How do you evaluate used oscilloscopes before purchasing?

[ramenbytes]

Update:

Sorry for the delay, I waited on posting this until I could proof-read and edit it.

I ended up getting a Tektronix 2246 Mod A for $100 locally, and from what I can tell it appears to be fully functioning or very nearly so. According to the seller, it had sat in storage since 1994. As a bonus, the seller threw in a HP 200CD Oscillator for free, which he would have otherwise thrown out. The oscillator doesn't appear to be functioning, and he said it'd had a tumble, so I guess that can be my first valve device repair victim. For starters, the power cord needs replacing. Someone floated it...

I appreciate everyone who gave feedback, I tried to take it all into consideration during the buying process. Below I'm including a list of various quotes from this thread and how they factored in or related to the process, mostly in hopes that the consolidation will be useful for anyone who happens upon this thread later.

Evaluating the condition:

Another one to check for before purchase is the availability of the manual (for common ones like the 2246 they are online but for exotic ones this might not be the case) as those includes the schematic and checklists/repair guidelines.

This was one of the first things I checked when I saw the scope listing come up, since from this and other discussions on the forum and the TekScopes group it seemed that old test equipment needs some caring for from time to time. I passed on some better priced scope listings specifically because I had issues finding operator or service manuals.

The most common issue in old scopes are aging capacitors and worn dials.

Yeah, the one I got needs a little cleaning of the potentiometers. I did managed to break a plastic knob during removal for cleaning, but some epoxy fixed that well enough. I've had a quick peek inside, though not a very thorough one, and I didn't see any obviously faulty caps nor any RIFA suprises. I did catch a glimpse of some motorola "ZM" diodes however, which from what I've read so far are candidate #1 for preemptive repair. I'll be looking into a possible pre-emptive recap as well.

Overall those old scopes are nearly indestructible and easy to repair so even if you find an issue just price it in.

Ease of repair was a definite factor in deciding on a scope, and one of the reasons I didn't follow up on a 2445A listing. It could be working, but for a first scope I don't want to risk U800 issues. If I join TEA then maybe... Before buying, I also checked both this forum and TekScopes for reports on issues that others had with the 2246, to make sure I knew what I'd likely need to repair and that there wasn't any failure prone unobtainium inside.

The quickest check would be firing it up and probing the square wave or ramp (some analog scopes have that), playing with the knobs to see all the adjustments are working:
- min/max is in the right place
- smooth/no jumping on the knobs

I played with the knobs to see that they had an effect, though I didn't know what all of them did at the time. I used the beam finder to show a basic grounded trace on all the channels, and pushed relevant looking buttons until the B sweep was configured the same as the A sweep and showing a trace. I also messed with the knobs until I was convinced that the B sweep could display a signal across the entire width of the screen. Since it was originally set with non-zero delay and auto-level triggering it was only showing a trace on part of the screen, which confused me at first.

- don't forget  to test all voltage/time ranges and AC coupling

The probes I was going to use from the seller to test things ended up being lost somewhere among three storage containers, so the only signal tests I did were with a multimeter probe acting like an antenna. I did try some other stuff with a 9 volt and also my DMM's resistance setting (as a current source), however at that time I still hadn't realized that I had the coupling set to "ground". When I was trying the antenna test lead I realized this, and was relieved to suddenly see EM noise from the antenna being displayed. This also indicated to me that the DC coupling worked on at least Ch 1. I may have also messed with the AC coupling too just to make sure it would switch over and show a trace.

With those analog oscilloscopes always look at the brightness setting.

The brightness looked good too. A quarter turn on the intensity was bright enough for use, so it seems the CRT isn't that worn.

Spend some time understanding the on screen annunciators, they’re extensive on that scope. Check that they track the different selections such as vertical voltage per div, time, 1/time, delta time, etc.

Out of respect for the seller's time, I didn't thoroughly test all of the readout related features. I mostly just checked that when I turned a knob the readout matched this with the appropriate changes to the Volts/Div and Sec/Div, that the focus control worked. When I was messing with the B sweep I was also able to see the readout showing other relevant info. I think I checked to see if the menu showed up too but no in-depth exploration of it.

