General questions about Isolation i.e. how NOT to blow up your 'scope follow-up?

[FakeDave]

Hi! I've been tinkering with electronics for several years now, and getting familiar with an old analog scope while saving for a new digital one, and I want to take it to the next level by repairing some devices that might require looking at signals at higher voltages than just 5v logic. I've been doing a lot of research about how to avoid doing damage to scopes, DUTs, and myself when testing power circuits. I'm all about safety so I looked up what Tektronix had to say about it and they don't really recommend an isolation transformer anymore, but if you *must* use one that you use something like their A6901 Ground Isolation Monitor in conjunction with it, and I can see why. Windings fail, $tuff happens and your life is worth a little more investment. Then I looked into scope isolators, like the Tek A6902B Isolator and wound up getting an OK deal on one, the trade-off being it was missing the probe for the second channel (if anyone has a spare p850 probe, holla!). I also found what some of you had to say about it in a very old post here, and some advocated for using both an isolation transformer for the DUT and separate protection for the scope! Lower amperage isolation transformers like B&K's are cute, but getting anything over 3 amps gets expensive in a hurry. So what are your thoughts on the matter? And please be gentle, this is my first post.  Thanks! 

[helius]

First of all, welcome to the forum.
You aren't real specific about what your needs are: if you are probing devices that live behind the secondary side of a power transformer, you don't need any special isolation. This includes the vast majority of electronics outside their power supply modules.

[FakeDave]

First of all, welcome to the forum.
You aren't real specific about what your needs are: if you are probing devices that live behind the secondary side of a power transformer, you don't need any special isolation. This includes the vast majority of electronics outside their power supply modules.

Thanks! I was a bit vague about that 'cause I felt like the post was getting long-winded. I want to be able to do it all, really (OK within reason). I want to go from looking at things like serial signals one day to big power relays and switches the next.  I'm a broadcast engineer who didn't get their start with EE like all the "old-school" analog guys did, and I find that I've learned things through experimenting at home (on cheap things) that have applied to work (and much more expensive things) and vice versa. So I'm trying to make up for lost time by cramming as much as I can into my brain. My poor brain. 

[Ian.M]

It doesn't matter whether you use an external battery + inverter or an isolating transformer, *ANY*  form of 'floating your scope' is strongly discouraged, unless the scope was designed for floating operation.   A scope is *NOT* a multimeter - a multimeter typically has operating controls with insulation designed to keep the user safe when probing a 1000V circuit, no matter which lead is 'hot' - a mains powered scope doesn't have that grade of insulation, and is designed with the expectation that its chassis will be grounded.   Competent engineers have died making measurements with floating scopes, so don't join them!   

[bdunham7]

A couple of general observations for you:

1.  It is a mistake, IMO, to overthink the general case while underthinking the specific one.  For example, you worry about your isolation transformer not being big enough or failing.  I find very few cases where a bigger isolation transformer would be needed.  You should give an example of DUT where you feel you need a bigger transformer and then we can discuss how you would approach that specific case.  Buying super-safe equipment so that you can go poking around randomly is not a very good plan.  You have to understand each DUT and how to safely approach it. 

2.  I don't think of isolation equipment primarily as personal safety equipment.  Of course, in the case of old hot-chassis tube electronics, perhaps it is.  But for the most part, there are different reasons for isolation circuits.  I have a scope isolator in the same line as yours (A6909) and an old Scopemeter.  A typical use would be to scope a component, such as an amplifier output device, where neither side can be referenced to ground.  I can essentially use the scope in the same manner, with no more or less care, as a DMM.  This works great up to about 600VAC (300 for the Scopemeter) and has a significant advantage over differential probes in terms of circuit loading and isolation. 

3.  If you go to true high voltage--lets say anything over 600VAC, but I'm thinking about things like CRT circuits, microwave ovens, etc--then you need to shift gears entirely.  Because of the energies and frequencies involved, you need a ground-referenced high voltage probe, or  you need to be an experienced expert that has thought the situation through.  There's no "doing it all" in this realm nor any place for vague generalities.  And you need to understand dV/dt and why it matters.

[rstofer]

One question:  why?

You can measure ground referenced higher voltages with a 100:1 probe.  This is fairly inexpensive.
If you simply must measure voltages that aren't ground referenced, use a differential probe.  These are NOT inexpensive but they solve virtually all of the ground loop problems and don't require the scope to be floating.

Differential probes:



Single ended versus differential probes



Note that when she uses two single ended probes, both probes are grounded and, of course, that is the problem we're trying to avoid.

Our host's differential probe:

https://www.eevblog.com/product/hvp70/

[bdunham7]

One question:  why?

