Let's teach Keri how to use an Oscilloscope!

[KeriJane]

Hello.

Well, I've gotten a bit used to my new 25mhz Owon scope. For low-frequency stuff, it's great. Now I hook it up to ATX Power Supplies and..... UH OH.

Great big gnarly spikes all over the place, even on the ground wires. Even with the little ground clip attached to the particular ground wire being tested. Multiple volts of P2P!

Worse, the amount of ripple voltage varies with the volts per division setting, indicating that the ripple/spikes don't originate with the circuit being tested.

Oddly, a really cheap 6 year old pull from an upgrade that weighs about ONE pound came up relatively clean with no motherboard attached, but started spiking all over just like a new Antec EA380 when attached to a motherboard.

Obviously I'm doing something really, really wrong here. So, some reading about scopes, PSUs, probes "ringing" and testing procedures later.....
Oh! I needed to attach a 10uf and a .1uf cap together in parallel across the probe tip and ground. Ta Da! no more giant spikes all over. Thanks, JonnyG.

However..... I still have some questions.

Someone somewhere stated that adding the decoupling caps across the probe tip is meant to simulate the caps on the motherboard. The great big spikes are evident with the PSU connected to a motherboard! Both an old one and a brand new one with zero hours on it. Both with lots of great big caps way bigger than 10uf. Wouldn't this kinda sorta indicate that maybe something else is going on? Like great big gobs of electromagnetic interference that transcends any number of caps and goobers up my probe?

Someone else stated that a "Differential Probe" is meant to be used for ATX PSU testing, but that it isn't really that big of a difference. I noticed that Differential Probes are kinda expensive. So are Active Probes that are specially designed to get rid of the annoying ringing spiking thing that I might be suffering from. Does anyone have any experience with these?

My Decoupling Capacitors obviously alter the characteristics of the probe, rendering it inaccurate when testing against the 1khz/5v reference signal. Doesn't this make it useless for testing PSUs as well? Or am I meant to use just the caps on a bare cable without any probe? I'm going to test this out further. I suspect that the caps are meant to be a DIY probe, not to be added to a 1X-10X probe.

Oh well. My sig is more appropriate than ever!

Have Fun,
Keri

[Bob Parker]

Re: Let's teach Keri how to use an Oscilloscope!

Oh! I needed to attach a 10uf and a .1uf cap together in parallel across the probe tip and ground. Ta Da! no more giant spikes all over.

I've been qualified in electronics for 33 years - I've never heard of that one before and I don't think I'll ever try it.

OK, one thought. Is this oscilloscope earthed to the same mains earth as the power supply? If it is, the resulting earth loop might be causing phantom spikes to be appearing on the waveform. If that's not causing it, I'm out of ideas.

I haven't been paying attention to your other posts so I'm not sure what your experience level is or what else to suggest. There are plenty of really smart people here who I'm sure will give you some ideas to check out. Good luck.

[KeriJane]

Re: Let's teach Keri how to use an Oscilloscope!

Hi Bob.

Here's one of the posts I mentioned.
http://www.jonnyguru.com/forums/show...01&postcount=5

The people over there seem competent and this post in particular details the differential probe and the decoupling caps. The picture he refers to is only viewable by members of that forum but is just a probe end with two caps attached to it. A 10uf/50v electrolytic and a .1uf ceramic cap are across the probe shield and tip in parallel with each other.

I copied it on one of my probes and it works well in the 1X position.

For the Ground Loop thing, I had tried 3 combinations: 1-both scope and PSU on same outlet. 2- on separate outlets not on the same circuit breaker. 3- scope on an uninterruptible power supply , PSU on outlet. No combination had any effect.

My experience level? Well, I made a thread asking for help so I'd be inclined to say "low to moderate".

Electronics and mathematics absolutely do not come easily to me. But I've always been very curious about electricity and everything electronic. (a tragic combination )
Everything that I've learned has taken years of intense study with frustratingly slow progress. Aptitude: 1/10, Persistence: 9.5/10

Have Fun,
Keri

PS. I have a similar tragic affinity for Chess!
(I like it a lot but play at sub-moron level)

[Bob Parker]

Re: Let's teach Keri how to use an Oscilloscope!

