Hi everyone, I bought the subject car almost two years ago and love it. The other day, I went on a short trip of about 200 miles. Normally, the car is driven around town. Along the way, I started to pay attention to the fuel gauge and realized it is very non-linear. It is not really a problem since it is very consistent, but it puzzles me why they designed it the way they did. In the case of my car, if the markings on the guage were evenly spaced instead of the way they are, it would be a lot better. (Here's a link I found using Google to a picture of the instruments
formatting link
Notice the 3/4 and 1/4 marks are closer to the full and empty marks respectively than they are to the center mark. In my car, however, it would make more sense if the opposite were true. I can drive about 40 miles (at 70 mph) before the guage gets down to the F line. It only takes another 40 miles to reach 3/4. However, the next quarter takes
95 more miles. The third quarter takes about 80 miles. The final quarter goes by so fast, you can actually see the needle moving. it only takes about 25 miles to get from 1/4 to just above the E mark. (The light comes on about 8 miles before that). The needle then stays just above E until I get gas. If I get gas when the needle first gets to that point, it takes about 9.8 gallons to fill it up, so there is almost 3 gallons left according to the manual.
Like I said, I am not really complaining, but I am curious as to why they would have placed the markings the way they did. Here are some possibilities:
The marking placement is left over from an older fuel tank that was shaped differently.
The engineers told the artists to place the 3/4 and 1/4 marks closer to the center, but they misunderstood and did the opposite.
The marks were placed to compensate for an oddly shaped fuel tank, but later, they also compensated electronically (via the computer?).
My car is the only one that does this and is therefore malfunctioning.
IMO, Miata fuel gauges have always been pretty bogus (though not as bogus as oil-pressure "gauges"). When my NB fuel gauge is on "E" I've got about 1/4 of a tank left. Quite a ways below "E" the needle pegs. At that point I've got about 1/6 of a tank left.
The speedometer is crap also: with the stock wheels/tires it's off by almost 10% at freeway speeds. The odometer is only off by about 4% which (judging my my sample of two cars) is quite the feat of precision instrumentation by Miata standards. ;)
Embarassingly, my parent's GM has a speedometer that is accurate to a fraction of a MPH at freeway speeds and an odometer that's dead-on.
I guess the way the fuel-gauge is designed is due to the fact of the irregular shapes of the tank. It is probably quite square in the middle with odd shapes on the top and bottom. The fuel gauge sender still is a floating device connected to a variable resistor. The gauge lets you see the resistance. With the odd shapes on top and bottom the fuel-level drops relatively faster creating the fuel-economy-curve you describe...
I'm not sure if the computer does compensate for this kind of behaviour...
Greetings,
Jeroen Technical Commissioner Mazda MX-5 Club of the Netherlands
Complaining about the accuracy of fuel gauges is older than I am, and I'm an "old geezer". Complaining about the accuracy of Mazdas and Fords Speedometers is also nothing new. Perhaps the newest thing is that GM did improve the accuracy of it's speedometers on at least some of it's makes and models.
The govt standards for Speedos used to list two different accuracy specs. Plus & minus so much, and plus with no minus. Ford & Mazda chooses to use the plus no minus standard. On the Miata, if you wish to disassemble things a bit, you can set the speedo needle to give a more accurate reading. (At least on the NA &NB's.)
The speedo on our NC is spot-on at 30 according to the speed indicator device mounted on a lamp post in our road. Haven't tested it any higher speed because the neighbours might get upset (and the car might end up in the village duck pond if I don't make the bend (curve))
Really?! What do you mean 'set the needle'? Is it as simple as removing it, turning the needle a certain # of 'degrees' and sticking it back on? In other words, it's not that it's generally out of calibration, it's just set to show you going slower than true speed all of the time?
Well now, that sounds like a fun challenge; see how high you can get the readout and still make it around the corner ;-) All you need is a closed circuit and a professional driver and you have all the makings of a performance car commercial (including the fine print disclaimer)!
These signs are quite common here in the UK. Some studies have indicated that they are more effective at reducing speeds than Gatso speed cameras.
Our local authority have one that they move around the area, attaching the device to a lamp post in various areas. As you approach it, if you are exceeding the speed limit, it lights up and tells you what speed you're doing. There is a fascination amongst the local boy racers when it appears on a straight bit of road to see how fast they can go through it, but (un)fortunately the LED display only has 2 digits :-(
Basically, you have two instruments that measure speed. The electric speedometer, and the tachometer(indirectly) We used a GPS that also gave a speed readout as a reference. You determine what the error is in the speedometer with the GPS or other means, and also note the tach reading and gear. Take the instrument hood and the faceplate(gauge cover) off. The best speed to use for cal& set is one that matches a normal highway speed and, on most models, is at the top of the speedometer, or about 1/2 the total range of the speedometer. You may see a white bar that was hidden by the faceplate and bezel. This white bar represents the range allowed when the factory put the speedometer on some sort of a test set.
Anyway the speedometer needle is removed gently with something like a fork. If you have a small needlenose pliers, it can be used to prevent the shaft from moving while you remove the needle. Using the GPS or another instrument such as a ScanGaugeII, you set the car to the desired speed. Cruise control helps. The Speedometer needle is gently pushed on the shaft so that it is the same distance from the speedometer face as it was prior to removal, and reads the correct speed. It's a good idea to set it to read just a bit high, since it's difficult to get it to be exactly correct. For safety you should have a knowledgeable passenger, and little or no traffic on the road. The OBDII digital outputs can be your friend, in that there is a speed and RPM output that can be read with the aforementioned scangauge II or similar OBDII devices. If the tires size is incorrect, the OBDII digital output will not be right. This forces the use of a GPS or highway calibration posts (if present), and a stopwatch. Requirements More or less straight and flat road that allows you to maintain a steady speed.
