Homemade gauge test tool, C-3826

-
67Dart273

67Dart273

Well-Known Member
Joined
Oct 14, 2010
Messages
60,035
Reaction score
33,300
Location
Idaho
I finally dug through the junk box, went down to Radio Shack, and built myself a copy of the Mopar C-3826 guage test tool. It's a little crude, but hey, I'm not going to "show" it.

According to what I can find, the three test resistances for senders (in the C-3826 tool) are

L = 73.7 Ohms (empty)
M = 23.0 Ohms (1/2)
H = 10.2 Ohms (full

when the gauge is properly supplied by a working voltage limiter at 5V

The diagram:

Now Rad shack has a fairly limited supply of resistors, and I wanted to be sure and get enough wattage. That's the reason for the odd setup, as well as the switch

With the switch in "center off" for "cold" or "empty" the resistance of 73 ohms is all that is in the circuit. This is made up of a 50 ohm, (2) 10 ohm, and a 2.7 ohm I had in my junk box all in series.

With the switch to one side, the original 73 ohms is now paralleled with the proper resistance to make the combination work. So on one side I added (3) 100 ohm, which in parallel with the original 73, gives me 23.2 ohms for "1/2 tank" or middle of the gauge range

With the switch to the other side, I added (3) 33 ohm resistors in parallel, which once again are also in parallel with the 73 ohm combination, to give me 10 ohms for the "full" or "hot" reading.

The actual measured end results are 72.6, 23.2, and 10 ohms

2qcfg9z.jpg


A shot at the interior. Glued the resistors down with "GOOP." No it ain't pretty. Yes, it works

I think I left enough room to later install a 5V regulator so I can substitute that for a suspected bad instrument voltage limiter

hwetf8.jpg


mcpt02.jpg
 
Not shooting you down or anything here but could you not have done the same thing with a multimeter? (ohm sweep)
 
No, the object here is to test the GAUGE, not the sender.

The way I look at it, gauge troubles stack up like this:

1 Poor connection in the 12-14V IGN supply, wiring, connections at the PC board connector

2 troubles with the limiter, IE bad socket connection, limiter not working properly

3 gauge problems, (sticking, out of cal) or connection problems at the PC board

4 sender wire problems/ connections

5 bad/ out of calibration senders, or sticky/ rusted tank sender

6 bad connection in the ground/ return path

You can check the supply voltage with a voltmeter

You cannot really check old -- style limiters (pulse)

You need a specified resistance to check the gauge(s) (the device I built)

You need an ohmeter to check wiring connections, and resistance of the senders and ground path.
 
Nice work! Can you show us how you use it?
 
Subscribed :coffee2: That is so cool and smart 67Dart273 :happy10:.
 
FISHBREATH said:
Nice work! Can you show us how you use it?
Click to expand...

The procedure is in all the factory shop manuals. All it does is substitute known resistance for the sender. So, say, if you think you have problems with the fuel gauge, you first make sure it has "solid" 12-14V to the instrument regulator, and that the regulator is probably OK.

Then you unhook the sender, and clip this from the sender wire to ground. There are 3 switch positions which give you the "empty", the "1/2" and the "full" mark on the gauge.

If that does not happen, you check the sender wire, and move the box up to the gauge cluster and retest, or check resistance and connections on the sender wire.

If that's OK, you are getting up to either a bad gauge limiter or the gauge itself.

What I've read indicates that all these vintage cars used the same gauge movements, so this box should work with fuel, temp, or optional oil gauges.


I plan, later, to build a 5V regulator into the box so I can also substitute the voltage limiter
 
I was thinking about make one when I was rebuilding the harnesses for my cuda. I ended up using resistors from the Shack instead. great tool and very handy for someone doing mopar gauge work. you can get the 5v regulator from them also. nice job!
 
67Dart273 said:
The procedure is in all the factory shop manuals. All it does is substitute known resistance for the sender. So, say, if you think you have problems with the fuel gauge, you first make sure it has "solid" 12-14V to the instrument regulator, and that the regulator is probably OK.

Then you unhook the sender, and clip this from the sender wire to ground. There are 3 switch positions which give you the "empty", the "1/2" and the "full" mark on the gauge.

If that does not happen, you check the sender wire, and move the box up to the gauge cluster and retest, or check resistance and connections on the sender wire.

If that's OK, you are getting up to either a bad gauge limiter or the gauge itself.

What I've read indicates that all these vintage cars used the same gauge movements, so this box should work with fuel, temp, or optional oil gauges.


I plan, later, to build a 5V regulator into the box so I can also substitute the voltage limiter
Click to expand...

