Frustrated and trying to understand ground.

AcousticBruce

Joined Nov 17, 2008
58
Perhaps the main confusion is about the term "isolation". The common definition for that in consumer applications is that the mains power and neutral (not safety ground) are isolated from the internal electronics, that is, there is no internal path (or voltage) from the mains power to earth ground in the electronics. That is true for many devices such as computers and TVs. The AC safety ground may or may not be connected to the device chassis. If the device has exposed metal it often is (indicated by the power cord having a 3rd prong in 110V U.S. power plugs).

Switching power supplies, such as used in computers, are isolated from the mains by the switching transformer as mentioned in Post #15. Because of the high switching frequency such transformers can be much smaller than line transformers for the same power level.

1:1 isolation transformers are used when you want to more safely work on a line powered device that is not otherwise internally isolated from the mains. It prevents electrocution between the line power voltage and earth ground (safety ground). It does not, of course, prevent electrocution between the power line voltage from the transformer output and circuit common.
Hmm so my hacked supply should be isolated? I did a continuity test from ground of scope and the ground of the ATX power supply and got continuity. So does this mean I have a bad supply?

I figured building one of these would be fun and safe. I would really like the grounds to not be connected. Maybe I need to build a new supply or buy one. I am trying to figure out the best route to study electronics without the ground being an issue. It is messing up my ability to do these projects and see them correctly on the scope screen. I am dealing with mainly under 10 volts.

kubeek

Joined Sep 20, 2005
5,724
Which brings us back to the PC power supply. The output ground is, for safety reasons already mentioned by crutshow, connected to the safety ground of your mains power. This is why you can´t simply connect two 12V pc supplies in series, because you would short one of the supplies.
The same thing happens with the oscilloscope, the scope gnd is connected to the safety earth the same way, so if you try to connect the scope in reverse, with gnd clip on the supply´s +12V and the probe tip to the supply´s ground, you will short the +12V rail through the connected safety earths.

Joined Dec 26, 2010
2,148
Note however that in case of a fault developing, it is possible for such isolation to be lost, and unless some part of the circuit is deliberately earthed this may not at first be obvious. It is however quite common (pun noted) to connect one side of a power supply output to mains ground, possibly as a safety precaution, and/or to enable the use of more effective mains input filtering. You must therefore be very careful about where you put your oscilloscope "ground", or there may indeed be expensive and hazardous fireworks.
@ AcousticBruce: I also tried to tell you about the earth connection, quite a few posts ago, but you did not seem to "get" it. This was part of the reason for my expressing doubt about advising you further.

Joined Dec 26, 2010
2,148
Personally I would just live with the grounded supply, and accept making ground-referenced measurements. Using an oscilloscope was something I did a lot of until fairly recently. Most of the time voltages are measured with respect to common anyhow.

thatoneguy

Joined Feb 19, 2009
6,359
Personally I would just live with the grounded supply, and accept making ground-referenced measurements. Using an oscilloscope was something I did a lot of until fairly recently. Most of the time voltages are measured with respect to common anyhow.
Same here. For a measurement across an RC Circuit, I just use channel A and B, invert channel B and then add the channels together. The single output waveform is the voltage across whatever the two probes are across.

AcousticBruce

Joined Nov 17, 2008
58
@ AcousticBruce: I also tried to tell you about the earth connection, quite a few posts ago, but you did not seem to "get" it. This was part of the reason for my expressing doubt about advising you further
.........................
Personally I would just live with the grounded supply, and accept making ground-referenced measurements. Using an oscilloscope was something I did a lot of until fairly recently. Most of the time voltages are measured with respect to common anyhow.
I didn't get it and still don't because I am new to electronics. I have been asking for a source to learn this for quite some time. Can anyone lead me to a source to learn grounding and safety with oscilloscopes and power supplies? Book, Web Page, Youtube, Course. Whatever I am willing to spend money to learn.

Same here. For a measurement across an RC Circuit, I just use channel A and B, invert channel B and then add the channels together. The single output waveform is the voltage across whatever the two probes are across.
If I buy a new power supply, one that is made for a bench top, will i need to use two channels on the scope to get one reading? Is this issue because I have a hacked power supply?

Is my issue because I have a hacked power supply from an ATX PSU or is it because all power supplies do this?

Wendy

Joined Mar 24, 2008
22,146
As I have said, a standard power supply is isolated. I would have to pin the blame on your hacked PSU. I have no problem with hacked PSU's, but you have to know what your getting.

thatoneguy

Joined Feb 19, 2009
6,359
If I buy a new power supply, one that is made for a bench top, will i need to use two channels on the scope to get one reading? Is this issue because I have a hacked power supply?

