Voltage Protection

I was thinking of a DC/DC converter for the PIC for isolation and regulation. Fuses shouls be near batteries or power sources.

greg

Yes I have a DC/DC converter for the power for the PIC itself, but there are also sensor lines to the unregulated power lines. The question is how to protect the PIC input pins that are connected to these.

I agree that fuses should be near batteries, but in my case (see above) this is impractical and disturbs measurement. I'm looking for an alternative.

The extra resistance in the fuses can be ignored since no significant current will flow through the lines towards the circuit. Besides, I think it is a better idea to use a Hall current sensor than a shunt resistor. The starter engine may draw more current than the shunt can handle.

Well, for the voltage sensing that is true. For the shunt in combination with the MAX472 it is a different story (I think). It doesn't measure the two potentials on both sides of the shunt, but uses an opamp to have a current flow through an external resistor. In both sense lines a resistor of 100 ohm is placed. The maximum current is 1,5 mA. I don't exactly know what the typical resistance of an inline fuse and fuseholder is, but I guess it might be a few ohms. That still is a few percent. Maybe I'll have to accept that, but I was wondering if there is a better solution.

Thank you for the hall sensor suggestion. It would indeed be better for the starter battery, but in my case I only measure current on the "light" battery, which doesn't have to cope with currents higher than

Jack, thanks for this good suggestion. I can understand that this will work!

Do you perhaps also have a good idea for the fuses in the current sensor chip?

Search for 'load dump'. Automotive electronics have heavy duty zeners/transzorbs that will take 60V for a while. The alternator and voltage regulator go nuts if the battery is accidently disconnected. This sounds like it could happen in a boat. I also vote for circuit breakers instead of fuses. Its a long swim to radio shack.

Walt Jung's "Op Amp Applications Handbook" has a chapter on device protection -- chapters of the book are available on Analog Devices website, or you can purchase the book online.

To protect ADC input pins you may want to buffer with an opamp in unity gain configuration. To the non-inverting input you attach two shunt diodes, a

Jung points out that glass rectifiers will experience a photo-voltaic effect -- small potatoes in your application but problematic if you really have to measure something.

Yes, after checking with a few other people it seems that I need transzorbs on all inputs that are connected to the unregulated power system.

Does anybody know which transzorb is suitable for vehicle/boat usage?

Well, a lead-acid battery being charged fast will have a terminal voltage up to about 14.6V, so I'd pick a transzorb where the minimum rating (they're rated min to max; there's a tolerance) is above that:

We used to use the 1.5KE series:

so, in this case, I'd probably go for the 1.5KE15A (1N6725A) or 1.5KE16 (1N6276).

If it's a "24V" system, I'd just go for the 28V or 32V 'zorbs.

And, of course, there are others - here's the full range at Vishay:

Have Fun! Rich

Thanks for this advice!

And how do you think about an extra resistor in series with the transzorb? I understand that it is better from a protection point of view, but is it necessary?

Huh?

A series resistor would raise the transorb's trip voltage unacceptably and uselessly.

Oh I'm sorry. I mean this

• -------- R -----+ ------- | - APPLIANCE ^ |

- ------------------+---------

A hash choke:

Good Luck! Rich

Do this:

• -------- L -----+ ------- | - APPLIANCE ^ | - ------------------+---------

;-)

Cheers! Rich