Improving the T-962 Reflow Oven – Part 4

In previous posts, I show you my implementation of the 2-channel thermocouple amplifier for my T-962 reflow oven. Now I finally got the time to do the 4-channel version. For those who read Werner’s original hack for the improvement of the cheap Chinese made T-962 reflow oven,  you’ll notice that his original implementation used four thermocouple amplifiers, so that you get more accurate measurements at 4 different places within the oven. Werner’s code already has support for 4 channels disregard whether you use MAX31855 or the one-wire version MAX31850 thermocouple amplifiers.

After I submitted the board to my usual PCB house and waited for two and a half weeks, the boards arrived on my desk.

P1050322

It’s just a two-layer PCB, nothing fancy, here is what it looks like.

P1050324

Not much has changed from the two-channel version, still the usual separate analog and digital ground. I maintained the same PCB size but I squeezed in two more MAX31855. For this version, I have to used a smaller version of the terminal blocks. I believed that these terminal blocks are the smallest you can have, yet still allow the original thermocouple connections to fit. Also, I moved a few passive components to the bottom side of the board, so when I reflow this board, I’ll have to do the bottom side first.

Here is the closed up of a populated board.

P1050327

Here is the comparison for the 2 and 4 channels boards when fitted on top of the original controller board. Click on the picture to enlarge.

P1050328P1050329

After I powered up the system with the new 4-channel board installed, the firmware detected the extra channels and start using it in the measurement. With no thermocouple probes fitted, the new channels read zero, so it will not affect the measurement if you do not have any probes attached to the input of the extra channels.

Now I’ll need to get 2 new thermocouple probes and to figure out how to route the probes around within the oven. There are a few places on the side where I think I can squeeze the probes through into the main oven compartment, but I need to find ways to secure the probes yet still allow the drawer to move in and out freely without placing any strains on the probes and wires.

I’ve updated the schematic to show the extra channels. I haven’t figured out how to provide .zip files download on this site, therefore I can’t upload my project files here, but you can order your own copies of the board here.  here

–Update 26 Mar, 2015–

Bill of materials

Here’s the full bill of materials for this board.