For a sub hundred buck used scope* just make sure you get a trace and can view a calibration waveform on all channels, a reasonably sharp and bright trace can be had without maxing out any controls, and fiddle with it a bit while subtly watching for nervous behavior from the seller. Unless you know the guy it's kinda rude to spend a hour or more checking everything for such a low buck transaction.

That ended up being the gist of what I did. Your post and the others that advocated for managing my expectations of used equipment helped me adopt an appropriate mindset. Prior to that I think I may have been expecting too much. The meetup with the seller actually ended up being around an hour, though most of that was just chatting with the seller and him showing off his collection of electronics stuff (I think I saw an old scope labeled "Pacific Bell"?). Actual testing time was a fraction of that.

CRO & DSO trade offs:

Don’t get too enamored with old crt based scopes, they may be fun to learn some basics with but the functionality compared to a modern digital scope is night and day. Add in the portability and noise levels and a modern digital scope wins in every category every time.

I've definitely had fun learning the scope so far, and using it to look at PWM signals for Arduino and Pi has been enlightening. I'm also getting a better idea of the limitations of CROs (mainly the lack of sophisticated recording and analysis functions), however I don't think I'll hit any walls because of that for a good while, so I should have plenty of time to save for a decent DSO. I ended up going with a CRO instead of a DSO because the price was good enough (counting the oscillator as part of the purchase) and while I may have wanted some of the DSO features like a computer interface, I don't really need them and the lack of them wasn't a deal breaker. That, and forgoing the DSO for now means I have some budget room for non-scope stuff like the components for a small project.

I don't see the problem with starting with an analog scope. It's IMHO a good way how to start, especially if you can buy some advanced one for a good money. But ultimately, it all deppends on the application.
For amps and audio stuff, analog is still great and might be even better than some cheap DSOs. For debugging digital circuits (Arduino, robotics) it's good to have a DSO, because you want to be able to check non repeating signals (like communication on any bus). There are ways around this (like changing the program to generate trigger signals), but they are almost always pain in the...

If you are debugging digital waveforms, then use a digital domain tool such as a logic analyser. Basic ones are remarkably cheap, and more expensive ones can be much more effective than a scope.

I also pulled the trigger on a $9 saleae clone, which in conjunction with the CRO should cover just about everything I'm doing with Arduino's and RPi's, at least for now.

In which case, if I probe a DC source like a AAA battery with the scope AC coupled, then I shouldn't see anything other then a zeroed out signal right?

Correct. Doing that same thing with a power supply can show the ripple and noise of the supply.

Spend some time understanding the on screen annunciators, they’re extensive on that scope. Check that they track the different selections such as vertical voltage per div, time, 1/time, delta time, etc.

Don’t get too enamored with old crt based scopes, they may be fun to learn some basics with but the functionality compared to a modern digital scope is night and day. Add in the portability and noise levels and a modern digital scope wins in every category every time.

Don’t spend more than $50-$75 on such a scope. That money would be best added to upgrades and accessories for a modern scope. Remember that an analog scope cannot display a one shot event, the signal you’re examining must be continuous/repetitive. For what is dead easy to do today with a modern digital scope, required a rather expensive and not so common “storage” scope back in the day.

That said, many analog 'scopes with a storage function spent almost their entire, very productive, lives with the storage function used maybe two or three times.
In much work, displaying a "one shot" event is a rare requirement, whereas in other activities it is an everyday thing.

The important question is, what  will the OP use the Oscilloscope for?

For right now, I'm looking at just audio work (valve guitar amps and effects pedals) and digital stuff with Arduino and RPi's. Unless I missed something, I won't be hampered by the lack of storage for quite a while. Even when I do need storage capabilities, projects like this one should cover a lot of it: https://hackaday.io/project/171961-a-digital-camera-for-analog-oscilloscopes

Pricing:

Don’t spend more than $50-$75 on such a scope. That money would be best added to upgrades and accessories for a modern scope. Remember that an analog scope cannot display a one shot event, the signal you’re examining must be continuous/repetitive. For what is dead easy to do today with a modern digital scope, required a rather expensive and not so common “storage” scope back in the day.

