Analog vs. Digital: The Ongoing Debate
The debate between analog systems vs. digital systems is an age-old one, perhaps originating since digital systems were first developed or conceived. The debate usually centers around accuracy, reproduction, control, and other parameters. This article will address those, but from the standpoint of automotive instrumentation.

Analog and Digital Explained
As a start, let's differentiate between analog and digital. In a general sense, an analog system is one that may have virtually infinite levels within a given range, whereas a digital system has a finite number of distinct levels within that range. Most any information we capture in the real-world is analog. For example, the temperature in a room could be 76 degrees Fahrenheit, or 76.1, or 76.106472601193..., etc. It can actually be that value. But is it meaningful to represent it to that precision? Well, if someone asks you what the temperature is, and you read the mercury thermometer on the wall, you will probably round off the answer to "76 degrees". In other situations, someone else may round off to the nearest 5 degrees, and in industrial situations, the rounding may be to a fraction of a degree.

What do we do in "real life"?
In the situations above, even though the "natural" data being measured was inherently analog, it was subconsciously converted to a digital representation. "76 degrees" tends to imply that the next closest values would be 75 and 77 degrees. We do this with other data as well, such as rounding off time to the nearest minute or second, or speed to the nearest mile-per-hour. These are all digital representations of some naturally-analog parameter. The implied (or stated) separation between the possible levels is called the "resolution".

Instruments
Switching our focus to a vehicle now, there are quite a number of different pieces of information or data we need to monitor. As an example, using temperature again, we want to know if the engine is operating outside of a temperature certain range, and even within a "normal" operating range, we would like to know how it changes with respect to some other influence. So not only do we need to know if the temperature rises above 230 degrees, but we'd like to identify a change from say 195 to 198 degrees. The ability of an instrument to detect a small change is called its "sensitivity". Once it detects it, we expect it to reproduce the measured data in a way that we can identify the change. This again, is the resolution of the display/instrument.
At this point, there is another important definition that needs to be stated, and that is "accuracy". The accuracy of an instrument is its ability to repeatedly reproduce the measured data on it's display. For example a gauge that reads a temperature changing from 95 to 98 degrees is considered accurate if it shows this on a display from 95 to 98 degrees, but not as accurate if it showed this as 93 to 96 degrees. In both cases, the gauge is equally sensitive, but the accuracy differs.

The Life of a Piece of Data

Before we can fully identify whether analog or digital is better, we need to consider what happens to the data during it's lifespan -- ie: from the point that it is captured (input) to the point it is displayed (output). For an automotive instrument, we can summarize the stages as Input/Capture, Transmit, Modify, Transmit, and Output/Display. The explanations and issues associated with each of these is explained here:
  • Input/Capture: This is the stage where the real-world parameter is measured by some electronic and/or mechanical device and converts it to another form such as a varying voltage. The typical problems at this stage is sensitivity and linearity (being able to produce a voltage that is a constant proportion of the parameter being measured).

     

  • Transmit: The captured signal is sent to the instrument, typically along wires. Typically, this stage adds errors due to voltage drop in the wires, and EM (electro-magnetic) noise induced from other sources. A vehicle is a harsh environment from an EM perspective.

     

  • Modify: At this stage, the data is modified, typically to remove noise, amplify it to match the required range of a display, convert to another signal type if required by the display, or to compensate for some non-linearity. Modification adds its own non-linearity errors, which may or may not be compensated for in various grades of instruments.

     

  • Transmit: Again, the signal is transmitted, but this time to the display. Even if the transmission distance is much shorter here, there are significant sources of EM noise which may need to be accounted for.

     

  • Display: At this final stage, the modified data is displayed in a form legible to a viewer. For an analog meter movement, this is typically a bi-metallic coiled strip of metal that expands rotationally as it heats up due to a voltage being applied. With one end of this strip secured, the other end moves and has a pointer attached to it, which is the moving needle. There are many sources of errors here, starting with the non-linearity of the metal strip/coil. The non-linearities of both the sensor and the meter movement unfortunately do not cancel each other out nicely, but rather tend to generally compound each other to produce a greater error. There are also other mechanical limitations to be considered such as the weight of the needle, which although relatively light in weight, does still suffer from vibration due to the movement of the vehicle and it's engine. To compensate for this, some dampening mechanism is usually added to the meter movements, and this adds a delay in display response. Usually it is not a problem, but for fast-moving parameters (such as boost) it is noticeable. With lesser dampening, the inertia of the needle would cause it to overshoot past some actual datapoint, resulting in additonal errors. One more source of error here is the parallax error that comes from reading the gauge at an angle. The further the needle from the backing plate (with the numbers marked on it), the greater the potential for parallax errors.

