Temp gauge mystery

-
My 1973 Dart resto is now in the home stretch and I'm revisiting my temperature gauge problem. I'm using a version of the tester C-3826 that I have, where you connect one test lead to the sender terminal connector and the other test lead to a good ground. The FSM then says to turn the ignition on. My tester supplies power at different voltages to the terminal connector(3v, 4.5v, 6v, 7.5v, 9v). When I turn the tester on, the gauge needle moves the correct amount only when I leave the ignition OFF. When I turn ignition ON the gauge needle goes to zero, all the way to the left. Does this make sense? I'm confused as to why the needle drops with the key on. Does tester C-3826 supply power like mine does?
20250707_124949.jpg
 
I've never heard of this. (The idea that the tester applies power??)

The only gauge testers I've seen are nothing more than switched resistors. Typically, one for the "low" point in the gauge, one for half scale, and one for the "high" point. All the gauges more or less, have the same scale. In other words, electrically, you could take a non-Ralley fuel gauge and sub it for a temp gauge and they should read the same.

Also, a "round the corner" way to sort of test gauges, would be to "lets say again," a non-ralley cluster----rig the temp and fuel gauge so they are in series, and hook to a 1 1/2V battery, and the two of them, because they are in series and identical, should both read the same.

This is not a Mopar built tester, but it works the same. The yellow print was added by the original poster, not I. It simply shows the resistances used for the 3 readings.

So far as I know, the factory testers are the same way

c-3826-jpg.1714848381
 
Last edited:
Also, LOLOL I just found this on a post from C bodies.....how to use the dash dimmer rheostat in an old headlight switch for a gauge tester!!!
(You would have to set the resistance to the three values noted in the photo above, using your multimeter)


"Stan I have used the headlight switch to test fuel, temp and oil pressure gauges. If you hook the B2 terminal on the headlight switch to ground and the I terminal to the sending unit terminal of the gauge or the wire to the sending unit, pull the switch to the park light position and rotate the knob the gauge should go up and down accordingly"
 
Here it is. Seems to work for testing a gauge. Hook the leads directly to the gauge posts, (or the sender lead) and turn it on. 3 volts moves the needle just north of 120. 4.5 volts moves the needle above halfway. 6 volts would peg the needle all the way to the right but I shut it off before that happens. My big question remains: why does it stop working when the ignition is turned on?

I plan to make one as described, thanks for that info.
20250707_150235.jpg
 
What you have there is a power supply, plain and simple. DO NOT backfeed that into the gauges

You may/ will likely blow up the gauge regulator/ limiter

If you have the cluster out/ disconnected, and if you also have the IVR pulled out, you could wire it to the gauges to see if they move, but you have no values to justify what power level causes what action at the gauge(s)

One thing you could do is to check that the gauges are equal reading as I said before. What you would do, is, with the cluster out and disconnected, and with the IVR removed, hook either power supply lead to one of the gauge sender studs. Connect the other lead to the remaining gauge sender stud. This will feed power in series through both

Select the lowest voltage setting, and let them sit for maybe 10 seconds, see what they read. If they both move, they should read the same. If they are below half scale, step the voltage up one notch, and repeat. BE CAREFUL that you do not get too much power to them. THEY DO NOT operate on 12V. The 12V feeds the IVR which outputs roughly 5V equivalent in normal operation.
 
Last edited:
. My big question remains: why does it stop working when the ignition is turned on?
Because you are supplying voltage to the pin that gets grounded by the sender AND the car is providing voltage then the pin that it normally does

Take alight bulb and apply 12v to the base and the center pin.... No light. There is no delta between the voltages.

Watch this video and you will be able to test your gauges like a pro.


 
I'm super busy right now guys. Hopefully I'll get to respond tomorrow. Im not ignoring your posts. I really appreciate all the great information you are giving me that will help to eventually solve this problem. Please hang with me.....thanks!
 
What you have there is a power supply, plain and simple. DO NOT backfeed that into the gauges

You may/ will likely blow up the gauge regulator/ limiter

If you have the cluster out/ disconnected, and if you also have the IVR pulled out, you could wire it to the gauges to see if they move, but you have no values to justify what power level causes what action at the gauge(s)

One thing you could do is to check that the gauges are equal reading as I said before. What you would do, is, with the cluster out and disconnected, and with the IVR removed, hook either power supply lead to one of the gauge sender studs. Connect the other lead to the remaining gauge sender stud. This will feed power in series through both

Select the lowest voltage setting, and let them sit for maybe 10 seconds, see what they read. If they both move, they should read the same. If they are below half scale, step the voltage up one notch, and repeat. BE CAREFUL that you do not get too much power to them. THEY DO NOT operate on 12V. The 12V feeds the IVR which outputs roughly 5V equivalent in n

Life is getting in the way of everything. Nothing bad; just too much happening.

