The following info comes from the 1981 Light Truck Shop Manual:

• The write up on the fuel gauge does not give resistances, either for testing its readings or of the gauge itself. Klricks - Where did you get the info on the sender showing 10 ohms at full and 73 ohms at empty? That looks to be right, but I'm just curious.
• The oil pressure gauge should show mid-scale with 22 ohms, and at the L, meaning Low, with a 73 ohm resistor. Apparently this replicates the Rotunda 21-0015 tester's values. And, the gauge itself should show between 10 and 14 ohms.
• Temp gauge is tested with a 10 ohm resistor and a 73 ohm one. The 10 ohm should put "the side of the pointer closest to the H no more than one pointer width away from either side of the H". The 73 ohm resistor should put "the side of the pointer closest to the C within one pointer width away from either side of the C". In other words, the 10 ohm resistor should make it point at H and the 73 ohm should make it point at C. And, the gauge itself should measure from 10 to 14 ohms.

So, it looks like 10 ohms is full scale, 22 ohms is mid-scale, and 73 ohms is the bottom of the scale, such as Empty, Cold, or Low. And, the gauges themselves should read as between 10 and 14 ohms on a meter.

Which brings me to an admission - I goofed! I forgot the resistance of the gauge in my calculations of voltage in the previous post, so I will be editing that post ASAP to account for the added resistance. And, I'll post here when that has been done. The previous post has been corrected - I hope.

 

My .02 worth.

Being that you have the resistance of the gauges sending units,Make a simulator with variable resistors to make the gauges read full,empty, etc.

You have an extra plug for the cluster. Wire your simulator up and set your resistance for the gauges to read a certain reading..

One that the manual says should read , say Full.

If you make a voltage regulator that is adjustable, you could then adjust the voltage to whatever it needs to make the gauges read what they are supposed to ...

Would that work...?? Trav...

 

Yep, that's the way to do it. Let me say it another way, just to make sure. Set up a cluster with the 8 to 9 ohm resistor on the input, and two resistors for each of the 3 gauges, a 10 ohm and a 73 ohm, selectable by switches. Fire it up and see that the gauges read what they should read at both ends of the scale with the ICVR operating. Then, take the ICVR out and put a variable-voltage regulator in its place and vary the voltage until the gauges read correctly at full scale. Then, check the readings at the other end of the scale. If they are right you are done. Go build a fixed-voltage regulator.

 

That is the deal.
If you want to get fancy, you could put it all in a plastic box and use the switches for set resistance or a variable resistor and dial up anything from low to high...to check out the full swing of each gauge...


At least we're on the same page....Later...Trav...

 

Don't see any need to put it in a box. In fact, as I think about it there is no reason to test all the gauges at the same time, so one set of resistors is all that is needed. But, that assumes the regulator can handle all the gauges showing full scale at the same time.

So, the question becomes how much current will be pulled with all gauges at full scale? Going back to a previous post that worked out that the voltage across the gauge/sender combo would be 6.6 volts if all gauges were at full scale - although the ICVR wouldn't be giving a 100% duty cycle. Since I = E/R, the current is equal to 6.6 volts/7.3 ohms, which is .9 amp. But, that would more than peg the gauges if left on continuously, so the voltage will be something less than that, and the current will be less as well. IOW, one of the 1 amp regulators will be adequate.

However, we are talking about using a variable regulator so we can pick a non-standard voltage. The LM317 is one such regulator and it has a 1.5 amp capability. Having said that, it has a minimum 3 volt drop from the input voltage to the regulated voltage if I read the spec sheet correctly. In other words, if we need something like 5.5 volts to the gauges we'll need at least 8.5 volts to the regulator. But, with the resistor in the wiring harness of about 8.6 ohms, we will have a voltage drop across the resistor of 4.3 volts at .5 amp, which will bring the 12.8 volts of the battery right down to the minimum allowable input voltage to the regulator of 8.5 volts. In fact, if we pull .75 amp that will drop 6.45 volts and at 14.4 at the battery we would only have 8.0 volts to the regulator. So, I'm thinking that the resistor in the harness has to be bypassed.

 

This is one of those trial and error deals..
Is it fairly easy to bypass the resistor in the harness?
Check with resistor in and with resistor out...

 

I don't know where the resistor is, so I don't know the answer to that question. Agree, it will be trial and error. I'm used to that.

 

You aren't alone with being used to trial and error..
For a long time, I thought I was the only member of that exclusive club.

Gotta go. I have to put new water pump on the 351..Started leaking out of weep hole on bottom...Fun time tomorrow...Later...Trav...

 

There is NOTHING fun about replacing a water pump on the small block. There must be four different types of bolts holding it on, with differing lengths. And one with a totally different head. Further, some of the are highly likely to twist off due to corrosion. Good luck. Take a lot of pictures and make notes.

 

In this picture(which I think is correct) http://imageshack.us/a/img507/4813/solsel.jpg The diagram shows the whole setup somewhat like a voltage divider. You need to think of this circuit like two different circuits.

The one circuit comes from the battery, goes through the 8 to 9 ohm resistor, through the heating element in the IVR, and then to ground.

The other circuit follows the same path, but flows through the gauges.

Let's forget about the IVR blinking on and off, l want to address the concerns about the sending unit positions affecting the voltage level and the readings of the gauges and or the IVR's reaction time.