Thanks. It seems so simple. I have wondered in recent years about making some sort of device that would give accurate gallon readings on the gauge. When I worked at the Grumman Low Speed Wind Tunnel, we had to calibrate the balance by taking readings (data points) of measured loads on to discern the historesis in the structure. The forces at the data points became the new "zeros" in the system. Of course, we had analog strain gauges, which fed into an A/D converter, and then into the data acquisition computer.

Is it possible to use such as system to accurize a gas gauge? The system could be calibrated by loading one gallon at a time and taking a data point, which would correspond to "one gallon" on the gauge. Of course, this would require some sort of computer chip that could take the non-linear ohm output of the sending unit and convert it to a linear input for the gauge. I wish I had something like this during the "gas crises" in the early and late 1970s because I would have had a more accurate read on my potential range on the interstate highways when too many gas stations closed early. I should dust off my old electronics lab and text books.

Is this feasible?
 
I think what you propose would be "chrome plating a pig." The sender itself, even original factory ones, are pretty cheaply made, and it defies logic that they lasted as long as they do/ did. Of course we have all read about how bad the repop senders are.

But even if you got a really repeatable sender, the tank itself? What do you do about up/ down hills, acceleration braking, etc?

IF the sender was a quality piece, even as an analog resistor, and non - linear at that, it would be an easy matter to build an analog - to digital converter, (several chips available to do just that) and of course for some cars you can already buy aftermarket clusters. My neighbor rebuilt an original Ferd Broncheroid a few years ago, and installed a really cool all - digital dash cluster

Frankly, in the situation I'm in, being that much more accurate is not something I want to spend time on, but I can certainly see where many would be.

But Ma isn't the only culprit in this mess. YEARS ago, I bought S&W gauges for my old FJ-40 Landcruiser, which then had a 340/ 4 speed. I bought the "matching" S&W fuel gauge and sender. The sender, and the tank had the old - style GM (and others) bolt in flange, so the sender dropped right in. I followed the destructions to the letter for calibration, but the damn thing was never really very accurate

The gauge used a "bridge" circuit, which should have been great, but the needle was not damped at all, so just going down the road, with, say, 1/2 tank, the gauge would flop between 1/4 and full. Considering the money I paid, it was almost useless. I almost ALWAYS carried a Jerry can of gas on the rear.
 
Well that's one way to do it, or you could replace the 3 - 33 ohm resistors with one 10 ohm that happens to be a tad high in resistance. I guess it all depends on what you have lying around the shop.
 
cudajim said:
I guess it all depends on what you have lying around the shop.
Click to expand...

That and the switch. It's center off, which dictated the parallel arrangement, as well as some of the resistors I had.

The other problem with Radio Sh$$ parts is their resistor selection is really pretty poor. It was just about as cheap to find parallel combos as it would have been to find resistors individually, which themselves would have to be combinations due to both selection and wattage.

And I could have ordered 'em form Mouser or somewhere, too.
 
Thanks, Dart, you are probably right about my idea being a chrome plated pig. When I did those interstate driving trips during the fuel shortage days, I kept a mileage log. The all-wise federal government forced the speed limit down to 55mph, but I did 60mph (1 mile per minute) so I could guesstimate my available range and time to the next potential gas station (I-95 was not completed back then and there were many diversions onto US1 and 301 through small towns). Ah, the good old days!

Thanks for the info.
 
I got a replacement gas tank sending unit for my 67 dart and it doesn't make sense when you look at how much it says is in the tank and how many gallons you add to fill it.
 
This is a pretty well known problem with the repop senders which may well all be coming from the same little hump-backed Asian with poor eyesight working by oil lamp in the primitive mountains of Northern China. ---figuratively speaking

They are not the correct resistance range. I have not seen a definitive fix, other than bending the arm to get the MT resistance correct.
 
Old thread guys! But I have a question about the tester.

MOPAR always refers to a gauge tester to check oil press. fuel level and coolant temp.

Would that homemade tester be able to test a coolant temp. gauge???

I have a 73 Dodge truck, temp gauge doesn't even hit 120 degrees when it's warm. (160) Gauge goes from cold to hot if you ground the sender wire. Sender reads approx. 86 ohms. cold and was reading 31 ohms when I had it around 160 but the gauge was not even at 120.
 
nothingbutdarts said:
Old thread guys! But I have a question about the tester.

MOPAR always refers to a gauge tester to check oil press. fuel level and coolant temp.

Would that homemade tester be able to test a coolant temp. gauge???

I have a 73 Dodge truck, temp gauge doesn't even hit 120 degrees when it's warm. (160) Gauge goes from cold to hot if you ground the sender wire. Sender reads approx. 86 ohms. cold and was reading 31 ohms when I had it around 160 but the gauge was not even at 120.
Click to expand...


