#Data Acquisition System for BAJA car

1 messages · Page 1 of 1 (latest)

stable geode
#

Hi, currently I am and undergrad EE student working on the universities Baja SAE team and trying to increase the teams cabapilites for data collection to help make design decisions. Our team has not run anything like this in the past so I didn't have anything to work from previously. I designed this PCB with the hopes of collecting data for variables such as wheel speed, brake temperature, brake pressure, clutch temperature, IMU, steering rack position and many more. All of the data that is logged will be stored on a microSD card for further processing post testing. The system is powered from an M18 battery and is then stepped down to 12, 5, and 3.3V to power the various IO. I choose to go with an ESP32S3 Wroom microcontroller for the base of the system and this should communicate with all peripherals necessary. Attached is the current standing of the project if anyone has feedback on ways to make improvements or parts I just did downright wrong that would be appreciated. This is my first time attempting a PCB of this scale with elements such as the usbc port, onboard esp32, micro sd card slot, and external adc.

broken tree
#

can you post/send me a zip of the project directory or point to a git repo? i'm bad at looking at pictures

stable geode
#

@broken tree Here is the zip file for the project directory, thank you in advance for any feedback

shrewd path
#

What is your expected current load on +3.3V, +5V, and +12V rail?
You will get a very big voltage drop over the LDOs, it might get hot.

stable geode
#

The expected current load isn’t significant with the biggest current loads being the brake light on the 12v rail being about 500 mA at peak but this is not constant, and the esp32 on the 3.3v rail, the 5V rail should not see a significant load through the sensors attached with the biggest load we could have plans of attaching being either an IMU or a 2x16 lcd through an i2c backpack but I can’t imagine either of these components having a significant current draw

#

Some heat would be ok the longest possible time we would be running this system at one shot would be about 4 hours during an endurance race in which the car will be consistently moving allowing for airflow depending on how the enclosure is designed

shrewd path
#

ok... I assume M18 battery is 18V nominal? This means on 12V LDO you will get P=(18V-12V)*0.5A=3W of heat already. This is quite a lot

stable geode
#

I do agree that 3w is a good bit of heat already, and Yes M18 is 18 volts nominal with full charge being about 20.1V and low charge being about 16V, the 500mA draw is a high estimate we calculated I believe the actual could be lower we just went high to have some wiggle room in case, would something such as a heat sink help with the potential heat? We have access to making custom aluminum heat sinks need be

shrewd path
#

ESP32-S3-WROOM datasheet says the 3.3V supply should provide min. 0.5A. This means 7.35W in worst case dissipated on LDO. Such load will probably not be for long time, but still, sounds like quite a lot of power lost.

stable geode
#

That is true I believe that current draw from the esp32 comes from using the wireless capabilities such as WiFi’s and ble which we don’t have a current plan to implement, and I think with the capacity of the m18 batteries being 5-8ah depending the worst case power loss would be acceptable due to the limited time a system like this would be running

#

The main reason chosen to go with LDOs is due to our lack of success when experimenting with our own buck converters on pcb’s many candles have been made from various sources of error

shrewd path
#

Ok, so you will not use the Wi-Fi / Bluetooth at all?
Because the placement of the module on board is incorrect, and the board could interfere with the PCB.

stable geode
#

Yes correct no wifi or ble will be used at all and the chosen esp32 package has the external antenna socket on it should we ever choose to utilize those functions, as this would allow us to place the antenna in a more optimal location

shrewd path
#

Ok, so the -1U variant.

stable geode
#

Yes

shrewd path
#

One thing for sure, you can add a series resistors and filter capacitors on ADC_INx lines.

stable geode
#

I was thinking about that possibility for an optional low pass filter, but I wasn’t entirely sure if it would be necessary due to the relatively short runs to sensors due to the size of the vehicle, I probably will add those in before final revision and use jumper 0ohm resistors if they are not needed

shrewd path
#

And what is exactly on the other end of "Clutch Cover Fans" connectio?

stable geode
#

Those are just additional lines attached to the 12v rail that we wanted the option to control in case we end up putting small 120mm fans to circulate air in our clutch cover as we are testing current solution to our clutch temperature issue

shrewd path
#

ok, but this line will not provide power, only PWM control or something?

stable geode
#

As of right now they will not be populated rather future proofing to a degree, and we would like them to provide power if needed with the mosfet acting as a switch in this case I am unsure if this was the best way to do this as I am between a relay and mosfet

shrewd path
#

There is 10k resistor. So you will get 12V with this resistor in series. You will nto be able to output any serious power thsi way

stable geode
#

The 10k was a place holder value I figured the actual value should be lower just didn’t calculate the value as I was between mosfet and relay as previously mentioned

#

The real value would likely be in the range of 100-250ohms to allow around 100ma through the resistor if needed

shrewd path
#

with 100ohm and 100mA, you will get 10V drop on this resistor. And also 1W power dissipated.

stable geode
#

Ah that’s right I was thinking about it it wrong at first

shrewd path
#

This is not the right transistor topology to power external load like fan. You can use transistor as simple low-side switch. Or e.g. 2 tranistors as high-side

stable geode
#

Ah ok I wasn’t sure if I was using it correctly thanks for pointing it out I would likely use the 2 transistor set up for a high side switch or find a viable normally open relay to use I’m thinking relay might be more viable for this application now

shrewd path
#

I understand this M18 battery is already something "smart"? WIth protections? Otherwise you might add some kind of fuse, just in case

stable geode
#

The M18 is the standard Milwaukee battery from their line of 18v power tools, the battery does have some protections but the adapter we already use has a built in inline fuse

azure jungle
#

As long as you follow the datasheet there should be no issues

#

They are quite straightforward, at least the ones that have few connections except but to the key supporting components

stable geode
#

I’ve tried my hand with following the data sheet on a few different models from various brands one of my main issues seems to be with inrush current on initial startup, when used with a current limiting bench top supply they will start fine but when run off a battery they seems to short on startup, would you have any recommendations on easy to implement switching buck?

azure jungle
#

but inrush current shouldnt be that big of a problem if your not using massive caps

stable geode
#

I wasn’t using massive caps I think the biggest was 10uF and I also had an issue with the chip size on one in which I had issues soldering the chip properly due to the small size of pads

azure jungle
#

sounds like you mustve had other issues then because 10uF is not a lot

stable geode
#

Yeah I’m not entirely sure, when I am back at my computer I can upload a few test boards that are just the buck converters I was experimenting with

azure jungle
#

pretty much a fully integrated module

#

internal inductor

stable geode
#

That one looks like it could be a viable option, I’ll have to mess around with it and make a separate board first to test before integrating into this design

#

The only thing I notice is it wouldn’t allow for a 12v rail