One waveform One picture

Ask for and share advice on using the PicoScope kit to fix vehicles here.
ben.martins
Pico Staff Member
Pico Staff Member
Posts: 534
Joined: Tue May 16, 2017 1:02 pm

Re: One waveform One picture

Post by ben.martins »

Dead shorts are usually very simple to determine. Blown fuses, burnt wiring & melted insulation are all tell tale signs but what happens when you have a dead short but none of normal suspects?

One such example was a recent visit to a machine with a long list of fault codes all reporting an issue with the 5V feed to the majority of pressure sensors fitted to it. The main fault code also gave a level 4 warning which limited the machine in a number of ways one being slew. Fault codes in machines are usually accompanied with an audible warning which can be silenced but not when it is a level 4, nothing annoying about that when trying to diagnose!

As normal a quick check of the Level 4 fault code in the technical information was less than useful. DHX1MA gave a cause of the pump controller being defective and reports that as this is an internal defect, troubleshooting cannot be performed. Helpful....

Using the scope to back pin the level 1 fault codes for the sensors, sure enough there wasn't a 5V signal to any of them. The wiring diagram gave us the knowledge that all the sensors reporting an issue share the same 5V reference from the hydraulic ECU, which was conveniently buried in the cab. After much wrestling we confirmed that the ECU was not sending out a 5V signal. With things starting to point towards an ECU we made a quick phone call to the parts supplier. This made us rethink the cause due to the fact they hardly ever sell ECU's for this machine, it was over £2000 and would have to come from Japan.

In order to prove it isn't the ECU and knowing we had a 5V reference issue, we turned back to the sensors on the 5V circuit. In order to rule out shorted wiring and sensor problems the plan was to disconnect the sensors to and remeasure at the ECU. Fortunately a number of junction connectors split the 5V from the ECU and shared it across the majority of the sensors. Unfortunately the connector is mounted in such a position it would have meant stripping the cab.

With nothing else for it, the dreaded process of disconnecting each sensor began. By removing each one from the circuit we could see if the fault codes changed when switching the ignition on. This wouldn't help us if there was a wiring issue between the ECU and the sensor but, whilst extremely time consuming, it would give some direction. Luck was most definitely on our side for a change as the first sensor we disconnected removed the Level 4 warning and when looking through the live data list we had pressures from the sensors previously listed with a fault code! This doesn't happen often and I did decide to do the lottery that night! I must have used all my luck up though on this one job...

Using the break out leads we set up a capture in order to see what was happening when this sensor was connected.
Shorted Pressure Sensor.png
As we can see, there is a big in rush of current indicating a dead short but what is interesting is there is almost a holding phase where voltage drops to just under 2V and 280mA of current is being allowed to flow whilst the ECU is seeing if it will rectify itself. Then after a period of just 90ms, the ECU shuts off the 5V supply which in turn knocks out all the other sensors.

As both Neil Currie and James Dillon mentioned, before condemning a component always think about what else you could check. If there is something, regardless of how long or difficult it is, we should be doing it to ensure the correct diagnosis.

I hope this helps.

Kind regards

Ben

muttnjeff
TwoWaves
TwoWaves
Posts: 72
Joined: Tue Jul 21, 2020 7:19 pm

Re: One waveform One picture

Post by muttnjeff »

Nice post.

Steve Smith
Pico Staff Member
Pico Staff Member
Posts: 1578
Joined: Sun Aug 25, 2013 7:22 am

Re: One waveform One picture

Post by Steve Smith »

The challenges we face with diagnosis not only require investment in test equipment, training and hours of ongoing study but also the need to manage the diagnostic task at hand. The following study is one such example where the combined costs of diagnosis and the parts required exceed the value of the car.