Part Value Device Package Description
C1 47uF CAP_POL3528 EIA3528 Capacitor Polarized
C2 100n CAP0603-CAP 0603-CAP Capacitor
C3 10n CAP0603-CAP 0603-CAP Capacitor
C4 100n CAP0603-CAP 0603-CAP Capacitor
C5 0.01uF CAP0603-CAP 0603-CAP Capacitor
C6 0.01uF CAP0603-CAP 0603-CAP Capacitor
C7 10n CAP0603-CAP 0603-CAP Capacitor
C8 100n CAP0603-CAP 0603-CAP Capacitor
C9 100n CAP0603-CAP 0603-CAP Capacitor
C10 10n CAP0603-CAP 0603-CAP Capacitor
C11 10n CAP0603-CAP 0603-CAP Capacitor
C12 22uF CAP_POL3528-KIT EIA3528 Capacitor Polarized
C13 22uF CAP_POL3528-KIT EIA3528 Capacitor Polarized
C14 100n CAP0805 0805 Capacitor
C15 100n CAP0805 0805 Capacitor
C16 100uF CAP_POL1206 EIA3216 Capacitor Polarized
C17 22uF CAP_POL1206 EIA3216 Capacitor non-Polarized
C18 22uF CAP_POL3528-KIT EIA3528 Capacitor Polarized
C19 100n CAP0805 0805 Capacitor
C21 100n CAP0603-CAP 0603-CAP Capacitor
C22 10n CAP0603-CAP 0603-CAP Capacitor
C23 0.01uF CAP0603-CAP 0603-CAP Capacitor
C28 0.01uF CAP0603-CAP 0603-CAP Capacitor
D5 MBRA140 DIODE-SCHOTTKY SMA-DIODE MBRA140
D10 1N4148 DIODESMA-ALT SMA-DIODE_ALT Diode
FB1 Ferrite Murata BLM18PG471SN1D 0603 Ferrite Bead Farnell # 1515745
FB2 Ferrite Murata BLM18PG471SN1D 0603 Ferrite Bead Farnell # 1515745
FB3 Ferrite Murata BLM18PG471SN1D 0603 Ferrite Bead Farnell # 1515745
FB4 Ferrite Murata BLM18PG471SN1D 0603 Ferrite Bead Farnell # 1515745
FB5 Ferrite Murata BLM18PG471SN1D 0603 Ferrite Bead Farnell # 1515745
FB6 Ferrite Murata BLM18PG471SN1D 0603 Ferrite Bead Farnell # 1515745
FB7 Ferrite Murata BLM18PG471SN1D 0805 Ferrite Bead Farnell # 1515745
FB8 Ferrite Murata BLM18PG471SN1D 0603 Ferrite Bead Farnell # 1515745
FB9 Ferrite Murata BLM18PG471SN1D 0603 Ferrite Bead Farnell # 1515745
FB10 Ferrite Murata BLM18PG471SN1D 0603 Ferrite Bead Farnell # 1515745
FB11 Ferrite Murata BLM18PG471SN1D 0603 Ferrite Bead Farnell # 1515745
IC1 MAX31855 MAX31855 SO08  TC Amplifier SPI
IC2 MAX31855 MAX31855 SO08  TC Amplifier SPI
IC3 MAX31855 MAX31855 SO08  TC Amplifier SPI
IC4 MAX31855 MAX31855 SO08  TC Amplifier SPI
IC5 SC18IS602BIPW SC18IS602BIPW TSSOP16 I2C-bus to SPI bridge NXP
IC6 MIC5216-5.0 V_REG_MIC52055V SOT23-5 V_REG MIC5219
IC7 MIC5216-3.3 V_REG_MIC52053.3V SOT23-5 V_REG MIC5219
J1 Thermo-04 1X2-3.5MM 1X2-3.5MM 3.5mm Terminal block-2pos
J2 Thermo-03 1X2-3.5MM 1X2-3.5MM 3.5mm Terminal block-2pos
J3 Thermo-02 1X2-3.5MM 1X2-3.5MM 3.5mm Terminal block-2pos
J4 Thermo-01 1X2-3.5MM 1X2-3.5MM 3.5mm Terminal block-2pos
JP1  SPI M04PTH 1X04 Header 4 RA
JP2 Test Points HEADER-1X6 1X06 Test Points. Do not populate!
JP3 Test Points HEADER-1X6 1X06 Test Points. Do not populate!
JP4 SysFan_Out M02POLAR JST 2-pin JST XH Series 2-pin 0.1″ header, vertical
JP5 SysFan-In M02POLAR JST 2-pin JST XH Series 2-pin 0.1″ header, vertical
JP6 SYSFanCTL M02PTH 1X02 Standard 2-pin 0.1″ header. RA
JP7 DSTemp M02PTH 1X02 Standard 2-pin 0.1″ header. RA
LED1 3v3 LED0603 LED-0603 Yellow LED
Q4 BSS138 MOSFET N-CHANNEL BSS138 SOT23 Common N-MOSFET
R1 10K R-US_R0603 R0603 RESISTOR
R2 1K R-US_R0805 R0805 RESISTOR
R3 10K R-US_R0603 R0603 RESISTOR
R4 10K R-US_R0603 R0603 RESISTOR
R5 10K R-US_R0603 R0603 RESISTOR
R6 10K R-US_R0603 R0603 RESISTOR
R9 10K R-US_R0805 R0805 RESISTOR
R10 10K R-US_R0805 R0805 RESISTOR
R11 10K R-US_R0805 R0805 RESISTOR
R13 10K R-US_R0603 R0603 RESISTOR
R21 100R R-US_R0805 R0805 RESISTOR
R22 100K R-US_R0805 R0805 RESISTOR
U1 DS18B20 DS18B20 TO92 Dallas DS18B20 1-wire Digital Thermometer

 

46 thoughts on “Improving the T-962 Reflow Oven – Part 4”

  1. Oh wow, you’ve been busy, very nicely done, both of the boards!! I have been busy with other work stuff but will definitely continue supporting this setup. I can’t promise that the functionality won’t break as I can’t really test it, but if you find something strange just yell! 🙂 You noticed the #define to allow using all four channels and taking the hottest one and regulate using that? Very handy with a setup like this one if you have sensitive components..

    Like

    Reply

    1. I thought I’ll give it a go since your code already has support for 4 channels. It seems like fun to spin another version just to try it out. Also, when I review the 2-channel drawing on screen, I notice that I have not made the best use of space available, so I try to see if I can squeeze in two more channels. The result is what you see here, same board size but with 4 channels. I have a few spare boards left and I can send them to you. Just let me know where you want me to send the boards to.

