Electric gauges are inaccurate, but do they have to be?

[Jeff73Mach1]

I can't count the number of times I've heard of, read of or have personally seen gauges that were no better than idiot lights at telling drivers what was going on with their vehicles.

If someone says "My car runs hot." It has gotten to the point that my first response is "How did you determine that? Is it puking coolant? Have you used an infrared thermometer?"

So here is my thought/question/point

Why aren't automotive electrical gauges and their sending units capable of a much higher degree of accuracy than we get?

What brands are the most accurate, and is their accuracy high enough to consider reliable?

Is there any electrical gauge sending unit combo that can match a mechanical gauge/sending unit combo for accuracy?

If there is nothing in the industry that it highly accurate, why aren't we looking outside the industry for solutions?

I'm spending a good chunk of change to build a high performance engine-I'd like to be able to accurately monitor it, but I do not want oil lines running to the interior of my car. What are my best options?

 

[CZ-75]

Sometimes a gauge that looks like it can read a broad range of tempurate or pressure may be hooked up to a sending unit that is designed for an idiot light. In most of these instances, the gauge will read in the middle instantly.

 

[bristang]

I was at a Ford school in hte early 90's when the instructor showed us the wiring diagram for the "new" gauge configuration coming that model year. It was an oil pressure gauge and the gauge was wired suspiciously like an idiot light. Seemed that Ford was tired of the complaints from owners regarding fluctuating oil pressure gauges at idle, so the "new" gauges would read smack dab in the middle of the range anytime oil pressure was over (i think) 7 psi.

I have an Auto meter electric oil pressure gauge in my Mustang. I have a very hard time pumping hot, pressurized oil into a plastic tube inside the passenger compartment. While I have not had it happen personally, I have been off the beaten path when a buddy's plastic tube broke under the dash of his Land Cruiser. Luckily he was not alone and we were able to get him back to the shop.

Mechanical gauges may be more "responsive", but I will never fully trust hot vital fluids under pressure in the cabin.

 

[CZ-75]

Yep... made the gages work just like an idiot light... I saw it in my 2002 Dodge Pickup truck too

 

[MechEng]

What this really comes down to is the fact that electrical gauges are a “relative” reading, meaning they fluctuate and move with the supply voltage that is being provided to them. This is why if you look at our cars, there is a voltage regulator on the back of the instrument cluster that provides the reference voltage for all the gauges.

The accuracy and consistency of this voltage is what is going to determine how much your voltage and therefore readings, are going to wander.

Another “problem” with electrical gauges is that many of them work off the principal of electrical resistance (water temperature sending units specifically) which means that any change in resistance along the wires leading to and from the gauge, are going to play into the reading. This would stay fairly constant in a more mundane environment, but a car is not so mundane. As we just stated, resistance is a function of temperature, and as your car gets hot and cold, the wires leading between all these points, experience changes in their internal resistance. Now these changes are fairly small, fractions of an ohm, but in a system operating at 5V in what is supposed to be a 0 resistance set up, these small changes can at times be noticeable.

If you want really accurate gauges, there is a way to do it, and it’s something that I’ve been toying with. It would be to switch from Analog to Digital. Basically change out all your analog sending units with digital ones, they aren’t susceptible to voltage fluctuations or temperature fluctuations, and they’re much more accurate.

You then have to make a decision though, you can either than take that digital signal and convert it to an analog level as close to the gauge as possible, or switch to gauges that can accept that raw digital signal. For the sake of keeping a stock look, let’s say you go with option 1. This would mean having a microcontroller located in the dash that all the gauges would talk to, the onboard D/A converter would take care of turning the digital signals into analog levels. These analog levels, because they’re being produced “absolutely” (the input voltage into the microcontroller has no effect on the output levels of a D/A) they are more accurate than pure analog signals, and because the cable run is much shorter, there’s less wire to be affected by temperature and small resistance changes. Another added benefit is that the microcontroller is incredibly high impedance on these outputs so you’ll get very clean quality signals.

So this would give you highly accurate gauges, without the need to have mechanical ones. I know this would work because it’s how I did a number of laboratory set-ups at one of my previous jobs. The only difficulty here would be finding digital sending units that would fit our cars, and you’d need to add wires for each unit.

