Battery / Volt Gauge discrepancy

[jagwit]

Having just fitted a modern 130A to my Series 3 V12 (http://forum.etypeuk.com/viewtopic.php?f=6&t=9767), I was rather disappointed to find that the Battery guage/ Volt meter still went well below the "13V" indication, deep into the "Off Charge" zone.

I also could not understand why the Volt gauge presented a reading so much less than what my Fluke multi-meter was telling me....

Having checked voltages with my Fluke multimeter, I concluded that the Volt gauge was basically grossly under reading. It was on the basis of this incorrect gauge readings that I decided to upgrade the alternator in the first place!!! Having now studied the wiring diagram of the car, and making a few more measurements, I concluded that it was measuring the voltage at a spot which was rather INAPPROPRIATE from an electronic perspective.

First though, a tiny lesson in electronics:

Volts = Current (amps) x Resistance (ohm)

So:
If you tiny amps multiplied by tiny resistance, you get tiny Volts.
Eg. 0.1 Amps x 0.1 Ohm = 0.01V

End of electronics lesson. Back to the Volt gauge discrepancy:

1) The Volt meter measures the voltage on the 12V input terminals of the "voltage stabiliser". This is not a good place to measure "Battery" volts because;
2) If you follow the source of that 12V back to the terminal block next to the battery, there are 24 (Twenty four!!) electrical contacts in the circuit from the terminal block (next to the battery where the alternator output connects to) to the volt meter. This is a problem because EACH one of those 24 contacts has a tiny amount of resistance (remember the lesson??);
3) The volt meter itself has a resistance of just 120Ohm. This is bad as the volt meter alone draws approx 0.1A from the power source. This 0.1A is no longer "tiny", this falls more into the "smallish amount" category. This current passes through all 24 of those contacts (each with tiny resistance) causing a measureable voltage drop over each of those contacts which adds up;
4) Not only the Volt gauge draws current though! The Fuel tank gauge draws current, the Coolant sensor draws current and the oil pressure sensor draws current - all from the same terminal on the voltage stabilizer and through the same 24 contact points!. If you combine these currents with that being drawn by the Volt gauge, the voltage drop over those 24 contacts combined becomes quite substantial. (I am still ignoring the voltage drop over the wires and over the fuse).

So, if we go back to the electronics lesson:
Lets assume the other gauges also draw 0.1Amp (collectively) so we have a total of 0.2Amp being drawn over 0.1Ohm contacts. Now we have 0.2 (amp) x 0.1 (ohm) = 0.02V PER CONTACT. So how is this a problem? Well there are 24 of them so that 0.02V must be multiplied by 24 to get the total voltage loss to the Volt gauge which comes to 0.48V!! SO....... If the alternator is putting out 14V, the gauge would only measure 13.52V!!
(The actual numbers above is not exact, used only to demonstrate the principle.)


So what is the solution?

1) Remove the terminal with the green wire from the Volt gauge and tape it up so it can't accidentally make contact with anything else;
2) Install a wire terminated with FULLY INSULATED 6mm FEMALE spade terminals at both ends from the Volt gauge to the RIGHT HAND SIDE of fuse block as shown in the picture below. Its the thin green wire with the red terminal.

20161120_193122.jpg (169.39 KiB) Viewed 4670 times

This position is STILL not ideal as there is now still 10 contacts between the 12V source and the measuring point, but at least the number of contacts is now less than half and likewise the effect. I still find a 0.15V difference being measured and the actual voltage (as measured by the multi-meter). But at least the "Battery" gauge is now reporting a situation consistent with what my Fluke is telling me.

I have also found the Volt gauge to be a very crude bi-metallic based instrument. This is why the readings being displayed are not very accurate or consistent. Tapping the lens can cause the needle to move higher up the scale. This is what it looks like inside:

tn_20170301_101014_Richtone(HDR).jpg (122.62 KiB) Viewed 4670 times

[jagwit]

Having opened the Battery Condition gauge, and cleaned the lens, I now have to do the same for the remaining instruments....