      As for the #define you mentioned, do you mean the MAX_SPI_DEVICES in max31855.c? It’s set to 4 by default in your code, right? I still haven’t figure out where best to place my extra probes. Seems the oven is only good near the center and I don’t get many good results if I place the boards near the corners. Where do you place your extra probes?

      One nag I had is that during these few cold weeks we have here, the oven is having problem following the profile initially. The oven lives in the cold workshop and I need to have a few dry runs before putting the board in. Do you think we should include a pre-heat profile in the firmware which brings the oven temperature up to 40-50C degree, say, before the actual reflow profile starts?

      Like

      Reply

      1. I’d love to get one of these boards, then I can run a really funky setup with both 1-wire and I2C->SPI TC interfaces simultaneously! I’ll ping you over email!!

        I was thinking about the MAXTEMPOVERRIDE define, that’s the one that will allow you to limit the temperature to the highest readout, not an average of the first two channels. I attach the two additional probes to the circuit board(s) I reflow (when I want to profile the actual temperature if I have sensitive components). The newly added serial logger/grapher makes it very handy to get the actual board temp plotted together with the desired profile.

        For pre-heat of the oven I would recommend utilizing bake mode. That could definitely be hacked into terminating when reaching a certain temperature for a certain period of time. I have also noticed that getting the entire oven up to temperature really helps the initial ramp-up, I can only imagine that being even worse when the oven is in a cold place!

        Liked by 1 person

  2. Great upgrade, I have the small T-962 reflow oven and I am now on your firmware V0.5.0. Makes definite improvement in usability and button response. Fantastic. Whilst I noticed your mention of preheating due to cold weather, I cannot complain about cold, being cold in London UK. My problem is, that whilst the oven seems to be drawing approx 650W at 240V and lamps seem to be on 100% from beginning of the ramp-up cycle, it never reaches the preset temperature. Your nice additional tool – ramp test and PDI test shows inadequacy of my oven. In your 250degC ramp test I get up to 175 degC and PDI test does not reach reach the set max temp. OK, that’s one thing if you would have an idea what is going on. This gets me onto the wishlist, if I can bother you. Can you show the max temperature recorded during the cycle and leave it on the display with the graph, until pressing “S” button? Often the cycle finishes at 60-90 deg C, bleeps and the info is gone.
    I was thinking about adding a small fan inside the chamber, above lamps, to re-circulate the air, as in true recirculating oven, but we have not fully enclosed chamber, with fan at the back to do the cooling and perforated bottom to exhaust. The recirculation would not have to draw any fresh air in. Another thought is whether to leave lamps exposed, or shield them with a solid, or perforated sheet metal, therefore shielding IR, or allowing some direct IR heating. I think the lamp power is not sufficient for a pure convection heating.

    Like

    Reply

    1. I agree with you about the problem with the oven’s temperature ramp-up rate. I have the same small T962 model and I was surprised to find that the two heating elements are incapable of bringing the oven temperature up to the rate a reflow oven supposed to. Those elements are only rated for 300W at 220VAC each! I guess this may due to the big thermomass of the oven, which is the result of the material used and poor insulation. I was told that people got better results with modified toaster ovens because toaster ovens are often made with thinner material and they have a fully enclosed chamber! My experience with this oven is that if you set the idle speed of the cooling fan too high, the oven will not follow the ramp-up profile at all, hot air just get pushed out from the bottom. So I’m not sure if adding a small fan above will help, apart from finding enough room between the heating elements and the 2 thermocouples, you’re just facilitating more heat loss through the bottom opening.

      My current solution to the problem, though still not fully eradicated the problem, is to shield the drawer bottom with tin foil but allow some perforated opening for the hot air to escape during the cooling phase. I’m still trying to figure out where the best opening should be, you may want to give this a try too.

      As for the firmware features you requested, you really should direct this to Mr. Werner Johansson of Unified Engineering, he is the man to go to. Although he often read this blog, you may get a quicker response if you contact him directly, try submit a new issue at

      https://github.com/UnifiedEngineering/T-962-improvements/issues

      Like

      Reply

  3. Thanks for your suggestions and link! I just read about the thermocouple calibration and realised that I screwed up – by adjusting blue pots to room temperature! The /link issues mentions calibration, I will try to attempt it. I will also try your method of blanking the tray vents. I found that the most efficient setting of the fan is 12. But first is to actually calibrate it. There is suggestion of bake mode and setting higher temperatures. Do you know the best way to calibrate from scratch, or shall I go to the /link and submit an issue? I am all new to this.

