First SMD reflow attempt

The LED boards turned up from Seeedstudio today. Here’s one:

My first board!

My first board!

That didn’t turn out too bad. The text for the resistor values is too small, I’ll know next time.

I thought I’d better get on with it to see what happened so I started by testing the board for continuity where I expected it. That seemed fine. Next up was applying solder paste to one of the boards to give this process a trial run. That took a little while and I learned a few more things not to do – like putting your hand on part of the board that you’ve just applied solder paste to πŸ™„

Here’s timelapse video so it’s not too tedious to watch…

I guess I wasn’t too happy at that point, I seemed to have large variations in the amount of solder paste and none of the components were particularly straight. Still, I wasn’t going to be too disappointed if this one didn’t work out. Time to load up the oven:

Before reflow

Before reflow

Not the best of photos but you can probably get a sense of the general placement of the components. It was time to hack the controller firmware (which I still haven’t got around to finishing properly) and see what happened…

I watched the whole process through the magnifying lens of my desklight. At around 200ΒΊC some of the paste started to melt and turn shiny and, miracle of miracles, the resistors started to rotate into perfect alignment. Surface tension rocks πŸ˜†

At aroundΒ 235ΒΊC all the solder seemed “done” so I shut down the oven and opened the door – the oven is relatively slow to warm up so it had been close to that temperature for a while. I hoped it would be “long enough”.

Cooked and cooling down

Cooked and cooling down

I’ll admit to being gob-smacked at this point. It really looked like everything had gone the way it should. The fairly random amounts of solder paste end up giving fairly similar results at each joint. All looking ok really.

Neat!

Neat!

I’ve got to be pleased with that. But does it work?

Red

Red

Green

Green

Blue
Blue

It works! 😯

I’ve labelled the Blue and Green pins the wrong way around so I’ve maintained tradition by cocking something up. Have to admit to being really pleased with the result – my first ever circuit board, my first ever reflow attempt – and it actually works!

I’ll practice with the rest of the boards at some point and then get around to designing something a bit more useful. Has to be worth chucking 3 transistors (and resistors) into a breadboard and PWMing the nuts off this with an Arduino though.

Posted in Reflow oven | Leave a comment

Improvements to the “Man Lab”

I’m usually struggling with the fact that my computer is on a different desk to where I do soldering and generally put bits together. Yes, I can create “dead tree” versions of datasheets quite easily but I do find paper to be, mostly, a pain in the arse – especially when I print on A5 to save space and then can’t read the bloody things because they’re too small πŸ™

A while ago, I backed HDMIPI on Kickstarter – this is an excellent project to produce a 9″ HD screen for the Raspberry Pi. Unfortunately, it’s not going to be ready for a couple of months and, to be honest, I’ll probably not use it for what I bought it for. The plan was to use it as a screen on my work bench that I could display PDFs on. It’s probably too small for that but it’s a nice screen so I’m sure I’ll find something to do with it.

Meanwhile, we’re upgrading our desktops at work and I’ve been clearing out the buildup of crap obsolete IT equipment that you inevitably get. Since I look after hardware selection, purchasing etc. etc. then I tend to make a point of offering any unwanted kit to anyone in the office and I also tend to make myself last in the queue – it wouldn’t feel right if I just cherry-picked all the best bits before the others got a look in.

This time, there were 3 old monitors that I decided were no use any more (a blurry 21″ and two OK but only 19″). One of the 19″ ones went immediately, followed by the 21″ a few minutes later. I’d started to think I might like one of the 19″ ones but didn’t hold out much hope…

A couple of days later, it was still there! So, how did that turn out…

A Subtle Modification

A Subtle Modification

I wanted the monitor on its side to give me a better view of an A4 format page and also because it makes better use of the space.

Fits in nicely!

Fits in nicely!

Another view...

Another view…

 

I’ve got the Pi powered from a USB Hub with a 2A PSU. There’s also the dongle for a Kingston wireless mouse, a Rii Mini “Elegance” wireless keyboard and plenty of scope for expansion. I’m driving the monitor through a cheap HDMI to VGA adaptor – a few changes to the config.txt file were necessary to get this to work:

# This is for 1024x768 19" TFT Monitor, placed on its side
hdmi_group=2
hdmi_mode=35
display_rotate=1

That takes care of setting the resolution to 1280×1024 (native for the TFT panel) and then rotating the display through 270ΒΊ to get the orientation I need.

Datasheet

Datasheet

 

Electrickery!!!

So that gives me a lot more flexibility on accessing information when I’m working at that desk – it will certainly save a lot of twisting around to look at stuff on a screen behind me and no more need to keep shuffling bits of dead tree.