It should. They recommended the same tester to use for all of the gages. I used to have one but it got stolen when I was moving from Detroit and Miller tools has since obsoleted it.
 
Both fuel, temp (and sometimes optional) oil gauges have the same gauge movement, so same sender resistance. Your 73 should be the same.

In case you missed it, the reason I used those resistances is because I had some, and because of the switch I used.

Probably an easier better way is to simply buy resistors or combinations of resistors to give you the three needed, then solder each resistor or resistor group to a RadShack clip lead which you have hacked in half.

For example, you can buy four - 100 ohm, 1/2 watt resistors, wire all four in parallel, and that will give you a 25 ohm, 2 watt resistance for the 1/2 scale check.

I've not checked what the gauges draw at full or empty, but I would not buy just 1/4 or 1/2 watt single resistors. I'd think at least 1 watt, but you might get by with 1/2 considering the short time needed.
 
67Dart273 said:
I finally dug through the junk box, went down to Radio Shack, and built myself a copy of the Mopar C-3826 guage test tool. It's a little crude, but hey, I'm not going to "show" it.

According to what I can find, the three test resistances for senders (in the C-3826 tool) are

L = 73.7 Ohms (empty)
M = 23.0 Ohms (1/2)
H = 10.2 Ohms (full

when the gauge is properly supplied by a working voltage limiter at 5V

The diagram:

Now Rad shack has a fairly limited supply of resistors, and I wanted to be sure and get enough wattage. That's the reason for the odd setup, as well as the switch

With the switch in "center off" for "cold" or "empty" the resistance of 73 ohms is all that is in the circuit. This is made up of a 50 ohm, (2) 10 ohm, and a 2.7 ohm I had in my junk box all in series.

With the switch to one side, the original 73 ohms is now paralleled with the proper resistance to make the combination work. So on one side I added (3) 100 ohm, which in parallel with the original 73, gives me 23.2 ohms for "1/2 tank" or middle of the gauge range

With the switch to the other side, I added (3) 33 ohm resistors in parallel, which once again are also in parallel with the 73 ohm combination, to give me 10 ohms for the "full" or "hot" reading.

The actual measured end results are 72.6, 23.2, and 10 ohms

2qcfg9z.jpg


A shot at the interior. Glued the resistors down with "GOOP." No it ain't pretty. Yes, it works

I think I left enough room to later install a 5V regulator so I can substitute that for a suspected bad instrument voltage limiter

hwetf8.jpg


mcpt02.jpg
Click to expand...

Hey, just coming across this thread. The photos are no longer available. Would you happen to have new links or Be able to provide new photos? I’m interested in getting/building one of these for when I fix my wiring
 
Actually I would NOT necessarily follow this "plan." These were made from as I said "the junk box" and they involve parallelling odd values to make up the total. Much easier to just buy the correct individual resistors. Here is a photo of a test tool that someone has edited in the correct values in yellow

c-3826-jpg-jpg.jpg


Those three, for empty, 1/2, and the full mark are all you need. I would get at least 1 watt or larger resistors, as there is some current going through there
 
67Dart273 said:
Actually I would NOT necessarily follow this "plan." These were made from as I said "the junk box" and they involve parallelling odd values to make up the total. Much easier to just buy the correct individual resistors. Here is a photo of a test tool that someone has edited in the correct values in yellow

View attachment 1715537911

Those three, for empty, 1/2, and the full mark are all you need. I would get at least 1 watt or larger resistors, as there is some current going through there
Click to expand...

Great! Thanks. I need to figure out what’s up with my temp gauge and gas gauge (temp has no reading even though it seems to be hooked up and gas moves linearly, but can go from full to 3/4 in a few miles and then functions as expected below half) and this should help.
 
There are many threads on these issues. Some of them are

1....broken/ loose harness connector pins at the PC board. Clean them and solder them to the board traces

2...Ditto the contact fingers for the IVR (instrument voltage limiter) Solder jumpers from the contact fingers to the board traces

3...Poor connections at the gauges studs. Loosen/ tighten the nuts several times to "scrub" the connectors clean, and consider replacing the nuts

4...Flakey / bad IVR replace with a commercial solid state replacement or google search one of the many DIY circuits

5...Add a grounding pigtail to a common point on the cluster and bolt it to the dash frame/ column support

6...Don't discount poor connections in the power circuit feeding the cluster, or to bad connections leading to the senders. This includes the left kick panel connector (tail harness gas tank sender) and the bulkhead connector (temp sender and oil if used) as well as poor connections at the senders that is the wire end itself.

And then of course the gauge units themselves can be bad
 
You must log in or register to reply here.
-

Back
Top