    Like

    Reply

    1. I had the idea to replace the original controller right from the beginning. I had the prototype done on a perfboard. I even had the PCB made which I reuse part of the thermocouple circuit to go with the new firmware. Ever since Mr. Johansson brought us his improved firmware, I gave up on the original idea and improved the original controller with the Maxim’s thermocouple amplifiers, MAX31855. Mr. Johansson had kindly added support for my version of hardware in his firmware since v0.3.0.

      During the early days of testing, I did not go the old school way to calibrate the thermocouples. Instead, I measured the analog outputs against the digital outputs from the MAX31855. Before I built the first prototype, I bought two ADAfruit’s MAX31855 breakout broads which I used to check against the original circuit. My results indicated that the readings from the analog circuit is way off at high temperature, even though I attempted to calibrate the blue pots to return good readings at low to mid temperature. However hard I tried, the on-board analog circuit always return lower readings at high temperature. Like you, I eventually messed up the settings causing the controller not performing correctly. That’s why I went down the digital thermocouple amplifier way and never look back.

      Like

      Reply

      1. Yeah, the analog TC interface has “some” issues to say the least. It can be off by quite a big amount, but there are several people that uses them with good results (together with the cold-junction mod where this thing started back in November 2014). An impressive amount of code from several different places have been put in the firmware since then! I think the improved firmware have all the functionality of the original shipping one at this point. Inching towards a 1.0 release.. 🙂

        I just added an attempt to explain the calibration steps to perform on the analog TC interface to get the gain (and offset) adjusted (here https://github.com/UnifiedEngineering/T-962-improvements/issues/57 ).

        Like

  4. Is it correct to say that the temperature control is distorted by a lack of proper cold junction reference?

    Would it be easier/cheaper to actually provide the cold junction using a Peltier unit?

    Like

    Reply

    1. The working of thermocouple comprises of two dissimilar metals. These are joined together to form two junctions, which are maintained at different temperatures respectively. The thermocouple cannot be formed if there are not two junctions. Since the two junctions are maintained at different temperatures an EMF is generated within the circuit and it is the function of the temperatures of two junctions.

      As Mr. Johansson described in his Wiki, The original firmware did not take the temperature measurement at the cold junction of the thermocouples. Instead, they made the assumption that the cold junction is always fixed at 25 degree C, which is incorrect. Since the controller is placed inside the oven and during normal usage, the temperature above the oven compartment will rise. Therefore, the calculation of the actual temperature within the oven will be incorrect.

      I’m not sure what you mean by using a Peltier unit, but the ideal case would be to take readings directly from the cold junction of the thermocouples. I guess the original designer tries to save a few components and come up with this bad design. So the simplest and cheapest solution Mr. Johansson provided for us is to use the Dallas DS18B20 1-wire temperature sensor to measure the temperature near the cold ends.

      Like

      Reply

  5. I see that this board also includes the PWM fan mod built in. Pretty nice. I am wondering if there is any difference in your SPI->i2C based design over the 1-wire design? Obviously the PWM mod works with your design and the current firmware, but what about the 1-wire design? Im just trying to get a better understanding between the two and understand any limitations.

    Awesome work BTW!

    Like

    Reply

    1. Thanks, I’m glade you like it. There is no reason why 1-wire design cannot be used here. My board was designed based on SPI verion of the TC amplifier because I already have those devices on hand, therefore, I need an I2C-to-SPI bridge to make the devices accessible by the controller. Obviously, if 1-wire devices were to be used, the analog part of the board will need to be redesigned appropriately.

      Software wise, there is no different. Mr. Johansson’s firmware will scan all known devices upon boot-up and make use of what’s available. If you look at the code, I believe the 1-wire devices has priority over the SPI devices, if non of these devices were found, it will fall back to use the original on-board analog design. As for limitations, both 1-wire and I2C are multi-drop bus design so I guess the only limitation I can see here is that you’ll need to solder one extra wire!