BTW: The position of the needle is where it sits after a cold start and having tapped the lens. The voltage was 14.4V on the battery at this time. As the alternator warms up, the voltage does drop back to around 14V and then the needle sits on the "N" of "ON CHARGE" or between the O and N of On CHARGE.

But at least it now stays there while idling with Aircon, fans and headlights on. Only when I pop it into Drive and RPMS drop to below 500 (on the tacho) does it move lower but then the voltage does drop to below 13V as per the multimeter.

[42south]

Good post, thanks for that.
I have just upgraded the alternator on my S3, and am experiencing similar symptoms.
I think heat might have also had an influence. After a 3 hour run with 30 minutes of stop/go at the end, in warm conditions, I noticed the voltmeter at 12 volts or less, dropping into deep discharge with my aftermarket fans on high. Normally this does not happen.
I couldn't check the voltage at the battery on that occasion, but will do next time it occurs.
Then I'll make your mod.
Cheers

[PeterCrespin]

Having just fitted a modern 130A to my Series 3 V12.... I was rather disappointed to find that the Battery guage/ Volt meter still went well below the "13V" indication, deep into the "Off Charge" zone....Having checked voltages with my Fluke multimeter, I concluded that the Volt gauge was basically lying (grossly under reading). It was on the basis of this incorrect gauge that I decided to upgrade the alternator in the first place!!!

I've got no problem with Ohm''s Law but am struggling a bit with your logic Philip. If the location of the voltage tap and the interposition of two dozen connections between 12V source and gauge are the fundamental 'structural' problem you suggest, surely almost every S2/S3 voltmeter equipped car should suffer 40+ years after manufacture? The fact that most E-type VMs seem to work adequately, suggests you have overcome an issue with your particular car?

Electrical non-expert caveats apply, of course, so feel free to put me straight. But then if I knew nothing about lubrication or cooling systems (say) I'd be more inclined rather than less to check OP and temp with independent instruments before rushing into a rebuild. It's bad enough to find a used and fairly easy alternator swap made no difference to your readings, but imagine if faulty oil and water gauges still read low after 5-10 grand's-worth of engine work!

[jagwit]

but am struggling a bit with your logic Philip

You're right Peter, logic was somewhat absent from this exercise. I was a bit disappointed in myself having stormed into the alternator upgrade so unquestioningly. It was rather un-Engineer like. I do have a BSc Electronic Engineering degree, but that is a bit like saying my car has had an MOT done 30 year ago. Combine that with my Blond & Blue eyed nature, add a bit (Ok, a lot...) of OCD-ness and a heap of prejudice and there you have .... me.

With hindsight, In this instance, prejudice probably played the biggest role. I somehow had it in my blond mind that the Butec was good for only 45Amps and that it was probably rather bad in delivering current particularly at low rpms. It was only the day before yesterday that I saw the stamp on it that it was rated for 60A.

The fundamentals of an ideal Volt meter are:
1) It should have infinitely high resistance;
2) so that it does not impact in any way the circuit it is measuring.

The meter in the E, does not meet either of these requirements.

When all the electrical contacts are fresh, the downstream point where it measures is probably OK too, but after 45 years of oxidation? So yes, it could very well be that I've overcome a problem on this particular car (without yet identifying that actual problem....). As for other E's, perhaps others are more accepting of its readings than I am? OCD remember....

The principle stands though, if you want to measure battery voltage, you should measure as close to the source as possible. I would have connected the Battery gauge to the bottom left fuse blocks (where permanent 12V lives with even less connections in series) but having realised that it draws 0.1Amp, it would drain the battery in only 3 days or so. There is a 0.15V difference (OCD )between where it measures now (after changing its measuring point) and the input terminals on the bottom left fuse block.

There is another matter that also comes into this. Gauge calibration. The gauge in my car is certainly not calibrated correctly showing 13V when the voltage on its terminals are 13.45V. (OCD again...) Seeing in another thread that it can be calibrated (to some extent), that will be next on the agenda. First I need a variable power supply though.

[JagWaugh]

it would drain the battery in only 3 days or so. There is a 0.15V difference (OCD )between where it measures now (after changing its measuring point) and the input terminals on the bottom left fuse block

Wouldn't you have to have the ignition switched on for those 3 days?