I’ve also got the Pi on my desk with a camera on a USB Snake Light mount – well the light was pissing me off but a bit of insulation tape on the USB contacts soon sorted that out πŸ˜‰

I should be able to take pictures of whatever I’m building and even do movies and time-lapse quite easily now.

It looks fairly cool too! 😎 πŸ˜›

Posted in Raspberry Pi, Uncategorized | Leave a comment

Cascade LED tube investigation

Today I decided to see if could figure out how those cascade LED tubes I mentioned recently actually work. Here’s one:

IMG_20140125_135428

 

They’re simple to take apart (I’ll leave re-assembly as an exercise for the reader πŸ™‚ ) – just remove the clear plastic heatshrink from the top and it all falls to bits:

IMG_20140125_135634

 

Here’s a closeup of one of the LEDs:

Nothing special then

Nothing special then

The first test I did was with the voltmeter. Connecting to “-” at the top of the board and then to each pin of the LED in turn I expected to find one that read 0V at all times. Nope. Both of the LED pins have a fluctuating voltage relative to the incoming “-” pin with one bigger than the other. Not sure what that means but we’ll try to find out. Incidentally the 5V incoming also has a fluctuating voltage relative to “-” πŸ™‚ – the PSU really is a piece of shit by the look of things.

Right, now for some more investigation…

Look at one LED

Look at one LED

I connected the end LED up to a couple of probes on the Open Logic Sniffer (and the GND probe to the “-” connection). That was as clear as mud. It looks like there’s some kind of PWM going on but that’s about all I can see really. I wasn’t able to relate it to what the lights were doing and there’s definitely a signal on both pins. Weird.

Taking a closer look at the chip…

mystery_ic

Hmmm….

Well that was useful. So it’s some kind of “no name” chip with 8 pins and a ground/heat dissipation pad underneath.Β Time to dig a little deeper…

Looking at circuit board traces

Looking at circuit board traces

As you can see, it’s not possible to see where the tracks go to try and associate pins with traces since they appear to duck under the chip. Not sure what good that would have done anyway, since I don’t know what the pins do but, whatever.

I wondered if sampling 4 LEDs at once would shed any more light on things…

4 LEDs, 8 pins to measure

4 LEDs, 8 pins to measure

This time, the answer was conclusive…

As clear as mud...

As clear as mud…

That’s a “No” then πŸ™

Time to put this away and not worry about how it works. I think I’ll just appreciate them for what they are πŸ™‚

Posted in Uncategorized | Leave a comment

Reflow oven performance

Not had much time today for one reason or another…

Didn’t have much luck with automatic tuning of the PID parameters for the oven. Did manage to knock it into shape manually though – there were some horrendous overshoots before tuning, now it pretty much hits the mark.

What we need to achieve, according to the datasheet for the solder paste I bought is:

Temperature profile

Temperature profile

So, after faffing around with tuning I decided to compare the oven performance with what was required of it. This was done by simply setting the target temperature (Setpoint) to 240Β°C and then turning the oven off and opening the door once that point was reached. How does that do?

Required vs. Reality

Required vs. Reality

So, the oven is too slow to meet the required profile. Problem? I suspect not to be honest, most other profiles I’ve seen have large error bands rather than sharp lines. The ideal profile hitsΒ 240Β°C and then cools down – I’d be tempted to have this oven hit, say, 235Β°C instead to limit the exposure of components to high temperatures. The only way to see if this works is to test it and I have some circuit boards on the way to do just that.

I’ll finish the software for this in the next few days then it can operate stand-alone in the garage when I need to reflow stuff.

I’m using a Raspberry Pi on my work desk to allow me to access the PID controller. I must admit that I’ve been getting very frustrated with the arse that is X11 on a slow machine with only a fast ethernet port. Also the editor in the Arduino IDE is, in my humble opinion, a steaming pile of shite. That doesn’t make for a happy Dave πŸ™

The solution was easy enough:

  • Tell the Arduino IDE to use an external editor
  • Dump /home/pi onto my server and NFS mount it back to the raspberry pi
  • Edit locally using Sublime Text (with syntax highlighting and git support)
  • Use the “Upload” button in the Arduino IDE (which still runs on the pi) to compile and upload to the PID controller

When I get around to it, I’ll switch over to Arduino Makefile – may as well compile the damn stuff locally too. All I then need to do is to add a “remoteupload” target to ssh in to the pi and run avrdude from there on the hex file compiled on my linux box. Sorted, and faster, and not at the mercy of the pi lunching it’s SD card…

Sublime!

Sublime!

That was worth spending the time on. I can see me needing the “Bench Pi” more and more often so having a better workflow is a must.

Posted in Reflow oven | Leave a comment

Reflow Oven – final hardware steps

Ok, picked up some block connectors at the local hardware store today. They didn’t have any of that fibreglass clad cable so I just went with what I had to hand. It’ll do πŸ™‚

An easy way to connect

An easy way to connect

The above shows the use of a 5A connector – that should be more than enough for a 240V 620W oven.

