Constant Voltage Regulator

[kevin_tbird]

Back to my vehicle. After adjusting the CVR it is running at about 50% duty cycle and cycling between 0v and 12v. The question is why the GAUGE performs differently when the CVR is grounded.

I'm wondering if I could have a short to ground inside the gauge, but am not sure how to tell. I cannot find any specs for internal resistance in the gauge.

To be clear 12v is supplied to the CVR 100% of the time and 12v is supplied to the gauge about 50% of the time.

Kevin

[Jimz Bird]

Keven,

Sorry about rambling on

Have you tried the 6V Lantern battery to see if that cause the gauge to sweep? Actually 3V should give you about a quarter of the way up the scale and 4V about 3/4 of the way. The lantern batter is just a bit easier to use due to the spacing of the terminals.

What is the resistance between the terminals of the gauge.

It should be in the range of those listed.

"CAUTION, this unit must be installed and grounded with mount screws prior to operation check or applying voltage. If power is applied without a ground the unit sends 12V to the fuel and temperature gauges causing both gauges to fail."

[simplyconnected]

...The duty cycle we are talking is adjustable on the stock style CVRs. The average voltage was below 5v on two differ CVRs when bench tested (using a 12v bulb as a load and analog gauge.

My understanding is, the factory CVR's are sealed and not adjustable. Here is one sold by rockauto.com:

I could be wrong buy why produce a CVR that outputs anything but 6-volts?

...The gauges both work on the bench using the gauge itself as the load and tapping a 12v source on and off with the sensor wires connected to ground.

Kev, that's not how it works in your car. The gauge is simply another resistance in series with your fuel tank sensor. The tank sensor is like a turn signal flasher unit, with a bi-metal strip that opens and closes.

...So instead of throwing parts at the problem (it may come to that) lets try and figure out why a ground at the CVR frame is necessary and why it would make a difference to the gauge output...

I totally agree. Inside your CVR is a bi-metal strip (a heater that flexes at regular intervals, like a turn signal flasher). It needs a feed and a ground to complete the 'flasher' path. Also on the bi-metal strip is your output wire. When the power is on, the strip makes and breaks a set of grounded contacts based on current flow. When the contacts open, your gauges see 12-volts. So let's check out the current to ground on your CVR.

Back to my vehicle. After adjusting the CVR it is running at about 50% duty cycle and cycling between 0v and 12v. The question is why the GAUGE performs differently when the CVR is grounded.

I'm wondering if I could have a short to ground inside the gauge, but am not sure how to tell. I cannot find any specs for internal resistance in the gauge.

I think Jim gave you resistance values. BTW, your gauges are totally isolated from ground, so do a resistance check from the gauge post to ground.

You used a light bulb for a load? What kind of bulb? How many amps were going through it?
Use your meter to; check current coming into the CVR, current flowing through ground, and output current. Do these tests with the CVR connected to the proper sending units. An empty tank will have the highest resistance, but a full tank should have nearly zero resistance.

To verify proper currents, use 4 'D'-cells in series or a lantern battery to power your gauges with the ammeter connected in series.

...To be clear 12v is supplied to the gauge 100% of the time and 12v is supplied to the gauge about 50% of the time...

Huh? Is this a possibility?

[kevin_tbird]

Resistance readings at 72 degrees ambient. CVR and Gauge disconnected from vehicle.

CVR
IGN terminal to case 60.5 ohm
output terminal to case 60.5 ohm
IGN terminal to output terminal 0 ohm

Fuel gauge
post to post 12.6 ohms
no continuity to ground from either post

Temp gauge
post to post 12.5 ohm
no continuity to ground from either post

Next tests were performed using a Lantern battery outputting 6.16v.
Fuel gauge moves to just past 3/4 full
Temp Gauge moves to max.

Using two batteries to obtain an output voltage of 11.87v. Wiring the CVR in series with the gauge and grounding the CVR case to the battery the CVR produces 7v.

I.m now looking for some resisters to continue testing.

[LuckyJay]

It may be impossible to create modern solid state accuracy while using 50+ years old production mechanics. I've been following this thread with interest. My take is that the fuel gauge is at best, simply an indicator of the tank's remaining contents. My practical experience with most vintage auto instruments does not allow me to trust running things on empty.

[Joe Johnston]

I had a couple of originals. The metal cap can be pried off to see the innards. Looks like they are somewhat adjustable to me as the screw pushes on the contact arm. I don't remember if the new unit had a screw like that or a LED sticking up from the fiber back.

[simplyconnected]

It may be impossible to create modern solid state accuracy while using 50+ years old production mechanics...

Jay, you would be surprises at the accuracy you can get out of our old gauges. That technology hasn't changed much.

You're right about part of it; if the CVR is off from the factory and the voltage is funky, then the currents cannot be trusted. That's where a solid state CVR shines (with a modern fuel tank sending unit). So in this case, modern is much better. The old stuff will work, but it only gives a rough estimate.

Kev, resistance readings don't mean much when measuring heating devices because resistance goes down as they heat.

Ex: I have a 60-Watt GE household light bulb (a resistive device) in my hand. It measures 17.6-ohms resistance. Wait a minute, using Ohm's law; Volts over Resistance = Amps (115/17.6=6.5-Amps) That can't be right... 60-Watts/115-V only equals 0.52-Amps. <--that's more like it.