Below we have a 2004 Audi 1.9 TDI @ 203303 miles
Customer complaint was lack of turbo boost and grey smoke under acceleration accompanied with fault code “462D” Boost pressure control (Above control limit)
In plain English we have a turbocharger over-boost condition which is well documented across many forums and case studies

The fault indeed was sticking vanes in the (VNT) turbocharger assembly which was confirmed with a vacuum gauge applied to the turbo control actuator. However, I have always been curious about the level of boost these turbochargers can achieve before “over-boost” conditions are met

Below we monitor a number of inputs relative to boost pressure and capture the over-boost event which I am sure would have increased further still if my measurement hose had not been forced out of the inlet manifold under the shear pressure
Image 1
Image 1
The focus of the above waveform falls upon channels B, C and math channel negduty(F)

Note how manifold pressure continues to rise at 13.41 seconds (Scope view 2) yet the vacuum to the turbo control actuator has returned to near atmospheric (emergency operation)

With no vacuum applied to the turbo control actuator, our boost pressure should decrease in proportion, even though the gas pedal remains to the floor and engine speed is increasing

As we can see above, there is no decrease in boost pressure (quite the opposite) where manifold pressure continues to increase to 3.247 bar (gauge pressure) before my hose connection failed

The turbo control actuator vacuum switching valve (N75) and the engine ECU are both performing as expected given the negative duty of the PWM signal has fallen to 5% (minimal current flow) and the vacuum applied (to the actuator) has reached near atmospheric pressure (0 bar gauge pressure)

Interesting to note the vacuum applied to the turbo control actuator (Ch C) & the negative duty cycle of the N75 vacuum switching valve (Ch “negduty(F)”) are a mirror image of one another

So how much boost is too much?

Unfortunately, I don’t have a categoric answer but the boost pressure above is impressive and testament to the strength of these engines.

The typical peak manifold pressure for a 2-litre diesel engine is approx. 2.8 bar (absolute) or 1.8 bar (gauge)

For the engine above (1.9 TDI AVF) the peak manifold pressure is 2.4 bar (absolute) or 1.4 bar (gauge)

In our waveform above we have captured peak manifold pressure at 3.247 bar (gauge) using the WPS500x pressure transducer

To put this into perspective, if peak manifold pressure should be 1.4 bar (gauge) & we have captured 3.247 bar (gauge) our over-boost equals 3.247 bar -1.4 bar = 1.847 bar or 26.788 psi (gauge)

Now I understand why my hose blew off and next time I will secure with a hose clip (school boy error on my behalf again)

Regarding the picture in the waveform, this can be found in the VW SSP 190/10 which is well worth reading (like all SSP’s) I thought it best to show an image of the internal “workings” of the VNT turbocharger rather than a mundane picture of a turbocharger in situ as authorization to replace the turbo was declined.

I hope this helps, take care……..Steve

Steve Smith
Pico Staff Member
Pico Staff Member
Posts: 1578
Joined: Sun Aug 25, 2013 7:22 am

Re: One waveform One picture

Post by Steve Smith »

There are a number of key questions to ask and observations to make prior to diagnosis as these often provide vital clues and can prevent hours of diagnostic pursuit.

The four key questions I try to adhere too are:

1. How long has the problem been evident?
2. When did you first notice the problem?
3. Has any work been carried out on the vehicle recently?
4. When do you experience the problem?

I am sure there are far more but too many questions may lead to a “stone-wall” feedback from the customer

Regarding observations, a visual inspection (inc. a road-test) is invaluable and reveals far more about the task ahead. I often include the following paragraph in case studies……

“Confirmation of specification is of the utmost importance when it comes to diagnosis as
there is often a temptation for customers to modify their vehicle with fashionable accessories that lack the fundamental quality control and engineering that was originally intended for the vehicle.”


Here is one such example https://www.picoauto.com/library/case-s ... hesitation that I will summarize below:

Our customer complaint is severe hesitation under load.

Thanks to a basic visual inspection it was noted the vehicle was equipped with a “performance” accessory kit installed to the Fuel Rail Pressure Sensor.

Such a device is designed to convince the PCM that the fuel pressure is lower than expected. This results in an increased output from the common rail diesel pump and, theoretically, an increase in performance.
Image 1
Image 1
Of course, removing the performance accessory kit cured the problem and with hindsight I should have inspected the VCV and PCV in order to understand the pulsations we have captured towards the end of WOT and beyond. (I guess the cure did not warrant further diagnosis)

To conclude:

Whilst the performance accessory kit was discovered early, during the basic diagnostic process (basic visual inspection) it would have been very tempting to simply remove the device, road test, and assume everything was in working order.