For a sub hundred buck used scope* [. . .]

*Any more and you're likely better off saving your pennies for a new DSO, between the feature set and questionable dependability of aging test gear. At some point you end up more of an oscilloscope repairer than user, not fun if it isn't your intent.

You can do all the testing you want--and you should--but in the end your best bet is to get as good a deal as you can on the least beat-looking scope possible and hope for the best.

I think I compromised a little on pricing, paying $100 instead of the recommended $75 max. I did pay $1/MHz though, so bang-on what the recommended CRO pricing seems to be. Given that I didn't have to pay for shipping, I was able to evaluate it before buying, I got a free oscillator that I can at least sell for parts if not repair, the scope appeared to be in good condition, the last 2246 sold in three days flat, and $100 seemed to be about the ebay price floor, I was okay with a bit more than the recommended price ceiling. Now that I've had it for a couple days and can see that it definitely does appear to be functioning well (I haven't done a performance test yet though, and still need to recap and re-diode), I don't think I'll have much of an issue getting my money back later if, worst case scenario, I need to sell it to finance a DSO in the future. Hopefully I won't need to though.

Final thoughts:

Thanks to everyone who contributed, both through posts and DMs. I now have a functioning 4-channel 100MHz scope + a valve-based oscillator to practice working on before moving to guitar amps.

Edit June 28, 2021, 12:18:18 am
So the 2246 was sold to someone else a little earlier while I was replying to this thread  . I'm still considering other CRO options and appreciate further advice on evaluating them, however I'm also going to be seriously considering the option of just getting an entry DSO and putting off any other non-essential expenses for a while.

Don't worry: there will be another one along soon

Do worry: get the right class of probe for you applications, and understand how to use them without blowing up your scope/equipment or yourself.

Do worry: the cost of probes can be non-trivial!

It looks like I'll need to get a higher-voltage (around 1kV?) probe at some point for checking waveforms on valve pins. A decent one from probemaster looks to be around $90, so yeah non-trivial indeed. Because I didn't go with a DSO though, I should have room in the budge later if I end up needing such a probe. I'm going to wait until I have a measurement I can't make without it before getting one though. Depending on the state of the oscillator, that may end up being soon.

Before working on the oscillator though, I'm going to build a TS808 clone as a budget compromise between guitar and electronics and to get more experience with low-energy circuits before learning high-voltage valve stuff. I plan to use the scope to better understand the circuit, and to mod it. Thanks again for all your help and input!

Edit at July 12, 2021, 1:30:38 am
Fixed the formatting on one of the paragraphs, lines are not hardwrapped at 80 columns now.

[tggzzz]

It is always a pleasure to see someone that asks, listens, understands and thinks

Now you can, as I taught my daughter, "have fun - safely". There are a few ways you can hurt yourself with a scope, especially where high voltages and consumer equipment is involved. I suggest you have a look at some of the references in https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/

[jmelson]

Edit June 28, 2021, 12:18:18 am
So the 2246 was sold to someone else a little earlier while I was replying to this thread  . I'm still considering other CRO options and appreciate further advice on evaluating them, however I'm also going to be seriously considering the option of just getting an entry DSO and putting off any other non-essential expenses for a while.

I think one of the newer digital scopes like the 1054z would serve you better as a first scope. I have seen Amazon sell that scope as a open box return for $350 on occasion. I prefer the Siglent 1104x-e, but that is a pretty solid $500. The older analog scopes like Tek 465 will likely need repairs from time to time, and the later model 4000 series analog-digital are very complicated to work on and often tough to get parts for. They make a better second scope if you can find a decent one cheap enough. Just my 2 cents.

Well, I got a 2465 some years ago that I have been really happy with.  a much higher-grade  instrument than the 22xx series.
I have no problem with analog scopes, but the digitizing ones can do more.

