Reflow oven and Solder paste?

http://www.ebay.com/sch/i.html?_from=R4 … te&_sop=15

It is very easy to use, the only downside is the lead content (so not ROHS compliant), personally I’ve probably absorbed far too much lead over the years to start worrying now, and of course slowboat wait.

I have no particular seller to recommend, since I haven’t bought any recently, but I suspect they are one as good as the other, and since you are only risking a couple of dollars, and are protected by Ebay/Paypal I wouldn’t loose too much sleep over this sort of transaction going pearshaped.

I’m going to get some paste locally this morning from Jaycar, but its far more expensive that that stuff you linked to, so I’ll order some of that today.

PS. I splashed out on a cheap reflow oven, as I gave up on converting a toaster oven, as the cost to convert it, looked only slightly less than buying a pre-made one

I the oven is a T-962 from eBay.

However, I’ve already reflashed the replacement (open source) firmware, as the stock firmware isnt that good.
I still need to fit a DS18B20 (onewire thermometer) to do the cold junction compensation, and then I need to look at calibrarting it.

So I’ll let you know I I get on

However, I’ve already reflashed the replacement (open source) firmware, as the stock firmware isnt that good.
I still need to fit a DS18B20 (onewire thermometer) to do the cold junction compensation, and then I need to look at calibrarting it.

not actually used mine as yet…

stephen

https://github.com/UnifiedEngineering/T … provements

The DS18B20 OneWire thermometer is required, as the unit does not have cold junction compensation. Which means that its thermocouple temperature readings, of the 2 x ThermoCouples (TCs) in the oven, will be incorrect if the unit is not operating in an enviroment that is at 20deg C.

I added the DS18B20 yesterday, however it looke like the calibration of one of the two TCs was not correctly calibrated in the factor, because with the new firmware it outouts debug over the serial ISP header on the control board, and I can see one of them is reading 25 too high.

There is an Offset and Gain correction screen for the TCs in the new firmware, but my TC is at the limit of its correction range, as it only lets you change the offset by +/- 25

I can of course rebuild the firmware, in fact Ive already does this, and Im using the version I compiled, so I can change these limits to allow more than 25 deg, but I wonder if i need to adjust the opamp gain on the board using one of the trim pots. However Im reluctant to mess with the analogue settings at the moment.

Im not sure if I would have known about this problem with the original firmware, as I dont think it displays the individual temperatures anywhere.

I will try swapping the two inputs to confirm if its an analogue signal conditioning problem on the control board or a faulty TC.

I also have an SPI thermocouple interface module which uses a dedicated chip (M31855) which interfaces via SPI.
So I may try connecting the erronous TC to the interface and read the temperature using a Maple mini. Im pretty sure there is an Arduino library for this chip.

BTW. Im also using a standalone digital thermometer which has a thermocouple probe, which I have positioned in the middle of there the PCB would lie, inside the oven.
But I need to double check the accuracy of this meter in melting ice water and boiling water

Well, at the moment its virtually useless.

One thermocouple is still reading at least 25 deg higher than the other one.

I swapped over the actual thermocouples and the left channel still reads at least 25 deg too high.

So I tried connecting both the left and the right thermocouples to a digital multimeter I have, that has a thermocouple input, and both thermocouples are fine.

So the problem seems to be the analog electronics that conditions the signal from the thermocouple.
I found a jpg image of a schematic that someone had drawn up. I’ve attached a bit of the schematic (sorry I can’t post it all but its too big and the quality is bad because it seems to be a jpg only)

t962_tc_inputs.jpg

t962_tc_inputs.jpg (82.33 KiB) Viewed 799 times

T962.png

T962.png (47.64 KiB) Viewed 796 times

I would assume that when working correctly, the opamp output should read somewhere between 0 and 3v3 depending on the gain (RW3 or RW4 presumably being used to control this) and the vale of the thermocouples.

If the opamp is bad, then you wont see any change when you fiddle with those pots. BTW do you not have a scope? This would perhaps be more sensitive than your DMM. The PigScope might even do the trick given the voltages we are dealing with here. Also double check the wiring to your thermocouples. The opamp may be one of these.

I did try adjusting the pot but It didn’t seem to make any difference to that channel.

I took the PCB out of the oven, and tried to measure whether it is broken or not, and it appeared to be OK, however its hard to tell as I was measuring its value while still on the board.

Someone else on Hackaday had a similar issue, and his opamp was faulty, so I wonder if I have the same problem.
These units are built as cheaply as possible, and I’d not be surprised if they were using reject parts etc.

I’m probably going to use an alternative fix which is already supported by the new firmware.

