(KiCad 3D rendering of preliminary design)
Altimeter features:
|
History and Current Status (updated 2025-Jun-02)
In May 2012 my brother, Ben, convinced me to try again to design an altimeter with an on-board USB-UART interface and a coin cell retainer. However, squeezing all this functionality into a board smaller than ½"x1" (which has always been the goal for my altimeters, so they can fit into small rockets with a 13mm body tube) would require using some very tiny SMT components that are extremely difficult to assemble with a soldering iron. Instead, I used ExpressSCH and ExpressPCB to design a larger .68"x1.75" initial prototype using the USB-enabled TI MSP430F5508 µC in an LQFP-48 package, and a 4-pin Spy-Bi-Wire connector for programming and debugging without a jig. The next month a former co-worker assembled this prototype altimeter for me. But I had problems getting it to work, and subsequently abandoned it.
Over the next 10+ years I designed many other altimeters (both USB and non-USB), had PCBs made for some, even had some assembled. The prototype of the tiny ALT2430G-290x310 had its parts (on both sides) installed with a pick and place machine and reflow oven by a small assembly shop in Kimberley BC in December 2014. I got much of its MSP430 firmware and the host software (written in VB6) working. But none of these altimeters were finished or flown.
In November 2023 I started learning and using KiCad for PCB design. My first major KiCad project was a custom USB-UART-RS232 module, which I had manufactured and assembled by JLCPCB in China. The second version has SMT parts on both sides of the board -- the build quality looked good! So, on May 8, 2024 I decided that since I had now found a place that could assemble double-sided SMT boards (with tiny components) for me at reasonable prices, I would again try to design an altimeter with on-board USB, and abandon my other out-dated non-USB versions.
My new altimeter would use an MSP430G2553 microcontroller in a TSSOP-20 package, the FTDI FT234XD USB-UART in a DFN-12 package, an all-surface-mount USB Micro-B connector, but a thru-hole coin cell retainer for a CR1220 (as the solder joints would be more rugged). I managed to squeeze everything onto a tiny board only .45"x.99" (11.43 x 25.15mm) in size! Some of the components I selected were not stocked by JLCPCB, so I had to pre-order them. After they arrived at the factory, I placed an order for 5 panels, each containing 2 altimeters. The 10 assembled boards arrived on December 29th -- a late Christmas present to myself!
This altimeter doesn't have a Spy-Bi-Wire connector (there wasn't room for one), so I soldered wires to one board to connect it to my TI MSP-FET430UIF device programmer for the MSP430 for development and debugging. I also started designing a jig that will be needed to program the MSP430 in the other 9 altimeters. The jig will consist of a sandwich of 6 different PCBs that I will order from OSH Park. Four pogo pins will contact .050" round test pads on the altimeter. I have created a crude preliminary sketch for it.
But as I started working on the firmware, I ran into some problems. First, in my rush to get the altimeter manufactured, I forgot to check whether the MSP430's UART can invert the TXD and RXD signals, like the FT234XD can, so they will be 0V when idle, which is ideal for when the MSP430 is powered by the battery and the USB disconnected -- no, it can't. I should have used inverters on these lines, such as the Nexperia 74LVC1G14GX4Z in an extremely tiny (0.6mm x 0.6mm) X2SON4 package.
Then, I couldn't get the MSP430 UART to transmit any data. When I tried to run some test code on my two TI MSP-EXP430G2 LaunchPad boards, that didn't work either, and I discovered that both were damaged. It didn't help that I initially forgot to change the TXD/RXD jumpers from "software" to "hardware" orientation. This caused contention, and might have damaged the MSP430G2553, and maybe even the MSP430F1612 uC in the emulation section? It's also possible that the grounding of the anti-static mat on my workbench had failed or was intermittent.
So, I put everything on hold for a few months to work on other projects, and to build a new, better workbench. As of today this still isn't finished. I also ordered several more LauchPad boards for testing. Stay tuned for future updates.
Last revised 2025-Jun-02 16:08 PDT.
Copyright 2014- David C. Wiens.