Maybe you've seen nice clean waveforms in magazines or online, how come they look so good? Perhaps the guy in the bay next to you doesn't seem to have the same problem with his scope. What's going on? There are several factors that contribute to this "noise" so before you stuff your labscope in your toolbox to never be seen again, let's examine them.

EMF: Electro-Magnetic Fields

First we must understand what this noise represents. Noise in electrical signal represents EMF. The amount of electrical noise is determined by the strength of the EMF. As you would guess, the underhood environment is an extremely noisy place with a high level of EMF.

So, what creates these strong electro-magnetic fields? The ignition system is the main source but there are others such as solenoid circuits, alternators, and more. EMF is even produced by the shop AC power, fluorescent lights and other electrical devices. Anything electric can produce EMF.

The problem with EMF is that if it is introduced into your testing it can invalidate your measurment or make it hard to interpret. The goal is to prevent this from happening. There are several steps that you can take to  minimize this problem.

OK, so noise is ever present, but why does it show up on your labscope or graphing meter? The first reason has to do with the meter itself. Some have a noise problem and while others don't. Two meters, the UEI ADL-7100 (also sold as the Sun LS-2000) and the Snap-on Vantage, seem to be very susceptible to noise. Both are very powerful and capable of displaying extremely high speed "glitches" in a signal. They sample the signal at incredibly fast rates and always display any major changes. This strategy is often referred to as peak-detect or min/max. 

These meters are designed to operate this way for good reason; it makes them very powerful. After all, we want to capture those glitches don't we? When we want to capture a problem with a potentiometer such as a TPS or a vane airflow meter or when we look at high-speed signals such as ignition or injectors this peak-detect capability is essential.

The downside of this becomes apparent when measuring low voltage, low current, and low frequency signals. Have you ever tried to view a signal from something like a coolant temperature sensor or an oxygen sensor and only saw a waveform that seemed to be growing hair? What you are seeing is all of the EMF that I mentioned earlier. It can make a signal very difficult to analyze with any accuracy. It’s the peak detect-function that causes the meter to respond to the noise. 

You won't experience this problem with some meters of course since they don't have this peak-detect feature. This is fine, but then you may not have as much ability to capture legitimate problems. Still other meters, such as the Fluke 98, have the ability to turn this peak-detect feature on or off. This allows you to have the best of both worlds. Fluke calls this the smooth function. With meters such as the ADL-7100 and the Vantage however, you have no choice but to operate this way all of the time so you must find ways to work around it.

Shield Yourself

The second contributor to noise is the test leads used to connect to the signal. Many meters come with test leads that are virtually identical to what's found with most multi-meters. They may seem perfectly adequate at first but in reality they act just like long antenna picking up all of the environmental EMF. This isn't a factor with a digital multi-meter because it samples relatively slowly and averages the value to be displayed. The leads may pick up the EMF but the multi-meter just flat out misses it; it's not a factor. With your labscope or PGM however, it's a different story. We rely on it to measure and display high speed signals accurately. This high-speed measurement ability also means that the EMF may be a problem. 

Shielded test leads reduce this problem. They are constructed of a coaxial type cable with the signal conductor surrounded by a braided wire shield. This shield is connected to the vehicles ground and serves a dual purpose as a ground conductor and a shield. EMF is absorbed by the shield and passed through to the ground conection. Unfortunately, like I mentioned earlier, many meters only include non-shielded, multi-meter style test leads. I assume that cost is the motivating factor when manufacturers choose not to include shielded test leads with their equipment. (Please note that meters such as the Fluke 98 and the Interro PDA do come with shielded test leads).

How much difference do shielded leads make? Look at figures 1 and 2.  They show signals from the Snap-on Vantage and the UEI ADL-7100 using the stock multi-meter type test leads. In each example the ground lead is securely attached to an engine ground and the signal lead isn't even connected to anything yet. All of this noise is being picked up from just the underhood environment! Now look at figures 3 and 4. They show the same meters with shielded test leads hooked up the same way. The voltage and time base has remained the same. Notice that there is substantially less noise present on the signal. It's clear that shielded test leads greatly reduce the effects of EMF on your testing. This alone makes them a worthwhile investment.

But wait, there’s more

So it's clear that shielded test leads are advantageous, but are they enough? Most of the time they are, but not always. The vehicle wiring too can act as an antenna for EMF. The engineers that design engine control systems employ certain strategies to overcome this so it doesn't normally have any effect on the engine operation. It can affect your testing though, even when you use shielded test leads. No matter how good your test lead is shielded, you may still encounter noise. To help eliminate this many meters employ strategies to overcome noise such as the smooth function found on the Fluke 98.

 Some meters, such as the Fluke 98 and Interro PDA, also specify the use of a 10:1 filter when testing low frequency signals such as oxygen sensors. The 10:1 filter is essentially a ten mega-ohm resistor in series with the test lead that doesn't allow the unwanted noise to reach the meter. This means that the signal is now represented 10:1 on the display. In other words, one volt is now represented by 100 milli-volts. In most cases, the meter accounts for this change while in other cases you must make the appropriate adjustment to the voltage scale yourself. Meters such as the Vantage and LS-2000/ADL-7100 already have a ten mega-ohm input impedance though so the 10:1 may not be the best choice.   

There is still another noise reducing option in addition to the 10:1 filter. This option is the low pass filter. It works great with tools such as the LS-2000/ADL-7100 and the Vantage. The low pass filter is placed in series just like the 10:1 but it doesn't affect the signal voltage. The low pass filter cleans up the signal by removing the high frequency portions and thus reducing the noise. Take a look at figures 5 and 6. They both show an oxygen sensor signal captured with shielded leads and then with a low pass filter. The difference is obvious with both meters.

There is no disputing that the low pass filter is effective at removing unwanted noise. This makes the signal much easier to analyze and measure accurately. Be cautious though, the low pass isn't for everything. You must keep in mind that it works by eliminating high frequencies. This is fine for observing relatively slow signals such as oxygen sensors and thermistors. If you’re looking at a high-speed signal though, the low pass will give you inaccurate results. You don't want to use it to observe square wave signals, ignition, injectors, etc. In addition, you wouldn’t want to use the low pass when looking for glitches in a potentiometer, such as a TPS, either. The low pass should only be used for low speed, analog signals when high-speed changes aren't a concern.

The bottom line

So what does this all add up to? Well, if you're serious about your testing then you are most likely concerned about signal noise. You've spent the money on the high-tech equipment so why not get the most out of it? Purchase a set of high quality shielded test leads and maybe even some other noise-reducing accessories such as a low pass filter. You won't regret it. No matter what type of test leads you are using though, take extra precautions to isolate them as much as possible from sources of EMF such as ignition components. The key, as always, is to understand the systems you are testing as well as the characteristics of the test equipment that you are using. Knowing these things will make your testing a much more pleasurable experience and enable you to take your abilities to the next level.

Written by Chuck Walker

Chuck is an applications specialist here at AES. He is an ASE certified Master Automobile Technician and holds a California Smog Check Technician license. He has extensive field experience solving tough driveability problems and diagnosing & repairing automotive electrical, electronic and computer systems.

Copyright 2000 by AES (Automotive Electronics Services)