How to test a switch mode power supply with an analog oscilloscope

[tautech]

Never is this method safe and should never be suggested on a public forum.
Users with limited knowledge of scope use and without the appropriate probes can kill themselves, I will not have that on my conscience. Get the proper equipment or give the repair to somebody that knows what they are doing.

Point taken, note added to the post

I have not find in any book or user's manual th indication to use an isolation transformer. It is very hard to find isolations transformers on the market, there must be a reason for that.

The safest method for measuring/testing elevated non ground referenced voltages is with a Differential probe.
Get one or let someone else do the repair that knows what they're doing. 

[james_s]

Isolation transformers are readily available, I would not attempt to service a SMPS or anything with a hot chassis without one. If you can't find one ready made, you can wire two identical transformers back to back.

[ataradov]

I have not found in any book or user's manual the indication to use an isolation transformer.

What books have you read on the topic?

It is very hard to find isolation transformers on the market, there must be a reason for that.

What are you talking about? Here is the first one out of a dozen appearing on a first page of amazon search: https://www.amazon.com/Tripp-Lite-Isolation-Transformer-Outlet/dp/B00006HPFH/ref=sr_1_2?ie=UTF8&qid=1488935050&sr=8-2&keywords=isolation+transformer

[j57H8I4]

Isolation transformers are readily available, I would not attempt to service a SMPS or anything with a hot chassis without one. If you can't find one ready made, you can wire two identical transformers back to back.

That is what all people say on forums and videos, but, as I already said, I have not found that indication in any book, nor I have ever seen an isolation transformer manufactured by a primary electronic equipment producer.

[ataradov]

I have ever seen an isolation transformer manufactured by a primary electronic equipment producer.

What is "primary electronic equipment producer"? Tripp Lite is a fairly well known brand for this kind of stuff.

You are talking about specialty tools. You won't find one make by Sony or Samsung.

[j57H8I4]

What are you talking about? Here is the first one out of a dozen appearing on a first page of amazon search: https://www.amazon.com/Tripp-Lite-Isolation-Transformer-Outlet/dp/B00006HPFH/ref=sr_1_2?ie=UTF8&qid=1488935050&sr=8-2&keywords=isolation+transformer

Tell me a book or an oscilloscope user's manual where it is indicated to use an isolation transformer.

[ataradov]

Tell me a book or an oscilloscope user's manual where it is indicated to use an isolation transformer.

It is called common sense. This is not specifically oscilloscope-related. It is about general safety rules when working with high energy circuits.

Show me a book that tells that it is a bad idea to shove fingers into the wall outlet. The same book will be applicable here.

Also, if you want to kill your scope, or yourself - go right ahead. There is no real point arguing here.

[j57H8I4]

What is "primary electronic equipment producer"? Tripp Lite is a fairly well known brand for this kind of stuff.

You are talking about specialty tools. You won't find one make by Sony or Samsung.

Have you ever seen an isolation transformer produced by Tektronix, or Rigol, or Keysight, etcetera?

[ataradov]

Have you ever seen an isolation transformer produced by Tektronix, or Rigol, or Keysight, etcetera?

Why would they? You do not isolate the scope. You isolate the device under test.

Also, isolated inputs is a feature of high-end scopes with a huge price tag.

[ataradov]

I'm sure it was posted in one of your threads already. But I'll repeat. Watch this:

[j57H8I4]

Why would they? You do not isolate the scope. You isolate the device under test.

Also, isolated inputs is a feature of high-end scopes with a huge price tag.

If the use of an isolation transformer was so important you should find as many isolation transformer on the market as many oscilloscopes, because for each oscilloscope sold also an isolation transformer should be sold. Instead you can very hardly find isolation transformers on the market and you find them on eBay, not on primary electronic dealers shops.

[ataradov]

because for each oscilloscope sold also an isolation transformer should be sold.

This is not true. Not everyone works with non-insulated mains-powered devices. Furthermore, people purposefully design things to be isolated, if possible. This is a consumer safety requirement as well. But sometimes it is unavoidable.