Pretend the IVR is on all the time. In standard voltage divider design, the current flowing through the divider is supposed to be 10 times the amount that flows through the "tap". If it's designed like this, then any load variation in the "tap" circuit will not significantly affect the voltage divider's ability to hold the voltage at a constant level.

To put the above in simpler terms, I bet(have not measured it myself) the current flow through the IVR to ground is so much larger than what the gauges actually use, that any variation from a full tank to a empty tank, or hot or cold engine, etc will not affect the IVR's operation.

 

I got the info from '86 Ford F150-F350/Bronco shop manual.

I just now went back and looked up info on each of the 3 senders.
There is a paragraph written up for each sender showing how to do a manual test (without using a special tester).
Fuel and temp gauge: use a 10 and 73 ohm resistor.
(At 10 ohms the pointer is 2 needle widths above F and 73 ohms puts the needle 2 needle widths below E).

Oil gauge use a 10 and 22 ohm resistor. The 22 will move the pointer to the high band which is really mid scale on the gauge.

Note that the senders are not linear.

The gauges themselves should be 10-14 ohms.
So that pretty much verifies what you found.

Note that the most of the wiring diagrams in my manual do not spec the ohms except for 1 drawing for the F250-F350 diesel motor. That one does show 8 and 85 ohms. I don't know if the senders are actually different or maybe that is the absolute full scale of the senders?

Also note that in '87 Ford went to a 16 ohm empty 158 ohm Full system:
Welcome to Classic Instruments


Here is a free program that allows you to simulate electrical / electronic circuits. It takes some getting used to in order to place components but once you get the hang of it ..... it works very well. You can place multiple volt and amp meters anywhere in the circuit..... change R values and see what that does to the current / voltage etc.
Solve Elec - Educational software

.

 

Dave - This is one of the rare times I will disagree with you. My ICVR's measure about 64 ohms for the heating element in them. And, the other leg, the one with the gauges and senders, will vary from 7 ohms to 20 ohms depending on the temp/pressure/fuel load. So, it is exactly the opposite to standard voltage divider design as the current flow through the heating element is very small in comparison to that in the gauges. In fact, the load on the 8 - 9 ohm resistor will change almost by a factor of 3, which means the voltage drop across it will vary by that amount as well.

Would it help if I drew the circuit up, leaving out all the connectors, references to the tachometer and fuel tank valve, and added the resistance values? I think both of us understand the circuit, but maybe for others?

 

Kl - That does pretty much confirm what I found, with the minor exception of where the needles will be when using the 10 and 73 ohm resistors. The 81 manual says the edge of the needles should be within one needle-width of the high and low, but yours seems to indicate the center of the needle should be two needle-widths above on the high side and below on the low side. I THINK there is a minor difference in 1/2 of a needle width in those two descriptions.

And, I'll play with the simulator and see what I can come up with. Thanks.

 

If you want PM me and I will send you my source files already drawn up.
I have one to simulate the existing system and one to simulate using a fixed voltage regulator.

 

The only problem with that I see is the resistor wire is not exact. It's between 8-9 ohms. Your truck it may be exactly 8.0, Franklin2's 8.3, Truckertav's 8.5, klricks' 8.7, and mine 9.0 as examples. This is due to the length of the resistor wire and how it was cut for the wiring harness. It's not perfectly cut between each vehicle, although it's close it's not exactly the same.

Most people would not notice if you design the ICVR giving account of a resistor wire at a 8.5 average.

Yes the input voltage does vary, as well as the resistance from the gauges and sending units, and the ICVR would need to compensate for all of that as well.

This is how I understand the circuts...

Think of the ICVR as operating like a glorified Thermal Turn Signal Flasher. When the resistance of the lightbulbs heat up the flasher, it breaks contact. When it cools down it closes and alows the light to burn. This is what causes the bulbs to flash. Without the resistance the flasher never breaks connection etc.

The factory ICVR works in much the same way. Resistance creates heat, and the ICVR cycles on and off in rapid succession and changes due to the resistance values. When they fail, the contacts stick and the gauges peg. When you hit the "dash", it unsticks and works again for a little while... etc...

Where it differes from a thermal turn signal flasher is that the circut is wye'd off. One portion goes to the gauges and one goes through the heating element to ground, as already discussed. It also cycles a heck of a lot faster.

This is why when you have a bad ground to the ICVR, or forget to screw it back onto your printed circut like another forum member did, the ICVR pegs the gauges.

It does work like a tap, but most of the current as you say goes to the gauges.

With the heating element with it's dedicated wye hooked to ground the variable resistance of the senders and gauges and the system in general is compensated by this. The more resistance caused by the average of the gauges and sending units combined, the more heat and the faster the cycles, the less resistance, the less heat, the slower the cycles. Just as Klricks states.

In this way the gauges get a average of 5.5 volts no matter what the sending units resistance levels are, as the cycles of the ICVR's heating element cycle at different rates to compensate.

A electronic ICVR design should alow for the same variables. If it's output is fixed, this would theoretically become a problem when the resistance values change. It needs to be variable.

As for bypassing the resistor wire, not a good idea as it goes from the fuse panel to the instrument cluster with another section wyed off and terminated in the harness to reach the required length.

I would want something that was plug and play, no wiring mods and be as accurate as original.