Jon

[ramenbytes]

Now you can, as I taught my daughter, "have fun - safely". There are a few ways you can hurt yourself with a scope, especially where high voltages and consumer equipment is involved. I suggest you have a look at some of the references in https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/

Will do! I've already started going through some of the stuff (I found the link in a previous post of yours), and it's been helpful so far. The high voltage references I've seen in there appear to mostly deal with floating measurements, however I haven't had a chance to read through them all yet. Do they also cover general operator safety with high voltage equipment?

[james_s]

As far as analog scopes go, the 2465 is a heck of a nice instrument. You just have to hope it doesn't break down because I think those are the ones that use a bunch of irreplaceable hybrids.

[BrokenYugo]

I don't really have a reference to point you at but I must say one must be VERY careful working on tube gear, both for personal safety and the safety of your test equipment. Oscilloscopes and probes are not idiot proofed nearly as heavily as say, a multimeter. The biggest gotcha being that the scope measures everything relative to earth and the ground clip is earthed. Ask if you don't understand how to safely probe something.

[tggzzz]

Edit June 28, 2021, 12:18:18 am
So the 2246 was sold to someone else a little earlier while I was replying to this thread  . I'm still considering other CRO options and appreciate further advice on evaluating them, however I'm also going to be seriously considering the option of just getting an entry DSO and putting off any other non-essential expenses for a while.

I think one of the newer digital scopes like the 1054z would serve you better as a first scope. I have seen Amazon sell that scope as a open box return for $350 on occasion. I prefer the Siglent 1104x-e, but that is a pretty solid $500. The older analog scopes like Tek 465 will likely need repairs from time to time, and the later model 4000 series analog-digital are very complicated to work on and often tough to get parts for. They make a better second scope if you can find a decent one cheap enough. Just my 2 cents.

Well, I got a 2465 some years ago that I have been really happy with.  a much higher-grade  instrument than the 22xx series.
I have no problem with analog scopes, but the digitizing ones can do more.

Any working scope is better than no scope; the search for perfection is a great way of delaying action and exhausting the available budget.

Bearing in mind the aphorism in my .sig, I'd say the OP has plenty of skill and imagination. He will be able to find ways to use it for everything except one: capturing signals that cannot be made to repeat.

[Berni]

Sometimes its actually a better learning experience to use old gear.

Having more direct control of things such as analog scopes forces you to think in therms of how the scope works and how to get it to do what you need. Working your way around the gears limitations makes you think about things that you otherwise would not. It can teach you tricks that are sometimes even useful on modern digital scopes later on.

Also with old test gear you can actually get full schematics in a service manual along with an explanation how it works. This lets you see how it all works without even taking the cover off (Tho i do recommend taking the cover off anyway, its cool to see what is in there and how it is made)

Besides there there will always be a better scope out there somewhere than what you already have. No point in worrying about it. You can always buy another scope later on.

[Old Printer]

Artek manuals is a good source for Tektronix user and service manuals. Tek made some of the best service manuals of anybody in the business and Arttek Manuals has done an excellent job of scanning and saving them in pdf format. Most have been converted to searchable text format, which can save a lot of time. Prices are very reasonable IMO. I have a Tek 475 and a 2225 and I bought a full set of manuals for each. If/when things get squirrely they will save you a lot of time. Good luck with your new scope.

[wizard69]

Looks like you got your scope, so no sense in going into depth here.   However with used electronics realize that they could run another 10 years or blow a cap tomorrow.  The difference between new and used is often nothing more than a warranty.    For your indicated audio work any scope or scope like object would likely be OK.  I can remember when people in the hobby would have loved a 5MHz scope.   

I’m not trying to dismiss the value in modern DSO’s here.   They can be wonderful in many cases but also complete overkill. 

[james_s]

Normally I'd advocate buying a new DSO, but if there isn't the budget there isn't the budget, and as has been said before, the best scope for you is the one you can get your hands on. When it was new the 2465 cost a fortune and today it is every bit as capable as it ever was. People designed and debugged all manner of stuff using instruments that are primitive by todays standards.

