IVR pulse at 4 volts gauges read low

[1969383S]

Working out those minor details. 69 Barracuda

My IVR pulse is only about 4vdc, gas, temp and oil pressure gauges are all working but all seem to read lower than expected. Ohm readings on sensors seem to be ok.

Before I yank the cluster back out and do the solid state upgrade looking for some opinions on the voltage output. Would this lower voltage cause the lower readings and just be corrected by getting the correct 5v signal, or would it be wise to replace all the gauges while the cluster is out?

I took voltage measurements with a quality analog meter from the three sensor wires. Battery voltage at 1000 rpm reads 14v. All wiring harnesses are new M&H pieces.

 

[harrisonm]

I restored a 69 Barracuda several years ago and ran into LOTS of electrical problems including instrument problems; to include the IVR problem. I did a TON of research and learned a lot. I can tell you that the current pulse coming from your IVR should be about 5 volts, so at 4 volts your instruments will run low. I assume you know that the IVR in your car is built into your Gas gauge. That is why the gas gauge has three terminal on the back, and the Temp and Oil Pressure only have two. The Alternator gauge has two terminals also, but is not on the gas, oil, temp circuit. I can go into great detail on how the Gas, Oils and Temp circuit works and how to test your gauges, if you would like. I would highly recommend D&M restoration to either repair the IVR in your gas gauge or to help you with a digital IVR upgrade. You can call them and they will be glad to talk to you about it. If you want my rather lengthy iteration on the gauges, let me know. http://dandmrestoration.com/

 

[Pete in NH]

Hi,

You can't really measure the IVR output voltage with a conventional voltmeter. The original IVR is a thermal/mechanical device that slowly switches between open and +12 volts. On a conventional meter you would see this pulsing, not a steady voltage. I'm puzzled if you are indeed seeing a steady 4 volts. The gauges are thermal devices that average this pulsing voltage into a heating value of something around the equivalent heat value of a steady 5 volts.

If you are seeing a steady voltage , either someone has already done an electronic conversion or something is messed up somewhere.

 

[1969383S]

I let me know. http://dandmrestoration.com/

Thanks harrisonm,

I am aware the IVR is in the gas gauge and that it will have to be disabled to add one of the solid state solutions. My thinking is if the voltage was at 5v then the current flow would increase slightly based on a standard resistance value (say 26 ohms roughly 1/2 tank of fuel). This should raise the gauge a tad higher if my math is correct. The same should hold true for the temp and oil pressure.

I hate the need to pull the cluster back out but it is what it is. I also hate to spend another 500? to replace them when they are functioning. Guess I need to decide to just do the IVR or do all of the gauges while I have it out.

I will look into repair with D&M.

 

[1969383S]

Hi,

You can't really measure the IVR output voltage with a conventional voltmeter. The original IVR is a thermal/mechanical device that slowly switches between open and +12 volts. On a conventional meter you would see this pulsing, not a steady voltage. I'm puzzled if you are indeed seeing a steady 4 volts. The gauges are thermal devices that average this pulsing voltage into a heating value of something around the equivalent heat value of a steady 5 volts.

If you are seeing a steady voltage , either someone has already done an electronic conversion or something is messed up somewhere.

Its a pulse from 0v to 4v. I used a tripplet analog meter to read it diconnected from the sender.

My DVOM (Fluke) will not pick it up.

 

[Inertia]

What is the IVR voltage with the engine running at 14V..???

 

[1969383S]

What is the IVR voltage with the engine running at 14V..???

I did not check the input voltage to the fuel gauge if that is what you are asking. I will do that tonight and post the result.

 

[Inertia]

I'm suggesting that if your reading "about 4vdc" at battery voltage of 13v or likely less,, then running,, it's gonna be closer to 5v..

Being factory guages,, they are more "indicators" than true, dead-on measurements..

 

[1969383S]

I'm suggesting that if your reading "about 4vdc" at battery voltage of 13v or likely less,, then running,, it's gonna be closer to 5v..

Being factory guages,, they are more "indicators" than true, dead-on measurements..

Got you on the quality.

I took all readings with car running and the idle set to 1000 RPM. Last time the car was roadworthy the gauges all read much higher than they do now (over25 years). I have about 35-40lbs of oil pressure at idle and the gauge used to read about 1/2. It only reads about a 1/4 now. I removed all the aftermarket gauges on the restore and do not want to put them back in even though they were much more accurate. I used to run them in tandem and they seemed pretty close only slower to re-act.

 

[Pete in NH]

Are the senders still connected while you are making your measurements? If so, you have a voltage divider being formed by the gauge itself, something around 20 ohms, and the sender. For example: 80 ohm sender/ 80 + 20 ohm gauge = 4/5 of original voltage pulse.