Is my issue because I have a hacked power supply from an ATX PSU or is it because all power supplies do this?
Ok, I'll try to put it in a form that Everybody in this thread will agree with, the SMPS special cases, the Microwave oven special cases, battery circuits, and benchtop powered circuits. Everybody should be happy with the following list of problems and why they occur, along with suggested solutions. Not only that, but simple enough for even beginners with a scope can understand it (I hope).

Scope ground problems occur when:
1) Your scope is grounded, the ground clip of the probe is earth ground.
2) Your ATX Power supply black wires/negative are connected to earth ground.
3) Appliances that plug into an outlet will have voltages in them at any point with a potential different from earth ground. In other words, if you clip your scope probe to it, current will flow to ground. There are a few cases where this isn't true, but never gamble. It is best to always make measurements with the black scope clip from Earth ground. This is why the alligator clip on a probe snaps on and off, leave it off when working with AC and all measurements are from ground.

You will NOT have any problems when using your oscilloscope with:
1) Battery powered circuits, there is no current path from the battery to Earth Ground.
2) Automotive uses if you drag your scope outside, the car isn't grounded (unless a tire is removed and the disc brake is embedded into moist dirt)
3) 100% TRANSFORMER ISOLATED CIRCUITS, No Earth Ground present. This includes some AC - DC adapters, but check with a multimeter first!

SOLUTIONS

1) Buy a benchtop power supply with fixed or variable outputs. All quality supplies are not grounded, or have an earth ground option (little tab that goes to a green stud which is connected to earth ground). Read the specs before purchasing to ensure the outputs, both positive and negative or V+/GND are fully isolated from Earth Ground.

2) Only work on battery powered circuits, then you can clip your ground wherever you like and measure like you would with a DMM.

3) When working on AC, make all measurements from Earth Ground, be sure the item being worked on is plugged into a GFCI outlet to prevent electrocution in the event you become a path to earth ground, rather than your scope. Using the A-B probe trick is very effective in these instances.

5) *SUGGESTED* Use a 1:1 120V 1500VA Isolation Transformer between the GFCI outlet and THE DEVICE BEING MEASURED (Not the Oscilloscope, though you can do it that way as well, you run into the reverse problem of having "floating hot scope leads" which is worse). You could buy an isolation transformer for your ATX hacked supply and it would be "floating", so your scope would work on circuits powered by it as easily as if they were connected to a battery. However, isolation transformers are not cheap, so I'd suggest purchasing a benchtop isolated power supply in addition to an isolation transformer for the times when you need to work on 110V Equipment, such as LCD/CRT/Plasma displays that go on the fritz (common occurrence, once you have a scope and know how to use it in addition to a foundation knowledge in electronics, you can fix a LOT of stuff). While getting an isolation transformer, get one of the variable output type, so you can reform caps by bringing up the AC voltage slowly. Again, rarely used, but absolutely required when working on equipment that hasn't seen power in a quarter century.

Always have the scope and equipment under test plugged into GFCI outlets, especially if you are wearing a grounded anti-static wrist strap (I avoid the straps after brushing against one too many improperly wired Non-UL Listed appliances from China). They may get annoying by tripping a lot due to a fault you are attempting to troubleshoot, but they help you avoid acquiring the nickname of "Twitch".

Last edited:

AcousticBruce

Joined Nov 17, 2008
58
Ok, I'll try to put it in a form that Everybody in this thread will agree with, the SMPS special cases, the Microwave oven special cases, battery circuits, and benchtop powered circuits. Everybody should be happy with the following list of problems and why they occur, along with suggested solutions. Not only that, but simple enough for even beginners with a scope can understand it (I hope).

Scope ground problems occur when:
1) Your scope is grounded, the ground clip of the probe is earth ground.
2) Your ATX Power supply black wires/negative are connected to earth ground.
3) Appliances that plug into an outlet will have voltages in them at any point with a potential different from earth ground. In other words, if you clip your scope probe to it, current will flow to ground. There are a few cases where this isn't true, but never gamble. It is best to always make measurements with the black scope clip from Earth ground. This is why the alligator clip on a probe snaps on and off, leave it off when working with AC and all measurements are from ground.

You will NOT have any problems when using your oscilloscope with:
1) Battery powered circuits, there is no current path from the battery to Earth Ground.
2) Automotive uses if you drag your scope outside, the car isn't grounded (unless a tire is removed and the disc brake is embedded into moist dirt)
3) 100% TRANSFORMER ISOLATED CIRCUITS, No Earth Ground present. This includes some AC - DC adapters, but check with a multimeter first!

SOLUTIONS

1) Buy a benchtop power supply with fixed or variable outputs. All quality supplies are not grounded, or have an earth ground option (little tab that goes to a green stud which is connected to earth ground). Read the specs before purchasing to ensure the outputs, both positive and negative or V+/GND are fully isolated from Earth Ground.