    There is another class of analog gauges that uses an electromagnet configuration similar to a motor, but with a coiled resistance acting against the shaft. These exhibit much better linearity than the bi-metallic strip meter movements, but still suffer from the mechanical and parallax errors.

    For digital gauges, the display is typically some form of electronic numerical display device, such as an LED or LCD. These displays do not suffer from the problems associated with analog meter movements above.

Analog Instruments
For an analog gauge, the sensor is perhaps the first major point of error. The typical sensors used in automotive applications have a noticeable amount of non-linearity. There are industrial-quality sensors that can be used (but with a gauge to match) that have excellent linearity and accuracy below 1%, but they would be prohibitively expensive for a non-pro racer. Even sponsored professionals would find it difficult to justify the costs of some of these sensors.

Add the slew of errors associated with the traditional analog meter movements, and that most analog gauges do not have any form of data modification to compensate for the non-linearities, and you can understand why analog gauges are not the best way to monitor the parameters of a well-tuned performance vehicle. Note that there are some gauges that claim to convert/modify the data electronically, then output this to a motorized meter movement, but we have not seen this in production yet. They would still have the mechanical and parallax limitations of mechanical instruments.

Digital Instruments
Digital instruments convert the sensor output signal to a digital form where it can be manipulated without any additional source of noise or error. As long as the digital form includes enough distinct levels to provide the resolution we need from an instrument, then they will outperform analog gauges.

Digital instruments typically use the same type of sensors as traditionally used with analog gauges, which is possibly due to ease of transition and low cost due to popularity. However, with digital electronics and especially microcontrollers, it is a simple matter to map the sensor output with that of the real-world parameter to compensate for any non-linearity. The mapping usually includes compensation for any losses during transmission. In addition, transmission along wires be made to exhibit less losses by lower loading, as the sensor signal is now not a direct part of the display drive circuit.

For display, digitial gauges will typically use a numerical digital display such as a 7-segment LED or LCD display. Whereas this solves most of the problems associated with mechanical/analog meter movements, it adds a new source of errors -- reading and interpreting the numerical value is not as easy as with an analog needle, especially if the numerical value is changing rapidly. One work-around to this is to use a bargraph display instead, but due to hardware limitations, the number of distinct steps that can be displayed with a bargraph limits its ability to reproduce an accurate value. When used however, it is also easier to spot a trend in the changing parameter.

One other issue surrounding digital gauges is its visibility in bright sunlight. However, this issue has been solved more recently with the use of high-intensity LEDs, which are becoming more efficient with emerging technologies. LED's traditionally served a role as indicators, with incandescent and other bulbs used for illumination, but LED's are becoming increasingly popular as illumination devices nowadays.

The VEI Solution
So analog systems have their issues, mostly with inaccuracy, but are easier to read and interpret. Unfortunately it's of little use to be able to easily read and interpret an inaccurate value.

Digital numerical gauges can accurately generate a value and display it, but for fast moving data, interpretation becomes a problem.

Digital bargraph gauges can also accurately generate a value and display it with a graphical display, but are limited in distinct levels. These work very well for fast-moving data, and quickly identifying trends, but not so well if an accurate numerical value is required.

So is there a solution? Yes there is, and it comes from a merge of the two digital options, by incorporating a digital bargraph AND a digital numerical display in a single display. With this system, a quick glimpse gives you an instant interpretation of what is happening with the graphical bargraph, yet you're able to gather the full and accurate information with the numerical display. Of course there is significant complication with creating a digital instrument that has a microprocessor, a numerical display, a bargraph display, and power circuitry, etc, all in a small enclosure. It's not a new concept, but due to design and technology issues, have only been used on larger-sized instruments, such as 3-3/8" speedometers and tachometers. But the VEI design team has figured out a way to do this for all of our smaller (2-1/16") instruments as well! Finally, to address the visibility issue, we are using super-bright LED's AND anti-glare filters in our instruments, resulting in visibility that would rival most analog gauges.

Bonus Features
While we're in there, and already have a power microprocessor, we can take advantage of further computer processing power to add other useful features not possible with traditional analog gauges -- upper and lower user-settable alarms for proactive notification of an error condition, adjustable bargraph ranges, and more. See the features document for more information on these bonuses

Copyright (c) 2025 VEI Systems. All Rights Reserved.