For example, if you work with digital circuits and power them from your Korad power supply, you are fine without any isolation transformer. Your power supply acts as one.

But if you want to look at the guts of that power supply, then you need one.

[james_s]

Most people are not using scopes to poke around in line powered switchmode power supplies, and those that are have an isolation transformer. The transformer isolates the DUT, *not* the scope. The scope should *always* be earth grounded.

Walk into any TV repair shop around, if such a place even exists anymore and I guarantee there will be an isolation transformer on the bench. Any lab where switching power supplies are designed will have one, we certainly had one in the lab at the last place I worked that did hardware. It is a mandatory safety item, it protects you as well as your test equipment.

[james_s]

It is called common sense. This is not specifically oscilloscope-related. It is about general safety rules when working with high energy circuits.

Show me a boot that tells that it is a bad ides to shove fingers into the wall outlet. The same book will be applicable here.

Also, if you want to kill your scope, or yourself - go right ahead. There is no real point arguing here.

Heh that makes me visualize a light socket with a warning label that says "Do not insert tongue"

I shouldn't even joke about it, people expect to be warned and protected from everything these days, common sense is not so common.

[djnz]

Rather than risking you life, why don't you ask all the questions you have about isolation transformers here on the forum? I am sure people will be willing to help!

Lots of nasty outcomes are possible here - from loss of life to causing a fire to causing damage to your oscilloscope or the power supply. Why not take the time and learn to do this safely? This is also about habits and developing some respect for electricity. Perhaps you will be lucky a few times and nothing so bad will happen. But if you do not develop the attitude of safety, some day your judgment and your luck might both fail  and cause real damage.

There is a third option apart from an isolation transformer and differential probes, which is using an isolated channel oscilloscope. All isolated channel scopes I know of are battery powered, but not all battery powered scopes have isolated channels, so be careful. I bought a used Tek THS720A  recently, was around US$200.  Even with that, all the standard precautions like probe de-rating with frequency apply.

Good luck.

[james_s]

Even if I had a fully isolated scope with a differential probe I'd still use an isolation transformer. One slip of the finger and you could end up with 340V with lots of current behind it going through your body. Working on non-isolated mains powered gear without an isolation transformer is a deathwish.

[Ian.M]

Unfortunately the safety provided by an isolation transformer vanishes the moment you connect a non-isolated scope ground clip to the primary side of the SMPSU under test, and worse, it prevents any upstream RCD/GFCI tripping.  To retain the safety advantage, either use a fully differential HV probe, or an isolated input scope or a CAT II or better battery powered scopemeter, and beware of sneak grounds introduced by other testgear, etc.

Otherwise its live working rules:

• Start with a clear and tidy workbench and make sure there is nothing grounded where you may accidentally come into contact with it.
  .
• Remove all rings, bracelets, metal watches, phone or music player headsets etc.
  .
• Keep one hand in your pocket to avoid shocks across your chest.
  .
• Test before touching!
  .
• Don't hold the probe in place with your hand.
  .
• When ever possible switch off before moving the probe.
  .
• Use a large heavy insulating mat, checked for damage, to isolate yourself from ground.
  .
• Do not work alone.
  .
• Do not work if tired, distracted or intoxicated.
  .
• Do not allow pets or young children in the room.
  .
• Have a clearly marked isolating switch to cut power to the bench.
  .
• In case of electrocution, make sure that all personnel are trained to cut the power before touching the casualty.
  

That's how techs in the TV repair trade worked without killing themselves before GFCIs became common.

