Running a microcontroller in car - power supply aspects

[stj]

Re: Running a microcontroller in car - power supply aspects

btw, the best thing about these modules is the pinout and spacing matches 78xx linear regulators!

[stj]

Re: Running a microcontroller in car - power supply aspects

the mod i do on some is to replace the pot with 5k (original is something like 200k)
and cut that track - then solder a resistor across it i salvage from the top for 5v.

this gives me something like 5v-5.5v ajustment range!

[Dannyx]

Re: Running a microcontroller in car - power supply aspects

Ahh thanks I forgot which one you are using. That module uses MP2315 IC which has a low power mode set by a resistor to GND on pin 1. It's claimed to draw under 1mA and you get 10mA? I haven't tried one. You cut the tiny jumper?
This guy found it has an instability and he did some mod. for low voltage I think https://www.youtube.com/watch?v=yyrgW5dNUIo
I'll bet a bigger capacitors in/out would fix it.

Yeah, OK, I never noticed the instructions when I got mine (the original link is gone anyway). I just assumed that track's just part of the board and is supposed to be there all the time, so I didn't mess with it, meaning I was getting BOTH the pot AND the fixed voltage resistors connected at once.....which means the same is going on in that monitor right now *fail*

Ok, knowing that, I'll probably give it another go and report back. Keeping the power consumption of this whole thing as low as possible is very important.

[redwire]

Re: Running a microcontroller in car - power supply aspects

Ahh thanks I forgot which one you are using. That module uses MP2315 IC which has a low power mode set by a resistor to GND on pin 1. It's claimed to draw under 1mA and you get 10mA? I haven't tried one. You cut the tiny jumper?
This guy found it has an instability and he did some mod. for low voltage I think https://www.youtube.com/watch?v=yyrgW5dNUIo
I'll bet a bigger capacitors in/out would fix it.

[Dannyx]

Re: Running a microcontroller in car - power supply aspects

THESE are the "green" ones There was a link the first time I mentioned them, but it probably got buried in the post and you missed it

[redwire]

Re: Running a microcontroller in car - power supply aspects

What's a "green one"? They simply might have put the wrong resistor in there, to give 4V not 5V. Or the IC has a different reference voltage or is defective. What IC does it use.

[Dannyx]

Re: Running a microcontroller in car - power supply aspects

the pot is connected by default, there is a trace to cut if you dont use it

Are you sure ? I don't see any traces that need cutting, other than those solder pads on the back which were obviously not bridged out of the box until I did so, unless I'm missing something. The pictures on that Ali link show the underside too.

[stj]

Re: Running a microcontroller in car - power supply aspects

the pot is connected by default, there is a trace to cut if you dont use it

[Dannyx]

Re: Running a microcontroller in car - power supply aspects

do the green ones have a pot on top?
if yes then you need to cut a track to isolate the pot before using the fixed resistors

Yes, it does have a pot, but the tracks on the bottom are "pre-cut" and you have to put a blob of solder across the ones you want to "activate": 1.5v, 3.3v all the way up to ADJ which is the pot itself, so I don't think it's connected otherwise.

[stj]

Re: Running a microcontroller in car - power supply aspects

do the green ones have a pot on top?
if yes then you need to cut a track to isolate the pot before using the fixed resistors

[Dannyx]

Re: Running a microcontroller in car - power supply aspects

Not having the knowledge and tools (scope) to do in-depth analysis, the most basic check I did on them was measure the idle current draw, with nothing connected to the output, an input voltage of 12v and the output set to 5v. The smaller green ones were slightly disappointing, at least out of the box: the output is a full volt lower than what you set it to and the current draw is 10mA. These have some solder pads on the bottom which you bridge to set a fixed output voltage, so I bridged the ones for 5v and only got 4v. Not sure whether this can be improved by adding a beefier cap on the output or tweaking those resistors. There IS an "ADJ" setting as well, which inserts a small trimmer pot into the feedback loop in place of fixed resistors to control the output, but I was advised against using pots in the long run.

I used one of these small ones to "fix" a monitor whose 5v (or 3.3v - can't remember) power rail had died and the switching IC was too expensive to replace, so I botched one of those in and it works to this day, so I think the monitor was happy even with 1v less than the rated input on that power rail, OR that one happened to have better regulation - can't remember precisely.

