Oscilloscope

[Th3_uN1Qu3]

Re: Oscilloscope

Because in the primary the "ground" reference is the negative side of the filter cap(s), which is rectified mains. If you connect the ground of the scope to the negative of the filter caps, you make a direct short across the mains (hot to earth) and you blow the power supply's fuse.

[godin]

Re: Oscilloscope

Thanks, I guess the safe thing to do is always check to see if there is a voltage between what you are about to clip to and the ground clip!

[PCBONEZ]

Re: Oscilloscope

Without isolation you CANNOT probe in the primary of a SMPS with an earthed chassis scope - only with one powered by batteries and without a metallic frame.

That a regurgitation of the same old 'ancient knowledge' -what- that's been repeated since at least the '50's.
[You back when much equipment and even most buildings didn't have grounds...]
.
As I already expressed I can't think of a single -why- for that -what- when it comes to modern grounded equipment.

.. And it seems neither can anyone else. - Including Tektronix..
.
Please explain your -why-.
.

[godin]

Re: Oscilloscope

I must admit I am accepting it at the moment but not truly understanding. I will have to think more. Having looked at an smps schematic, the bridge rectifier is before the transformer which is not something I have encountered before.

[PCBONEZ]

Re: Oscilloscope

I must admit I am accepting it at the moment but not truly understanding. I will have to think more. Having looked at an smps schematic, the bridge rectifier is before the transformer which is not something I have encountered before.

Basic SMPS block diagram flow path.

Input filter
Rectifier AC-DC [High Volts]
Switching Section [makes HV AC at Higher Frequency]
Step-Down Transformer
Rectifiers AC-DC [OP Volts]
Output Filters

In modern gear there is a ground in just about every stage.
.

[PCBONEZ]

Re: Oscilloscope

Because in the primary the "ground" reference is the negative side of the filter cap(s), which is rectified mains. If you connect the ground of the scope to the negative of the filter caps, you make a direct short across the mains (hot to earth) and you blow the power supply's fuse.

And adding an additional transformer on the AC side of the rectifier changes that - how?
.

[Th3_uN1Qu3]

Re: Oscilloscope

Adding a transformer means there is no more direct path to the mains hot and neutral, the outputs of the transformer are floating and you can connect whatever side you want to earth ground. Wire the negative side of the primary filter caps to earth ground without an isolation transformer and tell me what happens.

[godin]

Re: Oscilloscope

Adding a transformer means there is no more direct path to the mains hot and neutral, the outputs of the transformer are floating and you can connect whatever side you want to earth ground. Wire the negative side of the primary filter caps to earth ground without an isolation transformer and tell me what happens.

Do you mean an isolation transformer that floats the ground too and just leaves a floating AC with no ground connection to the neutral?

[PCBONEZ]

Re: Oscilloscope

Adding a transformer means there is no more direct path to the mains hot and neutral, the outputs of the transformer are floating and you can connect whatever side you want to earth ground. Wire the negative side of the primary filter caps to earth ground without an isolation transformer and tell me what happens.

.
Thanks for making me think but I'm still not buying the story.
- If I have a GCE somewhere in my thinking then point it out.
Here goes...
.
Wire either side of the Iso-Xfmr secondary to ground and tell me what happens.
You seem to be trying to tell me it magically isn't going to pass AC current to ground.
I'm thinkin' it will.
.
Your whole theory is based on two misconceptions.
#1 - One side of a rectifier isn't -always- 'at ground'.
#2 - A negative voltage -IS- a 'HOT'. - If it will shock you, it's a HOT.
.
.
Looks to me like:
"Wire the negative side of the primary filter caps to earth ground"
is going be the exactly same in a modern SMPS regardless of an Iso-Xfmr.
.
There is a chassis/earth ground between CY1 and CY2 before the rectifier so adding a transformer is not going to change anything at all.
All the Iso-Xfmr seems to be doing is forcing the neutral before the rectifier to be below ground potential. If anything that's MORE dangerous.
.
Either way [Iso-Xfmr or not] the DC negative side of that rectifier is below earth ground potential. [A negative voltage to ground is STILL a 'HOT'.]
- The CY caps arrangement sets the ground reference and ensures that.
Either way [Iso-Xfmr or not] connecting ground to either side of the rectifier it will cause a short because BOTH sides of the rectifier are HOT relative to earth ground.
[BOTH sides of the rectifier are HOT. - WHY would anyone connect a ground straight to a HOT?]
.
.
Avoiding the problem you think is there is easy.
Always connect the scope ground leads to earth ground.
If you need to check a + HOT to a - HOT [or any two HOTs + or - ] then use two leads and use 'add'. [Invert as necessary.]
It's not that hard.
Since it's DC in this case using the right tool [a DMM] would be even better.

That's how I was taught and probably why I couldn't see where you thought there was a problem.
I never connect a scope ground to a HOT [negative or positive] in the first place so the problem is a non-issue.
.

[Th3_uN1Qu3]

Re: Oscilloscope

Do you mean an isolation transformer that floats the ground too and just leaves a floating AC with no ground connection to the neutral?

Yes.

Always connect the scope ground leads to earth ground.
If you need to check a + HOT to a - HOT [or any two HOTs + or - ] then use two leads and use 'add'. [Invert as necessary.]
It's not that hard.
Since it's DC in this case using the right tool [a DMM] would be even better.