An interesting question though. Thanks for asking!

 

[Jeff73Mach1]

Damn! I actually understood what you were saying! I personally would not mind changing out the gauges to something that accepted a digital input. So to add a layer of confusion, are digital sending units available for the automotive industry at this time and if not what industries use them?

 

[mdan575]

I don't know what you are expecting. When talking about the stock gauges, what do you really expect from 40 year old technology. Couple that with a low amp alternator and yes you are going to have imperfections. New gauges really aren't that much of an issue. Give them a good ground and constant voltage and they are quite accurate in my experience. I use them all the time at work and on my vehicles. Fluctuate yes they do, funny thing about that. Coolant temperature is constantly going up and down as the thermostat opens and closes. Most only measuring from one point on the engine so they will fluctuate. Alternator is completely dependent on the engine RPM. A/F mixture is constantly changing also. Oil pressure same thing.

Stock gauges suck bottom line. But I really haven't seen an issue with after market gauges being noticeably inaccurate. Now if you buy cheap gauges expect issue. Glow Shift and Autometer are not in the same realm of quality. Personally my plan is to switch to Dakota Digital gauges in the future. From there I can change to more modern sending units and better technology. Upgrade my alternator to a 3G and life is good.

Personally I agree with the digital gauges they work quite well when programmed with the right formulas. The best part is the formulas can be adjusted for accuracy.

Also remember this isn't brain surgery or flying. If your worried about one or 2 degrees or psi then maybe you need to rethink the setup and fix the underlying issues.

 

[MechEng]

Off the shelf wise, something that would straight drop in I'm not finding much. There are some units from Dakota Digital but I don't know that they would work with anything other than their gauges.

There are units out there that would work for our application but some light fabrication would be required.

Units like this are used pretty heavily in industrial applications so they're fairly common.

 

[Jeff73Mach1]

I expect that stock gauges are not going to be accurate. I expect that a temperature gauge should fluctuate to show a thermostat opening and closing.

I am confident that some aftermarket gauges are better than others, and that I don't know who makes a good product and who just makes an expensive one. They all make lots of impressive sounding claims, but without understanding the science, I feel ill prepared to critically examine those claims.

I know other industries use more expensive and more accurate testing equipment that could be an upgrade over that commonly used in the automotive industry. I just don't know enough to start searching and finding reliable source material to learn from.

I do not need a gauge to be accurate to 1 degree, but I am interested in learning about which gauges work best and why, so that in making changes and upgrades I am informed and knowledgeable.

I know there are some very smart people here with backgrounds, education and experiences that I do not have. I am trying to tap into that for my own education.

When I finish my build I expect everything to operate properly. That has been my general experience on most past projects. Still monitoring conditions accurately helps to maintain peace of mind.

for an example, earlier responses to this question lead me to find this

http://auberins.com/index.php?main_page=product_info&cPath=5&products_id=134

and while their available gauges are horribly ugly, it is a starting point in my search to learn more.

 

[mdan575]

Glow Shift works but is cheap and ricey. Also breaks easily. Same goes for many of the off brands sold on Ebay.

I have had good luck with Autometer analog and digital gauges. Long life with minimal issues as long as they have good ground and power source.

I am currently using an Aeroforce Analogic Gauge in my Kenne Bell powered truck. It is able to monitor multiple sending units at the same time with 2 displayed. Amongst other features. But it also doesn't look right in classic cars unless the dash gauges are converted to digital.

http://aeroforcetech.com/products_analogic.html

That is a nice sending unit you posted problem is many of the companies make the gauges specific to a voltage reference. So they are not all compatible. One thing I like about the new units like Aeroforce that are digital programmable.

 

[Totalled]

I'm putting electric autometers in mine. I studied Austin Vert's thread and am planning something similar, without cutting up a stock gauge setup.

 

[midlife]

MechEng's description is dead on.

I'll give another example. When battery voltage finally reaches the ignition switch and is tapped for a voltmeter, guess what it reads when the car is running and the battery reads 14.5V? Yup, 11.5V at the voltmeter. Why? Resistance along the wires plus resistance at the connectors due to corrosion and poor contacts. Very normal, as we checked two different cars.