You're right in that for most purposes an active differential probe will allow the OP to poke around randomly with relative safety as long as the DUT is not particularly high voltage.  The issue with differential probes that I think maybe is overlooked a bit is that they introduce circuit loading that may be more than some anticipate.  A typical differential probe will have an input impedance of 10M, but that is split as 2 separate 5M+5p (or so) connections to ground.  If you are measuring something small floating on a voltage that has significant potential to ground--then you have introduced a total load of 2.5M+10p to ground.  So, for a HV or HF circuit where either of those loads would matter, the DP is not ideal. 

There are no simple solutions that allow amateurs to safely poke away at high potential circuits without understanding them.  However, these may be the closest thing to that simple solution.

http://www.micsig.com/html/41.html

[vk6zgo]

First of all, welcome to the forum.
You aren't real specific about what your needs are: if you are probing devices that live behind the secondary side of a power transformer, you don't need any special isolation. This includes the vast majority of electronics outside their power supply modules.

Thanks! I was a bit vague about that 'cause I felt like the post was getting long-winded. I want to be able to do it all, really (OK within reason). I want to go from looking at things like serial signals one day to big power relays and switches the next.  I'm a broadcast engineer who didn't get their start with EE like all the "old-school" analog guys did, and I find that I've learned things through experimenting at home (on cheap things) that have applied to work (and much more expensive things) and vice versa. So I'm trying to make up for lost time by cramming as much as I can into my brain. My poor brain. 

First up, we need to get some terminology sorted out.

Broadcasting is such a diverse field, & is notorious for throwing the word "engineer" round with abandon, ranging from the bloke who operates the mixing desk, who may be very knowledgeable in his field but not know much Electronics, to the Studio Tech who looks after the minutae of recording & reproduction of sound, to the Transmitter Tech who has a close relationship with the "big bities".(maybe not so much in these solid state days, but tube equipment had serious HT voltages, from 2kV, to 10kV & upwards, depending upon power level.)

Or to the "big kahuna"--- the Chief Engineer, upon whose desk, the "buck stopped", & many others in between.

Some of these guys may have EE or other degrees, others had, (in the past) Govt certification, others still, started out in the military.

Which kind are you?

[FakeDave]

I thought maybe a 100:1 probe might be enough but I wasn't sure, so that answers your question.

[FakeDave]

Yeah, I figured some of the "true" licensed engineers would object to me throwing that term around.  My station doesn't lump technical directors or equipment operators into that category, so if it takes more than a simple reboot or if something needs to be fixed they call me.  I started out long ago as a network field tech, so as I said I didn't get much EE exposure and I don't claim to know much about it, although I know enough to be dangerous. I'm the lead "engineer" but I don't call myself a "Chief engineer" because as you said, they are the big fish and that title takes a bit more knowledge and experience, but I'm chipping away at it bit by bit. I've been in the industry for over 5 years now so I know a bit more than your standard cable pusher. My goal is to be able to repair more things, and I like to learn by trying things and failing. I just don't want to damage any equipment in the process.

[FakeDave]

A couple of general observations for you:

1.  It is a mistake, IMO, to overthink the general case while underthinking the specific one.  For example, you worry about your isolation transformer not being big enough or failing.  I find very few cases where a bigger isolation transformer would be needed.  You should give an example of DUT where you feel you need a bigger transformer and then we can discuss how you would approach that specific case.  Buying super-safe equipment so that you can go poking around randomly is not a very good plan.  You have to understand each DUT and how to safely approach it. 

2.  I don't think of isolation equipment primarily as personal safety equipment.  Of course, in the case of old hot-chassis tube electronics, perhaps it is.  But for the most part, there are different reasons for isolation circuits.  I have a scope isolator in the same line as yours (A6909) and an old Scopemeter.  A typical use would be to scope a component, such as an amplifier output device, where neither side can be referenced to ground.  I can essentially use the scope in the same manner, with no more or less care, as a DMM.  This works great up to about 600VAC (300 for the Scopemeter) and has a significant advantage over differential probes in terms of circuit loading and isolation. 

3.  If you go to true high voltage--lets say anything over 600VAC, but I'm thinking about things like CRT circuits, microwave ovens, etc--then you need to shift gears entirely.  Because of the energies and frequencies involved, you need a ground-referenced high voltage probe, or  you need to be an experienced expert that has thought the situation through.  There's no "doing it all" in this realm nor any place for vague generalities.  And you need to understand dV/dt and why it matters.

So that's your idea of gentle, huh? I kid. OK, let's start with this last bit first, I kept it general not because I'm willfully ignorant of the finer details of the subject but because I see myself working on a lot of different things in the future, and I don't want to have to buy something new in a year or two because what I use now isn't capable.

