ViaEMS PCB Designs
In order of creation...
- Analog/ADC Core board based on a DIP TLC2543
- All throughhole
- Dedicated 5V supply
- 4 Thermistor inputs
- Inputs for EGO, BRV, MAP, TPS
- 5 Spares
- IO core board
- 4x ignition outputs
- 6x Low side drivers
- CAN driver
- Digital inputs
Do not use because:
- 3v supply for CAN doesn't work at all
- the FETs without gate drivers has poor performance from a 3.3 V microcontroller
- No VR inputs
Original all-in-one board for ViaEMS
- STM32F407VGT
- 8x Low side outputs (fed by gate drivers)
- 4x ignition outputs
- 2x High side drivers
- Dedicated MAP and AAP sensors on-board
- AD7888 SPI ADC with thermistor and core inputs, one spare
- CAN driver
Shouldn't be used anymore, but really only because the outputs chosen for test trigger vs real triggers are different in firmware. The board can be used again with some minor tweaks (and I have used this board for years)
Experiment board to prototype use of the MAX11632, and some knock sensing equipment. The ADC samples fast enough to dedicate two inputs to knock sensing, which uses an op amp input filter. Provides core inputs, a dedicated MAP sensor, and an additional 8 spare inputs.
The board works fine, but I'm unlikely to commit to the MAX11632: since it seems incredibly finicky with SPI noise, and it doesn't provide a good way to know if there is an issue. It requires a setup command. The knock sensing filter works great though. Ultimately issues with this chip pushed me to investigate the TLV2553.
Also, its way too big.
** Do not use! ** Simple breakout board as an upgrade to IO V2:
- IO Core board
- 5v supply (and fused output)
- 4x ignition outputs
- 8x Low Side drivers (including 4 injectors)
- 2x Digital inputs
- 2 VR inputs
Experiment with replacing the 2-port gate drivers with the 4-port MIC4469 variant. This board was made mostly to experiment with various FETs.
This board is missing inline resistors between the MIC4469 and FETs. This is incredibly important for the VND7NV04 -- without it, the negative pulse from injectors destroys the gate control logic and causes it to latch partly on.
Ultimately this board is also way too big.
Another experiment, this time for the TI TLV2553. This ADC has only 11 inputs, and samples at 200kHz (vs 300 kHz of the MAX11632). However, its control is identical to the TLC2543, which has self test functionality. It is still fast enough to allow for two inputs to be sampled at 40 kHz (and the rest at 10 kHz).
- Analog Core board based on TLV2553
- Small form factor (78x34 mm)
- 2x SPI headers to make debugging easier
- dedicated 5v supply (and fused output)
- 1x MAP input
- 2x knock sensor inputs
- 3x Spare inputs
I decided (after the fact) to make all the other boards the same form factor
The full contents of IO V2 could not fit on a single 78x34 board, so it is split into two -- one for the low side drivers, one for everything else. Other lessons learned from IO V2 were to add gate resistors for the FETs, and that indicator LEDs are worth the effort to plan into the design.
- Low side/Injector Core board
- Small form factor (78x34)
- 4x Injector outputs
- 4x LSD outputs (or an extra 4x injectors)
- Indicator LEDs
An upgrade from IO V2, with gate resistors fixed, indicator LEDs on all outputs and trigger inputs
- 5v supply
- 2x VR inputs
- 2x Digital inputs
- 4x Ignition outputs
Generic testing board for stm32f427 or gd32f470. This is the same pcb dimensions as the prior 3 boards, and comes with:
- 100 LQFP footprint for microcontroller
- USB Type-C (with FS-only pinout)
- Optional soic-8 spi nor flash footprint on SPI1
- Full 20 pin SWD+ETM trace connector