noise

Ask any questions relating to the PicoScope hardware or kit contents here.
If you have any questions prior to purchasing the kit post them here.
josegumby
OneWave
OneWave
Posts: 31
Joined: Sun Aug 27, 2017 3:46 am

noise

Post by josegumby »

Hello, I'm scratching my head on what seems to be a peculiar behavior I'm seeing on my scope screen.

A couple weeks ago, I was doing an alternator ripple test on a '98 Ford Mustang, and I noticed some noise in the signal, spikes which seemed periodic. Curious, I connected my TA018 (20/60 A current clamp) around the ignition feed wire to see if this is noise related to the ignition system. Sure enough, every single peak, lines up perfectly with the collapse of the ignition coil magnetic field (where the current ramp stops).
Ford_Mustang_1998_Petrol_20200111-0001.psdata
(412.48 KiB) Downloaded 618 times


After studying this waveform, I noticed that these peaks (caused by the ignition system) are over 1 volt, exceeding the limit of the scale that I was on. So last night I did the same capture, except the leads were connected directly to the battery instead. At first, with no current clamp, the noise from the ignition appeared way lower (around 200 millivolts or so).
current clamp disconnected.jpg


Next, I connected the current clamp to the ignition feed wire, and suddenly, the spikes go way up! As soon as I remove the current clamp, they go back down. I then very carefully connect the current clamp, making sure not to touch any wires or any part of the car, and the spikes go way up again. It's almost as if the presence of the current clamp is somehow amplifying the noise from the ignition system on the OTHER channel. These spikes from the ignition system are greater than 3 volts in many cases.
current clamp connected.PNG


How could this be? Is there something wrong with my scope? Or is there something special about AC coupling that I'm not aware of?

I realize that I changed from the alternator to the battery, and that might be causing inconsistent results from a couple weeks ago, and I realize that it's going to be normal to see some noise in the battery voltage. But why does simply adding the presence of the current clamp cause the noise from the ignition voltage spikes to suddenly amplify? That's what I'm trying to work out. Again, I was very careful not to touch the wire at times, yet it still had a dramatic effect.

Any ideas?

Thanks,
Curt

martinr
TwoWaves
TwoWaves
Posts: 53
Joined: Tue Mar 06, 2018 4:06 pm

Re: noise

Post by martinr »

Hello Curt

I was just browsing the forum; I don’t visit anyhing like as often as I should.

Did you get anywhere with this?

If not, a couple of thoughts:

Your photos show cables resting all over on the engine. Ideally you need to get your cables routed as far as possible from the engine and especially any ignition components. If necessary, place the scope and laptop well away from the engine, routing cables along bulkheads or getting them away from the engine area as quickly as possible.

Also, did you try grounding the earth screw on the scope to battery negative?

One last thing to try is disconnecting any mains power supply to the laptop, letting it run on its battery. That is definitely worth testing in general; I have one laptop power supply where I have marked NVH with a red cross through it on the mains plug to remind me never to use it with NVH. Generally, and routinely, I look at a signal and then unplug the power supply to the laptop and look for any changes.

Iver
TwoWaves
TwoWaves
Posts: 182
Joined: Mon May 27, 2019 2:55 pm

Re: noise

Post by Iver »

Hi Curt

Everything Martin has said is valid. Just to add, a battery charger connected to the car can also cause noise, the environment you work in, old dodgy electrics around the building wont help.

victor2k
TwoWaves
TwoWaves
Posts: 337
Joined: Thu Jan 28, 2016 7:17 am

Re: noise

Post by victor2k »

Hi,
Try to make a ground connection to your clampmeter and check again the alternator ripple test.
The 4xx5 series scope use TI 4521 differential amplifier for input signal and is not grounded inside of scope.
Sometimes the faster signal from a channel will alter a slower/little signal from other channel.
Best regards
Last edited by victor2k on Mon Mar 02, 2020 1:47 pm, edited 1 time in total.

victor2k
TwoWaves
TwoWaves
Posts: 337
Joined: Thu Jan 28, 2016 7:17 am

Re: noise

Post by victor2k »

Please delete this,I made a mistake and quoted instead of edit my post.
Thank you.

josegumby
OneWave
OneWave
Posts: 31
Joined: Sun Aug 27, 2017 3:46 am

Re: noise

Post by josegumby »

Thanks for your replies everyone.

The 4425 picoscope, USB cable, and blue test lead is hanging are all hanging from the "S"-hook at the front edge of the hood, so the cables are going straight down, and not actually resting on the engine, although that might not be apparent from the pictures. The amp clamp wire is going across the top of the mass air flow sensor, so next time, I will route that upward and away from the engine as well; although it's difficult to keep that away from ignition wires, since that is what I am actually measuring.

Next time I will be more careful about routing ALL the wires away from the engine, and I like the idea about adding the extra grounds; I'll try that next time.