Additional blue wire + SSR

Additional blue wire + SSR

I stole the neutral lead out of some flex I had kicking around. It looks a bit on the thin side but I can’t help but feel it must be enough for this application. Doing it this way keeps the relay well off to the side where it shouldn’t really get that hot.

SSR control wires

SSR control wires

I’m trying to use up the stranded wire I bought so I have yellow as GND and brown for the “+” pin which turns the relay on. There doesn’t seem to be much danger of the low voltage and mains coming into contact, unless the relay melts, or something.

A simple hole drilled in one of the indentations in the case and some hot glue for strain relief and that’s the control part taken care of.

I’ve tested everything – nothing blew up πŸ™‚ The oven is now controlled by a high signal on the brown wire. Time to get on with sorting out the software – mostly adding the PID library and implementing a state machine to cover reflow oven functionality. Of course, I stll need to add connections to the controller box to control the relay but that’s just “plug and play” so I’ll get it done soon.

All looking good so far πŸ™‚

Posted in Reflow oven | Leave a comment

smARtMAKER All-in-One Kit + extras

I finally received my stuff on Tuesday – quite a wait seeing it was paid for in November 2012!

All-in-One + a few other bits

All-in-One + a few other bits

I’m not entirely sure what I’m going to do with it! I’ve learned a fair bit since it seemed like a really good idea. I’ll tackle lots of things now that I wouldn’t back then – like the reflow oven project – and I’ve even got my first circuit board “in the post”.

Still, I’ve got it so might as well figure out if I can use it for quick prototyping – less wires would be good, I hate wires πŸ™

I took the smARtCORE L (Arduino Leonardo compatible) and sat it on a “QUAD” board with the free RGB LED and the 5 LED board. A quick hack of some code on Gist that I found a link to and here we go…

[embedplusvideo height=”800″ width=”600″ editlink=”http://bit.ly/1aU5oeI” standard=”http://www.youtube.com/v/w0mDPBqmzU4?fs=1″ vars=”ytid=w0mDPBqmzU4&width=600&height=800&start=&stop=&rs=w&hd=0&autoplay=0&react=1&chapters=&notes=” id=”ep1302″ /]

I still feel there’s some potential here for simplifying interaction with Arduino type stuff – especially as it comes with Leonardo, Uno and Mega compatible boards. There’s also an Android ADK board which might be interesting to have a play with.

One item was missing from the stuff I was sent. That’s been posted out to me from China. Also, a good many of the additional boards I ordered are on “back-order” – unfortunately this includes the Electric Imp adapters.

I guess I’ll keep plugging away with it in the background to see what I can make of it.

Posted in smartmaker | Leave a comment

Reflow Oven – sorting out the 240V side…

Here’s a tear-down of the reflow oven. I was hoping to finish this tonight, but as you’ll see, it seems better to pick up a few bits I’m missing and do a far better job than the one I was planning. It is mains after all so no harm in doing it right rather than cocking it up.

Getting the oven apart wasn’t too bad once I had the correct security bits (which did turn up on Tuesday):

Side view

Side view

The side view show the power going in with red being the live wire.

Connection to timer switch and power setting

Connection to timer switch and power setting

The timer switch (on the right) has the live lead going to the bottom of it. The “output” goes to one side of the indicator bulb and back out to the ends of the heating elements on the right-hand side of the unit. The other side of the bulb connects to one side of the power connector and the incoming Neutral feed.

Removing the timer

Removing the timer

The timer knob pulls off to reveal the screws that hold it in place.

Timer

Timer

The real job here then is to replace the timer with a solid state relay:

Timer replacement

Timer replacement

Live + 2 x "Output" wires

Live + 2 x “Output” wires

The SSR needs to connect to the live as “input” and the other two wires (heating elements + bulb) as “output”. I was originally going to use a bit of Veroboard but I can’t get comfortable with that. It’s going to be a bit of a bodge job if I solder the wires and the SSR onto some crummy bit of board and then stuff it in the base of the unit somewhere.

I started looking at the side of the unit where all the power comes in. It can’t be any hotter out there and there’s much more room to place the SSR where I want. I’ll pick up some 5A connector strip and a piece of whatever you call that heat resistant wire then I can easily pull out the live wire and safely join those other two to a new wire going back to the inside of the plastic side panel.

Easier out here...

Easier out here…

It looks like I’ll be able to stick the SSR on the side out of the way (more hot glue then πŸ™‚ ) where I would imagine it would stay fairly cool being low down.

So, not finished yet but I think I have a plan…

Posted in Reflow oven | Leave a comment