New DS2102 Smoking!!

[Helipil0t]

Ok... So I'll be the first to admit I'm still a bit of a newbie when it comes to electronics.  I've been building the odd project here and there and am slowly gaining some confidence.  I built an ELENCO XP-720 Power Supply and it works great.  So now comes the time to invest in an Oscilloscope.  After doing tones of research I decided to get a Rigol DS2102.  (Yes I know the 70mhz version can be hacked but they were all sold out).  So I get this beautiful scope in the mail and decide to play with it.  I create a simple AC/DC bridge rectifier to probe.  I connected the ground clip of one probe to the ground in my rectifier.  The probe itself I clipped to the positive output of the rectifier(by the resistor).  As a base of comparison I decided to hook up my second probe.  So ground clip to common ground and main probe to Positive of the DC side of my ELENCO.  After a few seconds a saw smoke coming from the ground clip of the number 1 probe right by the diode (smoke could of been the from diode too).  Of course I unhooked everything quickly and killed the power.  The ground clip was really hot. 

I've attached a diagram of the ELENCO and in the hatched area is my created rectifier.  I've illustrated where I placed my probes.  I did watch Dave's "How not to blow up your oscilloscope video." And I was under the impression that as long as what your probing is floating, all should be well.  Well the power supply is floating as it's on the other side of a transformer that isn't grounded to mains earth.  I'm sure it's a rookie mistake but can someone point out what that mistake is and help point me in the right direction?    I would be forever grateful.

First post on the forum by the way.   Love this blog! 
Cheers!

[tautech]

When using 2 probes the GND's for each have to be at the same potential.
Often you use only one GND as a reference for both probes.

The only way that you could have done that test is with an isolated channel scope.(that is GND's are isolated from each other)

[gaijin]

As tautech said the channel ground for most oscilloscopes are connected.
You shorted both 6.3VAC windings though the bridge rectifier from channel 1's ground to channels 2's ground.

[Monkeh]

I did watch Dave's "How not to blow up your oscilloscope video." And I was under the impression that as long as what your probing is floating, all should be well.  Well the power supply is floating as it's on the other side of a transformer that isn't grounded to mains earth.

It's not floating once you connect it to ground with your scope leads, now is it?

[IanB]

I'm sure it's a rookie mistake but can someone point out what that mistake is and help point me in the right direction?    I would be forever grateful.

Yes, it's a rookie mistake. An oscilloscope is not a multimeter. You can't attach the probes freely to any part of a circuit.

In this case the problem is the ground clips as others have indicated. All ground clips are connected to mains earth, therefore all ground clips are also connected to each other. The path between two ground clips is a short circuit.

[max666]

It's not floating once you connect it to ground with your scope leads, now is it?

That's not the point. The point is, that everthing where you intend to connect a probe GND clip to, must be floating befor you connect GND. And you can't measure anything meaningful, until you "unfloat" the circuit you're measuring (or you have to use a differential probe).

@Helipil0t: just draw a connection from your 'Probe 1 GND Clip'-Point to the 'Probe 2 GND Clip'-Point in your scematic and you should easily be able to see how you shorted out one half wave of the 6.3 VAC windings through the bridge rectifier.

[Helipil0t]

Yes, it's a rookie mistake. An oscilloscope is not a multimeter. You can't attach the probes freely to any part of a circuit.

In this case the problem is the ground clips as others have indicated. All ground clips are connected to mains earth, therefore all ground clips are also connected to each other. The path between two ground clips is a short circuit.

I should of thought of that.  Seriously... not to think of both probe grounds as having continuity was pretty stupid of me.  So you can only measure two signals in reference to the same potential.  That kind of sucks.  Seems a bit restrictive. 

max666:  I had a quick look at differential probes out of curiosity.  The price tag on those are enough to make my head spin!!    Probes worth more than the scope!  Gonna just learn how to use what I have and be happy with it. 

Thanks to everyone for the quick responses.  I appreciate the help.   

[tautech]

I should of thought of that.  Seriously... not to think of both probe grounds as having continuity was pretty stupid of me.  So you can only measure two signals in reference to the same potential.  That kind of sucks.  Seems a bit restrictive. 