[bitwelder]

From the old Sam Goldwasser's Sci.Electronics.Repair FAQ:
https://www.repairfaq.org/sam/smpsfaq.htm#smpsint

(the whole chapter is worth reading, but below is the quote about isolation transformers)

Incredibly handy widgets
...
Isolation transformer. This is very important for safely working on live chassis equipment like line connected switchmode power supplies (primary side). You can build one from a pair of similar power transformers back-to-back (with their highest rated secondaries connected together. I built mine from a couple of similar old tube type TV power transformers mounted on a board with an outlet box including a fuse. Their secondary high voltage windings were connected together to couple the two transformers together. The unused low voltage windings can be put in series with the primary or output windings to adjust voltage. Alternatively, commercial line isolation transformers suitable for TV troubleshooting are available for less than $100 - well worth every penny.

There is absolutely no imaginable reason not to use an isolation transformer for troubleshooting SMPSs except possibly for the final test where confirmation is needed that the inrush from a direct line connection (which will have virtually unlimited instantaneous current capability) will not damage the newly repaired supply.

[dav]

Isolation transformers are harder to find in EU rather than in US.
If you want a 220V->220V fully assembled one, you can take a look to this:
https://www.reichelt.de/Labornetzgeraete/PEAKTECH-2240/3/index.html?ACTION=3&GROUPID=4952&ARTICLE=101937

[vk6zgo]

Why would they? You do not isolate the scope. You isolate the device under test.

Also, isolated inputs is a feature of high-end scopes with a huge price tag.

If the use of an isolation transformer was so important you should find as many isolation transformer on the market as many oscilloscopes, because for each oscilloscope sold also an isolation transformer should be sold. Instead you can very hardly find isolation transformers on the market and you find them on eBay, not on primary electronic dealers shops.

When I worked at a TV Studio,we had dozens of Oscilloscopes,& about 3 isolation transformers.

Most of the time the 'scopes were used on equipment which did not have power supply faults.
For all other circuitry, the Switchmode Power Supply provided isolation, as it is designed to do.

As I fixed a lot of Picture Monitors, where SMPS faults were not uncommon, I had one semi-permanently on my bench.

Isolation transformers have no moving parts & last for years, so they are pretty much "niche market " items, just as desoldering stations were for many years.
For this reason, they are rarely found in Electronics stores like Jaycar, or Altronics in Australia, (& equivalents in other countries).

As for your previous comment, it is fairly unusual for such firms as Tektronix, Sony, etc, to manufacture devices like this, even for internal use in the factory, instead, they buy them from companies who specialise in such equipment.
They use forklifts, too, but with the possible exception of Hitachi, don't have "own brand" ones.

[vk6zgo]

Unfortunately the safety provided by an isolation transformer vanishes the moment you connect a non-isolated scope ground clip to the primary side of the SMPSU under test, and worse, it prevents any upstream RCD/GFCI tripping.

In fairness, the hazard of doing this is hardly different to testing the high voltage secondary circuit of a tube type device using a conventional transformer type supply.

Many thousands of such circuits were probed with normal Tektronix or other probes over tens of years, with vanishingly small numbers of injuries.

One thing,though, is the possibilty of the SMPS having a "common" side which takes up a larger part of the board.
It would obviously be wise to place the earth clip on that side,as the "hot bits" would be smaller & harder to touch. ( Of course, this when using an isolation transformer---just in case some " newbies" missed that bit)

  To retain the safety advantage, either use a fully differential HV probe, or an isolated input scope or a CAT II or better battery powered scopemeter, and beware of sneak grounds introduced by other testgear, etc.

Differential probes were, until comparatively recently, too expensive, so Techs had to rely upon a "functioning brain".

Even now, such probes are far from cheap.

Otherwise its live working rules:

• Start with a clear and tidy workbench and make sure there is nothing grounded where you may accidentally come into contact with it.

Really good idea,& hobbyists should embrace it,but realistically, this was pretty much "honoured in the breach" much of the time in workshops.

.[/li]
[li]Remove all rings, bracelets, metal watches, phone or music player headsets etc.
.[/li][/list]

This becomes more important in the current day,where people are wont to festoon themselves with all sorts of crap.

• Keep one hand in your pocket to avoid shocks across your chest.
  .
• Test before touching!
  .
• Don't hold the probe in place with your hand.
  .
• When ever possible switch off before moving the probe.
  .