2) Only work on battery powered circuits, then you can clip your ground wherever you like and measure like you would with a DMM.

3) When working on AC, make all measurements from Earth Ground, be sure the item being worked on is plugged into a GFCI outlet to prevent electrocution in the event you become a path to earth ground, rather than your scope. Using the A-B probe trick is very effective in these instances.

5) *SUGGESTED* Use a 1:1 120V 1500VA Isolation Transformer between the GFCI outlet and THE DEVICE BEING MEASURED (Not the Oscilloscope, though you can do it that way as well, you run into the reverse problem of having "floating hot scope leads" which is worse). You could buy an isolation transformer for your ATX hacked supply and it would be "floating", so your scope would work on circuits powered by it as easily as if they were connected to a battery. However, isolation transformers are not cheap, so I'd suggest purchasing a benchtop isolated power supply in addition to an isolation transformer for the times when you need to work on 110V Equipment, such as LCD/CRT/Plasma displays that go on the fritz (common occurrence, once you have a scope and know how to use it in addition to a foundation knowledge in electronics, you can fix a LOT of stuff). While getting an isolation transformer, get one of the variable output type, so you can reform caps by bringing up the AC voltage slowly. Again, rarely used, but absolutely required when working on equipment that hasn't seen power in a quarter century.

Always have the scope and equipment under test plugged into GFCI outlets, especially if you are wearing a grounded anti-static wrist strap (I avoid the straps after brushing against one too many improperly wired Non-UL Listed appliances from China). They may get annoying by tripping a lot due to a fault you are attempting to troubleshoot, but they help you avoid acquiring the nickname of "Twitch".

This was an excellent write up. Thank you for taking the time to help me understand more. I created an electronic channel on YouTube and I am going to be posting vids every time I learn something. Safety and bench setup is a video I want to make sometime, so all this knowledge is excellent. I am actually going to install a GFCI now. Thanks. It looks like I might have to buy a new power supply, so I am going to get saving.

In the meantime is it possible to find 1:1 120v transformers in equipment new or old? Like a TV or a Microwave or something like that?

thatoneguy

Joined Feb 19, 2009
6,359
There's a guy selling some 1kVA 1:1 toroidal isolation transformers for $150 each, you'd need to provide the enclosure, but that's a great deal. They weigh 26lbs, so probably add$50 for shipping as well.

Keep an eye on eBay for name brand types being sold by somebody getting divorced, and the medical grade ones typically add $200 to the price just to get that stamp of approval. So, expect to spend$200-$300 on one, not much you'll be working on will be drawing over 10 amps, so a 750VA or even 500VA would do for things like LCD televisions, allowing you to get into the sub$200 range. Do not skimp on the bench power supply. You really don't need much current. I have an isolated 12V/30Amp supply for testing amps and such, but those aren't cheap.

If you can only afford one (Power supply or isolation transformer), get a good isolated variable power supply, no question about it. You want current limiting and preferably gauges, so $189 for a variable 5A 0-30V with current limiting and isolation as Bill showed above would do for now. If you don't plan on working with AC items such as TVs or motor speed controls, the isolation transformer can wait for a while, and your power supply can run the rest of the things you'd normally be running from a line adapter. If forced to work with AC, just remember the A-B trick if you want to check a Diac or something. sheldons Joined Oct 26, 2011 613 another item i still use now when repairing tvs etc is a variac-where you can bring up the supply to the item being tested during a repair whilst taking voltage/current measurements which is a big help sometimes Thread Starter AcousticBruce Joined Nov 17, 2008 58 Well I cant imagine working on tv's for a while. Right now I just want to do lab work and make small circuits with 5v or so, AC and DC. Im sure even 3A is enough. But what if I want to do this? http://www.youtube.com/watch?v=fxENMvspTsI Haha JK. I didn't realize that it was going to cost so much for this stuff. I will have to do the A+B trick for a bit until i run into a deal with the piling cash I am saving. I think current limiting would be an absolute necessity and I should probably buy a isolated power supply ASAP. This is getting a little more expensive than i thought though. So I will work with what I have for a bit. Jotto Joined Apr 1, 2011 151 Think the first place to start is expressing that there are two types of grounds. Earth ground and common ground. SMPS do use both. I am using this example because its something I work with every day. If your troubleshooting a SMPS, you need to know the difference between the two grounds to be able to troubleshoot it properly. If your looking at the primary side of the PS you will need to use earth ground, the secondary is just common ground. Yes there are some that are a bit different, as stated before. Some units such as CRT units might have 4 or 5 SMPS in the unit. This is important for the reason of using a load tester, if you don't use the proper ground you will start troubleshooting a problem that doesn't exist. Earth ground will be obtained through the mains or the largest cap on the SMPS, common ground will just be a screw or chassis metal of the unit being tested. Earth ground is isolated from the secondary of the PS. thatoneguy Joined Feb 19, 2009 6,359 Well I cant imagine working on tv's for a while. Right now I just want to do lab work and make small circuits with 5v or so, AC and DC. Im sure even 3A is enough. But what if I want to do this? http://www.youtube.com/watch?v=fxENMvspTsI Haha JK. I didn't realize that it was going to cost so much for this stuff. I will have to do the A+B trick for a bit until i run into a deal with the piling cash I am saving. I think current limiting would be an absolute necessity and I should probably buy a isolated power supply ASAP. This is getting a little more expensive than i thought though. So I will work with what I have for a bit. It is a bit expensive to do it correctly, but it is a one time purchase. I have power supplies, signal generators, etc that have been with me a long time, and saved me hundreds of hours of time. Most "hobby stuff" will be run off 5V, 9V, or 12V anyway, so if you get a$30 switchable 2A switching supply at radio shack, you are covered, I THINK they are isolated, if not, you can get your money back.

