Oscilloscopes: safe use and testing

[Stephen1966]

Greetings tubaholics,

Isolation transformers.

I come to you as the proud new owner of a Tektronix 2236 analogue oscilloscope. A little preamble: In discussion with one of our moderators (Phil, aka pompeiisneaks) one of the difficulties this forum sometimes presents to a newcomer like myself is simply finding relevant information when we need it. Phil directed me to try the Google Search function because it is superior to the search function above it. And so, sure enough, this is where I found myself when I was looking for information on oscilloscopes. As a complete beginner, I am looking for anything and everything but unless you use semantic phrases in the subject line or users like me can turn quite generic search terms into something more specific, just browsing the topics can sometimes lead to dead ends. Not a complaint, you understand, I enjoy reading those dead ends and there is usually always something useful in discussion to pin for later. We don't have subject tags in the posts though, so I am trying to provide my new topics with semantic descriptions so that others like myself, might find them a little easier, when they go looking.

With that in mind, I want to start this subject on oscilloscopes, because as someone who has never used one before, but who has seen a broad variety of links to the subject in this forum, I think I can learn a lot. Especially in the field of tube amp testing and calibration. Where this thread goes, is anyone's guess at the moment but I am thinking that I personally, would like to approach it, from the ground up, from the basics and then post/link procedures from other parts of the forum as a kind of practical exploration of the topic. And this, of course, is something you might do as well. I enjoy the conversations here and it's really refreshing to find a forum where members are actually helpful, not to mention civil even when they disagree about ideas. You guys are inspiring.

So to get us started I wanted to ask your advice first, about isolation transformers. Being the proud owner of a new (old) oscilloscope, I have heard conflicting accounts about the need for isolation transformers and their use with oscilloscopes. I should add that I am still waiting for the new (CAT II/passive) probes to arrive, so this instrument has not been used for anything yet. I am healthily afraid of high voltages though and I don't really want to make a mistake here. It could really ruin my day. Some people seem to suggest that the DUT (device under test) should always be isolated but others say it isn't important as long as the metal chassis is always grounded and connected to the same point on it with the probes. When I was looking at this problem, I came across the Tektronix download at https://www.tek.com/document/fact-sheet ... lloscopes which discussed the method of making "A minus B" (pseudo differential) Measurements. It's unclear but it seems to suggest both the scope and the DUT are grounded, and it's unclear as well if, when it says 'Two oscilloscope channels are used to make one “A minus B” measurement,' this can be used to test two different parts of the amp circuit simultaneously (it suggests not as it suggests it is using the 'add' or 'both' functions in the vertical mode). But if we can safely test using a common ground point on the chassis with the probes and having the scope grounded as well through its mains connection, it means there is a lot we can do with a scope without the need for an isolation transformer. Given their high cost (more than the scope cost) I would prefer that. But I don't want to kill myself or fry the living daylights out of my equipment so, are there any caveats? Are there any things we should never do with the scope in regard to grounding?

It may be important to mention that I am an Englishman, living in the Czech Republic. Why? Because, during the Cold War when it was still Czechoslovakia, resources like copper were scarce here and so electricians used a system (I think they abbreviated it "TFT") which involved combining the neutral and the ground in a single wire and the cores were made from aluminium (aluminum, if you're Stateside), this meant they could use two wires in the cable (not three) and they overcame the shortage of copper by using an alternative that was cheaper and relatively plentiful. Of course, some of this domestic wiring is now over sixty years old, aluminium decays and power outages are relatively commonplace. We don't use that system for new property builds and restorations now but the first time I saw this, I was somewhat perplexed ("They did WHAT?" with many choice expletives to that effect) because back in Blighty it was something I had never seen in single phase, domestic wiring. A definite no-no, in fact. Even now, parts of this old house we live in have the two strand aluminium cable - a situation I'm rectifying as I go along. That (to my mind) presents a much lower impedence to ground than the UK system where (if I understand it correctly) the neutral and ground are connected at the distributing substation. I can't say if that presents a greater threat when there is a hot source in close proximity to the neutral/ground. My knowledge of it, is very basic, but it might be important.