[vk6zgo]

Update:


It looks like I'll need to get a higher-voltage (around 1kV?) probe at some point for checking waveforms on valve pins. A decent one from probemaster looks to be around $90, so yeah non-trivial indeed. Because I didn't go with a DSO though, I should have room in the budge later if I end up needing such a probe. I'm going to wait until I have a measurement I can't make without it before getting one though. Depending on the state of the oscillator, that may end up being soon.

You aren't going to need a 1kV rated probe for anything in that oscillator---DC HT is unlikely to be much over 350v, if that.

The tube Finals on some big guitar amplifiers just  possibly may be up towards 1kV, but nothing much else in the "valve" realm will be anything like that high.

If you start playing around in tube radio transmitters of reasonable power levels, all bets are off, however!

[ramenbytes]

Update:


It looks like I'll need to get a higher-voltage (around 1kV?) probe at some point for checking waveforms on valve pins. A decent one from probemaster looks to be around $90, so yeah non-trivial indeed. Because I didn't go with a DSO though, I should have room in the budge later if I end up needing such a probe. I'm going to wait until I have a measurement I can't make without it before getting one though. Depending on the state of the oscillator, that may end up being soon.

You aren't going to need a 1kV rated probe for anything in that oscillator---DC HT is unlikely to be much over 350v, if that.

The tube Finals on some big guitar amplifiers just  possibly may be up towards 1kV, but nothing much else in the "valve" realm will be anything like that high.

If you start playing around in tube radio transmitters of reasonable power levels, all bets are off, however!

Right, for the oscillator I think the highest voltage I saw was -160 VDC, 315 VAC on pins 4 and 6 of V5 so -475 V_peak assuming the VAC doesn't mean V_RMS and that I did the math right. If it actually is RMS though, then I think that comes out just over 600 V_peak. The other reason for possibily going with a 1kV probe is not for the higher voltage rating, but because probemaster's x100 probes just happen to be 1.2kV rated, and going with x100 means better scaling on the scope display. That is, my scope at x1 goes to 5 V/div max and has 8 divisions, so a x10 probe on a ~300 V_peak signal would take up more divisions (12 total) than are on the scope screen. Perhaps I've missed something though. I'll have to ask some questions on the repair board if/when I need to actually probe signals that big.

[Berni]

Yeah more of a thing is the x10 of the probe. Not all scopes have the same input voltage range at the BNC. But the 5V/div is a pretty common one that most scopes go up to, so yes that gets you 30V of range on 1x. You can run the channel offset off the screen and get something to show up, but how far that goes depends on the scope.

I would recommend getting a high voltage 100x probes for working on tube gear. I like being on the safe side and not running probes right up to there limits. You don't really need one for each channel, having just one is plenty.

[james_s]

Keep in mind the voltage rating of a probe gets derated with rising frequency. If I were poking around in vacuum tube gear I think I'd either use a HV differential probe or be extremely careful and check with a DMM first.

[vk6zgo]

Keep in mind the voltage rating of a probe gets derated with rising frequency. If I were poking around in vacuum tube gear I think I'd either use a HV differential probe or be extremely careful and check with a DMM first.

The thing is, when everything was "vacuum tube gear", ordinary Tektronix X10 probes were used every day for probing around in equipment.

Most of the high voltage stuff in everyday equipment, as distinct from Radio & TV transmitters & the like was DC or 50/60Hz ac, so well within their ratings, even allowing for derating.

Maybe not within the ratings of "El Cheapo stuff", although those are often surprisingly good.

[tggzzz]

Keep in mind the voltage rating of a probe gets derated with rising frequency. If I were poking around in vacuum tube gear I think I'd either use a HV differential probe or be extremely careful and check with a DMM first.

The thing is, when everything was "vacuum tube gear", ordinary Tektronix X10 probes were used every day for probing around in equipment.

