-1

I am analyzing this circuit of an ECU of an automobile and I found it's difficult to understand it's LM7805 was protected from spikes. You could find that 7805 configuration block in the right bottom of the diagram.

Is that a zener technique to bypass the spike instead of inductor? Could somebody explain me the configuration?

schematic: http://www.vems.hu/files/genboardv3/vems_v31_r222_schematic.png

stevenvh
  • 145,145
  • 21
  • 455
  • 667
Standard Sandun
  • 1,695
  • 2
  • 23
  • 38
  • 1
    You tell us first what `GND5` is in this messy schematic. (Olin will love it! :-)) – stevenvh Apr 12 '12 at 11:50
  • stevenvh that's where 5V loads are grounded. including AVR MCU. – Standard Sandun Apr 12 '12 at 11:52
  • So that means that the 5V from the 7805 is actually 5.7V (per diode D15) with respect to this ground? And that `GND` is used as ground for voltage regulators and decoupling? – stevenvh Apr 12 '12 at 11:56
  • Man, there are `GND` connections, `AGND` connections, `GND` connections with a note text to them "this is analog ground". This schematic has **everything**. – Connor Wolf Apr 13 '12 at 09:15
  • @stevenvh - Oh, incidentally, the net labeled `GND5` is connected to the net labeled `GND`. – Connor Wolf Apr 13 '12 at 09:16

1 Answers1

3

What a mess! First you need to fire whatever slob drew that schematic. Sloppiness in one area invariably means sloppiness in other areas too, which has no place in electrical engineering. You don't want this guy designing your circuits. The "eh, who cares" attitude he so clearly exhibits will cause a lot of problems.

Anyway, here is the snippet you are apparently referring to:

I don't see any zener diode, except maybe the one way at left. However, D46 is blocking any current from the +12V supply to it. There seems to be nothing driving the cathode side of that zener. Again this schematic is a mess and the circuit looks the same.

For a zener diode clamp to work in clipping spikes, the signal must have some impedance. I wouldn't put a zener right accross the 12V line of a car since that impedance is very low and the spikes would just fry the zener. You could put a resistor and/or inductor in series followed by a zener clamp, depending on how much current you need. However, a better way is to use a regulator intended for automotive power. National makes one, but I don't remember the model number off the top of my head. You can still preceed it with a little resistance and then a capacitor to ground. Size the resistor so that it drops maybe 2 Volts at most at your maximum current.

Olin Lathrop
  • 310,974
  • 36
  • 428
  • 915
  • 1
    Is that a CAPTCHA centre-bottom? I can see that it begins with "U2" but from there it's anybody's guess ... – MikeJ-UK Apr 12 '12 at 12:56
  • sorry sir I'm new , I didn't see your this: http://electronics.stackexchange.com/questions/28251/rules-and-guidlines-for-drawing-good-schematics/28255#28255 I'm new here.Thanks to the guy who PM me about link. – Standard Sandun Apr 12 '12 at 12:58
  • LM2936 NatSemi automotive environment 5V regulator. | As Olin says, the zener as shown does NOTHING. And on the full diagram the Gnd5 versus Gnd traeatment is weird, at best. Do you have access to the 'designer' of this circuit? – Russell McMahon Apr 12 '12 at 15:06
  • @Russell - Perhaps he escaped :-) – stevenvh Apr 12 '12 at 15:59
  • I think there is something between the car source voltage, and the zener rail. If you look closely, you see that the input goes through the fuse and diode to the net-label `+12V_P`. The regulator input is connected to the rail `+12V`. However, I cannot find where `+12V_P` is connected to `+12V`. – Connor Wolf Apr 13 '12 at 09:10
  • 1
    Also, the person who drew that schematic does not seem to understand how eagle works. He seems to be feeding EVERY net label into a giant buss (except where he is not). Seriously, can we track him down, and demagnetize his computers hard-drive or something, so he won't inflict such atrocities on anyone again? – Connor Wolf Apr 13 '12 at 09:12
  • 1
    @Fake: Note that whatever is coming thru fuse F1 and diode D45 isn't connected to the top of the zener. It is only connected to two caps, C75 and C76, which go to GND5. It's just drawn sortof making it look connected at first glance. – Olin Lathrop Apr 13 '12 at 12:00
  • @OlinLathrop - The lack of connection between F1 and D45 and everything else was what I was trying to point out (I guess I wasn't clear about it). Agreed on the HV idea, though. – Connor Wolf Apr 14 '12 at 05:11
  • 1
    @RussellMcMahon - I went looking for that regulator. Google brought me to http://www.national.com/pf/LM/LM2936.html, which makes sense. I clicked "datasheet", which sent me to http://www.ti.com/lit/ds/symlink/lm2936.pdf, (which was mildly surprising) where the first page is a TI title page and the rest is the National datasheet. I know National was bought by TI, but this is the first time I've seen anything like that! – Kevin Vermeer Apr 16 '12 at 13:44
  • @KevinVermeer - yes -I've seen a number of datasheets like that alrady. – Russell McMahon Apr 16 '12 at 15:58
  • @KevinVermeer TI had acquired National Semiconductor in 2011. That's why we see both company names in the datasheets. – Nick Alexeev Apr 24 '12 at 05:15