      Like

      Reply

      1. Thanks for the clarification! I finally got around to loading the code in an IDE that linked all the symbols so that I could easily navigate it. I see the OneWire and SPI libraries that check for the devices. You are correct. The OneWire devices are checked first, and if absent, the SPI is then checked. I also see that the cold junction information is digitally read from the device, which explains why the DS18B20 can be ignored. However, I noticed that you have a mounting point for a DS18B20 device. If the cold junction temp data is digitally read from the SPI device, what would be the purpose of the DS18B20 in your design? I also see that it is un-populated, but I was just wondering why it was there at all.

        Also, I see that you provided a link to purchase PCB’s of your 2.5 rev board (THANKS!) from the PCB house. Can you also make available the BOM? Have you ever considered selling some prebuild boards? I think that you have done a great job with the layout and mounting considerations. It really makes it look like a clean installation.

        Like

      2. My original idea was to use the DS18B20 to measure the enclosure temperature, which can be used to adjust the PWM rate of the system fan. However, Mr. Johansson’s firmware took care of that by reading the cold junction information from the TC chips, which essentially does the same thing, so this part of the circuit is now redundant and should be removed in next revision. Thus, the reason it’s unpopulated.

        I don’t have much people asking about assembled kits or bare PCBs! I guess I can make them available, if there is sufficient interest. However, apart from plugging in the board, there are other modifications to the original controller such as replacing the tall filter capacitors with a low-profile type, removing the old TC screw terminals(not strictly necessary, but will allow the board to sit lower down) and several soldering joints to the required signals. Do you think buyers are willing to do all these additional mods?

        Like

  6. Right, at the moment the DS18B20 will not be used if any of the digital TC interfaces are detected. The idea was that the highest temperature detected from any of the cold junction sensor would be used (that meant that the DS18B20 could be mounted to the original controller board and attempt to sense the TRIAC temperature for instance). I have both the DS18B20 sensor and 1-wire interfaces fitted in my oven to be able to test the software with different setups. I haven’t had time to populate the SPI board yet but the idea is that I should be able to test all three.

    The PWM:ing of the system fan will use whichever cold-junction temperature the system detects and it works just fine with the 1-wire version of the sensors too (that’s what I use myself on a daily basis). The 1-wire board I designed was done before the PWM mod was suggested.

    Thanks for the interest in the “improved software” project, it seems to be useful for a lot more people than I initially imagined!

    Like

    Reply

    1. Measuring the TRIAC temperature, that’s a good idea. In that case I shan’t be removing the DS18B20 in future revision of the PC board.

      Let me take this opportunity to say thanks, Werner, for providing us with the ultimate solution for this affordable reflow oven. Like many others, I would have gone down the route, replacing the controller with an Arduino based design, which is what I was doing originally. It’s a shame to have to replace the original controller because it wasn’t that bad a design, it was let down by the bad firmware!

      Like

      Reply

      1. You are most welcome! I’m glad that others find it as useful as I do myself – I’m using it almost every day!

        Like

    1. Yes, that’s possible for larger oven (T-962A) which “ES Technical” used in their installation manual. However, I’ve a smaller version and it doesn’t have much room inside for me to play with. Also, I try to avoid doing any mechanical modification to the casing if possible. Unlike “ES Technical”, I’m using 4 channels here, the two new channels will be placed very close to the board that I’m “baking”. It would be nice to have those TC sockets but at the end I still need to place the probes back into the drawer in order to measure the temperature. So my wish was to have those probes loosely “attached” to the tray, yet with enough slack to reach all four corners and allow the drawer to be pulled out fully without imposing any strain on the connections.

      Like

      Reply

  9. Hi all. I am new to this board. Just did the Unified Engineering firmware upload to the small T-962: added the cold junction sensor and made also the electronics cooling fan on/off mod. They all seem to work fine.

    HOWEVER, I discovered something which I don’t know why is happening: the left side of the inside of the oven runs colder than the right side. And that was confirmed with two temperature probe measurement, placed just near where the oven probes are. The readings are pretty much the same for all temperatures (external probes and internal), but the concerning thing is that the left & right temperature difference inside the oven may be up to 50 Celsius! It can also be sensed by hand, after the oven cools down a bit, allowing to touch the opened tray.

    The heating elements glow just about the same, so I can’t suspect them. But probably the flow of air inside goes more to the left than the right;I wonder how, given the way the PWM fan acts inside, and the cooling holes in the oven…

    Gain adjustments does not compensate for this behavior (I don’t even know exactly what they do, beside the heating elements are front & back not left & right) or how to use them; I only compensate the left & right TC readings, so that they will be similar to what external probes measure.