Later on, I would like to look at some of the testing procedures as tracing the signal, measuring it adjusting the bias and so on. I'm quite eager to get started because I built the Tweedle Dee using only a multimeter and now, finally, I will be able to properly adjust the PI bias. Job #1.

Stephen

[sluckey]

I've been using high end scopes since 1968 and have never used an isolation transformer. I also don't connect a high end scope to cheap electronics that don't have an earth ground.

[xtian]

That's the main point I'd like to reinforce--the Earth ground. Our projects SHOULD/MUST have the chassis bonded to Earth ground. Our AC-powered scopes should also be properly Earthed. Thus, the chassis, and the ground clip on the o-scope probe are both at Earth potential, and there's never any fear of connecting the probe's ground clip to the chassis or other convenient ground point.

If they're both grounded, why would you bother with the probe's ground clip? When measuring small AC voltages, you may get a clearer view by grounding the probe clip, because the probe shield blocks RF when grounded.

When would you NOT want to ground the probe ground clip? Some solid state amps will fail spectacularly if you ground the output. (Are these differential amps? I don't know! I need more data.) But it's pretty exciting when you make this mistake.

[Stephen1966]

That's the main point I'd like to reinforce--the Earth ground. Our projects SHOULD/MUST have the chassis bonded to Earth ground. Our AC-powered scopes should also be properly Earthed. Thus, the chassis, and the ground clip on the o-scope probe are both at Earth potential, and there's never any fear of connecting the probe's ground clip to the chassis or other convenient ground point.

If they're both grounded, why would you bother with the probe's ground clip? When measuring small AC voltages, you may get a clearer view by grounding the probe clip, because the probe shield blocks RF when grounded.

When would you NOT want to ground the probe ground clip? Some solid state amps will fail spectacularly if you ground the output. (Are these differential amps? I don't know! I need more data.) But it's pretty exciting when you make this mistake.

And thanks sluckey... It makes sense to me ("why bother with the ground clip if the scope and DUT are both already grounded?") I have a little solid state VOX practice amp but I couldn't tell you if it is a differential amp or not (it sounds fancy - so probably not) but this scope is really for my tube amp builds. Do I guess correctly that it would be the same (unnecessary ground clip) when I measure the higher DC voltages? For example plates and grids. The scope is rated up to 400 V but I think I read in the manual somewhere that it can handle higher.

Also... I should really try to understand this stuff thoroughly, before firing back with a reply, correct me if I am wrong, you're saying we only really need the ground clip attached when we measure the small AC voltages. So any stray RF's in the higher voltage range (including AC and DC) are what? too small to be significant?

[pompeiisneaks]

Correct, I've found accidentally that you can measure most things w/o the ground clip connected. I found it out when I was doing measurements w/o any issues and didn't realize I'd pulled at the right angle and disconnected my clip...

It still read just fine. The ground clip is nice because it does eliminate noise that can couple into the probe, and is super important when dealing with super high frequency stuff as any stray noise can give you massively wrong readings.

But at audio signal levels, most of the time, it's not a big deal.

~Phil

[Stephen1966]

So I found this on YouTube
Mr Carlson's lab talking about isolation transformers and variacs. Perhaps the ground/neutral bonding is more common than I thought. He talks about the ground and neutral being connected at the breaker box. He mentions guitar amps. Overall it sounds as if using an isolation transformer has the potential (no pun intended) for more problems than it solves and since the scope should also be on the mains circuit bonded to earth, it provides a roundabout path to earth via the probe ground clip/BNC shield to mains earth. If that ground clip touches any hot live it has the same fatal effect as if there were no isolation transformer at all. Maybe an even greater hazard, because the exposed metal surfaces of the scope would then also be live in that case. I was also curious about the ability of variacs to isolate, but he gives that notion fairly short shrift.

[sluckey]

Most variacs are auto-transformers and are not isolated. You can spend a bit more and get an isolated variac.