Otherwise, just power circuits with batteries 3AA (4.5V) will run 5V logic no problem.

AcousticBruce

Joined Nov 17, 2008
58
Well I did drop 412 bucks on this.

So I think I should probably have a good Power supply also.

ThatOneGuy: I have a 9v 1 amp transformer from a wall-wart. I am going to build a variable regulated PS from it... why not? This should fix all this trouble. Its a shame I went through all that process making a beautiful hacked PS and am not going to use it for a while.

Jotto: That is so true because I have heard them both but never really understood. I am starting to grasp it.... getting there.

thatoneguy

Joined Feb 19, 2009
6,359
Well I did drop 412 bucks on this.

So I think I should probably have a good Power supply also.
At 3:15 in the video is what a power supply with optional ground looks like, so if you get one, you know.

The Green terminal is Earth Ground. The black terminal is Negative, the red is positive.

There is a little hook strap that goes between the green terminal and black to make the black also Earth Ground - Leave this disconnected, and don't use the green terminal for your project (don't hook up tab to black, don't hook power wire to it, etc, otherwise it will be just like using your ATX supply, with black being earth ground, and red being V+

AcousticBruce

Joined Nov 17, 2008
58
Oh thanks for that at 3:15. What you said about earth and negative makes me wonder this.

If I hook to +5 and +12 wouldn't that give me 7V that is not hooked to earth?

It is ground +3.3 -5 +5 -12 +12 in that order.

thatoneguy

Joined Feb 19, 2009
6,359
Yes, that would be 7 volts "floating", but I'd still put a 10k resistor on your scope probe ground clip, through your DMM in current mode, and ensure that no current passes through the ground probe clip when touching either output of your 7V (or ground, for that matter). You can actually do this for all your outputs so you know which ones are isolated, so you can get 9V with 3 and 12, etc. It should be zero. This is assuming you are using one red terminal as "-" (circuit ground) and one red terminal as "+" (V+ for circuit).

Remember, the maximum current it the lowest of the two, so if 5V is 10A and 12V is 30A, than 10A is the most you can draw at 7V and expect any sort of stability.

AcousticBruce

Joined Nov 17, 2008
58
Yes, that would be 7 volts "floating", but I'd still put a 10k resistor on your scope probe ground clip, through your DMM in current mode, and ensure that no current passes through the ground probe clip when touching either output of your 7V (or ground, for that matter). You can actually do this for all your outputs so you know which ones are isolated, so you can get 9V with 3 and 12, etc. It should be zero. This is assuming you are using one red terminal as "-" (circuit ground) and one red terminal as "+" (V+ for circuit).

Remember, the maximum current it the lowest of the two, so if 5V is 10A and 12V is 30A, than 10A is the most you can draw at 7V and expect any sort of stability.

This is good news. I am looking forward to testing this. I am wondering how you came up with a 10k resistor? Was this just a random value that will slow current enough to not damage anything?

AcousticBruce

Joined Nov 17, 2008
58
I am not getting much current at 10kΩ. So I tried 1k (990Ω) and this is the mA readings I am getting from scope ground to ATX PS.

earth - 0mA (haha that's a good thing)
3.3 - 3.26mA
-5 - -4.5mA
+5 - 5.17mA
-12 - -11.2mA
+12 - 12.17mA

Also I am getting continuity between the scope ground and the metal chassis of the ATX PS. Isnt this dangerous?

I get continuity from the Ground (green) and metal chassis. 300 mΩ
I also get continuity from 3.3v and metal chassis (10 Ω).

I guess this isnt good at all.