Most of the high voltage stuff in everyday equipment, as distinct from Radio & TV transmitters & the like was DC or 50/60Hz ac, so well within their ratings, even allowing for derating.

Maybe not within the ratings of "El Cheapo stuff", although those are often surprisingly good.

Beware of AC coupled scope inputs, where the full voltage is across the scope's AC coupling capacitor.

David Hess mentions this frequently, e.g. https://www.eevblog.com/forum/testgear/probing-smps-with-oscilloscope/msg3126680/#msg3126680

[jmelson]

As far as analog scopes go, the 2465 is a heck of a nice instrument. You just have to hope it doesn't break down because I think those are the ones that use a bunch of irreplaceable hybrids.

Yes, they sure do!  And, also some very intricate switch components, but my guess on the hybrids is they come with X thousands of hours of life, and I will be hard pressed to use them up.

Jon

[vk6zgo]

Right, for the oscillator I think the highest voltage I saw was -160 VDC, 315 VAC on pins 4 and 6 of V5 so -475 V_peak assuming the VAC doesn't mean V_RMS and that I did the math right. If it actually is RMS though, then I think that comes out just over 600 V_peak.

Looking at those quoted voltages in the HP200CD manual, you naturally thought they existed at the same time, & were additive.

This is not the case, however.
Pins 4 & 6 of V5 are the twin anodes of a 5Y3GT full wave rectifier tube.

This link shows the pinout:
https://www.radiomuseum.org/tubes/tube_5y3.html

The type of rectifier circuit used with these devices is seen in this link:

https://www.angelfire.com/electronic/funwithtubes/3_Simple_Power.html

You will notice that the power transformer used has a centre tapped secondary.
These transformers were commonly referred to as being "so many volts a side".
In the second link, the transformer is quoted as "300V CT" which I take to mean "150Vrms a side".

In practice, the secondary centre tap is usually connected as the "common" of the HT supply, & in most cases to the metal chassis.

In operation, as the ac cycle progresses, each winding outer connection becomes positive w.r.t. the centre tap.
As this happens, the rectifier section which has its anode become positive, conducts, completing the circuit from the centre tap, through the load then the rectifier, & back to the outer connection of that half of the winding.
On the other half cycle, the other rectifier section, load & winding repeat the process.

It is tempting to just imagine the transformer secondary voltage & current staying as a nice clean sine wave at the rated rms voltage, with all the non-linear rectifier stuff down stream in the rectifier & load.

This is not the case, & secondary voltage & current contain a considerable DC component, as well as waveform distortion of the ac voltage.

The result is that the same voltage waveform is interpreted differently by the DC & ac ranges on a meter.

To add insult to injury, many DMMs, & certainly, HP's 1950s VOM are quite happy to lie to you regarding the rms voltage of waveforms other than perfectly sinusoidal ones.

[Berni]

Yeah sort of case of mixed AC and DC can be quite tricky for most DMMs. 

There are some modern DMMs out there that can be set to measure RMS voltage of only the AC part or to including the DC part too. But this is a pretty rare feature to find, i think the only ones i have that can do it are the two Agilent handheld DMMs (Forgot model numbers, one is OLED and the other is the next generation up LCD one)

Still if i was poking around such high voltage gear i would get a probe rated for high voltage. The regular probes probably have a good deal of safety margin that they would survive pretty much anything up to about twice mains voltage (apart from perhaps poking the output of a high power RF amplifier). But for the sake of safety i rather would not push the specs on something like that. Tho working with high voltage its more about having respect for it and being careful, no fancy probe can save you if you end up sticking your fingers into the 350V supply rail.

[vk6zgo]

Yeah sort of case of mixed AC and DC can be quite tricky for most DMMs. 