    I wonder how to fix this. Is this happening to you as well? Any workaround, to level the temperature inside the oven?

    I was thinking of using “Reflect A Gold” to coat oven’s interior, so the IR radiation from heating elements would reflect more inside. But then, the flow of air might be the same, still dividing the temperature inside.

    I’m out of other ideas… Any suggestions?

    Thanks in advance.

    Like

    Reply

    1. Yes, there are imbalances between the left and right hand side of the oven. For my oven, the original TC readings can differ by 40 degrees at high temperatures (i.e.> 150degC), that’s why we use the extra TC probes in the mod and have them placed very close to the board you’re “baking”, my external probes only differ by a few degrees and the firmware will use that extra information to compensate any variances.

      What is your minimum speed setting for the fan? Try keeping it as low as possible, I set my to 5.

      Like

      Reply

      1. Thank you for your answer. I thought it may be something with the software, but it seems that the air flow does this. I am thinking how to prevent this…, maybe cover the left hole above the trail rail? Do you think Reflect A Gold foil coating may help with heat loss & thermal inertia inside the oven?

        Like

      2. I’ve never tried the “Reflect A Gold” foil method, so I can not comment, but let us know how much improvement you gain from that modification. Be careful how you cover the holes as the oven needs to ventilate during the temperature ramp down period and the fan is already at it’s max. OK, opening the tray during that cycle may help, but you’ll have to stand by the oven every time you “bake” a board.

        I often find that I get good repeatable results if I let the oven runs through a cycle after the first power on, particularly during a cold day, so that all parts of the oven have reached it’s proper “working” temperature before I start “baking”.

        Like

  10. Covering both holes does make a difference, but not in a good way. I also covered the left hole only, but there isn’t much difference in the readings.

    On top of that, SET temperature differs about +20 C from ACT, and the highest measured temperature, on the right, with an external probe, is almost +40 C from SET. Very confusing.

    Like

    Reply

  11. Wow, that’s a great project!
    I’have been following this since long time, but always been too lazy to fix/upgrade my oven, till yesterday that I literally burned a really expensive PCB.
    I have the T-962A, will your PCB fit on top? or is only for the T-962?
    Any chance you can share the Gerbers file via mail? My PCB manufacturer is way faster on delivery times.

    Thanks
    Federico

    Like

    Reply

    1. T-962A is bigger than T-962 with more space inside. You should be able to fit the PCB on top of existing controller, you’ll just need to get the right PCB stand-offs. You can get the Gerber files here.

      Just for interest, which PCB house do you use? How much does it cost and what’s the turn around time?

      Like

      Reply

  13. Hi

    First a big thank you for sharing the PCB and firmware to the rest of us 🙂

    I just applied power to my T962C after connecting the new PCB into the existing circuit – and I have a few problems :

    X1 & X2 works fine, the same goes for the DS18B20 and the Fan-PWM – but L & R temperature is still the internal (original) circuit and not the new PCB 😦 – could that be related to the wire on the 74HC04 on the pictures not explained in the text ?

    I have checked for short circuits and solderings for the right connections.

    I had a problem with the BOM not having all info – example ; what is the value of the ferrite beads – it would be much easier for the rest of us if you provided the ordering number for your supplier …

    Best wishes

    Like

    Reply

    1. I used the same ferrite bead as used in Maxim’s MAX31855 evaluation kit. Its part number is Murata BLM18PG471SN1D, Farnell part # 1515745. I’ve updated the BOM accordingly, sorry for the confusion caused.

      As for the patches around the 74HC04, they are just 10K 0603 pull-up resistors soldered onto pin 1 and 3 in the tombstone position with the other ends connected to Vcc pin of the 74HC04. This is to prevent the heater and the exhaust fan being turned on while the CPU is placed in firmware update mode. During that mode the CPU pins are floating, the heater and the fan are on at 100%, thus the 74HC04 has nothing to do with the new thermocouple amplifiers.

      The firmware should detect the present of either MAX31850 or MAX31855 via the SC18IS602B I2C-to-SPI bridge. If any of these devices are detected, the firmware will switch to these devices instead of the internal analog amplifiers. Do you have all 4 digital thermocouple amplifiers installed? It might be possible that the firmware will use the analog amplifiers if only 2 digital amplifiers are found.

      Werner, can you confirm on this?