Yes Sir, I know it's a power supply. It's been a Godsend over the years. I just figured it would help me to determine if my gauge is working, or not. I guess I did accomplish that because the needle does move up commensurate with how much voltage I supply. I do like the idea of hooking up the Power Supply to the two gauges in series to see if they move in tandem. Question: the two gauges(gas and temp) share a common segment of the printed circuit board, If I remember correctly. Is that the power side or the ground side? I've been doing my testing, wiggling, and jiggling with the IC installed. I think that ultimately I'll have to remove it because I believe my fault lies with something with the IC. My sender is good, I'm betting my gauge is good, and I've given the wiring up to the Circuit board a real good look(continuity). What's left: the IC! Funny thing, I think that gauge worked before I removed the IC to replace bulbs and clean it up last year sometime.
I've been dabbling and restoring old Mopes for 38 years and this is my first gauge conundrum. Thanks for sharing your knowledge. I appreciate you all.

Edit: I used a test light and found all 4 posts of the two gauges had power. I guess that was a dumb question, then.
 
Last edited:
Because you are supplying voltage to the pin that gets grounded by the sender AND the car is providing voltage then the pin that it normally does

Take alight bulb and apply 12v to the base and the center pin.... No light. There is no delta between the voltages.

Watch this video and you will be able to test your gauges like a pro.



That's a great explanation, and that did occur to me but had I had no idea of the physics involved. Heck, I didn't even know a C-3826 tool was using resistance only, until 67dart273 told me. Can't wait to watch the video so I can be a a pro, too. Haha! Thanks.
 
As mentioned, the Miller C-3826 special tool (referenced in most all FSMs from the time) is a dealer used factory tool and is simply a three-way selectable resistor pack, having the resistance values I added to that cropped picture above prior to posting it some time ago on one of the forums. It does not provide any power. Miller supplied most factory special tools used at dealers then. It can be easily recreated and used to test/calibrate most any factory sweep gauges from that time.
C-3826-2.jpg
C-3826.jpg
IMG_1683.jpeg
 
Last edited:
Is that the power side or the ground side?
you did not watch the video did you? com on tell the truth :poke: :rofl:

all gauges in our cars, the ground is provided but the sending unit. The temp gauge, the oil pressure gauge, the fuel gauge. the common trace comes from the IVR (whether external like on the Darts or internal in the fuel gauge on the Barracudas) it is the 5-6 volt power for the gauge.

The IVR MUST be grounded to work . The IVR takes 12V and converts it to pulsed DC in the 5-6V range (its all in the video :poke: :poke:)

The IVR works like a blinker of sorts, it opens and closes slower or faster to provide a steady pulsed voltage for the load

the gauges have a bimetal strip with a heater coil around it when the senders resistance is low like 10 ohms the heater coil gets more current and heats the heater up and the bimetal bends more and the needle moves farther to the hot , Full end of the gauge. if the sender has high resistance 72 ohms there is less current and the heater is less hot and the bimetal bend less and the needle moves les so it stays to the empty cold end of the gauge.


1752038935468.png
 
Last edited:
you did not watch the video did you? com on tell the truth :poke: :rofl:

all gauges in our cars, the ground is provided but the sending unit. The temp gauge, the oil pressure gauge, the fuel gauge. the common trace comes from the IVR (whether external like on the Darts or internal in the fuel gauge on the Barracudas) it is the 5-6 volt power for the gauge.

The IVR MUST be grounded to work . The IVR takes 12V and converts it to pulsed DC in the 5-6V range (its all in the video :poke: :poke:)

The IVR works like a blinker of sorts, it opens and closes slower or faster to provide a steady pulsed voltage for the load

the gauges have a bimetal strip with a heater coil around it when the senders resistance is low like 10 ohms the heater coil gets more current and heats the heater up and the bimetal bends more and the needle moves farther to the hot , Full end of the gauge. if the sender has high resistance 72 ohms there is less current and the heater is less hot and the bimetal bend less and the needle moves les so it stays to the empty cold end of the gauge.


View attachment 1716428038
You are correct. I have not watched the video yet. I'm really looking forward to watching it so I can be a pro, too. Ha! All kidding aside, I'm anxious to see it. That's a cool set up in the above picture. Thanks to 72 roadrunner for the picture of a 3826.
 
Wiggling and working suggests the pins on the cir board are not making a good connection.

Cur boards are available and OEM ones can be repaired
Not trying to hijack the thread, but I had a recent incident with the water temp sender wiring. While driving the engine suddenly quit and smoke came from under the dash. After I got it home I pulled the instrument cluster and found the purple temp sending wire at the bulkhead was melted. Very close to a fire. Our old girls have a lot of years behind them now and who knows the condition of the wiring... especially where we cannot see!
 