Following the procedure above, we have not only confirmed that the performance accessory kit was the offending component, but qualified the vehicle performed correctly after the fix, showing no signs of the initial complaint or consequential damage.

I hope this helps, take care……Steve

ben.martins
Pico Staff Member
Pico Staff Member
Posts: 534
Joined: Tue May 16, 2017 1:02 pm

Re: One waveform One picture

Post by ben.martins »

I may now get told off here by Steve for not sticking to the "waveform" rule but technically it is still a waveform just not with PicoScope software but NVH.

This is a relatively straight forward customer complaint with judder from the steering wheel when the pressing the brake pedal. However, after the 3rd set of brake discs and pads alarm bells were ringing and I was asked to take a look. As this was a vibration issue it is only right we apply the Pico NVH kit to the vehicle starting with the first capture in the vehicle where the customer experiences the complaint. This was done using all 3 axis of the accelerometer and using the magnet to attach to the seat base.

The complaint was easy to reproduce and sure enough we had a large "T" vibration during braking indicating that the component of interest was rotating at the same frequency as the wheels. As the steering wheel was the responder in this instance we were confident that the source was from the front axle. Using 2 accelerometers on either strut we could now determine which side was causing the issue.
One picture one waveform.png
From the above we have Channel A, blue, on the nearside front strut and Channel B, red, on the offside front strut. It is quite clear that the issue lies with the offside as the amplitude of the vibration is significantly higher. Removing the wheels and carrying at a run out of the disc we had a significant out of round.
Disc run out video.mp4
(5.59 MiB) Downloaded 239 times
However, considering that 3 sets of discs and pads still hadn't cured the fault it is only right that a hub out should be done as well. (The inserted image in the waveform does give the game away a little!)
Hub run out video.mp4
(16.01 MiB) Downloaded 236 times
As you can see the hub is also out of round which is heightened by the diameter of the disc. No matter how may discs and pads this vehicle would have had it would still have a vibration. The nearside hub was also checked and also found to be slightly out of round although not as significantly as the offside. Using the evidence captured from Pico the customer got 2 new hubs and another set of discs and pads and the problem is no more.

Steve Smith
Pico Staff Member
Pico Staff Member
Posts: 1578
Joined: Sun Aug 25, 2013 7:22 am

Re: One waveform One picture

Post by Steve Smith »

Following on from Ben’s NVH adventure above, the case study below looks at the importance of not only identifying the Frequency and Amplitude of a vibration but also the offending Axis.

In addition to the Frequency, Amplitude and Axis we add into the mix “Load” which we can be determine from the road and engine speed graphed within the signal history.

The capture below was taken from the Honda Civic case study here https://www.picoauto.com/library/case-s ... celeration
Image 1
Image 1
Above we have a clear T1 first order vibration from a component rotating at the same frequency as the wheel and tyre. Whilst this is not an earth-shattering discovery, the fact that it predominately offends the vehicle in a lateral direction (under load) is a piece of diagnostic gold. Whilst tyres are more than capable of generating a lateral T1 first order vibration, they rarely generate lateral vibration under load only! Typically, a lateral vibration (attributed to a tyre) is present on or off load

In the picture above we can see the offending component is our drive shaft inner tripod joint, where damage to the “ball case” results in sticking, dragging or locking against the guide under the transfer of load when accelerating. (Note there was no vibration present during deceleration)

So why the importance around identifying the offending axis?


