Can use diode?

Nano installed in car. Ignition power goes to 5v buck converter to Nano. Works well.

However, when I take key out of ignition, with no power to Arduino, I plug in USB cable from laptop and suddenly other ignition power items power up from the USB power (gaining power in reverse from Arduino) and the comms between laptop and Nano misbehave. Not sure why, maybe sucks all the current down. So I need to unplug ignition power to buck converter before plugging in laptop which is not convenient.

Can I simply add a diode between buck converter and Nano so that power from USB does not flow back to car? How do I spec that?

Thx

Yes. That would be a good idea. but you would have to turn up your buck converter a little to make up for the .7 volt forward loss.

Something like a 1A diode - IN4007 .

that would work if that's all you had laying around. but the 4007 is usually for higher voltages, but i think a diode is a diode.

The reason for selecting a 4007 is that since it does everything that a 4001 does, so there is really no point in stocking 4001s and they may be harder to get, even more expensive rather than cheaper.

On the production line, if it is working well, all the product will be fully specced for the 4007, so no point doing two different badges!

Thanks guys. I have a couple 4007 already so that works.

wildcat99s:
Yes. That would be a good idea. but you would have to turn up your buck converter a little to make up for the .7 volt forward loss.

Thank you for this point! I was not aware of the voltage drop and that would have been an issue for me.

If I put the diode on the input side of buck converter v-drop would become moot? 14.2v car running would put 13.5v into buck converter and still spit out 5.0v?

Yes, that is better.

If your power draw is fairly small, less than perhaps 200ma the voltage drop across the diode will be a lot less than .7 volt. Perhaps about .3 volts. I needed a .7 volt drop and discovered that I did not get it. Looking at the datasheet for the 1n4007 it shows the voltage drop starting at .6volts at 10ma. I don't remember which diode I was using where it was only .3v.

If your buck converter is a fixed 5v output, then the diode goes between the car and the converter. If your buck converter is output voltage adjustable, which a lot of them are by a small pot, the diode would be best between the converter and your MSU. that way the MSU would be isolated when the USB is plugged in... but either way works.

amdkt7:
If your power draw is fairly small, less than perhaps 200ma the voltage drop across the diode will be a lot less than .7 volt. Perhaps about .3 volts.

No, the voltage drops ~20mV for each halving of current, its a logarithmic function of current, so you might
see 0.7V drop to 0.6V or thereabouts

If you want 0.3V drop you have to use a shottky diode (which has a lower forward voltage than
a pn-junction diode)

Thank you for the helpful info. I put a 4007 on 12v side and it works.

HOWEVER - I have something new that popped up. I have a serial connection tied into OBD unit into TX1 RX0 pins on Nano.

When there is no ignition power, now the power led on Nano stays lit. It is getting power somehow through the serial. There is no parasitic current that is measurable, so I don't mind about any current draw, but in order to connect laptop to and communicate to Nano, I need to unplug OBD connection also otherwise it still has a headache.

Is there any kind of hardware fix to this or this is way it is? Or software fix, can I turn those pins on and off only when I call data from OBD so they are most likely off when I turn key off and no more ignition power?

I don't know about a software fix, but as for hardware, i think the easiest fix would be putting a DPDT switch on your TX1 and RX0 lines so you can turn them both on and off with a single switch throw