Not trying to hijack the thread, but I had a recent incident with the water temp sender wiring. While driving the engine suddenly quit and smoke came from under the dash. After I got it home I pulled the instrument cluster and found the purple temp sending wire at the bulkhead was melted. Very close to a fire. Our old girls have a lot of years behind them now and who knows the condition of the wiring... especially where we cannot see!
You bet. I'm constantly on alert for direct shorts like that. I recently restored a 1977 Dodge truck and decided to omit the bulkhead connectors. I ran all the wiring through a rather large rubber grommet on the firewall.

Just last week I had copious amounts of nasty smelling smoke rolling out from under the dash of this Dart. I found a giant rodent nest that was completely enveloping the 3-speed blower resistor(inside the evaporator box). I got lucky it didn't ignite. Given the time and effort I've put into this project, it would have been a huge bummer!

I also decided to disconnect the battery before leaving the shop.
 
You bet. I'm constantly on alert for direct shorts like that. I recently restored a 1977 Dodge truck and decided to omit the bulkhead connectors. I ran all the wiring through a rather large rubber grommet on the firewall.

Just last week I had copious amounts of nasty smelling smoke rolling out from under the dash of this Dart. I found a giant rodent nest that was completely enveloping the 3-speed blower resistor(inside the evaporator box). I got lucky it didn't ignite. Given the time and effort I've put into this project, it would have been a huge bummer!

I also decided to disconnect the battery before leaving the shop.
Wow, you too dodged a bullet. Those varmits are good at chewing thru insulation!
 
Okay guys, I watched your video. What a gas( no pun intended). I learned a lot. This being my first rodeo with a gauge problem, I haven't needed to delve into it much.
 
Dana67,
Because you know so much about gauge "tech", may I respectfully call you "Tech", like the little guy on the table who was cavorting with Ed.
 
Some other points to check in the cluster before you stick it back in.

Check the harness connector pins for looseness and clean, flux, and resolder if loose. Also solder a bridge between the PC board trace and the contact fingers for the IVR. Someone was just on about this the other day. Loosen/ tighten the gauge stud nuts several times to scrub the pc board and contact area. Consider replacing the fake nuts with real ones.

Check the ammeter, I would take it apart, and check the studs assembly for damage and tightness

Add a grounding pigtail to a common PC board ground point, then bolt that to the dash frame or column support

I would NOT retain the stock gauge IVR. Order one for somebody such as RTE

Clean the pc board where the dash lamps contact, as well as the contacts on the sockets. Bend them for better contact. Examine the bulbs and replace. Years ago I would have just replaced them, but quality bulbs are getting hard to find.
 
Been here more or less since 2007. Hashed this over a few times. Wild needle swings are a symptom of failing mechanical limiter.
Too much juice from the limiter wouldn't effect the temp gauge with a cold water signal. The fuel gauge is drawing it at that time. That needle climbs real good. When the water is hot the temp gauge draws all that juice.
 
Last edited:
I posted and then went backreading to see if you had been told b4. I found so many typing IVR. That is instrument voltage regulator. The one we would use this application is a solid state component that can take in 9 to 35 VDC and output 4.8 to 5.4 VDC constantly. The symptom you describe comes from yesterdays mechanical voltage limiter. I can't bad mouth yesterdays technology because it did work, very well in fact, for Chrysler, Ford, VW, etc... Nothing lasts forever.
 
Hi Red. First time I learned of one of those was way back in 67 before I joined the Navy, had a '60 falcoon for a short time. The IVR shorted and pegged both gauges. "Intuition" told me that they both cant be bad, felt around back there with one hand and jerked the IVR out. Found it mostly by hope, luck, and feel. Went downtown, bought a new one, and away we went. "Kid's luck" LOL

I think maybe GM is about the only outfit who did not use them
 
Some other points to check in the cluster before you stick it back in.

Check the harness connector pins for looseness and clean, flux, and resolder if loose. Also solder a bridge between the PC board trace and the contact fingers for the IVR. Someone was just on about this the other day. Loosen/ tighten the gauge stud nuts several times to scrub the pc board and contact area. Consider replacing the fake nuts with real ones.

Check the ammeter, I would take it apart, and check the studs assembly for damage and tightness

Add a grounding pigtail to a common PC board ground point, then bolt that to the dash frame or column support

I would NOT retain the stock gauge IVR. Order one for somebody such as RTE

Clean the pc board where the dash lamps contact, as well as the contacts on the sockets. Bend them for better contact. Examine the bulbs and replace. Years ago I would have just replaced them, but quality bulbs are getting hard to find.
Great stuff!
 
Finally "refound" the old thread concerning this. Way way down towards the end is the photo of repairing the IVR connector

Also I forgot to mention that there are/ were folks making repop replacement PC boards for some models. I have NO idea as to service or quality of these


Use ONLY electronics rate flux and solder, rosin core solder, lead based, NOT lead free. And do NOT use plumbing based solders.

One source of rosin flux is here:
 

-
Back
Top Bottom