Once I started using SMD components, it widened my horizons but it also started me down the slippery slope. I started with careful soldering and decided to move to hot air soldering. And I could really do with a microscope to lend a helping hand.
The whole thing was triggered when I had to solder ESP modules on to a PCB. Most chips like SOIC were OK. I did not have to deal with TSSOP or TQFP. I stick to 1206 passive components, going down to 0805 only if necessary. I did try soldering the ESP modules by hand but the results were abysmal. Time to go down the hot air route...
The ESP modules are about 10x11mm and have castellated vias on a 1.1mm pitch. Castellated vias are a bit more challenging than pins and the pads are so close that it is easy to short them out. And if you do, you can't tell. Not without a microscope.
I use one of those three bracket, spring-loaded PCB clamps. My PCB is 40x20mm and can be viewed in its entirety once the microscope is set up correctly. I used solder paste dispensed using one of those conical plastic nozzles instead of the metallic needles. The needles are colour coded by size. The plastic ones are called TT Conical Needles. I used a blue one, which is 22g in size. I have heard about people having difficulty squeezing out the solder paste but I did not experience this, probably thanks to the choice of the right nozzle. I did have a similar tube of flux but I did not use it - I hate the detritus and I did not miss the lack of flux. The hot air station was set to 300°C and the airflow at level 6 (out of 8). There was just enough paste on each pad. The ESP module has quite fine pitch so the paste is just laid out in a line. Larger pads need more paste. The components are then placed and pushed down a bit so the paste acts like an adhesive. This comes in handy when the air starts blowing but the chips must stay put.
With the Atten, you have to pick up the handle and wait for it to reach the right temperature. Then point it in the general direction and then zoom in into components. Keep switching your vision between the PCB and the microscope image - one of the few times the position of the TV is a problem but worth the larger image size. You can see the flux starting to sizzle. Then the solder melts and the magic happens - the components are pulled into ideal position! While this may work for smaller components, for larger ones like the SOT-223 voltage regulator and more so, the ESP module, accurate positioning to start with, is key. This is where the microscope is indispensable. With a previous attempt, I did find a way for the surface tension of the molten solder to pull larger components into place. Put a blob of flux under the large component. When heated, this molten flux allows the larger component to float and to be pulled into position.
With the wide range of PCB solder mask colours available, I tend to go crazy and order different PCBs in different colours. For this PCB, I had ordered two lots. For some strange reason, it is cheaper to order 2 lots of 5 rather than 1 lot of 10. And the chosen colours were black and white. Big mistake! All component leads, pads and solder are shades of grey and it becomes a bit hard to tell which is which. Next time, I will settle for red, purple or even boring green.
Atten 858D+ Hot Air Station
Pretty basic and cheap, I got one a long time ago on the recommendation of a video by the Multimeter Hunter. It has a digital display for the temperature and a dial for air speed. There is a stand on one side (it can be moved to the other, if needed). When the handheld tool is placed on the stand, the heater is turned off. The air keeps blowing to protect the heater and then switches off when the heater reaches 100°C. The display then switches to sleep mode. When you pick it back up, it slowly heats up till it reaches the preset temperature.
Eakins 48MP HDMI microscope
Rather than settle for less than perfect bundles, I went for individual components, all from the same dealer on AliExpress - Eakins Micscope Store (sic). The individual components were:
- 1080p 2K/4K 48MP HDMI microscope camera
- Aluminium alloy stand
- 63x-180x adjustable zoom C-mount lens with 0.35x and 1x Barlow lenses
- Adjustable 144 LED ring light
- IR remote control (CR2032 battery not included)
- Mini-CD, USB cable, HDMI cable, wall warts for the microscope and LED ring light
The camera comes with a US/EU power supply and an AU mains adapter. It has a single USB port to connect it to a PC. It comes with capture software on a mini-CD. There is a slot for a micro-SD memory card. It also comes with an IR remote. As it has only one USB port, it does not have the facility of hooking up a mouse. On some models, this gives you some additional functions - mostly around measurement. Roughly the same functions are available with the software for the USB connection.
There are auto-focus versions of this camera available. There is also a mind boggling number of models to choose from. The auto-focus models are 3-5x times more expensive. With a lack of documentation, it is hard to choose or even differentiate between different models. This is compounded by lack of model numbers, even from the same manufacturer. But I guess the price point triumphs them all. The main thing to focus on is probably the sensor. Support for a mouse is a feature that may be worth looking for for those who care. Not me!
The mini-CD has the USB video view/capture software with it. Hook up the microscope to the PC using the USB cable it comes with and you can see the image on the PC application. The HDMI output is independent so you can use both outputs at the same time. It allows snapshots and video capture. This was boon as I could not track down my collection of micro-SD cards nor the SD to micro-SD adapter. Also available is the measurement function. Once calibrated to the actual image on the screen, you can use it to measure a wide range of parameters - length, radius, angle, etc. As it needs to be calibrated to the image, you pretty much have to calibrate it against a known distance with each session. The image above was calibrated against the four pin connector (300 mils). Once that was done, you can see multiple measurements taken. Each is displayed with the value.