There are some modern DMMs out there that can be set to measure RMS voltage of only the AC part or to including the DC part too. But this is a pretty rare feature to find, i think the only ones i have that can do it are the two Agilent handheld DMMs (Forgot model numbers, one is OLED and the other is the next generation up LCD one)

Still if i was poking around such high voltage gear i would get a probe rated for high voltage. The regular probes probably have a good deal of safety margin that they would survive pretty much anything up to about twice mains voltage (apart from perhaps poking the output of a high power RF amplifier). But for the sake of safety i rather would not push the specs on something like that. Tho working with high voltage its more about having respect for it and being careful, no fancy probe can save you if you end up sticking your fingers into the 350V supply rail.

Of course, there is 350v & 350v!

The big fat high value electros in a SMPS just after the input rectifier can give you a much nastier "belt" than the fairly low current capacity power supplies in such devices as the HP 200CD.

If you touch the HT line on something similar to the latter, it will be quite disturbing, but seldom dangerous.

The real danger with old tube equipment is more centred on the "transformerless" design of domestic radios so beloved of 1950s/'60s North American, & to a slightly lesser extent UK & European manufacturers.
With those. you can get a nice "handful of Mains" if the Active & Neutral power connections are transposed.

[tggzzz]

The real danger with old tube equipment is more centred on the "transformerless" design of domestic radios so beloved of 1950s/'60s North American, & to a slightly lesser extent UK & European manufacturers.
With those. you can get a nice "handful of Mains" if the Active & Neutral power connections are transposed.

Or mains current, if there is a significant potential difference between neutral and the scope's earth. An HV isolating/differential probe might be a wise investment.

[james_s]

The real danger with old tube equipment is more centred on the "transformerless" design of domestic radios so beloved of 1950s/'60s North American, & to a slightly lesser extent UK & European manufacturers.
With those. you can get a nice "handful of Mains" if the Active & Neutral power connections are transposed.

They're not really that dangerous, the problem is the back cover is typically easy to remove and often missing, and many of them if a knob is missing the shaft can be energized. The transformerless design continued into the relatively modern age, most CRT TV sets have a "hot chassis" where "ground" is connected to one side of the line. The difference is that they have sturdy enclosures and any external metal parts are isolated. I have a couple of old transformerless tube radios and I use them. Just plug them into a GFCI and make sure they're not missing any parts.

[BrokenYugo]

Some are wired with one side of the cord straight to the chassis (and thus knob shafts) but as I understand it most transformerless series string radios and TVs float the chassis with a capacitor and/or resistor.

[vk6zgo]

The real danger with old tube equipment is more centred on the "transformerless" design of domestic radios so beloved of 1950s/'60s North American, & to a slightly lesser extent UK & European manufacturers.
With those. you can get a nice "handful of Mains" if the Active & Neutral power connections are transposed.

They're not really that dangerous, the problem is the back cover is typically easy to remove and often missing, and many of them if a knob is missing the shaft can be energized. The transformerless design continued into the relatively modern age, most CRT TV sets have a "hot chassis" where "ground" is connected to one side of the line. The difference is that they have sturdy enclosures and any external metal parts are isolated. I have a couple of old transformerless tube radios and I use them. Just plug them into a GFCI and make sure they're not missing any parts.

The last couple of generations of CRT TV sets all used SMPS.

There is absolutely no necessity to have a "hot chassis" as the output of the power supply comes from the secondary of a transformer, (albeit a HF one), so is just as much isolated as the secondary of a 50/60Hz mains transformer.

 All the SMPS powered TV sets I have seen in Australia are of this type, some had the SMPS as a separate module, whereas others had the power supply on the main board, but electrically isolated from the rest of the board.

The overwhelming majority of BW "tube type" TVs sold in Oz had mains power transformers (an "Admiral" "transportable" TV in the late '50s was the only locally made transformeress TV I ever saw)

The only "transformerless" tube radio sets around were a few "private imports" & very small quantities built for "niche" markets which still had DC Mains in the '50s/ early '60s.

The motive to build transformerless radios & TVs was not strong in this country.
Radio & TV sets commanded higher prices here, compared to some other countries, transformers were relatively cheap, & "series string" tube were not made here.

Such tubes would have had to come from overseas, at high cost, mainly from the UK or Europe, as series string tubes for 110v Mains would require an additional series dropping resistor.