Its possible to use the Maxim Max31855 thermocouple to SPI chip, however the unit does not have any free SPI pins, so the firmware communicates with the 31855 via a I2C to SPI bridge IC ( as I2C is already used with some NV storage on the board)

I have some 31855 modules, but not the bridge IC, but I do have an Arduino Pro Mini 3.3v which I should be able to program to emulate the bridge IC, as the firmware only uses one command in the bridge.

However, I doubt I will have time to write the code for a few days

I did manage to get a small amount of solder paste locally (on Saturday), but I did need to drive about 20km to find somewhere that had the “lead” rather than the “lead free” version.

As most places seemed to have sold out of the “lead” version, I suspect its more popular than the “lead free” (as I’ve heard its more tolerant to the temperatures)

Re: Reflow oven

Currently I’m wondering if it was a mistake to buy it. But I was going to try to make a reflow oven using a toaster oven, and I suspect that would have been even more problems than buying one that almost works, and then modifying it to make it work properly (i.e with new firmware and better electronics).
At least the mechanics, e.g. the heaters and heat insulation on the T-962 are not too bad and it generally seems to be able to follow a heat profile, where as I think the modified toaster ovens have difficulty getting to 270 deg etc, unless you make loads of modifications and add extra heating elements etc (which would be hard to do and expensive – as you’d need to buy another oven to extract the heaters)

I’ve started to build the replacement electronics for the problematic Thermocouple inputs, but using some MAX31855 modules (from eBay) and an Arduino Pro Mini (3.3V)

The reason I’m using the Pro mini, is that I need I2C slave functionality and its not implemented in libmaple

Anyway. I connected up the Pro mini to the I2C pins (soldered to some resistors on the board), and after initially wiring SDA and SCL around the wrong way, because of an issue with someone elses post on how they wired their’s up.

I can get the Pro mini to listen to the traffic that the firmware sends out on address 0x28

Initially I was not able to send any data back to the reflow oven, because the I2C onRequest even was not being triggered. Fortunately I found a posting about a bug in the twowire lib, with a patch (comment out one line)
https://github.com/helenarobotics/Ardui … 2b39b0fbca

So I applied this fix and I now receive onRequest events, which I should be able to use to send data back to the oven.

But I will need to carry on with it this evening when I get some free time.

Thanks.

I found the explanation for why the AVR lib was not working, on the Arduino.cc forum

http://forum.arduino.cc/index.php?topic=125704.new#new

The ESP thing looks similar, but at the moment, www.esp8266.com is not responding, so I’ll take a look later to see if it seems to be the same bug

A handy alternative to SMD oven is a hot air SMD rework station, like this one that I have:
http://www.aliexpress.com/item/scotle-8 … 69194.html

I have the same hot air reflow station, but I have wanted to build a reflow oven for some time.
But then I thought it was cheaper to buy one.

However in the end, I’m not sure buying a cheap one, was such a good idea

where does the 270 come from, or is it because it’ll need the ramping up rate for the ‘take to 230c’ phase
from the ‘soak’?

stephen

I think the general issue with simple toaster oven reflow ovens, is that they dont have enough insulator or enough heating power to quickly ramp up after the soak phase, to melt the solder.

Im sure it is possible to convert a toaster oven to do this, but I was hoping that a cheap commercial oven would be a better option.

BTW.
As a bit of an update to my progress in using Max31855’s with the oven, I have successfully managed to get the open source firmware to think it has 2 external 31855 based TCs, by sending dummy data from the Arduino Pro Mini.

However when I send real data from a 31855, the temperature values shown by the oven are jibberish.
I suspect I have a byte ordering issue with the 32 bit data from the 31855 when transferred via I2C Wire.write(buffer, len)

The issue with the data, was indeed the byte order of the 4 byte buffer which was read over pseudo SPI via the adafruit lib

The byte ordering seems to have been reversed somehow, so I wrote some code to swap buffer bytes 0 and 3 and bytes 1 and 2, and now the firmware in the oven accepts the data and seems to read the correct temperatures.

I still however need to make a change to the firmware in the oven, because the bootloader in the Arduino Pro mini runs for about a second, so I need to delay the I2C search feature in the oven firmware by about a second, so that the pro mini is running when the I2C bus is scanned.

I also need to tidy up all my wiring with the pro mini etc

But at least I think that I’ve addressed the thermocouple issue and I also will have the option to add at least 2 more MAX31855 based thermocouples if I want.

I’m tempted to put a thermocouple on the tray right under where I would put the PCB, because the thermocouples in the oven are currently at the top of the chamber, and read higher than the readings I see with my stand alone thermocouple thermometer which is taped (with Kapton tape), just above the base of the drawer onto which the PCB would sit.

But I’ll need to order another thermocouple probe and also some more MAX31855 modules

58p uk free shipping

i’ve max6675 on boards, offer expired though

MAX31855 on a board £4.10 ukp
bare chip ~£7

top of the heap adafruit £16-£17

stephen