Return To Main Menu


 Speedometer Gear

Speedo Check

Gauge Check




Jeff Gent wrote: Are there different speedometer gears available out there. I assume there must be to go with the different r-p options but don't remember reading about them. Maybe I wasn't looking close enough. I noticed last night that they are a pressed on unit so shouldn't be too hard to change. Do I need to change both ring and pinion or just the pinion (easiest for an assembled unit).

Also, I may find the answer when I put it back together soon, but the 'ring' speedo gear is smooth inside so it's not held onto the output by splines like I'd expect. Is it held by being clamped between two pieces or does it use some sort of oil-film viscous friction to keep it going?

Keats wrote: Years ago on a 74 CJ5 I had problems first with a slipping speedo gear on the output shaft of a Dana 20. I replaced it twice before I found out that a metal gear was available instead of the red plastic number I originally removed. Get the metal one if you don't have it already. The gear on the shaft in the cable housing dictates the ratio you end up with and it is only approx. as gears and wheel size make anything exact near impossible, but you should be able to get close.

Return to Menu


Mark C. Johnson wrote: Hello All, It has been bothering me the past few months that my speedo does not work. So finally tonight I got around to pulling it out and apart. Now that it is apart how can I determine if it is good, bad or can be fixed. Here is the problem, the speedo when hooked up makes a lot of noise and does no move in the least. Any ideas, comments.

Rick Stivers wrote: Mark, If you have the speedo out, you can check the cable by driving the jeep and watching to ensure the cable is turning smoothly inside the housing. If it is jerking or not turning at all the cable is probably bad (Could also be the speedo gear at the transfer case). If the cable is turning smoothly check the end of the cable to ensure that the end of the cable isn't stripped out. Most of speedo cables are square on the end of the cable and if the corners strip off they won't turn the speedo anymore. Finally check the insert for the speedo to see if it is stripped. You can usually take a small screw driver and turn the speedo to see if it will work. If you turn the speedo and it doesn't move the indicator will probably need to be rebuilt or replaced.

Return to Menu

Gauge Check

Josiah wrote: Hi, all, I have a gauge question. In my '55 CJ5, I have the speedo shown at the bottom of this link:

I don't know if it is original or not. I am getting ready to wire it up on 12 volts, but I don't know if the gauges are set up for 12 volts. Is there an easy way to tell? I have access to every kind of electrical test equipment, including variable power supplies, ohmmeters, oscilloscopes, spectrum analyzers, etc. If they are 6 volts, I will just build a regulator into the speedo housing. The jeep was 12 volts when I got it, but it did not have the original dash, the fuel sending unit was broken, and the temperature sending unit was missing, and it has a Buick 225. I got new sending units and an adapter for the Buick block, and I want the gauges to work.

Dr. Vern Wrote: Dear Josiah, I have been away from my computer most of the week, so your question has been festering in my mind. Like a big old juicy zit, here is what I've come up with. Pour yourself some coffee, a soda, hemlock or drink of your choice as this might be a long one.

I also realize this is probably a moot point as you have found another gauge, but dang it all, I've been thinking about this all week so you will have to listen. <g Not all of this information applies directly to you, but it may be of some interest to anybody else bored enough to read it all.

No skipping ahead, but to make a long story short, I don't believe there is a way to tell whether a gauge was designed for 6 or 12 volt without rigging up a simple test circuit. Checking the internal resistance with an ohmmeter will not be sufficient. If there was a part number stamped on the back of the gauge, you could cross-reference that. Some background on Gaugeology 101 would be helpful, so let's start at the beginning. First the earth was dark and formless, and then God spoke. Fast-forward several millennia, past the discovery of electricity, the internal combustion engine and aerosol cheese. Sorry, I'm digressing again...

There are two main types of gauge systems in use over the years. The way each system works internally is quite different but each displays the information in the same manner, as needle movement. (A third type of system is a digital display, but that is another bedtime story)

The first type, like my CJ2A's fuel gauge, uses a balanced coil design to move the needle. There are two opposing coils inside the gauge that move the needle one way or the other. If the voltage increases, it increases current flow and magnetic strength on both coils so this system is not especially voltage sensitive. Operating above design voltage is not the best for any gauge, but many folks have done this on a balanced coil gauge with no ill effects. This illustration from the CJ3B page shows some details about the balanced coil design:

BTW, the picture at the bottom of that page is inaccurate. The gauge body must be grounded to work properly.

A balanced coil gauge is very sensitive and prone to being jittery. Some types have a damper to smooth these oscillations.

An easy way to tell if you have a balanced coil gauge is to observe the needle as you shut off power. If it drops instantly, it is a balanced coil design. Keep in mind if you have a damper, the needle will drop slowly and mimic the second type of gauge, which I'll get to in a moment. But if the needle does drop quickly (or is jittery when driving) it is most likely a balanced coil design without a damper.

Now on to the second major type of gauge, the correct name of which escapes me at the moment. I'll call it a thermal type, but a subclass is called a bimetallic gauge.