Several years ago, we wrote the general proceedure up, and posted it on miata.net. I don't know if it ever was placed in the garage section. As I remember, the NAs & NBs were allowed by law to be off (high) by as much as
5% (of reading?) The "white bar" at 60 or 70 represents this 5%
The total responsibility/liability falls upon the driver/user. We disclaim any responsibility and or liability.
At first we thought the odometers were also set to read slightly high. however after several 500 mile runs, they were as accurate as anything else we used to measure the distance.
If you find that the Speedometer is not linear, in that the readings are way off in portions of the range, the meter movement is likely bad. Even with an exactly correct reading at 1/2 scale, the end of range readings (high and low) may still be off, but should be within 5% of the total range of the meter, and hopefully within 5% of the reading. There is even a way to correct this, but it can also easily turn the meter into scrap. The process uses a test set that produces 5 different readings, and involves cleaning the bearings lubricating them with a special lubricant. (Jewlers used to have the lubricant) and carefully setting the springs. It's been more that thirty years since I occasionally did this adjustment proceedure with lab "reference standard" meters. Even then we only went to the trouble when there was no reasonable replacement. When the last jug of real whale oil ran out, that would be the end of the effort. The synthetics were not as good, and dried out or gummed up.
As an example, after going through the adjustment process, my 99's speedometer at 70MPH, reads 1/2 way between 70 and the next small mark. (About 1 mile high) This is as close as I could set while driving at the same time.
Some have used another method that I do not recommend-- It's safer in that the car is stationary. You have established how far off the speedometer is at a convenient speed. With the instrument hood and clear faceplate removed, you gently push the speedometer needle to the desired speed. Clamp the shaft behind the speedometer needle hub with needle nose pliers. You can then gently push the needle in the desired direction.
A similar process can be used to set the Tachometer to match the OBDII digital RPM output with the ScanGaugeII. The 99s tachs usually read one to two hundred RPM high, somewhat masking early activation of the RPM limiter.
The NA & NB speedometer factory set was generally for a reading higher than the actual speed. My 99 and otheres read about 3 to 4 MPH high at 70MPH measured. The difference was between the OBDII digital output and the meter reading. The digital output agreed with GPS readings using original OEM Pilot tires. There can be about 1 MPH reading/speed difference due to variations in exact tire size with different brands of say 195/50/15 tires.
The ones here are on trailers. They are uncalibrated and thus notoriously inaccurate, but since their displays are not used as evidence, it doesn't matter. They set off radar detectors up to a mile away, and certainly make a lot of folks slow down.
Wow, that's more than I ever hoped to know about Miata speedos ;-) but hey, I ASKED! Thanks Chuck, I do appreciate you taking the time to tell me in such detail.
Be aware that in most larger cities there are speedometer shops that are often used by local police and highway patrol to set their speedometers and almost all will do the same for you, the price isn't too awfully high if I recall....
The least expensive test is to have them run the car on the rollers and give you a printout of error showing what your speedometer reads at an actual measured speed. If it's too far off, they can then do a calibration for you. Since they have to remove the speedometer it costs more but you can remove the thing and just bring in the speedometer head to be calibrated, once they ran it on the rollers they'll know exactly what the error is.
I can say that not _every_ Miata speedo is off. My NA's was near perfect when checked with mile markers and a stopwatch or a GPS. Ditto with my NC. I just re-checked it today on a drive down from Pennsylvania and it was dead on -- or as close to dead on as I can get reading a needle over relatively coarse markings. Certainly not more than 1mph off at 70.
The only complication is that the Miata uses an electric speedometer. There is some sort of conversion circuitry built in the instrument cluster circuitboard (as I remember) or possibly the ECU that converts the speed sensor pulse train into an analog signal for the speedometer. The speedometer needs to be set with the electronics in operation, not as a separate unit. (After all, the factory tested it at some point, decided that it was within an acceptable accuracy, and installed it.)
In theory, you might build a test circuit that develops the same output voltage (and current) that the speedometer expects. Since using conventional "radio shack" parts would give you a non temperature compensated circuit, the repeatable accuracy might be less than desired. So you would end up using a DVM to measure the Miata's analog output with the speedometer connected, and with the Miata at a "known" speed. If you bothered to do this at about five points spread around the range of the speedometer, and then used the DVM readings to measure the test circuit output, you should be able to accurately calibrate the speedometer using the test set.
Since my speedometer, as well as several others we tested read consistently high, we elected to use the on car at speed method. This provided enough accuracy to get the speedometers to read between the actual speed and about one or two miles high at all reasonable speeds. The highest errors noted were based upon a drag strip's instrumentation and maximum speeds reached during a 1/4 mile run. 112-114 Mph (Turbo'd, 8-9 Psi) or about 96 for a stock 1.8 Miata) We do have a few side roads that have virtually no traffic and enough visibility to safely reach three digit speeds, but a suprise appearance by the sheriff is always possible.
Some speedometer shops have the capability to mechanically drive the speed sensor, much as you would a mechanical speedometer head, and can use the car's electronics. They have books showing the ratio of sensor RPM to road speed. To me this would be the most preferred (and safe) method. Unfortunately there is no local area shop that has this capability. Local new car dealers use equipment peculiar to the brands they carry. There is no local Mazda dealer (went down the tubes years ago) and the closest Mazda dealer (thirty miles away) is basically a parts changer.
MotorsForum website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.