholder.

And also no varistors or other protection for the SSR, they are usually quite sensitive to voltage spikes and often fail shorted so that the oven would overheat and possibly be a fire hazard depending on what you put into it.

[Winston]

If they'd have used a transformer with a primary center tap and a back panel voltage selector switch, or to save money, just a jumper on the PCB, they could have easily made it 230/115V.

Anyone know of a controller, preferably Arduino-based, that has the only feature of the Beta controller that really interests me, the learning/calibration function?

[peter.mitchell]

Just wondering, whats the go with using SSRs in these things, compared to using TRIACS? Is it primarily just heat dissipation.

A SSR is typically just either a triac or a pair of back-to-back MOSFETS, plus isolation etc. in the same package.
I think some use photovoltaic isolators to drive MOSFETS, which probably reduces losses- a triac drops typically 0.8V, so produces a fair amount of heat when switching heavy loads.
SSRs also have stuff like isolation & approvals taken care of.

BTW I didn't like how it looked like that low-voltage wire could touch the fuseholder.

I just queried it because often in these kinds of things it seems like they switch the load slowly, but quite hard, so i don't expect the elements(or whatever else is being switched) to last long - though I have no idea if they do or not. I understand why they do it (the loop is quite slow so the switching speed is fairly inconsequential), it just seems kinda wrong when you could use some triacs and run the elements more "proportionally". (I hope you understand what I mean)

The point about isolation and approvals though is fair enough in itself.

Is there some other reason to use MOSFETS, I mean at 10A you need a MOSFET with 80 mOhm on-resistance (and 300+ volt capability) which is quite expensive, just to compete with the 0.8 V drop in the triac?

I don't know if it is just me, but I do generally think of SSRs as back to back MOSFETS (even though i know there is a lot of variety amongst them), rather than TRIACS because of their performance when used with DC. To me, the "relay" part of SSR implies isolated, switching of anything, be it AC, DC, pulsed DC, AC with DC offset ect.

[mikeselectricstuff]

If they'd have used a transformer with a primary center tap and a back panel voltage selector switch, or to save money, just a jumper on the PCB, they could have easily made it 230/115V.

..except the SSR would need to handle twice the current for the same wattage

[TorqueRanger]

Those boards look absolutely beautiful, aside from the silkscreen, such a letdown. however.... if you dont use silkscreen on your products, 10/10!

Yeah, just the poor silkscreen, bummer.

Ever think about selling the boards as a limited edition???

[Bored@Work]

Hi
Has anyone here tried a small infrared reflow oven such as  http://www.ebay.com/itm/T962-INFRARED-IC-HEATER-REFLOW-OVEN-SMD-BGA-T-962-d-/260799459468?pt=LH_DefaultDomain_0&hash=item3cb8dbf08c

For $245 USD it is about the same cost as this thermal controller+toaster.

Google T962. That thing is horrible trash.

[jancumps]

Nice video of the home (small business) reflow owen, especially showing the process and the traps there are (calibration, metal grill etc)..

I wonder though, if you should mentioned a bit more clear that you didn't actually purchase the owen and controller (was gifted), and that the entire set (including materials, delivery etc) would have cost close or over 300 Euros.

it must have been mentioned somewhere. I'm too lazy to run through the whole 1 hour video to find out where though;)

[zapta]

Convection can make all the difference. The first run failed being undertemp, but notice the parts that didn't reflow were directly above the metal of the tray.

.. and here the the proof using my grandma's E4:

[DarkPrince]

Looks like 1500W is holding up pretty well for you. I noticed that after calibration the high temperature measurement was off. The controller thinks there was only a 6°C overshoot where as there was an 11°C one according to your reference (Agilent). Not a thermal conductivity/mass issue either as it also tapered off. I'd presume that is an issue where more than expected (beyond the thermocouple) non-linear relationship occurred. So before it was calibrated I wonder what the offset between true and measured temperature was at high temperatures (200°C+).

I'm looking forward to using a reflow oven. Tired of the hot-air rework which is a PITA with varying sized parts.

Thanks for the morning video!

[BravoV]

Just curious if using that kind of PID controller + a mounted hot air gun + an enclosure with some creative mod to circulate the hot air inside that spreads or circulates the air evenly without blowing off the board/components, will yield better result ?  Has anyone ever seen or tried similar setup ?