Yes I know that 'true' high voltage is much higher than average people think and it is something I have no interest being around, in fact I leave some projects that are say CAT3 to a bonded electrician. So please don't confuse my adventurous nature for bravado (I did say "within reason" after all.) That being said, eventually I want to experiment with Nixies and other tubes like geiger tubes so yes 300 volts or more isn't out of the question. And I know that most of what I'm talking about is usually multimeter territory but I want to see all the signals! so I do appreciate you mentioning a scope meter, they look very handy indeed and I've been reading a lot about them. So yeah, thanks for your 'observations'.

I would like to hear more about other people's favorite test gear while I'm on the subject. I saved an old Tek TM504 from the recycle heap and I've had a lot of fun tinkering with some modules, looking at all the different ones online and getting outbid on every single one I've ever bid on.    It is the perfect system for me, compact and easily reconfigurable for a variety of roles. I love it, I'm even already thinking about a TM5006 for my rack. Yeah, I know. I have a rack. I had to sneak it inside while my wife was away. I'll see you in TEA shortly.

[rstofer]

The issue with differential probes that I think maybe is overlooked a bit is that they introduce circuit loading that may be more than some anticipate.

I didn't watch the entirety of the 2d video I linked above but she is talking about differential probes having an order or magnitude more bandwidth than single ended probes.  If this is true, I suspect they aren't hobby grade in terms of cost!

[FakeDave]

The issue with differential probes that I think maybe is overlooked a bit is that they introduce circuit loading that may be more than some anticipate.

I didn't watch the entirety of the 2d video I linked above but she is talking about differential probes having an order or magnitude more bandwidth than single ended probes.  If this is true, I suspect they aren't hobby grade in terms of cost!

No, they aren't, there's a link in another comment to the one that the EEV blog's own Dave Jones designed, that's the one I would buy if I needed one. But like many other things in the realm of test gear, it's one of those things where the term "low cost" is not as low as you might think.  But in fairness they are a very specialized piece of equipment that is in demand and surely aren't that cheap to develop or make. Honestly one of those isn't even on my RADAR right now. My next several-hundred dollar purchase will be a Rigol 1054Z with all those additional pricey software keys now included. It's even gone down in price since I started watching them. In that case at least, it pays to be 4 or 5 years late to the party.

[0culus]

I thought maybe a 100:1 probe might be enough but I wasn't sure, so that answers your question.

Probes for truly high voltage tend to stop looking like standard probes, and be quite a bit more expensive too:

http://w140.com/tekwiki/wiki/P6009
http://w140.com/tekwiki/wiki/P6013
http://w140.com/tekwiki/wiki/P6015

That last one is pretty cool because it uses a dialectric fluid (freon) to achieve it's colossal rating. The later P6015A switched to a silicone based oil fill after freon got banned.

Personally, I'm going to to try to stay far away from anything that requires a probe like that. The only possible thing I might work with once I have room for a larger lab and can implement a strict safety protocol is CRTs in my old test equipment.

[BravoV]

Hi! I've been tinkering with electronics for several years now, and getting familiar with an old analog scope while saving for a new digital one, and I want to take it to the next level by repairing some devices that might require looking at signals at higher voltages than just 5v logic. I've been doing a lot of research about how to avoid doing damage to scopes, DUTs, and myself when testing power circuits. I'm all about safety so I looked up what Tektronix had to say about it and they don't really recommend an isolation transformer anymore, but if you *must* use one that you use something like their A6901 Ground Isolation Monitor in conjunction with it, and I can see why. Windings fail, $tuff happens and your life is worth a little more investment. Then I looked into scope isolators, like the Tek A6902B Isolator and wound up getting an OK deal on one, the trade-off being it was missing the probe for the second channel (if anyone has a spare p850 probe, holla!). I also found what some of you had to say about it in a very old post here, and some advocated for using both an isolation transformer for the DUT and separate protection for the scope! Lower amperage isolation transformers like B&K's are cute, but getting anything over 3 amps gets expensive in a hurry. So what are your thoughts on the matter? And please be gentle, this is my first post.  Thanks! 

Few advice from a fellow enthusiast too, I had very similar path with yours on probing mains line related devices, yes, I've read & researched a lot too similar to your journey.

I've walked thru the path of using isolation transformer, Tektronix Isolator A6902B (yes, I have two  ) and finally ended up with Differential HV Probe, and its the probe I used whenever I'm going to probe mains line connected/related DUT.

On A6902A/B subject, beware, its huge, here the shot to give you the idea on it's size compared to the scope, definitely not comfortable to use compared to current modern HV Diff probe.


Also on A6902A/B again, its all about the probes, it uses proprietary connector  , if you really want to buy one, simple, no probes, no deal, as you will face with the trouble finding the right probe for it.
I was lucky able to score with complete probes, here to give you the idea how they look like.


As you mentioned and fully aware of your self that it will be dangerous voltage, again, reminder, one mistake is enough to toast your scope ... or your life.