I didn't have a dual channel scope when i needed to check the primary. And i'm talking about seeing the switching waveform to the primary transistors - anywhere you put the ground clip in the primary, it's still "hot" re earth without an isolation transformer. Anyway, i do not own an isolation transformer so i just used an inverter and battery. Worked fine. As far as the safety aspect goes you have to be careful messing around in the primary anyway, but as long as the scope remains earthed there is no danger in touching the case of the scope... you'd have no business touching anything in the primary of the SMPS anyway.

[Peter9DO]

Re: Oscilloscope

I'm not going to lie the last few post between Th3 uN1Qu3 and PCBONEZ is great reading and entertaining but I'm so very very confused. I'm so confused I'm not even going to ask for an explanation in softer terms. But can you answer this for me please? Are we using the grounding clip or not and are we 86ing the isolation transformer or are we going to say "it's a good investment"?

[joshnz]

Re: Oscilloscope

I too am confused.
I am going to give a situation please provide answer.
1. Have only a single channel scope and need to measure waveform on primary side.

[PCBONEZ]

Re: Oscilloscope

I'm not going to lie the last few post between Th3 uN1Qu3 and PCBONEZ is great reading and entertaining but I'm so very very confused. I'm so confused I'm not even going to ask for an explanation in softer terms. But can you answer this for me please? Are we using the grounding clip or not and are we 86ing the isolation transformer or are we going to say "it's a good investment"?

With your 2 Channel scope you should never need one.
If you ever have to measure a HOT to a HOT use two probes both with grounds to earth ground.
Both + ADD with no Invert on CH2 will display the total.
Both + ADD with Invert on CH2 will display the difference.
.

[PCBONEZ]

Re: Oscilloscope

I too am confused.
I am going to give a situation please provide answer.
1. Have only a single channel scope and need to measure waveform on primary side.

Depends on which waveform you are looking at.
The PWM probably references to earth ground.
The OP of the switchers could probably be measured on the secondary side just as well - and do the math for the xfmr.
.
Beyond that you'll have to look at what you're working on in greater detail because they aren't all as 'ungrounded' as people seem to think they -all- are.
They can all be different. [See Pic].
The one in the Pic shows an earth ground right where Th3_uN1Qu3 says connecting a ground would blow it up.
That doesn't mean he's wrong for all PSUs though. - Or even most..
There's just too many different designs to generalize something like that one way or the other.
.
Worst case is you measure both sides to ground and do the math.
.

[godin]

Re: Oscilloscope

Me too, confused as hell. I can't see the problem with the smps but bow to greater knowledge.
How come if it is earth and not possible to clip to is it not a shock hazard on the casing? Thinking specifically of a PC power supply.
If you connect your scope clip to ground and not anything above ground what can be the problem. I am missing something!

If a filter cap is passing AC to ground then ground must be ground.... mustn't it?
If you float with an isolation transformer and float the earth then I can understand that there would be a voltage present that was not being returned and would use the scope probe earth as a route but not a short.

In PCBONEZ schematic, if the earth were removed from the emi filter between the 2 caps by floating/removing the earth connection as suggested, then the casing and ground of the equipment would become live as the caps become a voltage divider and if you connected a clip to the casing it would be above ground. The whole problem is caused by removing the earth connection.

Hope to have my confusion fixed by the knowledgeable

[joshnz]

Re: Oscilloscope

PCBONEZ I don't know how that psu even works if those grounds are in fact tied together half of the cycle is shorted.
I don't understand how this psu works.
unless L1 is the key.

[PCBONEZ]

Re: Oscilloscope

PCBONEZ I don't know how that psu even works if those grounds are in fact tied together half of the cycle is shorted.
I don't understand how this psu works.
unless L1 is the key.

It's DC at that point.
.
I don't get that one either.
The point was they aren't all the same and you have to look to see what's going on.
.

[godin]

Re: Oscilloscope

It's DC at that point.
DC doesn't cycle.
.

.....and you should be able to connect the ground clip to the negative grounded DC leg of that rectifier output. Shouldn't you??

[Th3_uN1Qu3]

Re: Oscilloscope

No, you can't. I've already shown above why.

Anyway to sum it up - if you're serious about troubleshooting SMPS you need a two channel oscilloscope.

[PCBONEZ]

Re: Oscilloscope

Me too, confused as hell. I can't see the problem with the smps but bow to greater knowledge.
How come if it is earth and not possible to clip to is it not a shock hazard on the casing? Thinking specifically of a PC power supply.
If you connect your scope clip to ground and not anything above ground what can be the problem. I am missing something!

If a filter cap is passing AC to ground then ground must be ground.... mustn't it?
If you float with an isolation transformer and float the earth then I can understand that there would be a voltage present that was not being returned and would use the scope probe earth as a route but not a short.

In PCBONEZ schematic, if the earth were removed from the emi filter between the 2 caps by floating/removing the earth connection as suggested, then the casing and ground of the equipment would become live as the caps become a voltage divider and if you connected a clip to the casing it would be above ground. The whole problem is caused by removing the earth connection.

Hope to have my confusion fixed by the knowledgeable

Some SMPS float the section between the first rectifier and the power transformer.
.
Before that section is grounded and after that section is grounded but that section itself is not.
.
What I'm saying is using an Iso-Xfmr to float the sections before and after -too- doesn't help anything or even make sense to me.
.