[Anks]

So can I have a uCurrent olive gold

[thereza]

I feel the need to chime in here.  I've been making prototype boards in toasters since '01.  I have made very complicated boards with all sorts of components including BGAs and mostly 0402 components.  In the past 12 years, I've had 1 board failure - it was a BGA that I didn't align properly.   The secret it to use an IR toaster - I have this one - http://www.amazon.com/Panasonic-NB-G100P-7-2-Quart-1300-Watt-Infrared/dp/B000063UZV (has been discontinued and they came out with a newer model; not sure what changed).  There is a black & decker one too but that one catches fire a lot. 

To reflow - put the board in the oven - turn it on full heat, watch it waiting for it to reflow (3-4 minutes), give it another 5-10 seconds to make sure that the undersides of the components reflow (thermal packages, etc)  open door and turn off.  DONE. 

Never measured the heating profile, never had a problem with the full-on approach.   Only failure was due to a misaligned BGA once. 

I use lead free solder from these guys - unrefriderated shelf life of ~3-4 years from my experience http://www.zeph.com/zephpaste.htm
make sure to get their nice plunger.  I also use their BGA flux for bga stuff.  everything else they sell is horrible.   And use the smallest tips (red ones).

If its complicated, I cut out a stencil using my silhouette stencil cutter.  cheap, works great.   Else just use the syringe to spread some paste on the pads.

[Chipguy]

Regarding the stencil, they definetly screwed that one up.
I just got a panel like yours from PCB pool some weeks ago and before I ordered that I actually called them and asked.
Their answer was clearly : "If you order a panel you will get a stencil for the panel".

It's not the first time they screwed that up. That was the reason I actually called them, because I was not sure.

[woox2k]

1. Would it be possible to use some blank PCB as a tray to avoid uneven temperatures on the board?

2. Considering that the temperature sensor fault is not linear, i think it would be best to set the offset while on maximum temperature to avoid any unexpected overheats (or underheat for that matter). The whole point would be to make it reflow nicely while not damaging the components. Wrong temperature reading on room temp is not important at all.

[Alana]

Have anyone used such oven to fix a laptop motherboard with chipset soldering fault?

[mtdoc]

1. Would it be possible to use some blank PCB as a tray to avoid uneven temperatures on the board?l.

How about using a small, toaster over pizza stone like THIS

It has some good thermal mass and works well to evenly heat a pizza crust ....

[mikeselectricstuff]

1. Would it be possible to use some blank PCB as a tray to avoid uneven temperatures on the board?

PCB is not ideal as it will not withstand repeated reflow cycles - a thin metal plate would be better

[MatthiasBerlin]

Severin, the maker of the oven, is a German family enterprise that manufactures smaller household devices. Their stuff is usually more at the lower end of the price range, but OK for the money.
Most of the products are not made in Germany, but in China. So most probably Dave's oven has been traveling around the world quite a bit yet .
The company claims world-wide availability of their products, but seem not to be present in the Americas and Australia, just roughly Europe, Middle East, parts of Asia and Africa. So you were right about (regular) availability in Australia, it seems. See http://www.severin.com/gb for English home page.

That mixture of English and German everywhere around the kit, including the control software, is funny .

[JOERGG]

At 11:55 in the video did i spot a unsoldered capacitor ?
And the long screw above the yellow connector looks misplaced, what is it used for?

[Pat Pending]

Don't use convection air if the oven is started from cold.

While the top surface of my PCBs (leaded paste) reached 230C, the underside 5cm lower was just getting to 200C.
I found it was better not to wire oven such that convection was always on.

After testing a layer of baking foil on the inside of the ovens glass door, the thermal ramp-up was much faster than without.
Aluminum flashing worked too and had a great appearance when installed. The thermal ramp-up was Ok, but the foil worked best even though it was crinkled.

I tried ceramic insulation between the oven and outer shell but this made cool down too long.

[EEVblog]

Looks like 1500W is holding up pretty well for you. I noticed that after calibration the high temperature measurement was off. The controller thinks there was only a 6°C overshoot where as there was an 11°C one according to your reference (Agilent).

