T-962 Reflow Oven anyone done the GitHub firmware mod?

[Fleetz]

Was heading down the toaster oven and modifying it with SSR’s and PID controller but eBay had a 20% off promotion and a seller had a T-962 yum cha reflow oven a silly price so grab it.

Was fully aware that the T-962 had some design issues which I have corrected with Kapton tape and proper earthing.

Was wondering if any members have recently done the https://github.com/UnifiedEngineering/T-962-improvements upgrade? In particular using v0.5.2 firmware that is now available?

I just recieved from Aliexpress a USB to TTL FDTI programmer to enable flashing the firmware but the micro USB connector has a pin that appears to have folded back on itself and has made connection impossible. Thought whilst I am getting the replacement would reach out to see if other members have completed the UnifiedEngineering GitHub firmware mod using the latest firmware?

Planning on using FlashMagic to flash the NXP processor.

Great to hear from any others that have gone the journey?

Cheers..

[at101]

Yep, I even went so far as to make my own profiles and re-compile the firmware. I did the cold junction compensation modification also since I was in there. It is also worthwhile to solder wires in and run them out a vent so you can program it with the case closed.

Make sure to clean up the masking tape mess while you are in there.

I know one side runs a bit cold but I haven't been using it enough to adjust things to compensate. It seems to work quite well and is generally within a few degrees of the set point over the entire run as long as I let it warm up before doing the run with boards in.

[Fleetz]

Yep, I even went so far as to make my own profiles and re-compile the firmware. I did the cold junction compensation modification also since I was in there. It is also worthwhile to solder wires in and run them out a vent so you can program it with the case closed.

Make sure to clean up the masking tape mess while you are in there.

I know one side runs a bit cold but I haven't been using it enough to adjust things to compensate. It seems to work quite well and is generally within a few degrees of the set point over the entire run as long as I let it warm up before doing the run with boards in.

Thanks for the reply.

Already done the Kapton and earth mods. Just wait for the replacement FDTI USB to TTL programmer to arrive to to do the flashing of the firmware.

Good idea, was thinking about fitting a socket to bring out the ISP to the rear.

Not been able to find out without flashing the firmware how many profiles are programmed in already to the v0.5.2 firmware? Could you please let me know what is the standard programmed profiles are?

Cheers...

[rs20]

The issue with the sides being a different temperature is really disconcerting. We had one of these burn a board, fill the room with smoke, and it's been sitting unused ever since. I've been using a simple frypan ever since and been getting much better results.

And yes, this was after the tape had all been replaced, firmware upgraded, and cold junction compensation implemented.

[KE5FX]

The issue with the sides being a different temperature is really disconcerting. We had one of these burn a board, fill the room with smoke, and it's been sitting unused ever since. I've been using a simple frypan ever since and been getting much better results.

And yes, this was after the tape had all been replaced, firmware upgraded, and cold junction compensation implemented.

What firmware were you running?  I haven't seen any toasted PCBs with the ESTechnical modifications, just a bit of uneven heating near the corners of larger PCBs.

[Fleetz]

The issue with the sides being a different temperature is really disconcerting. We had one of these burn a board, fill the room with smoke, and it's been sitting unused ever since. I've been using a simple frypan ever since and been getting much better results.

And yes, this was after the tape had all been replaced, firmware upgraded, and cold junction compensation implemented.

What firmware were you running?  I haven't seen any toasted PCBs with the ESTechnical modifications, just a bit of uneven heating near the corners of larger PCBs.

Is there another firmware modification available for the T-962? Are you talking about the https://github.com/UnifiedEngineering/T-962-improvements firmware or a different one?

If different could you point me to a link? Thanks.

Cheers...

Update: Found the modification you were referring to a UK based mod. Unfortunately the price is prohibitive at 185 pound by the time we pay shipping, GST tax, exchange rate etc it will be well over $420 AUD

[Fleetz]

Seem that at101 and rs20 have had totally different experiences?

Wonder if this simply different firmware versions with possibly rs20 version used having a bug which was later corrected and at101 is using that version?

[at101]

Not been able to find out without flashing the firmware how many profiles are programmed in already to the v0.5.2 firmware? Could you please let me know what is the standard programmed profiles are?

There are three solder profiles: AMTECH 4300 63SN/37PB, NC-31 LOW-TEMP LF, and AMTECH SYNTECH-LF. There are two more profiles for a ramp speed test and a PID control test. I replaced the NC-31 LOW-TEMP LF profile with one for MG Chemicals 4860P Sn63Pb37 since that is what I'm using right now.

You can grab the source and take a look at them in reflow_profiles.c. I've also attached it.

Seem that at101 and rs20 have had totally different experiences?

Wonder if this simply different firmware versions with possibly rs20 version used having a bug which was later corrected and at101 is using that version?