For the Ground Loop thing, I had tried 3 combinations: 1-both scope and PSU on same outlet. 2- on separate outlets not on the same circuit breaker. 3- scope on an uninterruptible power supply , PSU on outlet. No combination had any effect.

Hi again KeriJane,
Good to see that you thought about this possibility. I'm in Australia where mains power is considerably different to the USA so I have assume that safety ground wiring is much the same in both places.
Most likely you've still been getting the same ground loop regardless of which set of connections you tried. The UPS's power plug ground is probably connected direct to the ground pins on its outlets.
Ideally the thing to try would be powering the 'scope from an isolation transformer which has no ground connection on its outlet. Powering the 'scope from batteries would be even better, but one of your postings said you didn't have a battery pack for it.
There are a lot of things to consider if you want to get meaningful measurements from your new oscilloscope. Would it be OK if I contacted you directly to save boring everyone here?

Bob

[Gianni]

Re: Let's teach Keri how to use an Oscilloscope!

Hi Keri Jane

I've never heard of that one before and I don't think I'll ever try it.

Same for me. If you place a cap on the probe, you are not testing the "clean circuit" and in some circuit it could change the working mode. I think it is OK only to check ripple on PSU outputs.

What I can suggest you is to connect the ground of your probe very close to the point you want to test; I have seen people connecting the ground probe to TV Tuner because it is connected to GROUND and it is easy to attach the ground crocodile there, than they were checking signal on the board... you can't imagine what they were seeing on the scope instead of real signals.
Another thing you can do is to "twist" the ground probe around the probe tip, it helps to reduce the pick up noise.
Another thing you can do is to connect tip to probe ground, yes to short the probe, and the place it near the point you have to test: if you see on the scope a lot of noise and spike this can give you an idea what your probe is picking up. An ideal probe should show a plain line since it is shorted, but real probes always pick up noise.

I have never tested a MOBO with a scope but I think you will have a lot of spike due to the switching regulator and CPU operation because the working frequencies are high.

I have to go now, ciao
Gianni

[Bob Parker]

Re: Let's teach Keri how to use an Oscilloscope!

I have never tested a MOBO with a scope but I think you will have a lot of spike due to the switching regulator and CPU operation because the working frequencies are high.

Hi Gianni,
That is exactly one of the things I was going to write. Thanks for saying it first! If you are going to try to measure the noise coming from a power supply, you do not connect it to something which itself is generating massive amounts of high frequency electrical noise.
You need to have nice simple pure resistive loads on the power supply outputs so the only noise you measure will be coming from it. And not have ground loops either.

Cheers
Bob

[i4004]

Re: Let's teach Keri how to use an Oscilloscope!

>You need to have nice simple pure resistive loads on the power supply outputs so the only noise you measure will be coming from it. And not have ground loops either.

yeah, that's why i said what i said in ker's the thread about the scope.
i said the probe mods are kinda moot if you're not gonna have the load tester they use, and they use those expensive soonmoon(or whatever) things...

>There are a lot of things to consider if you want to get meaningful measurements from your new oscilloscope. Would it be OK if I contacted you directly to save boring everyone here?

eh..boring.
it's badcaps forum, and it's in the test equipment section, so scope talk is not prohibited.
that stuff would probably be interesting to many here....

[jpdoe]

Re: Let's teach Keri how to use an Oscilloscope!

>> that stuff would probably be interesting to many here....

I second that. My experience with oscilloscopes is very limited, so I find this thread quite interesting.

[PCBONEZ]

Re: Let's teach Keri how to use an Oscilloscope!

KeriJane

I think you need an isolation transformer to power the scope when using AC.

Does it do the same thing on battery power?

[KeriJane]

Re: Let's teach Keri how to use an Oscilloscope!

Wow!

Thank you for all the replies and suggestions.

So far as the scope probe goes, I did try winding the grounding lead around the probe with no effect. I also tried moving, straightening and coiling the probe lead and moving the equipment around the room. All with no discernible effect.