Consider to save money for HV differential probe, even the crappy one, still better than none, but buy new, imo, its worth the investment.

Once you owned it, probing mains related devices is very comfortable, an example scope that was probing the mains line that powered the scope it self, using HV Diff. Probe. Mine is Fluke, but current Chinese made are also pretty decent and "relatively" cheap, search around this forum, there are plenty of review/teardown etc.


My 2x1/4 watt resistors worth.

[DDunfield]

The desire to "float" a scope or DUT arises from the desire to measure something that is not ground referenced, and wishing to avoid a path to ground through the scope.

This subject comes up quite often, do some searching of the forums and you will find lots of opinions. In doing so you will find the words "safe" and "safer" mentioned a lot, as in "floating the DUT is safer than floating the scope". I prefer the term "less hazardous".

Below 40V or so isolation is generally OK - as coming into contact with <=40v is usually not very hazardous. At higher voltages things get dicey.

BAD ways to measure non ground referenced HV signals:

BAD#1: Floating the scope.

Isolating the scope and attaching it's ground to HV means that every exposed metal bit on the scope (BNC connectors, Calibration ground etc.) is now sitting at HV, and as you are used to doing things like connecting/disconnecting BNCs with your hands, there is a very real possibility that you will have a "brain lapse" and grab onto HV.  Also, although I would hope a TE manufacturer would assume the worst and provide really good insulation, you scope isn't spec'd for this purpose so this may not be the case.

BAD#2: Floating the DUT

Isolating the DUT and then grounding some HV signal (via your scope ground lead) means that all the normally ground bits of the DUT are now sitting at HV. Since the DUT was likely not designed with testing in mind and the manufacturer never intended it to be "grounded" this way, there are likely a LOT of them, and you may not immediately discern what is and isn't ground, and again "brain lapse"  will factor in .. there is a very real possibility that you will grab onto HV. And don't even think it has been acceptably insulated for this use case.

++ You will see a lot of forum comments saying that floating the DUT is "less hazardous" then floating the scope. This is because traditional scopes a large "grounded" metal boxes which you really don't want sitting on your bench at HV ... modern scopes are more insulated, so I'd say it's become more 50-50 ... what this means is that you remove "less" and end up with just "hazardous" no matter which end you float - AVOID DOING EITHER!


GOOD ways to measure non ground referenced HV signals:

GOOD#1 - use the "math" function of your scope to measure the difference between two channels.
  - Cheapest way (free)
  - Takes two channels for one measurement.
  - Limited by the HV ratings of your scope/probes.

GOOD#2 - Use a HV differential probe
  - Good HV-differential probes are expensive (and you WANT good).
  - Only uses one channel, allows all the normal measurements/math of your scope.
  - Can get HV-diff probes for really high voltages ($$$)

GOOD#3 - Use a properly insulated battery powered scopemeter
   - Pricey new, but can be easier to find older good ones used.
   - Must be designed for HV .. just because it's a scopemeter doesn't mean it's "less hazardous" to use on HV.

I personally have opted for #3 - I have two scopemeter I use for non-ground referenced measurements:

UT-81B .. ~$150-200 new, 8Mhz,  single channel, connects like a multimeter (test leads although there is a BNC adapter).
  - Rated like a multimeter 1000vdc, 750vac.

Fluke97 .. ~$200-$400 used. 50Mhz, dual channel, uses BNCs
  - Make sure you get original 10x probes with it - they are insulated and rated for 600v
  - Scope itself is rated up to 3KV with the proper x10 probes (as high as 25Okv with correct 1000:1 probe - I don't have one!)

Dave

[rstofer]

My next several-hundred dollar purchase will be a Rigol 1054Z with all those additional pricey software keys now included. It's even gone down in price since I started watching them. In that case at least, it pays to be 4 or 5 years late to the party.

I have the 1054Z and the first thing I did after unboxing was to unlock the advanced features.  Today, I would be looking at the Siglent SDS1104X-E and unlocking it to become the SDS1204X-E or, perhaps, just buying the 1204.  That extra 100  MHz of bandwidth would be nice and the UI is said to be much more responsive.

Sometimes the party moves...

[FakeDave]

My next several-hundred dollar purchase will be a Rigol 1054Z with all those additional pricey software keys now included. It's even gone down in price since I started watching them. In that case at least, it pays to be 4 or 5 years late to the party.

I have the 1054Z and the first thing I did after unboxing was to unlock the advanced features.  Today, I would be looking at the Siglent SDS1104X-E and unlocking it to become the SDS1204X-E or, perhaps, just buying the 1204.  That extra 100  MHz of bandwidth would be nice and the UI is said to be much more responsive.

Sometimes the party moves...

I can see why, that's a nice looking scope.