VISP-BREAKOUT is a small single sided breakout board with the 4 sensors in the right places brought out to SMD pads for soldering wires
VISP-I2C-DUAL is a small board with 4 sensors, and a 6-wire interface. 2-channels of I2C and VCC+GND
VISP-I2C-MUX is a 4-wire I2C interface with a I2C bus multiplexer to switch between the 2 sets of I2C sensors
VISP-I2C-XLATE was our first design. Single 4-wire I2C interface with all of the sensors visible on the same bus
VISP-SMART is a unreviewed completely 3v3 design with an atmega328 on board and useful pins brought out to SMD solder pads on the back.
VISP-SPI is an 8-wire interface to the sensors, using a 3-to-8 decoder to convert A0:2 into a Chip Select (where address 000 is unselected so the bus can be shared. 8 wires so an RJ45 jack can be used at the core controller for connectivity)
Even though I ordered "Super Swift" boards from @oshpark, they are going to take a while to get here, and we have questions we need to answer, now.
The barometric sensors we chose have pads that are simply far too small to air-wire them, without a microscope at least. These pads make 0402 components look huge.
I am going to have to etch a board.
I have some Blue Press-N-Peel PCB toner transfer paper from a project I did about 10 years ago (maybe longer). Three sheets left, I m going to have to make each one count.
Even though these sheets are old, they worked like a champ!
Normally I would use some Muriatic Acid and Hydrogen Peroxide to mix up a batch of Cupric Chloride. Problem is I'm all out of Hydrogen Peroxide. Off to the store.
Every store, every shelf, empty. Anything remotely related to disinfection. Gone. I live in southern NJ. Out here, things like Alcohol, Hydrogen Peroxide, and toilet paper seem to be really hard to come by right now.
I am going to have to get creative.
What else uses peroxide... Hair dye... maybe. Teeth whitening! Off to the oral care aisle to scour the shelves for a product containing what looks to be good amount of my essential raw ingredient. Mouthwash with teeth whitening has Hydrogen Peroxide, Score!
I have no idea how much the concentration is, so I am going to have to wing it here. Equal parts mouthwash and Distilled Vinegar seasoned with a large pinch of salt. Mmmm.
It's working, it's WORKING!
Results will vary, but here are my minty fresh BMP breakout boards!.
I just wanted to share this cutaway of Daren's design. If air is flowing left to right then the pressure going into the left tube should be higher than in the right tube. And if the air flows the other way then vice versa. I have some differential pressure sensors that just arrived so I'm going to try to see how sensitive they are by testing this idea below.
@djeddleman had two BMP180 development boards in his stash of stuff. He was kind enough to hook them up to one of the older generation pitot tubes we designed that used tubes still, and take some measurements.
The results were far less than promising.
We saw basically no deviation in the pressure differential read by the two sensors at flow rates of up to 3m/sec, and a static offset greater than that 'signal' by like 5 meters of an air column. That is outside the spec for even these. The new sensors are an order of magnitude more sensitive, but 10x nothing is still nothing.
I can only hope that perhaps one or both of the sensors was plugged or damaged during testing at this point. There was RTV involved, so it is a possibility and would explain the static offset.
@Steven.Carr plans on attempting to etch a board tomorrow to try the same test with our BMP388 boards, and hopefully have better results. The pads are truly tiny though for these devices, so... wish him luck.
@Steven.Carr dropped an Atmega 328 on the board and it actually routed.
Supports the existing OLED pins (same place) and it even has a TTL pin header for FTDI connector for serial/programming port.
It is an all 3v3 design with regulator. Has a variety of solder pad outputs for connecting to different kinds of motors (PWM, ETC), 6 analog inputs, and a couple of interrupt inputs as long as you are soldering wires to it instead of pin headers.
So that means we could have some smarts right in this device, and with just a buzzer and some inputs, we could have a standalone monitor, or even move the entire brains of the operation right here.
Three of the boards passed review, and were submitted to @oshpark today.
The huge pads on the back of the sensors are for hand soldering larger bypass caps if you don't want to place the 0402 ones next to the sensors on the front.
The three solder-bridge pads labeled "Regulator Bypass" bridge across the 3.3v regulator and the I2C lines for the first translator so they can be not installed if this is to be a 3.3v only device. The first translator acts as a level shifter too for 5v operation so it makes sense to combine these.
The "ADDR" solder-bridge pads are for shifting the address of the entire device (if the first translator is installed).
The snap-off support for the lower pitot tube worked, but was bonded a bit too well and took the edge of the tube with it. Moved that down a bit, made the tube one wall thickness thicker.
<EDIT> That worked. Snapped off perfectly now. We are dropping the latest one on a table saw to check the internal port geometry today.
</EDIT>
The board got some minor revisions as well, with the bypass capacitors for the sensors move next to them. They were probably close enough before but that gave us a little more wiggle room for routing...
The SPI version!
So we modified the 3D model to match, again, and that is printing now.
Daren Schwenke