The peeps that recommended the decoupling caps say to detach the ground clip because it will only pick up noise.
They appear to be right, because with the caps on the grounding clip has no effect. At least not on the P2P ripple measurement.... I didn't check mean volts. With the caps off of the probe the grounding lead makes a gigantic difference as one would expect.

The Motherboards both make a huge amount of interference. This was easy to spot by connecting the relatively "clean" one pound (weight) PSU to them and noticing all of the spikes and noise that resulted. Internally, that PSU is a joke. It's exactly the kind that JonnyG likes to blow up. The kind that says "300w" on the outside but might be good for half of that. However, the low power apparently results in less interference! Well, at least less that affects my scope.

My modest OWON scope does have a battery opening on the back. It is meant to be used with an 8000ma 7.4v Li-Ion Polymer battery and has a built-in charger. Holiday shopping has left me not feeling like spending the $90 for an OWON battery pack right now but I'm sure to obtain one soon.
A good question: Should I ground the scope when running on batteries? If so, where? And what to? The test ground on the panel or the probe ground clip?

And finally: The decoupling Caps that appear to work as intended. Both need to be on to have an effect, neither one works well by itself. One is electrolytic, the other ceramic disk. Why might that be?
Also, what exactly does "decoupling" do or mean in this instance? This sounds like another one of those basic concepts that takes me years to grasp. Like Algerbra

Thanks again,

Have Fun!
Keri

PS. Oh! go right ahead and PM if you want.
But this thread is kinda meant to be boring for people with a clue. So.... if it's relevant to teaching scope usage, other clue-challenged people like myself might want to see it.

[i4004]

Re: Let's teach Keri how to use an Oscilloscope!

I think you need an isolation transformer to power the scope when using AC.

why do i think none of the sites scoping the psus doesn't use isolation trafos?

didin't saw that mentioned anywhere.

however in some pdf linked from jonnyguru forum i saw it's really pita if you wanna do this properly..then again, it's more of a scientific paper, and there anything is bound to look complex...
here's that pdf
http://www.google.com/url?sa=U&start...PFP9m3cHbIyWJQ
poster there reffered to pages 60 and 61, if i remember correctly...

here's what i would do if i had (as much) interest in such type of scoping: ask somebody who's doing it.
oklahoma wolf is now doing jonnyguru's job, so he should be approached(either here or on jonnyguru forum) and asked about it.
what i miss about the pic(on j.guru forums) is the context; you have pic that shows zoomed "something".
and i would like to see few shots, so we could see the actual probe, the actual load tester etc.
not because i wanna scope that, but because one zoomed image is not enough...it just made me more curious.

when it comes to this particular case, i think biggest difference is load; they use pure resistive loads inside the load tester.
i don't think they have isolation trafo(offcourse, i'm just guessing...).

here something from there
http://www.jonnyguru.com/forums/show...5228#post45228

i'll try to PM o.wolf there to see if he can provide some data keri is looking for.

[KeriJane]

Re: Let's teach Keri how to use an Oscilloscope!

Here was my first clue that I might be experiencing some kind of interference...
I grounded the probe to a molex connector, then probed the other end of the same terminal to verify a 0v reading.
Surprise!
Giant spikes and lotsa ripple.

Grounded the probe tip to itself (with the clip thing) 0v as expected.

Here's a nice drawing to demonstrate>


Arrrgh! head hurt after that one!

Then I learned some more about frequency , ringing and interference with a capital "I".

Have fun!
Keri

[KeriJane]

Re: Let's teach Keri how to use an Oscilloscope!

Thank you in advance i4004.

That
DARN
PDF
File
From
Heck

MAKES
MY
HEAD
HURT

a lot.

Somebody can actually understand all that?
I give up
Maybe I'll go back to Terry Pratchett novels.

Have Fun,
Keri
I'm not really giving up, just a bit overwhelmed...

[jpdoe]

Re: Let's teach Keri how to use an Oscilloscope!

Hi, KeriJane. I think I can help you with the decoupling question. I have run into that question myself sometime ago.
When used regarding capacitors, decoupling refers to the use of such capacitor to isolate two circuits, at least as much as currents and voltages of frequencies that you're interested are concerned.

