How powering PIC in auto ?

[fratello]

I made some schematics with PIC (12F675=windows closer;16F648A=thermometer/thermostat) and I use them in my car. I powered them like in picture...but, sometimes, without any reason, the PIC's are damaged. If I replace them, in the sames boards, these works fine, for a while...(weeks, even month !) and...collapeses again ! Can somebody sugessting one better metod for powering PIC's in auto environment ? Thanks in advance for every reply and sorry for my poor english. All the best !

[mackrackit]

Can we see the remainder of the schematic?

Yes, that is a very basic power supply as one can get but I will bet you are getting a spike from someplace else.

[Archangel]

I made some schematics with PIC (12F675=windows closer;16F648A=thermometer/thermostat) and I use them in my car. I powered them like in picture...but, sometimes, without any reason, the PIC's are damaged. If I replace them, in the sames boards, these works fine, for a while...(weeks, even month !) and...collapeses again ! Can somebody sugessting one better metod for powering PIC's in auto environment ? Thanks in advance for every reply and sorry for my poor english. All the best !
Hi Fratello,
Maybe you are getting a back EMF from the relays or motors, relays are good for over 100v back emf,also; try some decoupling caps across the motor power leads. I am pretty sure you already are using blocking or snubber diodes, perhaps a faster diode.

[fratello]

Thanks for reply ! This is one of schematic that I use : module for up/down windows of my car. The powering is like in previous post. The capacitors for decoupling (10uF & 100nF) are SMD, near at pins of PIC's.
LE : For relays I use 1N4007.

[fratello]

In previous post VR1 is not correct...

[Gusse]

Hi Fratello,

Can you measure broken PIC's if the IO's have been damaged?
- are input resistances (IO to GND) at the same level as with new PICs?

Is there any possibility that ESD would enter the system when swithes are pressed?

In car environment, you can easily charge up to very high potentials and if you discharge it to switches, PIC it can be damaged.

1) You could add serial resistors to PIC inputs 4 & 5. Already small resistors (~100 ohm) will damp ESD currents.
2) 5.1V zeners could be added next to C1 and C2.
3) Or do both

BR,
-Gusse-

[fratello]

Thank You for reply !
I have just one damaged 12F675; the resistance from pin 4 to GND are about 9k7, from pin 5 to GND are 9k1, with PIC powering with 5V ; PIC outside from module : pin 4 to GND = 16k3 and pin 5 to GND 13k3. What reveals that ?
Unfortunate I don't have now one good 12F675 for measuring the IO resistance; I wait for postal messenger from e-commerce.

[Gusse]

Sounds pretty much broken to me. You could expect ~100kOhm input resistance or in that range (or even higher) for good parts.

Try to add small resistors and 5.1V zener for switch inputs.

BR,
-Gusse-

[Normnet]

I use a Digi-key 497-6469-1-ND Transient Voltage Suppressor (http://www.st.com/stonline/products/literature/ds/3068/p6ke.pdf)
on every outside lead connected to the circuit.
I believe they are much faster than old zener clamps.
Also try to add a 100 ohm resistor or better ahead of each TVS so it doesn't
have to pull the spike down in the entire system.
For the power supply their could also be a diode and a 100 ohm resistor in
series followed by a large capacitor and TVS in parallel to ground ahead of
the volt regulator.
The 100 ohm resistor can get hot depending on the current draw.
You may have to adjust to less than 100 ohm to maintain voltage at peak
current draw.

Norm

[fratello]

@ Normnet : Can You figure that on my schematic ? Thank You !

[Gusse]

If you are using surface mounted components (0603 or smaller), then it would be more wise to put ESD protection before any serial resistor. Resistor will burn or change value if you force ESD current to go thru it.
Old thru mounted comps can tolarate more ESD, but even with those I would recommend to do the same.

For ESD protection you can use: TVS, varistor or zeners. All those will give appropriate protection, if you have done your PWB layout correct way. Try to minimize all paracitic impedances (inductances & resistance) from ESD current paths.

BR,
-Gusse-

[fratello]

I understand...Thanks ! Just because it's cheap, can I use "Transient Voltage Suppressors (TVS)" VC 1206 05 D 150 (05 = 5.6 VDC Working voltage; D=0.4 J Energy ; 150=18 Volts Clamping)...It's a good choice ? It' s enough to use just one like in picture, or it's better to use one on every IO of PIC ?

[Normnet]

05 = 5.6 VDC Working voltage
Try a few. They may reset the PIC because 5.6 volts is very close to
operating voltage of 5.0 volts.


It' s enough to use just one like in picture, or it's better to use one on every IO of PIC ?
I use one on every lead external to the board.

I would place the TVS in front of the volt reg to protect it also.

Norm

[fratello]

If 5V6 it's not so good, what value do You suggest ?

[Gusse]

Working voltage of TVS (or any protection comp) must be a little over max DC voltage to avoid current leakage thru protection comp.

For de-rating you should use + 10-20%. This means that Vbr (breakdown voltage) value is important, it should be so much above normal DC voltage that you will never reach that in normal conditions.

5.6Vdc is usable (Vbr for that is ~6.5V or bit higher). With higher Vdc components (e.g. 7-10V) you will have less leakage.

BR,
-Gusse-

[fratello]

OK ! Thank you both verry much !

[amgen]

looks like the main reason for PIC is to check current to disconnect relay when window is all up or down ??
don
amgen

[fratello]

The reasons for using PIC :
-small circuit
-two different function (up and down) with two possibilities (long/short press)
-detecting over-current
-huge support from this forum (thanks again !)
-personal ambition !

[amgen]

you may want to add free wheeling diode across each relay coil, common practice because of the voltage spike generated by relay coil when released (switched off)

don
amgen

[fratello]

Please read post #4.