I'm running the v0.5.2 and the only issue I'm having is the oven temperature lagging the set point by a few degrees. I'd guess it's mostly a function of the oven design as it has only two lamps for heating and the two temperature probes are located at the top center of the oven between the lamps. The firmware does have calibration/compensation settings but I haven't gone to the trouble to set it yet.

I haven't burned any boards - they usually come out with a few of the larger components not quite wetted out. When developing my profile, I did run it a few times from both a cold oven and a warm oven to check for issues. I had to adjust the dwell times a bit to get the results I expected to see. After all that work, I am comfortable letting it run unattended in the garage for the duration of the profile.

[Fleetz]

 
[/quote]

There are three solder profiles: AMTECH 4300 63SN/37PB, NC-31 LOW-TEMP LF, and AMTECH SYNTECH-LF. There are two more profiles for a ramp speed test and a PID control test. I replaced the NC-31 LOW-TEMP LF profile with one for MG Chemicals 4860P Sn63Pb37 since that is what I'm using right now.

You can grab the source and take a look at them in reflow_profiles.c. I've also attached it.

[/quote]

Thanks for that appreciated.

I am new to SMD so the design and fab learning curve is ongoing. I am a retired EE having initially come from the valve era (might give my age away) and lived and worked through the solid state THP era, long time retired and now doing up projects which include SMB as a hobby, mainly for my grandchildren. Just designed my first SMD project in KiCad and have the board back fro JLCPCB. So please bare with me.

Been able to find the MG Chemicals 4860P Sn63Pb37 paste here in Australia and the review are very good which no doubt I should why your using that solder paste. So it would make sense to flash the version you have done into the T-962.

I am new to GitHub and source code compiling however don’t have a problem flashing firmware with the compiled HEX. Was wondering if I could respectfully ask for your compiled HEX v0.5.2 with the MG Chemicals 4860P Sn63Pb37 profile? Modifying the reflow_profiles.c source code and compiling is something a bit beyond me at this stage. If you could help with a compiled HEX that would be very much appreciated.

Thanks again for your earlier reply.

[at101]

I am new to GitHub and source code compiling however don’t have a problem flashing firmware with the compiled HEX. Was wondering if I could respectfully ask for your compiled HEX v0.5.2 with the MG Chemicals 4860P Sn63Pb37 profile? Modifying the reflow_profiles.c source code and compiling is something a bit beyond me at this stage. If you could help with a compiled HEX that would be very much appreciated.

I don't really know how to do GitHub but I've thrown it back up there to share: EccentricWkshp/T-962-improvements

I had quite a time installing the toolchain to compile this so I wouldn't recommend it as a first project. If you need a profile changed, let me know and I'll see if I can help out and recompile it.

[Fleetz]

I am new to GitHub and source code compiling however don’t have a problem flashing firmware with the compiled HEX. Was wondering if I could respectfully ask for your compiled HEX v0.5.2 with the MG Chemicals 4860P Sn63Pb37 profile? Modifying the reflow_profiles.c source code and compiling is something a bit beyond me at this stage. If you could help with a compiled HEX that would be very much appreciated.

I don't really know how to do GitHub but I've thrown it back up there to share: EccentricWkshp/T-962-improvements

I had quite a time installing the toolchain to compile this so I wouldn't recommend it as a first project. If you need a profile changed, let me know and I'll see if I can help out and recompile it.

You Sir are a gentleman!

Thank you for making the compiled 4860P v.0.5.2 HEX available, very much appreciated!

Regards...

[Fleetz]

I just picked up on the GitHub site this mod for expansing the number of thermocouple to 4?

https://github.com/UnifiedEngineering/T-962-improvements/wiki/Better-thermocouple-interface

What is not clear (well at least to me) is if this is mandatory for the v0.5.2 firmware upgrade? I wasn't planning on doing this modification and was hoping to use the two standard thermocouples.

Does v0.5.2 firmware support the standard T-962 thermocouples?

Cheers..

[at101]

I read this the other day then forgot to reply.

The firmware does indeed work with the standard thermocouples - no need to modify. In fact, I somehow missed that page until now.

[Fleetz]

Thanks for confirming that at101.

Just waiting on a USB>TTL programmer and will be doing the upgrade. In the meantime some MG4860P and MG4900 pastes arrived yesterday which I will try out on one of the JLCPCB prototype boards I am using to experiment with hand soldering using a 858D+ and the standard firmware T-962 but with the kapton and earth safety modifications completed.

Want to see what the difference is between soldering leaded and non leaded pastes are.

[Fleetz]

Whilst waiting for the programming interface to arrive I manually entered in a profile for MG4860P leaded solder paste.

I generated a time verses temp profile table based on the graph provided by MG Chemicals. Then inputed the profile into wave 7, one of two user definable profiles available on the T-962. Basically for every 10 seconds you input a target temperature until the 4860P graph is complete then save the profile.