Referring to the case study in the link above, our initial vibration measurement is always a 3-axis measurement taken at the driver’s seat bolt. Once we have identified the offending axis, we can remove 2 of the 3-axis signals (non-offending) from the accelerometer attached to the driver’s seat bolt (X & Y axis connected to channels A & B of PicoScope)

We now have 3 “free” PicoScope channels (A, B & D) in which to attach 3 additional accelerometers to measure the amplitude of T1 lateral vibrations (accelerometer Z axis) at strategic points about the vehicle (Zoning of our offending vibration)

Please note: when measuring a specific axis of vibration (e.g., Lateral) be sure to mount the accelerometer in the correct orientation as described in the NVH set-up wizard (Vertically with the screw thread facing forward)

To summarize, the focus of our attention when diagnosing NVH concerns surround?
• Frequency
• Amplitude
• Offending Axis
• Load related
• Road speed related
• Engine speed related
• Gear position (If entered)
• Additional PID data post49491.html#p49491
• Temperature
• Vehicle operating conditions
• Vehicle mode settings (Sport / Track power and suspension settings)

To assist with diagnosis and "evidence gathering", I have been taking advantage of videoing multiple Apps simultaneously; I have to say, it paid dividends in the above case study using a webcam under the vehicle alongside our NVH software!

The following link is an additional example of filming multiple Apps simultaneously topic22669.html

I hope this helps, take care…..Steve

Steve Smith
Pico Staff Member
Pico Staff Member
Posts: 1578
Joined: Sun Aug 25, 2013 7:22 am

Re: One waveform One picture

Post by Steve Smith »

Another NVH adventure here with a cheeky video as well as a picture (One waveform one picture one video)

Below we have a customer complaint of an intermittent metallic rattle when the vehicle rides over an uneven road surface or indeed hits a pothole

Often with such noises, we rely on product knowledge and experience to direct us to the source of a noise.

When raised on the ramp, “9 times out of 10” you will not see, hear or feel any component wear such as to warrant replacement and so where to from here?

A leap of faith may be in order to diagnose the concern but most certainly risky (not to mention costly)

The capture below has 4 NVH accelerometers connected to the LH & RH brake caliper mounting bolts and RH & LH front shock absorbers
Image 1
Image 1
Note how the accelerometer units are displayed in the NVH Time Domain as pascals (Pa) which is a unit of sound pressure rather than milli g (mg) for unit of acceleration!

This is simply to allow playback of the captured data within the software rather than exporting mg units as .wav audio files. We have discussed this “cheat” here viewtopic.php?p=92051#p92051

Returning to the image above, the area highlighted in the Signal History (Green rectangle) reports an increasing road speed from 37 mph @ an engine speed of 1534 rpm. Within the highlighted area of the signal history, we have a large energy spike typical of a momentary event such as riding over a pothole of drain grid

Above the signal history we have the signal from each of our 4 accelerometers where we can clearly see the energy spikes emanating from within the sign waves captured on Channels B and D (LH front caliper mounting bolt and shock absorber)

The sign waves are purely created by the rising and falling of the road wheel/shock absorber along an uneven road surface/pothole whist the high energy spikes within (that should not be present) are caused by the movement of the brake pad within the caliper (note brakes are not applied)

The energy spikes within are “momentary and sporadic” in nature and responsible for our metallic rattle sound

Note how the energy spikes relate only to the LH side of the vehicle and not present on the RH side. I can add here that when the RH road wheel ran over a pothole, the same sound could be generated (from the RH caliper)

Please focus on the gap between the brake pad and caliper carrier (inside the yellow square) in the above image and then refer to the video below
IMG_8692.MOV
Video
(4.59 MiB) Downloaded 228 times
New OE brake pads and discs were installed to the vehicle (which had recently been replaced) and now the complaint is cured

Using NVH to capture momentary noises as we describe here topic21061.html and here https://www.picoauto.com/library/case-s ... ing-column will allow for evaluation and triangulation of noises so eliminating numerous non-offending components from the diagnosis

I hope this helps with, take care…..Steve

Steve Smith
Pico Staff Member
Pico Staff Member
Posts: 1578
Joined: Sun Aug 25, 2013 7:22 am

Re: One waveform One picture

Post by Steve Smith »

Graphing resistance using the BNC+ Resistance test lead here https://www.picoauto.com/products/test- ... -test-lead came in very useful when chasing an SRS fault code for a 2014 B -Class Mercedes

The SRS ECU returned a very clear and concise fault code / description below:
“B005013-Driver side front buckle switch Open circuit”

A quick erase of the fault code, wiggle test of the wiring under the driver’s seat and re-scan confirmed fault code B005013 had returned

A visual inspection of the SRS connector (under the driver’s seat) confirmed no issues; however, “rocking” the driver’s seat belt stalk assembly back and forth would result in the repeated return of B005013

The diagnosis is therefore reasonably conclusive in that the buckle switch or harness attached to the seat belt stalk assembly is open circuit, but where?