All good!

• Use a large heavy insulating mat, checked for damage, to isolate yourself from ground.
  .

I've seen this done a few times, but  quite rarely----it sounds more like what you would do if testing heavy industrial switchboards.

• Do not work alone.
  .

  Very few TVs would have ever been fixed.

This is something we would have adhered to if getting into the power supply side of TV or Broadcast Transmitters, but small stuff like Monitors,TVs, etc were regarded as significantly less hazardous.

• Do not work if tired, distracted or intoxicated.
  .

  Intoxicated is certainly a "no-no",but  most Techs are tired most of the time---it's the nature of the job!

• Do not allow pets or young children in the room.
  .
• Have a clearly marked isolating switch to cut power to the bench.
  .
• In case of electrocution, make sure that all personnel are trained to cut the power before touching the casualty.
  

Yes!,Yes!,& Yes!

That's how techs in the TV repair trade worked without killing themselves before GFCIs became common.

[Ian.M]

The rubber mat thing is less important if you are on lino over a wooden floor, but if you are on concrete its strongly recommended.  Carpet normally insulates reasonably well but that can change if its dirty and the humidity is high + its an absolute PITA getting solder splashes out of carpet so you want a large mat down anyway.   With the mat, if you follow the one hand rule, the worst you will get is a shock across your hand or to your forearm.  Nasty, with a risk of burns or even deep tissue damage, depending on how much energy you get bit with, but with a minimal risk of killing you.  Not working alone was possible for all except 1 man repair shops and some hobbyists.  Its a rule I've broken a few times myself, but if you know there's no backup if something happens, and you have the intelligence to properly assess the risks, you tend to be a lot more careful and methodical. e.g. Alone, I will unplug and put the plug on my lap before moving the probe., but at the shop bench with colleagues nearby, I would frequently move the probe hook onto the next testpoint loop I'd pre-attached without shutting down.

Isolation transformers have no moving parts & last for years, so they are pretty much "niche market " items, just as desoldering stations were for many years.
For this reason, they are rarely found in Electronics stores like Jaycar, or Altronics in Australia, (& equivalents in other countries).

Isolation transformers - at least as bare transformers used to be readily available over the counter in high street electronics stores.  IIRC for about 50 bucks for a 250VA one.  However that was 25 years ago, when people in general still repaired stuff or at least were willing to pay to have it repaired.  If you aren't repairing or designing SMPSUs or working on lighting ballasts or AC power control circuits its an extremely niche market, and I rarely use one nowadays, though I still have the one with a 250VA toroidial I mounted up with a yellow painted double socket in an old wooden instrument case with a hinged lid, as an easily portable one for callout TV repair.

[stevelup]

Farnell (Element 14) list about 100 different appropriate types as do RS and therefore, I would expect, so would pretty much any component wholesaler.

It would be less than an hours work to put one in a nice box with a power inlet and output.

Most of them have split secondaries as well, so it would be easy to add a 240/120V switch too.

[David Hess]

Here is how I usually work with off-line circuits including off-line switching power supplies.

Neutral and ground are connected together at the breaker box which means any current drawn by the power supply raises the potential of neutral away from ground so ground of the oscilloscope cannot be connected to neutral, or any other point, of the off-line side safely or with good or even safe results.  Potentially neutral and hot could be reversed which makes things even worse.

One solution is to use a high voltage differential probe however this adds complications and may be expensive.  Further the differential probe performance could be improved if we could use a ground connection to limit common mode noise.

So what I normally do whether I use a differential probe or not, is to use an isolation transformer on the line side of the device under test.  Now ground, real earth ground, may be connected to any low impedance point like neutral or hot, with isolation they become the same thing, or even one of the rectifier outputs which is usually the common point for the line side circuits and the probe ground can be connected there as well without creating a massive low impedance ground loop.

[casinada]

On top of Dave's explanation there are other good explanations out there. As explained by others, know what you're doing, understand what you are doing.