V. reg grounded via diodes?

I don't think it's surge protection. Most likely, the designer felt a little more comfortable with a 5 volt supply closer to the upper edge of "normal" than the lower edge.

Jim

The designer may have needed a negative supply wrt IC ground for some reason. Can you post the entire schematic in a.b.p.s.e. ?

____ | | -------------|7805|--------------- 5V |____| | +--+---+-------------- 0V | | |+ - V --- ^ - --- | | | +--+---+--------------- -0.6V | === GND

One guess is that some switches are returned to chassis ground and the designer wanted to allow for the possiblity that chassis ground might not be quite the same voltage in different places in the vehicle. With TTL level inputs there's not that many mV to play with. Offhand I don't see the purpose of the second diode but maybe it's desirable if the product is mis-wired.

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

to

I

On Thu, 21 Aug 2003 6:15:36 -0700, Spehro Pefhany wrote (in message ):

What you've drawn below is all I have. I have no schematic, only the PCB I'm trying to troubleshoot. I sketch down parts of the circuit as I follow some traces. I noticed this v.reg configuration when I saw voltages far above 5v (with reference to system ground) at the ICs.

This does make sense, I guess. There are 10A relays on the pcb that are switched with 12v. These are referenced to system ground. Also, sensor switches (door, hood, trunk) switch 12v, so maybe the diodes provide a bit of isolation and protection from "undershoot"?

And yes, this type of auto accessory probably gets the power reverse-wired regularly during installation.

The reason for the diodes is a bit beyond my meager design skills. Just trying to understand the thoughts of the designer...

Thanks,

--
Dave C.
dave-usenet3016@mailblocks.com

Nope. Ground is still ground.

Wouldn't help at that point in the circuit.

Not even close. Thanks for playing.

Jim has it right

--and the 2nd (reverse-biased) diode is overkill.

No it's not, as I learnt during my apprenticeship, all those years ago, it's called a "Brother-in-law Diode". A diode that serves no useful purpose in the circuit but is included because the circuit designer's brother-in-law owns the diode factory!

Hope this helps, Peter

are

power to

I

The forward biased diode as shown will lift the regulator by whatever its forward voltage is with respect to ground. If 1N4005s are used I would expect the output to be around 5.6V. If not, I would take a look at the diodes and cap to see why. 5.2 volts tells me that something is wrong.

Ken

to

I 5.6V supplies are not uncommon. I would check continuity between the ground pins of the ICs and the 12V ground or measure the voltage between the output of the regulator and a known 12V ground. If if you have continuity or still measure 5.02V, something is wrong with the diodes or cap. It seems to me that if the ground of the circuit flows through the diodes to 12V ground, it will make the output of the regulator dirty. The current flowing out of the ground terminal of the regulator may be constant, but the rest of the circuit isn't. If the current changes, the voltage across the diodes will change, thus changing the output voltage of the regulator.

Ken

Also- the diodes allow the 5V logic to drive chassis referenced 5V relays with the normally reverse biased diode acting as a common clamp. Additionally, the logic can put 5.7V across the coils guaranteeing solid operation at elevated temperatures where coil current requirements increase. Isolation from chassis is still preserved. Please view in a fixed-width font such as Courier.

+-------------------------------------------------+------ | | | | | 5V REG | | +-------+ | | | | ckt o + o--+------IN OUT---+----+------+ / Vbatt | | | vcc | +-----> o /|\ +--COM--+ | '---' | | | big I | | + | | | |\ | | | | 5V === | |--| > -----+ | | | | | - | '---' |/ | | | \|/ | | gnd | | | | unreg +-------+----+------+ | | | | | | | +---||---+ 5V )|| 12V )|| | | | relay )|| relay )|| | +---||--+ | | - | | | | | | -------+-----------------------------------+-----------+------- | --- /// chassis heavy duty noise and transients from other junk on chassis does not enter/corrupt circuit

On Fri, 22 Aug 2003 6:37:08 -0700, N. Thornton wrote (in message ):

Yes, I can see this.

The purpose of this design is...?

--
Dave C.
dave-usenet3016@mailblocks.com

On Fri, 22 Aug 2003 4:48:13 -0700, K Wind wrote (in message ):

Voltages measure thus: 5.02 from negative terminal of v.reg to 5v rail all across the board 5.6 from system ground to 5v rail all across the board 12.8 (approx) from system ground to 12v rail all across the board

Voltages are clean, as viewed on a 'scope, all across the board.

This circuit, as designed, has worked for about 15 years. Never had a problem. Now some component in the circuit has decided to lose its mind. Could be one of the two custom ICs (40-pin microcontroller; 8-pin somethingorother), or a linear or non-linear component.

Thanks,

--
Dave C.
dave-usenet3016@mailblocks.com

On Fri, 22 Aug 2003 6:41:12 -0700, Fred Bloggs wrote (in message ):

I don't know how this effects your explanation, but the relays are 12v operated. The microcontroller drives hi-voltage (12v) drivers which drive the relay coils with 12v.

Thanks,

--
Dave C.
dave-usenet3016@mailblocks.com