A bit primitive but at least it give you an opportunity to create a profile. I tested the profile on a 100x100mm test PCB which I loaded component in the corners, edges and centre. It didn’t do a bad job with only one end of a resistor in one corner that didn’t solder.

Anyway some feedback for whatit is worth...

[Southerner]

Is there a better method/approach than using the T962?  For a while I saw a lot of chatter about converting toaster ovens for smt soldering and then it seemed to die.  Most were expensive for a hobbyist that does a board once in a while.  I was trying to find some of the projects on OSHPark's sharing site and could find boards but not the project that caused the board to be designed.

What is the current preferred method to make/get a cheap toaster oven for smt soldering now?

Thank you.

[Southerner]

I did go ahead and order a T962 oven.  The price of the T962A was more than I could afford and was not sure that I would need the bigger space.  800 watts is easier to run for power.  I am curious how much difference there is between the T962 and the T962A.  Also I can't quite understand what mods are needed to the stock units other than reflash to new firmware and replace masking tape with Kapton tape.  Are there other things that need to be done to them?  More insulation?  Where is it needed?

Thank you.

[Southerner]

I did go ahead and order a T962 oven.  The price of the T962A was more than I could afford and was not sure that I would need the bigger space.  800 watts is easier to run for power.  I am curious how much difference there is between the T962 and the T962A.  Also I can't quite understand what mods are needed to the stock units other than reflash to new firmware and replace masking tape with Kapton tape.  Are there other things that need to be done to them?  More insulation?  Where is it needed?

Thank you.

[jpanhalt]

I added the Maxim temperature sensor DS18x20 for cold junction compensation.  Be sure to get one that will work in "parasitic" mode (1-wire).  If you get the other version, all is not lost, but you need to pick 3.3V  from somewhere on the board.  That is not hard to find.

[shai]

Hi all, total newbie here to reflowing. I just got the oven and flashed the latest unified engineering firmware along with adding cold junction compensation. The screen shows it is detected. However, I'm struggling to understand how to adjust the temperature readings on left and right? At first my multimeter was showing lower than actual temp but then it caught up and eventually my oven was showing 160C while my multimeter was showing over 200C and so I shut it off fearing a thermal runaway event. How do you adjust the gain/offset settings in setup tab? I don't see any instructions anywhere?

[jpanhalt]

Get a decent TC temperature meter that you have calibrated against a reliable standard (I used mercury thermometers.  My Fluke was quite close.)

Then put that probe next to each oven probe.  (I inserted through the bottom grill.)  Run a profile and adjust offset and gain until it is reasonably close.  Do the other side.  Doesn't take long and you will find the agreement pretty good. Mine undershoots the screen's max temp, but is OK for Kester leaded solder.

Initially, I thought a low fan speed would help. But I ended up increasing the minimum fan to about 30% or so, as that gave better temp control.

[shai]

Get a decent TC temperature meter that you have calibrated against a reliable standard (I used mercury thermometers.  My Fluke was quite close.)

Then put that probe next to each oven probe.  (I inserted through the bottom grill.)  Run a profile and adjust offset and gain until it is reasonably close.  Do the other side.  Doesn't take long and you will find the agreement pretty good. Mine undershoots the screen's max temp, but is OK for Kester leaded solder.

Initially, I thought a low fan speed would help. But I ended up increasing the minimum fan to about 30% or so, as that gave better temp control.

Hi jpanhalt, thanks for the quick reply! What I'm struggling to understand is what is "gain" and what is "offset" ? I am assuming that offset is just temperature offset, but how about gain? How do you understand how to adjust gain? I don't want to play with the numbers randomly in fear that I'll overheat the oven. If you can provide a scenario and how you would adjust gain/offset, that would be super helpful. Sorry - complete noob.

At 234 Celsius on my oven, it showed 300 Celsius on my multimeter. My multimeter thermocouple was attached to a PCB in the center of the tray and my PCB is slightly elevated off the tray. This is 70 degrees difference... I'm unsure how to adjust gain/offset to fix this.

[jpanhalt]

Offset and gain work together.   Consider the equation for a straight line:  y = mx +b

This analogy is not perfect, but look at "gain" as the slope, m, and offset is b.

Starting at room temperature, I set offset to make the two readings (reference and, say, oven right) about equal.  Then ran a constant temp profile to about 150 to 200°.  At that point, set gain so they again match.  Had to repeat once once or twice making adjustments until they matched fairly closely.  Then repeat for the other oven TC.

Finally, I ran the leaded solder profile and compared the reported temp of the oven, which is an average of the two TC's with my reference put in the middle.  That is where I found it helpful to increase the minimal fan speed to get better mixing.

The final profile showed pretty good match between reported temp on the screen and the desired profile on the screen up to the final rapid heat phase.  At that point, reported temp lagged a little.  At its peak, reported temp was about 4 or 5°  low (I don't recall exactly), which was just fine for my Kester leaded solder paste.