Views on powering the ATMega328p in an automotive environment with a ZXTR2105FQ

Hi all,

I'm after your views on whether this regulator would be ideal for my application?

I am currently doing a project building an exhaust valve controller. I am going to be building a custom PCB with the option of adding GPS geofencing and RTC features.

I'm looking at powering it with a ZXTR2105FQ regulated transistor. I was going to use a buck converter but I heard that if running GPS you need a quiet power supply and looking at the ZXTR2105FQ is designed for powering microcontrollers in an automotive environment.

My Arduino is measuring 34-37mA at 5v with the valve controller program running, and the ZXTR2105FQ is capable of running a continuous current output of 89mA with a 12v input. Adding the GPS and RTC features I wouldn't have thought it would add much more mA in the mix.

What do you all think?

Mike.

Wtih 13.8V in (car running) and 5V out you have a Vd across the pass transistor of 8.8V.

At 37mA current, the Pd will be 326mW.

With an Rth(j-a) of 250oC/W and an ambient of 40oC (depends on where you mount it and operating conditions; cars can get pretty hot in the summer...) the junction will be at:

Tj = Rth(j-a) x Pd + Ta
Tj = 250 x 0.326 + 40
Tj = 121.5oC

The device's operating temperature range is up to 150oC. Technically, it should work but you have pretty small no thermal and power margins.

If you're really averse to switch-mode DCDCs I'd consider the use of a linear in a larger, more thermally-capable package that also gives more margin on current output and power handling.

It will make heat. You can put the converter in a grounded mesh cage to contain noise. Maybe use a 7812 or 12V zener diode to ensure the converter gets smooth input.

Have you ever seen those phone/MP3 chargers for car accessory (cig lighter) plugs? They cost $1 or $2 and provide 5V to USB. I dunno how much heat they make either but they're cheap and they work.

GoForSmoke:
It will make heat. You can put the converter in a grounded mesh cage to contain noise. Maybe use a 7812 or 12V zener diode to ensure the converter gets smooth input.

Have you ever seen those phone/MP3 chargers for car accessory (cig lighter) plugs? They cost $1 or $2 and provide 5V to USB. I dunno how much heat they make either but they're cheap and they work.

My original schematic was based on one of those USB chargers, which from what I understand is just a buck converter.

I need a regulator that can handle high voltage spikes as well as fluctuations and keep a stable, quiet 5v output.

Original Schematic:

Mike.

Blackfin:
Wtih 13.8V in (car running) and 5V out you have a Vd across the pass transistor of 8.8V.

I think it would be wise to allow for the voltage rising close to 15v on occasions, and not just for short bursts.

...R

A 7805 with standard decoupling capacitors, equipped with a heatzink powered, powered an 8 bit single board computer for 10 years without any failure.

Robin2:
I think it would be wise to allow for the voltage rising close to 15v on occasions, and not just for short bursts.

...R

That's what I was thinking. At idle the voltage is 14.4v and rises 2-3mV every so often. I have not tested it driving but I can imagine it spikes to 15v+.

Mike.

Railroader:
A 7805 with standard decoupling capacitors, equipped with a heatzink powered, powered an 8 bit single board computer for 10 years without any failure.

Now that you mention it, there is a 7805 sitting behind me that has been powering an Atmega 328 from an automotive type supply for 5 or 6 years 24/7 with just brief interruptions for battery switch-over. And my supply voltage regularly goes as high a 14.8 when the alternator is charging.

...R

s200bym:
I was going to use a buck converter but I heard that if running GPS you need a quiet power supply

You hear that did you?

Hmmm.

So the "Navigators" which all use a switchmode converter in the "lighter plug" don't work then? :roll_eyes:

Very strange. :astonished:

VIN - power input, connect to 3-5VDC. It's important to connect to a clean and quiet power supply. GPS's are very sensitive, so you want a nice and quiet power supply. Don't connect to a switching supply if you can avoid it, an LDO will be less noisy!

Adafruit GPS

I have not tested it driving but I can imagine it spikes to 15v+.

For circuits powered by or working in an automotive electrical environment, you must have protection against voltage reversals and substantial voltage surges, in excess of 125V.

Most people use TVS diodes, which are cheap, but handle over 1 kW transients. See this overview or this one (among many others on line).

Just to put my penny worth in
; I’ve used a 9v regulator feeding Vin on the board - splits the heat about ! I also use a series diode to drop another 0.6v on the input and give protection against reverse connection. You could look at using a different board , and drop the GPS to save a bit more current .
Can’t see why you need gps , anyway for an exhaust valve controller !! , just adds to the difficulty , getting a signal etc.

hammy:
Just to put my penny worth in
; I’ve used a 9v regulator feeding Vin on the board - splits the heat about ! I also use a series diode to drop another 0.6v on the input and give protection against reverse connection. You could look at using a different board , and drop the GPS to save a bit more current .
Can’t see why you need gps , anyway for an exhaust valve controller !! , just adds to the difficulty , getting a signal etc.

I want the GPS so I can do geofencing.

I did suggest putting the converter in a Faraday Cage to contain the EM noise.
Metal screen will do as long as it's grounded, cuts with scissors, can be shaped and glued/sewn/stapled together with craft tools.
A metal box will work and can sink heat too. An Altoids tin might have space for the controller but what does the GPS fit in? Where does it go that is not subject to EM noise?

Here's a thing about EM noise: the effect drops off with distance from the source cubed. Twice as far away is 1/8th the noise. 30 cm is 10x as far as 3 cm, 1/1000th the noise.

Other kind of noise is power noise. Easy to fix, set the converter to make 5.4V and put a 5.1V Zener between that output and ground. That should get very close to 5.1V, no? Or you could buck down to 6.6V and run that output through a 7805... would it be cleaner output?

That's 2 kinds of noise.

I think it would be easier to stick with the NCP1117 as used in the UNO. I'll run an M7 (smd 1N4007) and a 100R resistor in series. It'll drop the voltage from 12v to just over 8v and should help keep the heat down.

Schematic: