-
A little while ago I had inquired here about the BRTRO-420 reflow oven, and found only one other user. I decided to buy one and have soldered some 30 or 40 boards with it by now, so I guess it is time to share my experience:
Purchase:
I bought mine via AliExpress from „Charmhigh SMT Pick and Place Machine Store“. it cost 572 US$ including shipping via DHL Express to Germany. Customs charged 94,43 EUR for import value-added tax, plus 2,96 EUR in duty. No extra DHL charge because I have an account with them.
One caveat for non-commercial buyers inside the EU: The unit was not marked with a CE sign, so if customs decides to inspect the package this could mean the import is rejected. For a commercial buyer this also could mean extra hassle.
Build quality and electrical safety:
I opened the oven and had a look inside before using it for the first time. Overall, the build quality is pretty good. but the grounding of the enclosure is questionable. The ground wire from the inlet IEC receptacle is bolted down under one of the transformer mounting screws (typically an extra grounding bolt is required by regulations). In addition, the lower half of the enclosure is only grounded via the sheet metal screws that hold the upper half of the enclosure. Due to plenty of paint on all parts, this is a questionable contact. In my case there was continuity between top and bottom half, so I decided to leave it like that (and the room it is in is supplied via an 30mA RCD).
Proper Kapton tape has been used to secure the thermal insulation, so there is no nasty smell once the unit warms up (although there is a little „hot equipment“ smell).
Usability, reflow etc.:
The unit can store 8 different reflow profiles. One can set the parameters (duration and final temperature) for pre-heat, heat, reflow, hold, and cool down steps. The four button operation is ok, although modifying a profile takes a lot of button presses of course. The logic of the final temperature for the cooldown step is weird, though: It is also used as starting temperature for the preheat step.
The unit has IR heaters with a power of 2500W (according to seller, I didn’t measure the power, but had a look at the electricity meter while the heaters were 100% on, and the numbers seem about right), plus one air intake, one air exhaust, and one circulation fan. There are 2 thermocouples hanging in the air above the circuit boards.
The temperature indication is somewhat noisy, but good enough. I don’t know how the single displayed value is calculated from the two TCs.
The heater power is just enough to follow roughly a usual heating profile (6 min for a complete solder process). I would recommend to do one empty run if the unit is cold, after that the initial PreHeat and Heat phases are followed more closely (without that, the heaters are basically 100% on until the reflow temp is reached).
PID control is ok, but could be better, especially I don’t like that it does very short and unnecessary pulses on the air intake/exhaust fans once the temperature setpoint is reached. Cooldown becomes slow below around 120C and very slow below 100C, although the in/out fans are running. Typically not an issue, just open the drawer and wait 30sec for the final cool down.
It is annoying though that the drawer does not have a latch. The air intake fan seems to be a bit stronger than the exhaust fan and seems to spin up faster. This is not an issue during cool down (where the fans start with a lot of short pulses before being powered continuously), but it is when starting a new process before the oven has cooled to the final temperature of the cool down step, which is also used as the starting temperature of the preheat step.
So upon process start, the actual temperature is much higher than the set point, so the fans start full speed. This pressurizes the oven and the drawer opens about one inch in my unit. I started to lay a hammer in front of the drawer as temporary fix
. Would be nice if the process would just start at the current temperature to avoid the completely nonsensical fan spin for 30 secs or so. With the drawer not fully closed, the soldering results are bad because the front part of the oven will be colder.
The oven is not air tight, so it will blow out some warm/hot air. Flammable items should be kept at a distance, but my rubber desk top seems ok. But most of the solder fumes go through the exhaust, so all good there.
The controller shows the actual temperature curve, but updates a few pixels at a time, which is weird (maybe they don’t take a datapoint for every display pixel, but for a fixed time interval which is longer than one pixel duration in the plot). At the end of a process, the unit will beep and keep the fans spinning until stopped by the user.
It would be a nice feature if one could select to wait after each step until the temp. set point is reached before continuing with the next step, but thats not there.
Soldering results:
This is an IR oven with some convection added, but heating is still predominantly IR. I made some experiments and found that with the factory-supplied lead-free profile, green, red, and blue boards could be soldered just fine, although I had to add about 5 seconds to the reflow time to improve reliability with small inductors and such (228C target temp. for reflow).