I was getting that sort of difference between the Agilent and Fluke probes at high moving temps also. So there are likely several variable at play here. That's proabbly the best matching that can be expected with such ovens.

[EEVblog]

Regarding the stencil, they definetly screwed that one up.
I just got a panel like yours from PCB pool some weeks ago and before I ordered that I actually called them and asked.
Their answer was clearly : "If you order a panel you will get a stencil for the panel".

Yes, they just admitted they screwed that up, you usually get the full panel.
They have also said the silkscreen should have been better than that, and they would have offered a replacement in this case.
The EEVblog review curse strikes again!

[SPRX]

Please check below Dangerous Prototypes' video on a reflow oven. I found it very useful too:

http://dangerousprototypes.com/2013/02/14/workshop-video-55-infrared-reflow-oven-qinsi-qs-5100/

[jesuscf]

I have solder hundreds of boards using a cheapo B&D toaster oven with a homemade oven controller.  I also tested different toaster ovens with similar results.  A few things I have learned:

0) Nowadays you can cut your own high quality stencils using a craft cutter like the Silhouette Cameo.  All the juicy details can be found here:

http://dangerousprototypes.com/forum/viewtopic.php?f=68&t=5341

1) A LOT of heat escapes through the toaster oven window.  Covering the window inside with aluminum foil makes a huge difference.  If you need to see the solder paste re flowing cut an small hole into the aluminum foil!  Even with the aluminum foil there are two cold spots at the front on each side of the oven.

2) Put the temperature sensor directly on the board being re flowed.  Even better, put it closer to the highest thermal mass component that is closer to the window.

3) The convection fan doesn't help at all.  Actually, it makes things worst.  A 1500W toaster oven is capable of 1 oC/s.   Turn the fan on and you'll be lucky to get 0.5 oC/s.

4) PID controllers work quite nicely if they are tuned.   Change oven, ambient temperature, move the board around, etc. and the constants need to be adjusted.  I had much better results with a simple state machine that uses the code below.   The variables 'soak_temp', 'Soak_PWM', 'soak_time',  'reflow_temp', 'Reflow_PWM', and 'reflow_time',  are all adjustable using a keypad and LCD.  I adjust them depending on the solder paste used, the PCB manufacturer, and the profile of the components.  Attached is a screen capture of the temperature using the state machine controller.

5) Fumes.  Re flow in a well ventilated area.


switch (state)
{
    case 0: // Flux activation
        pwm=100;
        if (t_oven>=soak_temp)
        {
             state_time=0;
             state=1;
        }
        break;
               
     case 1: // Soak
         pwm=Soak_PWM;
         if (state_time>soak_time)
        {
             state_time=0;
             state=2;
         }
         break;
               
      case 2: // Ramp up
          pwm=100;
          if (t_oven>=reflow_temp)
         {
              state_time=0;
              state=3;
          }
          break;
               
      case 3: // Reflow
          pwm=Reflow_PWM;
          if (state_time>reflow_time) // Done.
         {
               pwm=0;
               state=4;
          }
          break;
               
      case 4: // Cool down
           pwm=0;
           if (t_oven<60) return;
           break;
}

[ShawnD]

..except the SSR would need to handle twice the current for the same wattage
[/quote]

Since the standard North American outlet is 120V 15A the existing 16A SSR will do fine.  Most appliances are 1500W or lower. 

[jwm]

Yeah that metal grill def caused an issue on the first attempt. Maybe some FR4 or similar as a baseplate to give more even thermals.

I was curious to see what a big piece of copper clad board would do acting as a heat shield and heat sink...

I have used a large sheet of solid copper on the tray between the board and the heating element to give signifigantly better results in a similar setup. Aluminum also worked and was cheaper but didn't develop nearly as pretty of a patina as copper did when heated.

     John

[HackedFridgeMagnet]

I thought I saw the metal tip of the thermocouple touching the pcb copper in the video.
If it was touching would this have effected the reading of the temperature, by being a 3rd dissimilar metal?

[berniwa]

Hey,

I have been using the beta layout reflow oven + controller for several years now (I got an old version of the controller - but it works the same) and I have to say, that the thing never let me down in all those years.