This is how the vast majority of scopes are constructed.
Use only one GND lead for 2 probe measurements, and you shouldn't get caught again.
If in doubt use a meter first to check GND potential

Get used to it and learn to live with it unless you have deep pockets.

[PetaVolt]

Dave's video explains this quite nicely:

[Monkeh]

It's not floating once you connect it to ground with your scope leads, now is it?

That's not the point. The point is, that everthing where you intend to connect a probe GND clip to, must be floating befor you connect GND. And you can't measure anything meaningful, until you "unfloat" the circuit you're measuring (or you have to use a differential probe).

Yes, that is exactly the point.

Circuit floating. Attach ground lead: Circuit NOT floating. Now you get to think carefully about the consequences of this change.

[SL4P]

The good news is that you probably didn't do any damage to the CRO
Before you do much more = check as well as you can that BOTH probes ar fully intact - shield and probe pin. And the two scope inputs.
A good lesson learned - hopefully without any real damage

[Mosaic]

Consider investing in an isolation transformer for the scope and  you can create a function diff probe using a two channel scope  and summing ch2 (inverted) from ch1. No grounds used there.

[vk6zgo]

Consider investing in an isolation transformer for the scope and  you can create a function diff probe using a two channel scope  and summing ch2 (inverted) from ch1. No grounds used there.

You don't need to isolate the 'scope to use the two probes in that way.

All each channel is reading is the voltage w.r.t ground at that point.
The relationship between two such voltages will show the voltage between them.

With a 2 channel 'scope,it does allow you to test two points which are above "ground",but it won't allow you to do what the OP wanted to do.

With more than 2 channels & three probes,you can connect up two probes in the manner you describe,to look at one point,& use the third probe and channel to look at the other.

[tautech]

Consider investing in an isolation transformer for the scope and  you can create a function diff probe using a two channel scope  and summing ch2 (inverted) from ch1. No grounds used there.

You don't need to isolate the 'scope to use the two probes in that way.

All each channel is reading is the voltage w.r.t ground at that point.
The relationship between two such voltages will show the voltage between them.

With a 2 channel 'scope,it does allow you to test two points which are above "ground",but it won't allow you to do what the OP wanted to do.

With more than 2 channels & three probes,you can connect up two probes in the manner you describe,to look at one point,& use the third probe and channel to look at the other.

A very sound argument for the 4 channel scope.
Invest in a Differential probe/probes or invest in an extra 2 channels and have added functionality as well.

[mcinque]

A lesson that I've learned with some sparks and my DSO is "ON A NON FLOATING CIRCUIT, NEVER CONNECT THE GROUND PROBE TO ANYTHING THAT IS NOT GROUND" 

[tautech]

A lesson that I've learned with some sparks and my DSO is "ON A NON FLOATING CIRCUIT, NEVER CONNECT THE GROUND PROBE TO ANYTHING THAT IS NOT GROUND" 

Yes most of us have done it. 
And we all have a meter that could do a quick check too!

[idpromnut]

Or another way of looking at it is, verify that the signal that you are expecting is there (DMM) and then go in with a 'scope and inspect it. 

[vk6zgo]

Or another way of looking at it is, verify that the signal that you are expecting is there (DMM) and then go in with a 'scope and inspect it. 

In the OP's case,I agree that would have been the best procedure.
In many circuits,though,the DMM may see either nothing,or a deceptively low voltage.
It would,however show the presence of a DC or low frequency AC voltage between two points.

In practice,measurements w.r.t. a common or "Ground" point (not always the same,as some circuitry does not connect the common to either Mains Earth,or any metalwork) are adequate for most work.
In the unusual case of needing to display the voltage between two points which are both above Ground potential,the method described by tautech is usually adequate.

[macboy]

...
Use only one GND lead for 2 probe measurements, and you shouldn't get caught again.

That may work OK for DC and very slow (kHz) signals, but for high speed signals, you can't get away with that. The divider in the probe needs to be grounded to a low impedance ground in circuit, or the signal will be shot.