Black PCBs were overheated, though - the silk screen became brownish - so I attached a TC to a bare board and measured 30C more than the oven indicated. So I had to set up a second profile for black PCBs with a lower reflow temperature (217C). With this profile, my black PCBs could be soldered fine without visible damage.
Because most of the heat is IR, there is shadowing. 0805 resistors right next to bigger parts (little transformer) were not soldered a few times. In addition: Don’t use the outer 5cm of the drawer, there are some cooler spots there (and more shadowing, in the center you are always „between“ two IR heating rods), so I got some un-reflowed components when placing a board there. Two boards 10x15cm plus a narrow one next to them worked fine (i.e. at least 20x20cm homogeneous solder area).
I have added photos of my first boards with the oven. The solder paste is lead-free Henkel/Loctite GC10 SAC305T3 (Sn96.5Ag3Cu0.5), printed with a stainless stencil, hand placed.
Little transformers like Murata 78253/55JC are soldered fine (beware of shadowing though, had one un-reflowed joint in 10 boards with specifically that transformer), inductors like Fastron PIS2816 work fine, also uSD-sockets and 40x0.5mm FPC connectors for displays. Huge inductors like Murata 6000B series don’t seem to work at all (at least I didn’t want to fry all the other components by increasing the reflow time sufficiently), no surprise given their huge thermal mass.
Summary:
I would buy this oven again. Sure it would be nicer to have a perfect oven, but that would require a much higher budget which is not justified for me. It does what I need, and is definitely an improvement over hot-plate reflow soldering. The temperature controller is usable as is and the power good enough to do lead-free soldering of regular components, but not for huge inductors or caps. But this is for small scale soldering, so doing them by hand is ok (at least for me).
-
This looks very similar to the T937, a bit smaller, and the smoke outlet is at the top which is better.
Can this one connect to a pc via rs232? -
There is no RS232 connector on the oven, although the controller board seems to have an unused connector with two optocouplers that looks suspiciously like an isolated serial communication interface.
-
Hi,
thanks a lot for your review. I am thinking about getting one.
How is your experience now, a bit later? Any news?
Were you able to solve the door issue?
thanks -
Hi,
no, I have soldered only a few more boards, nothing to add.
I haven't done anything about the drawer issue yet - in the long run I will install some kind of latch, but for now the hammer/weight works well enough. I have more pressing things to take care of in my workshop...one of them would be a permanent location and exhaust hookup for the oven, right now I have taped the flexible duct to the basement window.
Best regards,
Tassilo -
How about gluing two neodium magnets to the front sides of the drawer so when it is closed the magnets would hell keep it in place
-
Very interesting and helpful review.
Looking at your second board, there seems to be a solder bridge between pins 29 and 30 of the ARM chip. In your experience of this oven, how much touch-up do you need to do of fine pitch components? -
right now I have taped the flexible duct to the basement window.
Note that forceful extraction of air from the oven is extremely dangerous. Make sure you don't suck out the air from the exhaust. If you do, you keep the thermocouple cool while the board is getting 4th degree burn.
-
there seems to be a solder bridge between pins 29 and 30 of the ARM chip. In your experience of this oven, how much touch-up do you need to do of fine pitch components?
This has nothing to do with the oven. Rather this kind of defect indicates stencil printing problem.
-
Very interesting and helpful review.
Thank you. You probably mean pin 21 and 22 (the smaller dot is the pin 1 marking). That is intentional, there is a trace connecting the pads (these pins are VDDA and VREF+). Not my brightest idea to do it like that
Looking at your second board, there seems to be a solder bridge between pins 29 and 30 of the ARM chip. In your experience of this oven, how much touch-up do you need to do of fine pitch components?, in newer designs I have made the connection outside of the pad area, that makes inspection possible without looking at the design printouts.
The amount of touch-up depends on the quality of stencil printing and the steadiness of my hand when placing components. Paste printing is a compromise: I only do very few boards at a time, so I don't put enough paste on the stencil to print in one smooth motion. It involves moving the paste back and forth a bit. If one is not careful, it is easy to force some extra paste through the holes and under the stencil. The extra paste maskes solder bridges likely -> touch up required. It also does not help to smear paste between the pins when placing a QFP with a shaky hand
. Once paste printing is good and the chip placed without too much moving and correcting, solder bridges are very unlikely.