The oven works really fine (also over extended periodes of time) and the controller is simple and easy to use.
If done a lot of boards with the oven including all chip sizes down to large QFN packages (BGA will hopefully follow soon) and board ranging from 10 to 400 parts, all turned out really well.

Some cheap and easy tips:
- Get a piece of Metal Mesh from the local hardware store - the tray is fine for Pizza but not for small PCBs (see picture below)
- Get a spare board and attach the thermal sensor to that board instead of attaching it to the board you are currently soldering
  -> I found it really anoying to attach the sensor to board that already contained solderpaste, so I sacrificed one old PCB.
- Do multiple boards if you need to, but make sure to allign them centered between the two heaters
  -> As dave pointed out it is no convection oven, so there are temperature differences (but it's way better than cheap Infrared ovens from china anyway)

The guys at beta layout are quite nice (met them at serveral faires in germany) and produce amazing quality boards, not that cheap, but ok.
The thing I like about them is that they always deliver on time (what's really important to me) and have a great customer support.
They messed up one of my designs one time (accidently switch a layer) and after reporting it, they delivered a new PCB on the next day (1/2 day production + overnight shipping).

Greets
Berni

 

[lilshawn]

You can probably easily swap out the transformer in the housing for a 110v 6v AC transformer and/or adapter. It's just driving a LM7805. Not a big deal.
and with IEC connectors on the input and output, it'll easily plug in any cord from wherever you are from.

Don't order one because it's not available in 110 volts...order one because it's NOT!

[HackedFridgeMagnet]

Don't order one because it's not available in 110 volts...order one because it's NOT!

As mentioned previously there are increased current considerations for the SSR.

[Rasz]

That temp controller is not very good at all, 7 Degree offset and no decimal points in the reading or offset calibration value. I have previoualy designed thermocouple controllers myself, and you would need to really try hard to mess it up by that much!

Could you please make a follow up video where you trace out the analog part of the controller and check what voltage and temp reference they are using. Are they using the built in 10 or 12 bit ADC? In that case I would expect some white noise generator so that they can oversample and dither the reading.

from the way it acted when left on top of the oven Id wager a guess no cold junction compensation, just stupid opamp directly to the ADC ala cheap chinese soldering iron.

I already posted the datasheet in the video description.

I must be blind, I dont see it, is it inside video annotations or something? (those dont work for me on YT for some reason :/)

ps: Dave you blasphemer, using tablet instead of proper multimeter integrated display? Whats next? Multimeters with no integrated lcd at all? Chris would be proud

[robrenz]

I thought I saw the metal tip of the thermocouple touching the pcb copper in the video.
If it was touching would this have effected the reading of the temperature, by being a 3rd dissimilar metal?

The thermocouple bead touching another metal does not affect the reading other than possibly better heat transfer into the thermocouple than just surrounding air.

[gxti]

An isolated bit of metal touching the probe won't affect it but if there's some current path back to the measuring circuit you could have trouble. For example, even if the controller is floating, if you have the RS-232 connected to a PC it will be earth-referenced. This type of problem would result in wildly wrong readings, best case it shorts out half of the probe and you get half the temperature differential you expect. Worst case you get a bias voltage flowing through it and blow up your opamp. So floating is good. But since it's just a little bit off, it's probably not that.

[EEVblog]

from the way it acted when left on top of the oven Id wager a guess no cold junction compensation, just stupid opamp directly to the ADC ala cheap chinese soldering iron.
I must be blind, I dont see it, is it inside video annotations or something? (those dont work for me on YT for some reason :/)

Not so, it has CJC. Just expand the video description!

[trackman44]

I was wondering, is the oven controller schematic available for download, as in open source ( including the firmware)? Would be nice.

Will

[JoeN]

I have had it on my list to get a T-962A reflow oven.  But that model, while cheap ($420 on eBay), comes with some really bad reviews.  Among them people say, it smells terrible the first few times started up, contains cheap parts, is grounded poorly (safety issue!), the controller sucks (there are various upgrades from third parties to rectify this), and it heats unevenly.  From what I see from this review, you can make a very reasonable reflow oven with what is actually a very good controller at far less than the marginal T-962A costs.  If you care to, you can make it larger, just get a larger oven.  And there are toaster ovens on the market now at very low cost that include a convection fan.  I have one, it works great for pizzas.  I heard Dave mention that the 240V operating voltage is a problem for the USA, but most homes actually have 240V lines here for larger appliances.  Maybe the plug would have to be swapped out.  Not a problem.  I think I am going to have to give this a go once I find a place to source a 240V toaster oven with a convection fan.  I already see some for sale on Amazon for very reasonable prices.