I'm pretty sure from the picture that your CJ5 gauge is the thermal type. In fact I know it is because you mentioned you disassembled it and found the nichrome element. Although not an ironclad rule, separate round gauges (like the CJ2A fuel gauge) are typically balanced coil design. Rectangular gauges that are part of a cluster are typically thermal.

Both the thermal type, with a nichrome wire, and a bimetallic gauge operate on current flow through a resistive element causing heat. The great the current flow, the greater the heat. In the case of the nichrome element, the length changes depending on the heat from the current flow, and moves the needle. A bimetallic element is very similar, but has two layers of dissimilar metal joined in a strip. One layer has a greater expansion rate than the other, so the strip twists as it heats up, moving the needle. Unless attempting an internal repair, you would not need to know whether a thermal type gauge has a nichrome or bimetallic element. They each work is a similar fashion with current flow heating the element and moving the needle. The thermal element provides both a mechanical damping action and averages out any oscillations from the sending unit, preventing a jittery needle.

But this is where it gets very interesting, at least to me. A thermal gauge is very sensitive to operating voltage. If you increased the operating voltage, the current flow would increase and result in a false reading. That is why thermal gauges generally have an Instrument Voltage Regulator (IVR) mounted on the back. Depending on the vintage, the IVR can be either mechanical or electronic and provides a generally constant output voltage for the gauges. I think 5 volts output is a common industry standard, but there may be exceptions. The IVR is an inexpensive device, only a few dollars last time I checked. There may be some other vintage 6 volt vehicles with thermal gauges and no IVR, but I think all postwar Willys trucks and wagons had the IVR. (Remember the early CJ series had balanced coil gauges and an IVR was not needed.)

Now it is possible the IVR got tossed out years ago, or has shorted directly across so it acts as a solid conductor all the time. You might never notice the difference running thermal gauges designed for 5 volts on a 6 volt electrical system. But if that same 5 volt gauge were operated on 12 volts without an IVR, watch out. That may be what happened to your old gauge. During the gauge's very short 12 volt life, it would have read erroneously before doing an imitation of a cigar lighter.

Well, several kilobytes later, you are still staring at your gauge wondering whether it is for 6 or 12 volts.

From what I understand, all factory thermal gauges are designed to run on 5 volts. (My '84 F250 has thermal gauges running at 5 volts, for example) They could probably run on 6 volts, but for a 12 volt conversion it would require an IVR. I don't know if an IVR designed for 12 volts is any different than a 6 volt unit. The IVR's life is shorter on 12 volts but this may not be a major problem. When people say they did a 12 volt conversion (keeping the stock thermal gauges) with no problems, it is because of the IVR keeping the gauges at 5 volts. All bets are off for aftermarket gauges, as it would be easy enough to design a gauge for 12 volts and save the cost of an IVR. But the input voltage will fluctuate considerably whether on a 6 or 12 volt system, so an IVR helps keep the gauges accurate.

You could test my theory with a variable DC voltage supply. It sounds like you have some fantastic test equipment at your disposal, but I have used a model train power pack with great results. (Just watch out for cheap ones that leak AC voltage, or fancy ones with deliberate DC spiking for better slow speed operation)

Here is how to hook up a thermal gauge for testing. When you have zero resistance at the sender, it will provide maximum current at the gauge. Simply supply power across both terminals on the gauge. A thermal gauge should not be polarity sensitive. Start with zero volts and slowly increase the voltage. The needle should rise from its resting position. Increase the voltage slowly to allow the needle movement to catch up. Remember this is a simple test circuit that eliminates any added resistance from the sender, so the needle will travel full range when you reach the design voltage. When the needle is at full travel, measure the voltage under load and that is your magic answer. If it is 5 volts, then an IVR is required. If it is 6 volts, then the gauge came from a 6 volt vehicle that didn't use an IVR. (I don't think you will get a 6 volt reading, but I may be wrong when I thought all postwar Willys vehicles using thermal gauges also had an IVR) If you get a 12 volt reading, my IVR theory was way out in left field. However, I would be willing to bet large sums of your money that 5 volts is the magic number.

Keep in mind that you could not accurately figure the operating voltage by reading the internal resistance with an ohmmeter. The resistance of the thermal type changes so drastically with current that an ohms reading at rest is of little value. You could show continuity but little else. Even if you used an ubermeter capable of reading nano-ohms, the results would be meaningless unless you had a pair of different gauges (of known voltage) for comparison. But running DC voltage through a thermal gauge is quick and easy and duplicates real world conditions.

As far as how to determine operating voltage for a balanced coil gauge (which you don't have), I am stumped. I believe Willys stopped using that style when they went to the cluster instead of individual gauges in the CJ series, which predated the factory conversion to 12 volts. So I'm assuming (Danger! Danger!) that any balanced coil gauge used on a Willys was 6 volts.

Let's see, then it gets really interesting because 6 volts was still offered as an option after 1957. Those vehicles had thermal gauges and must have certainly had IVR's.

Wow, I really need to get a life. I'll go look in the shop and see if I can find one...

This page last updated 13 February, 2001

[Hit Counter]