Let's suppose you need to interconnect two circuits, each with a transistor biased with a specific DC voltages and currents. If you interconnect both circuits directly, the DC voltages and currents at which these transistors are set up will change. But if you use a capacitor of a capacitance such that its impedance is virtually zero for the frequencies you are interested in transporting from one circuit to another, and virtually infinite for direct currents (zero frequency), the bias point of the transistors will not change. A capacitor used in that way is called a COUPLING capacitor.

A decoupling capacitor is the inverse of that. I'ts used to avoid the entrance into a circuit of a signal of a specific frequency range. So let's suppose you have to feed an IC with a noisy power source. What you can do is get a capacitor of a capacitance such that it's a virtual short circuit for the noise frequency range, and connect it near the IC power pin, between this pin and ground. The DC current will go to the IC, and the noise will go to ground.

[PCBONEZ]

Re: Let's teach Keri how to use an Oscilloscope!

http://en.wikipedia.org/wiki/Isolation_transformer

[Oklahoma Wolf]

Re: Let's teach Keri how to use an Oscilloscope!

Just a quick comment... if you're getting a bunch of noise in the scope that's still there with everything switched off, you could be picking up something from elsewhere in the building. We have a TV set here that completely kills the ability to do my load testing when turned on, simply because it spews so much interference into the mains.

I don't use an isolation transformer. Reason being, I don't have one. Since the load tester is on a different circuit than the PSU being tested, I'd actually need two. I say that because when that TV set is on, the noise gets in there no matter which one is unplugged. Only if I unplug both load tester and PSU being tested does the garbage from that set go away, and of course that makes it hard to get any work done

I've learned to just make sure that TV is off whenever I start load testing something.

[PCBONEZ]

Re: Let's teach Keri how to use an Oscilloscope!

why do i think none of the sites scoping the psus doesn't use isolation trafos?

I don't know.
You thought they were using a great scope too.

[KeriJane]

Re: Let's teach Keri how to use an Oscilloscope!

Thank You JPDoe.

Now THAT's an explanation I can understand!

Apparently they want to shunt the irrelevant noise away from the probe tip and the capacitor pair seems to do a great job of that.


Hi Oklahoma Wolf!

Oh, I forgot to mention that the molex ground noise was only with the PSU "ON" not when it was off. "Off" resulted in no major interference on any wire including the ground ones.

Other circuits that I've tested here have had no such interference, including the PSU inside my LCD TV set. (I was looking at that for coil whine) Maybe that PSU works at different frequencies or power levels?

Have Fun!
Konfuse-a-Keri

[Oklahoma Wolf]

Re: Let's teach Keri how to use an Oscilloscope!

I don't know.
You thought they were using a great scope too.

I can't speak for other sites, but I use two scopes. There's the Stingray that gets all my screenshots for me and there's the big Matrix that sits on a table just below the load tester.

I've actually stopped using the Matrix for the most part. It never disagrees with the Stingray, and has largely become redundant. Most of the time I use it to burn in a PSU I've just taken apart and put back together because I don't want to waste time firing up the office computer just to get the Stingray online.

One more suggestion - if going by the ATX spec, set the scope for 10uS/sec or so. For reviews, I usually measure and capture at 25kS/sec on purpose just because it's easier to find and show low frequency ripple that way, but that's not to ATX spec. I'll often change it when I need to to show off oddities in the unit being tested, but the reality is that if I adhere rigidly to the spec I tend to miss stuff like the ripple I caught in last week's Seasonic M12D review.

[KeriJane]

Re: Let's teach Keri how to use an Oscilloscope!

Thank you again, Oklahoma Wolf.

You wrote that Seasonic review?
Hold on a second, I'll get some eggnog! Or maybe some NYE hats and noisemakers?



Seriously, I'm grateful that you took the time to stop by and try to help out a Mary Poppins fan.


Thinks OW is GAWD,
Keri


PS. Of course you can sing it BACKWARDS which is: Dosious-Ali-Expi-Listic-Fragi-Cali-repuS, but that's going a bit too far, don't you think?