With the seatbelt stalk removed, the BNC+ resistance lead was applied to the buckle switch harness in order to locate the intermittent open circuit. This is where the ability to graph resistance becomes invaluable as you can plot progressive or instantaneous changes in circuit resistance within the graph view of PicoScope
Image 1
Image 1
Above we graph the resistance of the seat belt stalk assembly (removed from the SRS system)

Note the resistance with the seat belt unlatched = 400 Ω whist latched = 100 Ω

Any movement of the seat belt stalk assembly would result in an open circuit due to a broken wire beneath the mild steel “cable grip” crimped about the stalk wiring harness (See included image)

Repeated movement of the driver’s seat belt stalk (over time) about the wire rope would inevitably result in an open circuit

Interesting how rectifying this fault also restored other faults on the vehicle including "Parking Assist" (Self-Parking) feature & Stop Start! Not sure I can see the link but "food for thought" if nothing else

I hope this helps, take care……Steve

Steve Smith
Pico Staff Member
Pico Staff Member
Posts: 1578
Joined: Sun Aug 25, 2013 7:22 am

Re: One waveform One picture

Post by Steve Smith »

Whilst creating a presentation for Autoinform in early November this year, the “brief” was to demonstrate “Practical applications and everyday uses for your PicoScope, highlighting scenarios where simply no other tool will do”

As you can imagine, these forum post and the content within featured heavily.

I also came across this gem of a Mini Cooper ABS fault which I added midway through a Tesla ASB case study.
The full story can be found here https://www.picoauto.com/library/case-s ... ing-lights but cutting a long story very short, we have a Mini with no speed signal from the left-hand front wheel speed sensor

The wheel speed sensor had been replaced with a reputable aftermarket component along with a wheel bearing containing a magnetic pole ring. PicoScope was used to confirm that a speed signal was present (using AC coupling) at both the sensor and the ABS controller. (See scope view 2 below ) However, there was still no wheel speed when using the scan tool to display live data and a fault code for “L.H Front ABS Sensor no speed signal” remained

The waveforms below were taken from the Mini’s front left & front right wheel speed sensors. On the surface, everything looks fine. We can clearly see the speed signal formed in the power supply’s (AC coupled) in Scope view 2, along with the current flow through the sensors via a x10 current multiplier (Scope view 1) It was only when comparing the current flow through each wheel speed sensor do alarm bells start to ring.
Image 1
Image 1
By using the reference waveform and scaling features in PicoScope, we can overlay the front left wheel speed sensor (green) and the front right wheel speed sensor (black) in Scope view 1 and see the immediate difference.

The signal rulers placed at 60 mA in Scope view 1 indicate the theoretical crossing point used by the ABS controller to calculate the frequency (approximately 6 mA). Each time the current level rises above 6 mA (divide 60 mA by 10 to obtain true current value) and then falls below 6 mA, the ABS controller measures the time taken between each of these events and calculates the frequency, so deriving wheel speed.

Note how the signal captured on channel C (green) never crosses the theoretical crossing point given the current fails to drop below 7.5 mA! The ABS controller simply cannot calculate the frequency and therefore no speed signal recorded via the scan tool for the left front wheel and the generation of the applicable fault code.

The cure for the Mini was to replace the aftermarket speed sensor for an original speed sensor. The wheel speed values returned and the fault was cleared.