[pieco]

I have used a large sheet of solid copper on the tray between the board and the heating element to give signifigantly better results in a similar setup. Aluminum also worked and was cheaper but didn't develop nearly as pretty of a patina as copper did when heated.

Another vote for using a copper sheet under the PCB! Works great.

[pieco]

The secret it to use an IR toaster - I have this one - http://www.amazon.com/Panasonic-NB-G100P-7-2-Quart-1300-Watt-Infrared/dp/B000063UZV (has been discontinued and they came out with a newer model; not sure what changed).  There is a black & decker one too but that one catches fire a lot. 

The replacement is the NB-GP110P. I have one I've been meaning to try, especially since it appears to have some built in sensors. Meanwhile, I've gotten amazing results with manually using a B&D Infrawave, that has yet to catch fire

I use lead free solder from these guys - unrefriderated shelf life of ~3-4 years from my experience http://www.zeph.com/zephpaste.htm
make sure to get their nice plunger.   

Sorry, have to disagree here. Zeph's prices, especially their shipping, is totally out of whack. They charge, on a weight-for-weight basis, more than 2x as other quality paste ($15.50/12g vs. say, $20/35g for ChipQuik). I'll agree that the paste is excellent quality, tho - perhaps some of the best I've tried.

Also - the plunger is total crap. Just awful. Now, I'm biased. See the last mailbag as to why that might be. Suffice to say, there are far, far better widgets out there that are in the same price range.

And use the smallest tips (red ones).

I would not go below 22ga with their paste, otherwise, you're straining the balls and dispensing flux, and doing it the hard way. Zeph paste is Type 3, so it has at least some particles that cannot push through a smaller tip.

[wigman27]

Hey All!

I have been thinking about building my own re-flow oven controller in the past few weeks. Perfect timing Dave!!

This is one that I really liked, http://forum.43oh.com/topic/3980-reflow-oven-booster-pack/.

Its open source and obviously comes with the code. But I was thinking if anyone here would have any suggestions on how best to write the code, obviously the code on the Beta example Dave shows is quite involved but I am mainly concerned with the main code to run the actual program. Would you sample in every cycle? would you adjust the on state at every cycle if required?

I would be really interested to hear people thoughts

Thanks again

[TheWelly888]

Yet another informative and interesting video, Dave! So many TV channels to choose from and nothing worth watching so thanks for the video! 

I wouldn't use the toaster to do both food cooking and reflow soldering - the flux fume residue will pollute that pizza, cornish pasty, pie or whatever!

[G7PSK]

Stand the board on some pieces of compressed vermiculite board, available from builders merchants, its used to line combustion chambers so will withstand repeated heating also it will not conduct the heat out of the pcb.
 

[Rory]

Isn't vermiculite a form of asbestos?

[angst7]

Isn't vermiculite a form of asbestos?

No.  http://en.wikipedia.org/wiki/Vermiculite#Asbestos_contamination

[Kiril]

Hi everyone,

If we look at the reflow profile for the solder, reflow is done and the temperature is under 175°C after ca. 210 seconds.
With the kit, reflow does not start until after 270 seconds and lasts over 10 minutes (600 seconds). The temperature also stays "forever" above ca. 210 °C (from 322 to at least 430; temp is over 230°C from 344 to at least 430) while in the reflow profile this time is limited to about 30-40 seconds.

So the question is how much this prolonged exposure to high temperature is something to worry about? I can imagine plastic parts could possibly deform slightly when heated, but what about baking the "electrons"? Does anything "bad" happen to component values? Or is this a non-issue in reality, resp. when would it become an issue?

Thanks Dave for showing us how this kit performs for real. Love the EEVblog and have learned a lot from you and the people on the forum.