      Like

      Reply

    2. I had to have a quick answer, so I downloaded the source code and discovered that the code demands that I connect both L & R temp input on the new PCB for it to display those channels …!

      I had tried to “bench test” it with the one extra thermocouple I have, in order to avoid to assemble the whole oven just to discover that I had made a mistake / assembly error – this is not possible unless you have two extra thermocouples to plug into L & R (SS0 & SS1) at the same time.

      Now I only have one issue left with the oven – thermal mass => finding the right insulation !!!!

      Like

      Reply

      1. I see that you replied while I was posting again – thank you for your quick reply 🙂

        I can see that I haven’t used the same ferrit beads in my first unit (I’m not planning more, but I had to order 10 PCB’s, so if need be …) but so far it seems to work OK after I discovered that I must connect at least two thermocouples.

        Like

      2. Yes, the thermal mass is an issue. I always run the oven through at least one reflow cycle with no PCB inside just to bring the rest of the oven up to temperature before I do the actual reflow, this works for me particularly during the winter months.

        Like

  14. What electronics parts supplier did you order the Bill of Materials listed parts from? Here in the US We have Mouser, Digi-Key and Newark. The list does not seem to generate many parts when uploaded.
    My T-962 is on its way and also have ordered the circuit board from Hong Kong’s DirtyPCB.

    Like

    Reply

    1. Hi Gary, sorry for the late reply, here in UK, I ordered majority of my parts from Farnell. I believe they’re called Newark in the US.

      Like

      Reply

      1. I’ll try them. I suspect these are pretty generic SMD parts for the caps and resistors. THANKS for the reply it helps.

        Gary K8IZ

        Like

  15. Hello. Is your board layout still shared on DirtyPCB? I was thinking about ordering one and building this for my T962 but I could not find it. The above link does not work for me. Thanks

    Like

    Reply

    1. Seems that something has changed over at DirtyPCB.com! The board is still available and I’ve updated the link accordingly. Click here for quick access.

      Like

      Reply

      2. Yes, you’re right, according to NXP’s website, the I2C bridge (SC18IS602BIPW,128) that I used has reached the end of life, but they have another one (SC18IS602BIPW/S9HP) in development. Not sure if the new chip will retain the same footprint and properties, but I guess I need to look into alternatives in case it doesn’t! Thanks for the info.

        P.S. My local dealer, Farnell UK, still has over a thousand in stock until it is exhausted!

        Like

  16. Hi guys,
    I succesfully made the modifications and finally I can say “bye bye” to that original awful controller.
    First of all I want to thanks very much to everybody has worked to this project and shared to the community.
    Then I’m here to give my 5 cents.
    After some measurements with another tehrmocouple thermometer, I have realized thet the real temperature in the reflow area is much different from that one measure dby the oven thermocouples, because they are much distant.
    So i’im doing some experiments.I hav epulled out the two thermocouples from their original locations and found a way to bring them on the plate.
    Need some tuning now with offset and gain parameters. Also need to be careful not to damage the thermocuoples when you open and close the drawer. Also you sacrifice some space on the drawer for the location of the thermocouples, fixed on a small PCB.
    But now you have a real measurement point….
    I have some pictures but don’t know how to post them here.
    Any comment is appreciated,

    Like

    Reply

  17. Hello,

    I recently runned tests on my T-962 and found the same results as reported here above:

    1) temp ramp too low (around 0.8°c/s when 2°c/s would be good)
    2) big difference between the measured temp and the temp of the board since thermocouples are around 1 cm above the board
    3) 40°c gradient from left to right

    The consequence being that only a small area can be used to reflow anything and that it is hardly possible to solder lead free at all. I did and the solder paste does not melt enough well.

    I don’t see the benefit of changing thermocouples or the controller board. The problem comes from the heating elements that are uneven and too weak.

    Question:

    1) Does anybody knows a replacement part for the heating elements? / a tutorial on this?

    2) What about adding 2 more heating elements disposed in reverse to get twice power and even heat ? I think the original heating tubes can be found at low price on the web. Will it be too powerful and burn the oven? I think the triac (rated 16A) should handle the current if an heatsink was added.

    Like

    Reply

  18. Hi ,
    Does anyone has the exact position of the fixation holes of the original controller board. I Would need to design a similar boar but with an wifi-UART bridge on it and I would like to not completely dismantle my over for now.

    Like

    Reply

Leave a comment