The message from the experience above is to use AC coupling with caution as signal fluctuations and ripples are centred on approximately zero volts, regardless of the fundamental signal level (removing all bias voltages). This does not mean that AC coupling should not be used, in this instance, it serves as an initial and quick measurement to confirm the activity from the ABS sensor only. (Perhaps to determine pole count) Here we can confirm that the power, ground, signal and pickup are functional but we cannot prove plausibility

However, to confirm the ABS sensor and circuit (plausibility) it is important to measure current flow while comparing the signal characteristics of an offending sensor with a known good (In this case, LH front with RH front).

I hope this helps, take care…..Steve

Steve Smith
Pico Staff Member
Pico Staff Member
Posts: 1578
Joined: Sun Aug 25, 2013 7:22 am

Re: One waveform One picture

Post by Steve Smith »

The recent case study below looks at how an engine ECU interprets an electrical malfunction of a throttle actuator.

The vehicle in question is a Skoda Octavia 1.0 TSI with engine code CHZD

Symptom: No response from the engine when depressing the gas pedal (MIL illuminated)

The DTC below was beautifully described to the letter (which makes for a refreshing change)

Throttle Actuator (G186)
P1558 00 [237] - Electrical Malfunction
Freeze Frame:
Fault Status: 00000001
Fault Priority: 2
Fault Frequency: 1
Date: 2021.12.05
Time: 12:37:45


What I found very useful was the included Time & Date stamp within the Freeze Frame data which is joy to utilize when the fault is “permanent” as with this vehicle

Long story short; erase fault code, cycle ignition, recheck fault codes and “P1558 00 [237] - Electrical Malfunction” returns accompanied with the relevant time and date stamp. (No need to run the engine)

Using PicoScope 4823, all 6 wires of the throttle actuator could be captured simultaneously whilst adding a 7th channel to monitor throttle motor current (N.B. below, some channels are hidden from view)

Image 1
Image 1

As can be seen in the image above, the fault with our throttle actuator was a broken return spring which is responsible for ensuring the throttle always returns to the closed position (i.e., fail safe)

Whilst this is plain to see with the throttle actuator disassembled, how did the engine ECU detect this as an “Electrical malfunction”?

In the image above I have added 3 reference waveforms from the faulty throttle actuator (“Magenta TPS 1 fault” “Lime Green TPS 2 fault” “Black Throttle fault”) and over-laid them against the new throttle assembly

Channels D & E capture the throttle motor power supplies (both positive and negative) from the good throttle actuator. Note, where there is a differential voltage (above 0 V) between these channels there will be current flow which is captured on channel G

Starting from left to right, the ignition and engine are turned off and then switched to “ignition on engine off

Refer to the following steps:

Step 1: At approx. 3.4 seconds, current flows through the throttle motors, partially opening both throttle butterflies from their “rest” position. Note the change to the throttle angles detected by TPS 1 & 2

Step 2: Between 3.45 and 3.55 seconds there is no current flow through either throttle motor, however, Channel B (Good TPS 2) and Channel F (Good TPS 1) indicate a change in throttle angle due to the action of the throttle butterfly return spring on the good throttle actuator.

In stark contrast, the reference waveforms (TPS 1 fault & TPS 2 fault) reveal how the fault throttle angle (of the failed actuator) remain in a fixed position and do not change!

The engine ECU will immediately recognize this condition as an electrical fault given the expected changing output of TPS 1 & 2 has not occurred.

Step 3: From approx. 3.55 seconds onward, the good throttle actuator receives additional current to set the throttle butterfly to the desired position followed by a holding current flow of approx. 850 mA

In contrast, the reference waveform “Throttle fault” reveals a different approach regarding the current flow behavior of the faulty throttle actuator. At approx. 4 seconds the throttle is “driven” to the rest position and held in place using minimal holding current.

Using the scope to capture all inputs and outputs from the throttle actuator (under load conditions) provides rapid confirmation of the circuit integrity whilst revealing the cause of our “Electrical malfunction” was of a mechanical nature

Many thanks once again to Kevin Ives at Ives Garage for another year of diagnostic twists and turns

I hope this helps, take care…….Steve

Post Reply