The larger DC-DC converters (which I pictured in one of my previous posts) only use up 8mA with an input voltage of 12v and 6mA with an input of 15v. These are advertised as going up to 30v. The datasheet for the XL400 confirms this + the input cap is rated at 35v....

[redwire]

Re: Running a microcontroller in car - power supply aspects

I would measure your 5V buck-converters and see if they are OK for low current use. Or change the Vreg IC to something low power and high enough input voltage to take transients in a car. The Pro-Mini LDO MIC5205 is a max. 20V input which is too low for a car, even with a clamp TVS. An 18V TVS clamps to 25V.

[Dannyx]

Re: Running a microcontroller in car - power supply aspects

Arduino Nano uses about 20-35mA, so I wonder about using a 5A buck converter? If it is OK with light loads and regulates OK- but it might waste power.

It's actually an Arduino Pro Mini - KiKad doesn't have a Pro Mini model so I just used whatever it had close, that's where you probably got the idea. I do have some smaller buck regulators too, but I'd probably have to add some electrolytics to it (both on the input and output) and I'm not sure how well those survive in a car where it can get very hot in the summer...

You can put the Nano to sleep and wake it up if IGN is on. The USB IC apparently draws most of the current.

No USB IC on the Pro Mini and I also had a look at THIS. I removed the LED and the regulator, but have not had a go with that sleep library. My hope is that it won't affect any of the functions running on it...

[redwire]

Re: Running a microcontroller in car - power supply aspects

It looks not bad, many chinese DC-DC have a 1A diode which quickly cooks.
Arduino Nano uses about 20-35mA, so I wonder about using a 5A buck converter? If it is OK with light loads and regulates OK- but it might waste power.
I think the spec. limit for a car is about 70mA quiescent drain before the battery goes dead when the car is parked for a few weeks. Subaru firmware bug made their ECU use 150mA, which quickly wore out and killed the car's battery and pissed off all the owners who had to buy a new battery.

You can put the Nano to sleep and wake it up if IGN is on. The USB IC apparently draws most of the current.
I use a small LDO LP2950 for my MCU car projects, and the bigger 5A TPS40057 modules for USB car charging cell phones. Those use about 50mA (with LED) at rest.

[Dannyx]

Re: Running a microcontroller in car - power supply aspects

I guess there are fake LM2596 boards out there. The IC runs at 50kHz, not 150kHz and put in a 1k resistor instead of the pot, for 5V out.
My main worry would be if the mystery IC can run at 30V input (during a spike)

My buck converter doesn't even have an LM on it - it's a XL4005E1 so this should be interesting

[stj]

Re: Running a microcontroller in car - power supply aspects

the advantage is you only power 1 relay to move the window, regardless of direction

[Dannyx]

Re: Running a microcontroller in car - power supply aspects

Good pointers on those caps and diode, though I went with a different approach to switching direction: I used a DPDT relay. The caveat is that I now need a second "master" relay to cut the incoming power, since the DPDT relay has no "off" state - it just crosses the + and -, so it'd always run one way or the other. The relay count would be the same (4 - 2 for each window), but I ended up going with your design instead, since 4 SPDT relays are smaller than two DPDTs.

[redwire]

Re: Running a microcontroller in car - power supply aspects

Maybe something like this sketch as a starting point.
For EMI, it depends on the motor's wiring (long run or not) and if the metal housing is grounded. The brushes arcing can make EMI as well.
I think a bi-directional TVS will limit the back-EMF from the motor and the relay contact arcing will be small.

[Dannyx]

Re: Running a microcontroller in car - power supply aspects

Just a bit big but oh well.

Yes - big indeed, especially at this high current.

A snubber or TVS across the motor will protect the contacts

R+C snubber like this ?

and I would add a couple EMI caps so the Arduino does not crash when the relay contacts spark, as they always do to some extent. If the relays are close to the Arduino it can crash due to the EMI from the contacts.

I was just thinking of that and yes, the arduino will be right next to the damn things, so I'll definitely have to do this. Where does this go ? Across the motor's terminals ?

[redwire]

Re: Running a microcontroller in car - power supply aspects

I guess there are fake LM2596 boards out there. The IC runs at 50kHz, not 150kHz and put in a 1k resistor instead of the pot, for 5V out.
My main worry would be if the mystery IC can run at 30V input (during a spike). I would bench test the module first. The zener on the output with a 1A fuse I think will protect the Arduino if the module shorts and gives say 12V for a while.