Kind regards,
Kiril

[Ment11]

Hi
Has anyone here tried a small infrared reflow oven such as  http://www.ebay.com/itm/T962-INFRARED-IC-HEATER-REFLOW-OVEN-SMD-BGA-T-962-d-/260799459468?pt=LH_DefaultDomain_0&hash=item3cb8dbf08c

For $245 USD it is about the same cost as this thermal controller+toaster.

I had one of those, and it made my boards look like a piece of toast using their profile for lead free solder. The boards still worked tho, and its pretty quick. I never took it apart, or checked how accurate the temperatures are, but even from outside the build quality was not that good. Also as someone stated before, the smell was absolutely horrible the first time I used it, and it set off fire detectors.

[EEVblog]

Dave, I was in Kmart (in Australia) today and they have a toaster oven for A$35.  It has the same control layout as yours does except I think it claimed 250^oC. Would the extra 20^oC be useful, assuming of course it is real?

If using a controller, no real difference. It not, well , you might have to be more careful.

[Winston]

If they'd have used a transformer with a primary center tap and a back panel voltage selector switch, or to save money, just a jumper on the PCB, they could have easily made it 230/115V.

..except the SSR would need to handle twice the current for the same wattage

The Sharp S216S02 they use is rated 16A @ 250V and is heat-sinked to the chassis. These ovens are typically rated 1500W.

[cybermaus]

LOL

If anyone in Netherlands wants to have 'the same' as Dave, I stumbled on this

(not for me though, I'll stick to my hot air gun)

[toomas]

Thank you Dave for this video. It got me inspired and I bought one tiny and cheapest possible oven to try if it can be used for reflow soldering. Pictures, measured temperature profiles and other details are here https://www.eevblog.com/forum/projects/el-cheapo-reflow-oven/

[Tioleco]

Can I use the kit in a small oven of 600w and with 9 liters of capacity?

[peter.mitchell]

Can I use the kit in a small oven of 600w and with 9 liters of capacity?

I dont see why not, but at 600w, it will have a long ramp up time unless it is well insulated.

[itdontgo]

I have T-962A and after many burned boards and partially soldered boards I can now use it to produce perfect repeatable results.  I've made several hundred pre-production boards in that oven.

The biggest problem, as with Dave's new oven, is the base is a solid piece of metal and it conducts all the heat away from the boards (especially 0.8mm PCBs).  To get around this I have some old FR4 boards I use to lift the corners 1.6mm (or sometime 3.2mm if it has a massive copper plane or big Module.  So long as there are no parts where you lift it it will be fine.

Also one of the heaters is better than the other so I have to put the boards towards the back to get it right... which can be helpful if you have lots of copper in one place and less at the other end of the board.  You can more evenly heat it this way.  They're not that shit when you know them and what beats the toaster oven is the cooling fan which cools them after reflow ready for the next batch.

Also you want to run it through once first time before you use it to warm it up!

I'd be concerned using any reflow oven like that in an office.  I take my boards from the office and reflow them in my garage as the fumes put me to sleep!

[B_S]

Hey guys.

didn't read all the posts on this topic, so sorry if this question already was answered.

what do you think, is this device capable of reflow bga components or would you have to heat up the µC so much to reflow the solder underneath so it would destroy the controller?

background to this particular question is: a while ago i was thinking of buying a quadcopter kit, but the controller board would require bga soldering, the fully assembled board would cost 150€ ( yep, i'm from germany ) more then the non assembled one.

now i'm wondering if this reflow oven could solve such assembly problems for those kits.


B_S

[david.barker]

I bought a BETA layout controller some time ago, although I use a different (convector) oven.  I wrote a Graphical User Interface (GUI) which interfaces to the BETA layout controller, making it much easier to configure the unit to match a specific solder profile and also to monitor and plot temperature in real time.  It's a free download, for anyone that would like to use it:

http://www.firewing.info/pmwiki.php?n=FirewingUser.BETA-Reflow-Controller

[M4trix]

Looks like they have updated the reflow controller to V3 Pro. Someone already bought it?

https://www.beta-estore.com/rkuk/order_product_details.html?p=613

[wavelengthinversion]

http://hackaday.com/2014/06/16/reflow-oven-controller-with-graphic-lcd/

Here is a project much like this that showed up on Hack-a-day.

This looks like it would be much more flexible the one you tested in this video.

Just thought I would toss this in there...