[pompeiisneaks]

Also if you want to understand the dangers of probing in circuit without an isolation transformer, then Watch Dave on EEVBlog where he discusses some of the gotchas and problems. Mind you this is all on modern solid state circuitry with much more complex setups, isolation between different parts of the circuit and bridging those can cause major issues etc. Especially things like Switch Mode Power Supplies (SMPS) where they have high voltage and low voltages separated by transformers and opto's and if you bridge the two, or ground things wrong you'll have a bad day. Nothing like this with tube amps by far.
https://www.youtube.com/watch?v=xaELqAo4kkQ

~Phil

[Stephen1966]

The engineer who sold me the scope found an isolation transformer 220-220 2000W but looking at the serial plate in the photo, it appears rated for around 10A. Would this be safe to use with our amps? I'm on 220-240V. The primary fuse on my Tweedle-Dee is just 1A which should protect the circuit if there is a surge.

[sluckey]

If your amps have a power cord that provides an earth ground to the chassis, you don't need an isolation transformer to work on it. If your amp does not have an earth ground connected to the chassis, then you need to correct that for your own safety.

[Stephen1966]

If your amps have a power cord that provides an earth ground to the chassis, you don't need an isolation transformer to work on it. If your amp does not have an earth ground connected to the chassis, then you need to correct that for your own safety.

Thanks sluckey they do (and always will ) I think I'm going to pull myself out of this rabbit-hole. I should stop buying stuff I don't need.

[Stephen1966]

Here's a link to the BOM for the 100W dummy load build I'm putting together from Aiken Amps' schematic
https://ampgarage.com/forum/viewtopic.p ... 84#p417184

Edit: here's a photo of the finished dummy load. The build is detailed on Martin Manning's 'Dummy Load' thread https://ampgarage.com/forum/viewtopic.p ... 57#p417557

Finished.jpg

[Stephen1966]

Here's a link to the signal generator I found on ebay
https://longweielec.com/products/tag-101
Nothing fancy, for sure, but I'm hoping it will do the job!

[Stephen1966]

Here's an introduction to oscilloscope basics from Aiken Amps:
https://www.aikenamps.com/index.php/the ... illoscopes
The section on triggering explains a complicated subject clearly. If you're a beginner like me, this is clearer than trying to make sense of the scope's manual.
[Edit: when Randall talks about the "polarity button," on my scope, it appears to be labelled "slope".]
Here's a picture of the control panel of the scope I'm going to be using.

2236(2).jpg

[Stephen1966]

Cap polarity, is this a big deal?

https://ampgarage.com/forum/viewtopic.php?t=29176

The consensus in that post seems to be that it probably isn't. However, I am curious to see just how the unmarked 716P orange drops I used in my Tweedle-Dee came out. The specs say that the 716Ps have an extended foil but nothing about having any kind of neutral polarity so I'm just wondering if it is "safe" to test them for polarity in situ. Naturally, before I or anyone else goes poking around in a chassis all the caps MUST be thoroughly discharged first.

Figuring out the negative pole with a scope seems easy. Please correct me if I am wrong.

• We connect a probe to each leg of the cap and look for the lowest amplitude waveform on the scope whilst holding the cap between our fingers to create a positively charged, capacitive coupling.

• The ground clip of the probe goes on one leg, the source clip of the probe goes on the other.

• The scope should be set around 5mV.

• Swapping the probe clips around indicates the shielded foil end of the capacitor when the amplitude is lowest: the ground clip of the probe indicates the shielded foil end.

• The shielded end should then be connected to the lowest impedence part of the circuit.

Asking whether it's safe, might not be the best choice of word (though safety is clearly an important aspect to bear in mind) instead, the question is really whether any other parts of the circuit will drain any capacitive coupling effect away so that the test is ineffective. As long as the caps are all fully discharged, we might rule out any positive charge coming from other areas of the circuit.

Mr Carlson's Lab again, has a video on testing caps' polarity but this and others like it never show the caps being tested in situ. It's a long video, but he describes the oscilloscope procedure around the 10 minute mark. Also, included is his schematic for a switching device that we can build to test the polarity used in conjunction with a scope.