If I measured Current, it would give me the real answer because resistance changes as it operates!
You need to measure real current. Most ammeters will read up to ten amps. You should expect under one amp.

Reading resistance of a gauge is not resistive, either. It is a winding that produces magnetism (pushing against a spring like a speaker), so it has impedence. Again, go for current at six volts.

When measuring the mechanical CVR, current will be high at first, then settle down and give more accurate cycles. Ever notice a turn signal flasher unit? The first couple flashes are faster, then it slows down and duty cycle becomes more consistant (especially in Winter). Same thing.

Great pic's, Joe. - Dave

[YellowRose]

Constant Voltage Regulator

The RT-Engineering IVR3 has an LED right above the Output connector that shows you whether or not it is operational. Of course, once it is installed and the instrument panel is back in place, you cannot see that LED. But if you were to take out one of your gauges, and it is mounted to one of the metal supports, you should be able to see if that LED is lit. It, as stated before, puts out a continuous 6v to the gauges. Here is a pic of it.

[scumdog]

I had an issue with my gas, oil and temp guages having high readings i.e oil and temp were always 'pegged-out' when the engine was running/warmed up.

And the fuel guage only started working after a 7 - 8 mile trip (although SOMETIMES it would work sooner) but once it was working it would keep going i.e. when stopping for gas and taking off again it would start reading straight away.

But it 'pegged out' when the tank was full, showed about 1/4 tank when almost empty.

Tried a new CVT but nothing changed.

Would an solid state one help?

[kevin_tbird]

Dave

Testing on the bench with the entire circuit set up and an ampmeter between the CVR and the fuel gauge I get .79 amp (digital amp meter).

The gauge reads zero with the amp meter installed.

[simplyconnected]

Dave

Testing on the bench with the entire circuit set up and an ampmeter between the CVR and the fuel gauge I get .79 amp (digital amp meter).

The gauge reads zero with the amp meter installed.

The ammeter adds nearly zero resistance. .79a is quite healthy. I assume your bench is using 12v with the CVR, is it a car battery? While this is going on, what's your voltage and duty cycle? I'm trying to determine if you are getting a big voltage drop across your bench batteries while under a load.

Yes, Tom. A solid state CVR would help a lot.

[YellowRose]

Constant Voltage Regulator

Tom, you might consider pulling your fuel indicator in the gas tank and checking it out. If it is like ours in the Squarebirds, the rheostat in it could be gunked up, dirty, bad, not functioning properly. Cleaning the tank fuel indicator unit might clear up your problem, or replacing it with a new one. At one point I was getting faulty indications on my fuel gauge and when we pulled it, I found that the float was full of gas! It had a pinpoint hole in it allowing the gas to fill up the float. That would be worthwhile checking also.

[kevin_tbird]

Dave

I am using two 6v lantern batteries. They output 11.54v.

With the circuit complete I measure 10.2-11.2v probing at the CVR ignition terminal. I have voltage fluctuating between 2.5v and 9v with a pretty steady sweep pattern. I'm guessing close to 50% duty cycle.


Add the ampmeter to the circuit (reading .78 amp and the voltage drops to 4.2v. The output voltage holds at a steady 4.2v also

Can you suggest how to measure duty cycle? My craftsman autoranger has a duty cycle setting, but I'm not sure how to use it. It is also the device that has my ampmeter - so I cannot use it for duty cycle at the same type as amperage.

[simplyconnected]

Kev, the reason your gauge needle isn't deflecting is because you're only supplying 70% of 6-volts. That's a 30% reduction in current (which is huge). Have you got an old car or motorcycle battery you can use on the bench? Current draw is under an amp but it's too much for your dry cells. Any heating element (like bi-metal strips) always draws a lot of amps to create enough heat to bend the metal.

You really need to be on the top side of 12-volts because your car runs at 13.5-volts when the gen/alt is turning.

It's GREAT that you have a sophisticated Craftsman VOM.


Duty cycle is, time on vs time off. This is important for controlling fuel injectors. We have a fancy name, Pulse Width Modulation (PWM), but it's really Duty Cycle.

Try setting the meter to 'duty cycle' and keep the prods in the volts/ohms & common sockets. Then measure like you measure, Volts.

Constant on = 100% duty cycle, and so forth.

I use my oscilliscope to measure duty cycle because I can set the time divisions. I don't have another instrument that reads Duty Cycle. My Fluke measures frequency and capacitance but not duty cycle. Hey, see if your autoranger shows 'average volts'.

Call me if you like. We don't live far from each other. 248-544-8834 - Dave

[scumdog]

Tom, you might consider pulling your fuel indicator in the gas tank and checking it out. If it is like ours in the Squarebirds, the rheostat in it could be gunked up, dirty, bad, not functioning properly. Cleaning the tank fuel indicator unit might clear up your problem, or replacing it with a new one. At one point I was getting faulty indications on my fuel gauge and when we pulled it, I found that the float was full of gas! It had a pinpoint hole in it allowing the gas to fill up the float. That would be worthwhile checking also.

How likely to 'create' a leak is it if I pull the fuel indicator the sender??

I'm a long way from help (and parts) if the sender unit does not re-seal when I fit it back into the tank.

I'm loathe to do anything to disable my pride'n'joy for any length of time!