0.15V difference is below the scale accuracy of the gauge. Despite having numbers printed on it what that gauge indicates is one of:
1) Charging system is not replenishing the battery. (needle left of midway)
2) Charging system is evaporating the electrolyte. (needle in red zone to the right)
3) Charging system is more or less normal. (needle between the red zones)
4) There is a poor connection somewhere (needle bouncing all over the place)

Calibrate it by all means, but remember that the basic point of the voltmeter is to tell you if you can concentrate on the other gauges, or if you need to get your Fluke out.

Andrew

[abowie]

Have you tried another voltmeter?

[malcolm]

The voltmeter on my '69 series 2 does exactly the same. I assumed I had a duff battery, or alternator, until I put a multimeter on the battery while running the car on tickover with all electrics (including the dual fans) on. The voltmeter was reading about 11 volts or less (dead battery zone!) whilst the true voltage a the battery was actually 12.5. When I ran the revs up to about 1500, the voltmeter showed 12 but the battery was really at 13.7
I just live with it, and use the voltmeter as a guage, looking for any changes from the wrong readings I'm used to! Might try the above modification to wiring though; it would be nice to know what is really happening.

[MarekH]

Firstly,

If you want to deliver 14v to the battery, or any other point, not as measured at the alternator, then put the alternator sense wire there rather than at the alternator. You can thus mitigate vltage drops in the wiring by making the alternator take up the slack by design. Typically it can be put at the battery or at the distribution point above it.

Secondly,

The bimetallic strip is a complete red herring - it simply means that the needle moves very slowly as the car shakes over all of those bumpy roads.

Thirdly,

What was your 0v baseline for this voltage? Was it at the alternator?; at the battery?;, or at the drop down panel in the middle of the car via the car chassis? You made no mention of the earth return path or its integrity. Consider some proper starter motor sized chunky earth cables between your alternator and chassis and battery.

kind regards
Marek

[jagwit]

Wouldn't you have to have the ignition switched on for those 3 days?

That point I had in mind has permanent 12V even with ignition off.

Have you tried another voltmeter?

No, but I'm willing to bet my life on the accuracy of this DMM.

If you want to deliver 14v to the battery, or any other point, not as measured at the alternator, then put the alternator sense wire there rather than at the alternator.

This alternator does not have a sense terminal. The upgraded alternator in my Jensen does and I do have such a sense wire coming from the battery to the alternator.

What was your 0v baseline for this voltage?

There is a whole bunch of earth wires that congregate onto the clock chassis. One of these had an unused female lug and that served as my ground reference. I checked today and with everything on, this "earth" had a 0.05V offset with respect to the body of the car - so its "a good" earth for the currents it has to support.

The permanent 12V point on the fuse block bottom left measured a constant 13.61V with my CTEK MULTI XS 25000 battery charger connected to the battery terminals and set to "SUPPLY" which gives a constant 13.63V on the battery terminals.

Again measuring on the Battery gauge + terminal, I was getting 13.48V meaning 0.15V has been dropped basically through the ignition switch, wires, terminals and connections between the Battery gauge and the battery terminal (Ignition switch in pos 1).

If I then reconnect the Battery Gauge to its original supply, the voltage drops to around 13.37, but it then varies as the Voltage stabiliser switches on and off.

I therefore maintain that my recommendation to move the monitoring point for the Battery gauge as per my first point remains valid.

[Whitact]

Responding to Mark's post I experienced similar symptoms during a (hot) summer run in Spain. Voltmeter began to indicate lower in the range as the day wore on. Next morning all seemed fine again, then in the afternoon a low reading. Eventually after returning home a clear pattern would repeat. Good voltage when cold, then low after the car had a run and been left then restarted. Low voltage indications were backed up by the ignition light being slow to go out as revs increased. Suspecting that the one or more of the alternator rectifiers were failing when hot I replaced the alternator and all was fine again.
It seems to me that the main benefit of the voltmeter and similar instruments is not the absolute value shown but a change in the pattern of the indications from previous trouble free periods.