The stand has 295mm of travel and a plain black 150x230mm table. It has a knob for vertical travel of about 50mm. It can be repositioned and tightened along the shaft and has a stop ring to stop it from crashing on to the table by mistake
The 1x Barlow lens is useless and can probably be used to protect the objective lens. 0.35x is too low. But these are the only options with the lens. With no Barlow lens, the working distance is about 90mm and all you can see is probably 10x10mm. I tried the 0.35x lens but just could not focus it. Turns out the working distance blows out to 400mm with the shaft position at 560mm - too much for the 300mm shaft. The solution? You can see that in the photo at the start. I flipped the sliding holder upside down along with the camera. I then placed the stand upside down on a shelf above my workspace and put a heavy object on the base. This barely gets me to the right height but it is usable. The working distance makes you forget the microscope is there when you are working. The LED ring light is powerful enough, even at this distance. The viewing area is about 50mm - enough to cover my entire PCB. The stop ring is still attached at the end of the post - I don't want the entire contraption to crash on to my PCB. A 0.65x and 0.5x Barlow lens are on order.
The setup menu can be accessed by either the buttons on the back of the microscope or the IR remote. As a manual does not exist, I will list the settings below. Push the Menu button to show the menu or hide it. You can also use the Menu button to go back to the main menu. Use the up/down buttons to select. The right arrow selects an option. For a multi-value option, use the left/right arrows.
- Language - a choice of English or Chinese or Chinese
- Line Set - Turn on or off crosshair with measurements. Also, you can place a number of lines on the screen with the ability to control orientation (vertical or horizontal), position, colour, thickness. You can set up to 16 lines. To make them go away, set them to Off in Direction one by one.
- Exposure - Use this to set brightness or set it to OFF for auto
- AWB - Set white balance. This is one case where I would override the default of auto
- Color - Choose between colour and black/white
- Frequency - Choose between 50Hz and 60Hz
- Record Size - Resolution of the video, perhaps. Choices are 4K(30fps), 2K7(30fps), 1080p(60fps), 720p(60fps)
- Resolution - Sensor resolution. Choices are 48MP, 38MP, 26MP, 20MP, 16MP
- Mirror - Choose between normal, flip vertical or horizontal and invert
- Monitor - Aspect ratio. Choose between 16:9, 16:10 and 4:3
- Bright - Brightness from 1 to 5
- Contrast - Contrast from 1 to 5
- Hue - I am guessing this is saturation, not hue - from 1 to 5
- Edge - Edge enhance (?) from 1 to 5 - seemed to make no difference
- Format - formats the micro-SD card
- Date & Time - sets date and time
- System - Choose between Factory reset, System info (Software version - Nov 11 2022. 13:00:41 VER: V2.1.2.48) and SD card info
The IR remote uses a CR2032 battery (not included) and besides the menu controls, has the following additional controls
- Digital zoom in and zoom out
- Flip vertical and horizontal
- Auto exposure
- Auto white balance
- Colour vs B&W
- Freeze - the image is frozen even if the object under the microscope is moved. Press again to release
- Play/pause
- Take a snapshot
- Start a video recording
- Show images and videos on SD card
My big screen TV is right next to my workspace so it lets me see it on a large screen while keeping my monitor for other stuff like the component placement image. Also, there is such a thing as too much light, even at this distance. Reducing the brightness can highlight details not visible in a washed out image. Turn off the Auto white balance and bump up the hue (4) and contrast (5) a bit.
Bloopers
No show is complete without a bloopers reel. Before this PCB, the previous attempt was a lot worse. It did not help that I stacked the odds against me. I used a 3 year old batch of solder paste. I did not have a microscope - delivery was a week or two away. I randomly chose a hot air temperature of 260°C. The failed version is on the top, the improved one is below.
In the failed version, the ESP module is misaligned. One side of it has vias straddling neighbouring pads. The repeated attempts at fixing it left a large amount of flux residue, making it harder to see. With no microscope yet, I just gave up and waited for it to be delivered. Under higher magnification, you can see patches of unmelted isolated solder. There are a few of micro-balls of solder. The beauty of the microscope means I can spot these - I now see one in the improved version as well.
For inspection, use the higher magnification. In my case, using it means using it at 1x. The working distance drops dramatically and I have to keep the PCB elevated on a box. Also, the depth of field deteriorates quite a bit. In the image below, the PCB is in focus but the top of the tantalum cap is not.
A closer look at the pins of the ESP module shows that it is perfectly aligned. There are no visible shorts. There is very little flux residue since I did not use any. Now to check if the PCB works...
No comments:
Post a Comment