 

[1969383S]

Are the senders still connected while you are making your measurements? If so, you have a voltage divider being formed by the gauge itself, something around 20 ohms, and the sender. For example: 80 ohm sender/ 80 + 20 ohm gauge = 4/5 of original voltage pulse.

No. Disconnected one a time and checked at the sensor input. Maybe I should unplug all 3 at the same time and re-test.

 

[Pete in NH]

Yes, I would disconnect all three sensors. Then you should see a higher swinging pulse. It will depend on how fast your meter reacts as to how far up the reading will go. Again, measuring the output of an electro-mechanical IVR with a conventional voltmeter is kind of problematic.

 

[1969383S]

Yes, I would disconnect all three sensors. Then you should see a higher swinging pulse. It will depend on how fast your meter reacts as to how far up the reading will go. Again, measuring the output of an electro-mechanical IVR with a conventional voltmeter is kind of problematic.

I will do it this weekend and video the meter readings. One other thing I noticed was every 4th or 5th pulse was quicker than the others.

 

[nm9stheham]

As said, part of the problem is that a voltmeter will not properly follow the voltage pulses. Even if the IVR is swinging to +12v, it is for too short a time for the voltmeter to typically show it. Occasionally, a digital voltmeter might show the true pulsed IVR voltage if it just happens to sample it at the right instant, and has the right meter response. The best you can do is to watch it for a minute and note the highest readings it measures, and consider that the pulsed IVR voltage. You need an oscilloscope or similar fast voltage-reading device, or an old, slow thermal voltmeter, to really know.

BTW, you CAN measure this properly by just disconnect the temp sensor wire from the temp sensor and measuring the voltage from the temp sensor wire. That connects to the IVR output through the temp gauge's resistance but with no current through the sensor, the gauge resistance's voltage drop is zero (i.e., it 'dissappears'), so the wire shows the actual IVR voltage. You don't need to disconnect all 3.

The IVR voltage being 'low' is indeed consistent with the low gauge readings. It is more likely that the IVR is 'short cycling', i.e., the 12 pulses that it is putting out are too short and giving you a low average voltage. If your Fluke meter is not showing anything, then I really suspect the short cycling of the IVR. The Tripplite analog meter is showing all that it can....

Getting to the IVR in your case is a pain; otherwise there is an easy way to check with a 6V battery.

 

[1969383S]

The IVR voltage being 'low' is indeed consistent with the low gauge readings. It is more likely that the IVR is 'short cycling', i.e., the 12 pulses that it is putting out are too short and giving you a low average voltage. If your Fluke meter is not showing anything, then I really suspect the short cycling of the IVR. The Tripplite analog meter is showing all that it can....

Getting to the IVR in your case is a pain; otherwise there is an easy way to check with a 6V battery.

Thanks for the response!

Care to elaborate on the 12 pulses? I am seeing erratic timing in the pulses. The readings I took were at each sensor one at a time disconnected one at a time. All of them read the same pulse voltage and irregularity. I do have a o-scope as well somewhere, and can use it. The analog meter picked it up well.

 

[1969383S]

The IVR voltage being 'low' is indeed consistent with the low gauge readings. It is more likely that the IVR is 'short cycling', i.e., the 12 pulses that it is putting out are too short and giving you a low average voltage. If your Fluke meter is not showing anything, then I really suspect the short cycling of the IVR. The Tripplite analog meter is showing all that it can....

Getting to the IVR in your case is a pain; otherwise there is an easy way to check with a 6V battery.

Got to thinking more about what you are saying. The fact my analog meter clearly showed the pulse from 0v to 4v @ about 1-1.5 second pulse seperation (more off than on) the short cycle theory sounds like the culprit. Which seemingly would make the gauges read low.

Next question would be has anyone just pulled the 12v feed wire from the barrell connector and re-route to the new solid state IVR. Thus bypassing the original IVR without pulling the gauge out and disabling it?

 

[nm9stheham]

Well if you have an o-scope, then use that to look directly at the voltage at any of the wires to the sensors. The 'ideal' duty cycle is around 17-18% with an input of 12v. (Contrary to the info on Allpar.com) At 14v, the pulse % may be a bit lower. The pulse rate of 1 second or more sounds fairly normal.

The way this IVR works, is that when power is applied, it sends 12v through to the gauge circuits, and at the same time, applies 12v to a nichrome heating wire in the unit. This heats and bends a bimetallic arm, and when this bends enough, is opens a contact at the input; that removes the 12v from the heating element and the gauge circuits. When this cools, the contact closes and 12v is applied again to everything and the cycle starts over.