Thanks,
Curt

PS - I NEVER have my laptop connected to mains. In this case, I was working mobile, and the car was in a parking lot far away from any buildings, so I was actually too far to be plugged in anyways.

martinr
TwoWaves
TwoWaves
Posts: 53
Joined: Tue Mar 06, 2018 4:06 pm

Re: noise

Post by martinr »

Thanks for the update, Curt. If by “next time”, you are planning to repeat the exercise in the near future, would you let us know how you get on? It’s just as helpful to know what doesn’t work as to know what does work.

Here’s another thought: whilst looking for the cause of your noise, try holding, for example, the current clamp, then its cable, then perhaps the ‘scope itself and see if it makes any difference to the noise level when you have physical contact; it might just help indicate the source of entry of the noise.

Another thought, instead of putting the clamp around the ignition feed wire, can you take out the relevant fuse in the fuse box and fit a fused loop to clamp around and read the current? That should take you well away from RF noise and that should tell you where the noise is or isn’t coming from.

Martin

josegumby
OneWave
OneWave
Posts: 31
Joined: Sun Aug 27, 2017 3:46 am

Re: noise

Post by josegumby »

Those are some good ideas, Martin. I'm not sure if I'll get to look at this car again or not. It belongs to my friend's son, who will be moving away soon. I've been trying to fix all the car's problems for him (it has had many) as this is his first car, and he doesn't have his driver's license (yet). I'm not sure what the fate of the car will be, if he will leave it where it is, or take it with him when he moves (and his girlfriend would drive it).

I have experimented since, regarding electrical noise, on a totally different, perfectly running car (MAP sensor), and there was all kinds of noise on all three wires of it; yet the car runs fine (2001 honda civic). My takeaway from that experience, is that there is a lot of noise under the hood, and the PCM probably filters it out, similar to how we do with our oscilloscopes. In the case of the Civic, the noise (from the ignition) was far worse, in all three of the MAP wires than on the Ford Mustang that this post was about. So... some of this noise is gonna be normal, and I need to learn to ignore it sometimes. I've attached the file with the MAP sensor noise on the civic for anyone's learning or amusement. Be sure to turn off the filters from all three channels to see the noise.

As far as the noise amplifying as soon as I connected the current clamp... if I get another crack at the Ford Mustang, I will try your suggestions and report my findings.

Thanks,
Curt
Attachments
Honda_Civic_2001_Petrol_MAP_Sensor.psdata
(16.98 MiB) Downloaded 592 times

martinr
TwoWaves
TwoWaves
Posts: 53
Joined: Tue Mar 06, 2018 4:06 pm

Re: noise

Post by martinr »

You’re absolutely right, Curt, there is lots of RF noise under the bonnet, which the ECUs filter out. I was watching a ScannerDanner Premium video only yesterday and that very topic reared its head as he connected his PicoScope to the power and ground wires to an ignition coil. He said to his students something like, the first time you see that that you’ll probably thing you’ve found a fault, but it’s perfectly normal, welcome to the world of oscilloscopes.

If you haven’t seen ScannerDanner, look him up. I’ve never seen a bad or even mediocre video of his, and Pico have incorporated 2 of them into their library

https://www.picoauto.com/library/training/picoscope-basics-video-part-i

https://www.picoauto.com/library/training/picoscope-basics-video-part-ii

If you have problems with those links (they don’t work for me because of filtering on my router), the originals can be found on YouTube. The second video really clarified a great deal for me on the topic of sampling rates.

https://youtu.be/81AGbZcgCZs

https://youtu.be/ntJJYhkmR4I


ScannerDanner’s work is so good that I’ve been a long time subscriber to his Premium channel.

martinr
TwoWaves
TwoWaves
Posts: 53
Joined: Tue Mar 06, 2018 4:06 pm

Re: noise

Post by martinr »

Lovely .psdata file, Curt; thanks for including it. I can see that your idle speed at the start was 860 rpm, that you took it up to 3297 rpm before taking your foot off the pedal, after which the engine was idling at 842 rpm as the trace ended.

Those spikes are ignition spikes. I don't know if they are directly from the collapse of the magnetic field in the primary circuit or the firing voltage at the very same time in the secondary coil circuit. My money is on it coming from from both circuits - notice how each spike generally has a positive and negative element. And it may well be nothing to do with RF pickup, but my guess is it's feedback through the circuit (ie if you could shield every cable in the car, it still would be there just as strong; that's my guess). But it's absolutely normal and in time they'd seem so natural you probably wouldn't give them a second thought, let alone want to filter them.

(If, instead of a total capture time of 5 Sec, you instead triggered off one of those spikes and set the capture time to a few ms, it might be interesting to see if you could make out anything else around the spike. I can see, zoomed in to the maximum, a tiny dip just after the spike. All of theoretical interest only, but a great way to learn some of the features on the 'scope. After years of playing with PicoScope, only the other week I finally discovered one thing it can't do: integration of a lowpass filtered signal in Maths Channels!)

Post Reply