This is a problem mostly for 0.5mm pitch TQFP packages. QFN seems more forgiving.
But even with bad stencil printing and smeared paste, with the reflow oven bridges seem to happen less frequently than with my old hot-plate soldering technique. -
No worries. Right now only the exhaust fan of the oven is used: without that running there is practically zero air flow through the duct (is is a basement window with a fly screen insert, originally intended to ensure air supply for the previous heating system of the house, and now about 1/3 of it is covered with a board to hold the duct). If the duct becomes much longer, an additional fan might be needed, but I would connect that to the existing one in the oven so that there is no flow during reflow.right now I have taped the flexible duct to the basement window.
Note that forceful extraction of air from the oven is extremely dangerous. Make sure you don't suck out the air from the exhaust. If you do, you keep the thermocouple cool while the board is getting 4th degree burn. -
The solder amount / fillets on the pads look great. Can you talk about stencil design - opening, thickness, etc.?
-
The solder amount / fillets on the pads look great. Can you talk about stencil design - opening, thickness, etc.?
I have set the stencil to 10% or 2 mil (whatever value is lower) smaller than the pads, mostly to increase the separation on the 0.5mm pitch pads a little bit. I am using framed stencils from Elecrow (www.elecrow.com). I haven't specified much, so they usually change the shape of the 0805 pads: They add a small notch on the inner side of the component, so that the stencil cutouts look more like a [ and ], this reduces the amount of paste that is under the component a little. I have some stencils where they did this and some where they didn't, does not seem to make a huge difference for my process. They also add additional support crosses for very large stencil cutouts.
Oh, the material/thickness: 0.12 mm thick stainless steel, no electropolishing (paste release is good enough for my purposes), I tried thinner and thicker, this seems to be the best compromise for my boards.
I have used laser-cut polyester/PET (Mylar) stencils before (from www.pololu.com and www.smtstencil.co.uk), but stainless stencils are so much easier to work with (and for 20$ for a 37x47cm stencil, no point in not using them). I like the framed ones. I have a very basic stencil printer for them, but before I had that I was basically taping the board to the table, place the stencil on top, secure it with two C-clamps, print paste, carefully remove the C-clamps and lift the stencil (flip one side up first). For simple boards with no fine-pitch parts, I sometimes just place the board on the table, put the framed stencil on top, hold it down with one hand and apply the paste with the other. The frame helps a lot to push the stencil down and prevent a gap between stencil and board.
-
I am using framed stencils from Elecrow
Could you post a photo of one of your framed stencils? I've been trying to work out exactly what a framed stencil is/looks like.
I'm working on a project where I will want to do some SMD work, so stencils are a new world for me. I've also been using Elecrow for my PCBs and quite happy with them. -
Hey TassiloH,
Could you please do a dry run for Ref at 245C and above please? Mine seems to not able to shut down the two front element after the reflow completed. V1.09
Thanks
-
Thanks for the details. Do you think T-961 or T-960 would be better than this for medium scale? Have you done 100pin tqfp chip? Do you recommend for 100-200 boards per day?
-
Sorry for the late response, just to busy...
@TomS_
See attached photo. A framed stencil is just the stainless stencil glued to a frame made from rectangular hollow aluminum profiles. If you need the frame depends on the way you do the paste printing - There are stencil printers that need the frame, others don't. If you do it by hand without a printing fixture, you have the choice, although I found the framed ones better for handling. If ordering from China, the frame will add to the shipping cost.
@jpliew
Right now the oven is not operational - I need to clear out the new location for it, but due to construction this was delayed and still is not ready, and it is not my top priority, because I still have enough assembled boards. I will try as soon as I get to it.
@istipb:
I have done 100 pin TQFP (see my original post), and also 144-pin. Soldering in the oven is fine, paste application is the critical part here. If you are talking about 100 boards 100x150mm in size, you can load a maximum of 2 I'd think, so you need 50 cycles. At 10 minutes per cycle with loading/unloading, this is more than 8 hours of continuous work. I think that is not what this machine is intended for (small runs, prototypes). If you have 10x20mm boards, you can load them all at once and it will be done in 15 minutes.
The T-961 and T960 are conveyor multi-zone ovens. A different category, and probably more adequate for 100 to 200 bigger boards per day, although I don't have any personal experience with them or ovens of this size in general.