The duty cycle is set at around 17% so that the RMS voltage (not the average voltage) to the gauge circuits is 5v. If your IVR is short cycling, then the duty cycle will be around 11% if the RMS voltage it is sending to the gauges is 4v. (But I don't think your 4v analog meter readings are giving you the true story.)

 

[cannucky]

Got to thinking more about what you are saying. The fact my analog meter clearly showed the pulse from 0v to 4v @ about 1-1.5 second pulse seperation (more off than on) the short cycle theory sounds like the culprit. Which seemingly would make the gauges read low.

Next question would be has anyone just pulled the 12v feed wire from the barrell connector and re-route to the new solid state IVR. Thus bypassing the original IVR without pulling the gauge out and disabling it?

I have talked to Ben aka DemonIVR about this quite a bit , he no longer sells the IVR directly he supplies a reseller in bulk now but he explained how to disable the internal unit to me in detail and the bottom line is because the circuit routes through the guage then feed the other 2 guages pulling the feed wire and re-routing to the new IVR either breaks the circuit completely or provides 5 volts to the wrong side of the internal unit and can lead to a problem from that . It sure is unfortunate he is no longer doing this kind of work but his retirement hobby became too much of a job for him , the tricky part is getting the gauge apart and back together without breaking that wafer material that holds all the terminals and internals .

 

[1969383S]

This is off of the RTE website.

The 12V input would most likely still be connected through the internal contact to the regulated terminal (Center pole). If the feed 12V was disconnected at the barrell then the 5V could back feed up to that point right. Would it hurt that the regulated voltage was going back up the orginal 12V trace to no-where? Maybe there is more to this picture that I can not see!

Might be grasping at staws but I really hate to have to pull the cluster back out. I am liable to scratch some of that nice fresh paint up.

 

[nm9stheham]

If you can get to where the 12v comes in, and disconnect that, it will not matter if the new 5v regulated feeds back to that contact. The 5v needs to feed to the gauge movement heating winding in the fuel gauge, and then on to the 2 other gauges. What you can't do is feed the new 5v into the old 12v terminal.

If you have the o'scope, I would do that first to make sure the old IVR is the issue. BTW, the pulse duty cycle may be fairly ragged and somewhat erratic but still be OK so you have to look at it for a while.

 

[1969383S]

If you can get to where the 12v comes in, and disconnect that, it will not matter if the new 5v regulated feeds back to that contact. The 5v needs to feed to the gauge movement heating winding in the fuel gauge, and then on to the 2 other gauges. What you can't do is feed the new 5v into the old 12v terminal.

If you have the o'scope, I would do that first to make sure the old IVR is the issue. BTW, the pulse duty cycle may be fairly ragged and somewhat erratic but still be OK so you have to look at it for a while.

Thanks.

Found my scope. Its at a buddys in macon. Won't have it back until next week.

 

[RedFish]

If you believe that bending the little thingy is a suitable method or remnants of the mechanical limiter will not effect a solid state regulators operation, By all means purchase a regulator that has unconditional replacement warranty. This is so the repeated R&R(s) will cost you only your time/labor ( providing you don't break something else in the R&R process ).

 

[1969383S]

Thanks for chiming in redfish!

Not my method for starters! It is just one thats out there for our viewing pleasure.

I have yet to locate a site that describes a different way of disabling it in detail and any pitfalls that would follow if not done a certian way. Do you have suggestions? It would appear that demonivr.com is defunct. I do see the IVR available on e-bay though.

 

[nm9stheham]

If you believe that bending the little thingy is a suitable method or remnants of the mechanical limiter will not effect a solid state regulators operation, By all means purchase a regulator that has unconditional replacement warranty. This is so the repeated R&R(s) will cost you only your time/labor ( providing you don't break something else in the R&R process ).

Some explanation of what you say would help, Redfish. If the old IVR is out, how can it effect anything? Are you assuming that somehow the heating wire for the old IVR is still in place? (I assume it would be cut or bypassed.) The IVR heating wire is in shunt with all of the gauge circuits, not in series in any way. Hard to see where the wires go in the pix.

 

[cannucky]

he's politely telling you that the old regulator parts have to be removed completely , the "thingy" that is bent out of the way has load of it's own in the coil of wire around the bi-metallic piece and cannot be safely left in there to bounce around contacting the remaining internal pieces of the guage .That is what I was saying in my post as well , you may also notice that the terminals and "Thingy's" are not mounted to the metal housing but the wafer material I also mentioned in my first post , this stuff is very very fragile almost like a soda cracker after 45 years or so eh making this work like brain surgery .