What's wrong here? The voltage on my meter (stock- center console) falls from near 13 to 11-and-a-bit, daytime, and has a corresponding drop from a slightly lower starting (mid-12 area) with the lights on. Turning on the lights, without other loads, barely deflects the needle down, maybe .3 volts--that's for 2 fr. corner, two taillights and two markers. Why would the brake bulbs drain so much more? -is there a short somewhere--and no, I haven't been blowing any fuses. Just wondered--don't want to get stuck in an after-dark traffic jam, sit with the brakes on and drain the battery. -C

if all of your connectors are clean and wiring is healthy, then i'd load test your battery to verify it's health.

everyone says, "ohhhh, my battery is only X years old..."

whatever. load testers are cheap easy insurance for your battery. and you'll drive with a bit more confidence and probably save one AAA use. =-)

Roger Good luck.

I'll check the battery--but that doesn't answer why so little voltage drop with the first position of the lights in the on position --and I might add, not much more drop when the fog/driving lamps are on as well, or enen (a bit more drop, but not much) when the headlights are on. Just the brakes draw it down a lot. -C

Check your wiring first. Make sure that the connectors arn't all rusted, etc. More resistance = more load = more heat and could eventually lead to burned wires.

Zach

Mine seems to do the same thing and the battery is new.

No reason to check ANYTHING! This is typical for 914! The volt meter is wired in the same circuit as the brake light switch and we all know how much power two brake light bulbs take to operate! This is one of the tech tips in the Tech Tips 700!! Typical 914 (we love their little quirks!)

As I always tell people, the voltmeter is there "strictly for entertainment". It even jumps around when you have your turn signals on, which makes me snicker.

If you want an actual indication of the health of your charging system, you need to measure with a decent voltmeter directly across the battery posts. Not the clamps, the posts themselves.

Once you get to know what your stock VM does under different conditions, you may be able to use it to know when a light bulb has burned out. Apart from that, it's pretty useless...

--DD

I have to agree here. Mine does the same thing and I used to worry about it all the time. After never having a charging problem, I just learned to ignore it.

Thanks, George. I read the book, cover to cover, in two straight stints while I was waiting to see if they would call me in as a substitute for jury duty (they never did). Guess I should go back and take another look at the ones I flagged or earmarked as important--I'd probably retain more w/a second reading.
So, good doctor, is there any modification to the hook-up that would render the gauge more accurate/useful?

Yeah... if it were run between the voltage regulator positive and battery negative (or engine ground). That would give an accurate reading of your voltage output.

Any downside to that? Do others do this modification?

I always looked at the voltmeter as a "mechanical turn signal indicator".

The damn thing is fairly worthless for actual voltage readings.

Just for fun, I took my 914 voltmeter to work and verified its calibration with a laboratory power supply. Within its typical operating range (say, 10V to 14V), the meter's error was significantly less than 0.25V. Outside of that range, the error was less than 0.5V.

OK. You won't be making any precision measurements with this kind of accuracy, but it's good enough for use on the dash of a 914.

So if the meter is so decent on the bench, why is it so darned awful when hooked into the 914 wiring? The answer is this: the problem isn't with the meter -- it's with the wiring. The meter is accurately reporting the voltage at the FUSEBOX, not at the BATTERY. So if the voltmeter twitches by 0.5V in synch with the turnsignals, it's because the voltage at the fusebox is twitching by 0.5V.

So why does this happen? OK, just some rough numbers: 2 21W turnsignals will pull roughly 3.4 Amps. Given this current, a 0.5V twitch requires a wire resistance of only 0.15 Ohm. The circuits probably had that kind of resistance when they were fresh out of the factory -- 30 years of oxidation on the terminals has only made the resistance greater.

The fundamental problem is one of design -- the engineers neglected to ensure that the voltage sense lead carries little current.

So what's the solution? Run a new voltage sense lead all the way from the battery positive terminal. Connect it to the voltmeter by passing it through the contacts of a relay. Switch the coil of the relay with the previous voltmeter lead. There you go. The resistance of our new sense lead isn't significantly less, but because none of our lighting currents flow across that resistance, we'll no longer have the voltage error. The key here is to use the sense lead ONLY for sensing -- if you have any other load being fed by the lead, you'll corrupt the accuracy of the meter.

Or you could just tolerate the twitching needle...

Good writeup Dave. You are spot on here. I like the idea of using the old volt meter connection to energize a small relay. This also will save on draining the battery over long periods (although that is a very small drain).