[jagwit]

Further observations:

1) The volt gauge actually draws 120mA (0.12 Amp);
2) The oil pressure gauge draws a constant 105mA (0.105 Amp, engine running, gauge almost at max);
3) The coolant temperature gauge toggles* between 100mA and 0 mA (engine hot, needle on M of NORMAL)
4) The fuel gauge toggles* between 50mA and 0mA (gauge at 1/4). (I'll want to re-measure this with full tank.)
5) The clock in my car is not working, so I do not know how much current that draws.

* It seems to me the Voltage "Stabiliser" should actually have been called Voltage UNstabiliser because it does not output a constant voltage - hence the current into the coolant and fuel gauges are not constant. Once my notebook PC is fixed I'll plot a trace of this voltage to see what it actually does.

The resistance from the battery terminal to the 12V input of the Voltage stabiliser (Key on Pos 1) varies quite a bit depending on how much I fiddle the ignition key. Lowest I could measure was 4.8 Ohm (typically 6-7 Ohm). With the key on Pos 2, the lowest resistance I measured was 12 Ohm (more often around 20 Ohm). So it seems poor contacts (high resistance) in the ignition switch could be a significant contributor towards low Voltage readings.

[MarekH]

1/ Stick a relay in to bypass the ignition switch - use the ignition white fed wire to power the relay coil and the brown wire to actualy deliver the power to the circuits.

2/ You can use a 7810 in place of the bimetallic strip "10v" IVR, which is really a 12v pwm unit running 5:1 on:off to give an average of 10v.
See http://www.jag-lovers.org/snaps/snap_vi ... 1287914415
I had to do this because I run LEDs in my dash and they strobed when the old 10v regulator was fitted.

3/ All of the gauges are the same item as far as I know - just the faceplates differ.

kind regards
Marek

[jagwit]

Good suggestions Marek!

It should be interesting to see if the constant 10V supply brings about any different behaviour on the Temp and Fuel gauges.

[PeterCrespin]

This sounds like calculating tree height to the nearest millimetre - a more or less impossible exercise in chasing spurious precision.

I think the instruments and senders are generally of matched imprecision. The slow, damped, analogue movements are suited to an on-off duty cycle regulator. Substituting a perfect semiconductor VR, or a reference grade digital instrument or sender, upsets the balance of mediocrity and probably engenders a false sense of improvement. I'd guess that you should either change all three components and get very accurate readings (for whatever use they might be), or just fix things that actually break, using the instruments as monitors of changes from their previous norms rather than delivering precise numerical measures.

I do understand the constant desire to tinker and improve, of course. I myself am thinking of changing my eye-liner and mascara brands...

[jagwit]

Having implemented the 7810 and now having genuine stabilised 10V supply to the temp and fuel gauge, I can confirm that they collectively draw 100mA. (temp on N of NORMAL, fuel at 1/4)

So, the 4 gauges (excl clock) draw more than 300mA.

This amount of current is not insignificant. If there should be just 5 Ohm of resistance between the battery terminal and the gauges, 1.5V is dropped, eg from 14V to 12.5.

This, is IMHO the reason the Battery gauge tends to read low.

[MarekH]

You can easily confirm that theory by disconnecting the wires to the other three gauges and seeing whether the voltage gauge moves up.

kind regards
Marek

[jagwit]

Well I'm amazed!! Our great grandfathers knew a thing or two when it came to building cheap instruments out of a few pieces of metal and wire....

Decided to calibrate the Battery gauge (which is a Volt meter really) and found the meter to be amazingly accurate (where it matters), even after 45 years. My respects (in this regard).

Gentle tapping of the instrument housing was required for the needle to find its true position. Probably taken care of by the vibrations in the car.

12V

12V.jpg (151.19 KiB) Viewed 4710 times

13V

13V.jpg (144.76 KiB) Viewed 4710 times

14V

14V.jpg (150.81 KiB) Viewed 4710 times

So, again, it is not the gauge itself to blame, but the position at which it measures and the effect the gauges' current draw (including this gauge) has on the voltage being measured at that location (12V supply to the voltage stabiliser).