Digital oscilloscope

[ppppenguin]

If you're mucking around inside earlier scopes please remember that they often have EHT of over 1kV derived from a mains transformer. Potentially lethal. I can assure you it hurts if you touch it. How do I know this? My arm ached for a couple of days from the violent muscle contraction which threw me clear and quite likely saved my life.

[Mark Hennessy]

Definitely the case for the older valve 'scopes, but not the Tek or Philips.

The Philips is quite interesting as it can run from 24V DC. The PSU is a conventional 50Hz transformer, rectifier and smoother making a nominal 24V DC, and that is fed - via a DC input socket - into a switcher that generates all the internal rails, including the HV and EHT stuff. Being able to run from external DC is "potentially" quite useful. Also, some are Class II - all our PM3215s were, but our PM3217s are Class I (I'm not sure if that's official, or a mod). Class II operation is obviously useful when earth loops exist, but potentially a safety hazard... From the manual, Rob's PM3305 appears to be Class I.

The Tek is a conventional linear supply, making 100V and +/-8V.

All that notwithstanding, I would hope that Rob knows not to run these with the covers off, and definitely do not adjust anything inside.

[ppppenguin]

Definitely the case for the older valve 'scopes, but not the Tek or Philips.

The scope that bit me was a Telequipment D61. All solid state. Don't think many scopes after that had mains derived EHT.

The switcher based EHT on newer scopes may sting a little but it's not lethal.

Your advice on not mucking around inside scopes, or any other kit, unless you know the hazards bears repeating.

[Mark Hennessy]

Yes, it's true that some early solid-state 'scopes did that. It feels like a transitional thing - not unlike early solid-state audio amplifiers that used transformers - as designers got used to the new technology. Fortunately, Rob's Tek and Philips 'scopes are well after that era, but it's definitely something that could catch you out - assume nothing! To be honest, I'm afraid of the circa 100V rails that many 'scopes have for their X and Y output stages

[Bushbaby]

I will hang on to the Tek scope but will be letting all of the telequipment scopes go now.

Best Regards

[BusyBee]

While agreeing totally with the above guidance on the high voltages I think it can be a good assumption to make that most anything with valves (which with a scope there is an obvious large one) will have some potentially nasty voltages about. Another point I would make is that in my experiences I have noticed that the high voltages seldom kill (but take heed as they can). It is the instinctive reaction which results in nasty gashes from metal chassis and similar which are the most common problem. From that point any higher voltage can be an issue.

I will re-iterate the original fact that 1kV mains derived EHT is likely to be lethal as that is important to know. Personally, although I seldom work with them, some valve equipment with 250V and higher supplies will not be good either. Maybe the danger here though is to make the false assumption that solid state equipment is low voltage. That is often not the case, as in, for example, scopes, CRT TV and higher power amps (particularly RF but thats another subject in itself).

Tracy

[Mike Watterson]

The Hameg scope I have uses 485V AC on the transformer and a diode multiplier to get the approximately -2000 V EHT. There is an opto-isolator for Z-modulation as the cathode circuitry is about -1950V. The brightness uses a 62V "Zener" to drop this. I discovered this when it failed. A 741 op-amp failed at the same time.

There is curiously all the circuitry except the rear panel BNC for Z in. There is a test point on Timebase section via about 560 Ohms, so I also added a BNC on that. The rear panel is designed for three BNC connectors. I'm tempted to put a buffer to Y1 just before the chopper/invert/Alt part of the circuit so as to drive a frequency counter and/or DMM as I often want to measure frequency more accurately than counting divisions on the screen with a signal not suitable for the counter.

I'm also tempted to build a powered very low capacitance probe to look at tuned circuits without pulling them off. Some scopes have an option for this. The 10:1 probe loads too much and 100:1 probe would not be enough signal, even should it be low enough capacitance.

I have a fear of exploding electrolytics. One on a valve TV exploded putting bits of aluminium in the wall when the BY127 failed, about 1972. But even a 16V tantalum can put a glowing core across the room. Sometimes I step aside before inserting mains plug!

[ppppenguin]

Rob has made a sensible decision to keep the Tek 935 and get rid of the older scopes. The 935 will do everything he needs to do with a scope and do it well too. I'd be perfectly happy with one for vintage radio and TV work. I have more advanced scopes because I'm doing professional design work. I sometimes look wistfully at the lovely Tek 547 with 1A4 plugin https://w140.com/tekwiki/wiki/547 that sits under my bench. Lovely instrument but almost never comes out to be used. Probably worth more stripped for valves (lots of ECC88) than as a scope. Would be sacrilege to do that.

As Mike has pointed out, the cathode(and grid) on most scopes is at a high -ve voltage. This is so the deflection plates can be at a sensible voltage, rather than a high +ve EHT. A few pre-war TVs with electrostatic tubes ran the deflection plates at +ve EHT so that the video could be coupled to the cathode or grid at a low voltage. It meant using 4 high voltage capacitors to couple the timebases to the deflection plates. OK in a TV with fixed timebase speeds, not OK in a scope.

Mike also mentioned adding a Y output socket. Some more sophisticated scopes such as Tek 24xx have a buffered Y2 output (Why Y2 and not Y1? I've no idea) . I often connect mine to a video monitor so I can see video that I'm probing. I lock the monitor to my lab SPG so there are no sync problems.

THose little bead tantalums can go bang with remarkable violence for their size. I've seen one hit the ceiling 6 feet above the bench.

[Mike Watterson]

Also a low ESR regular electrolytic on a programmable linear regulator giving 13V doesn't go bang when the regulator decides to "sing" at a Megahertz or so. The 35V bead tantalum lit up bright orange.

So maybe I should use Y2. Though the channels are identical. It's differential there, the output of the ancient uA733 amplifier? Not sure if a buffer on one "leg" will do or if I should put something exotic. A flat DC to 20MHz buffer may want more than a FET follower!

[Mark Hennessy]

The Hameg scope I have uses 485V AC on the transformer and a diode multiplier to get the approximately -2000 V EHT.

That's a good point - I'd forgotten Hameg did that. The only paper manual I have here is for the HM203-6 from the mid '80s, but that one definitely was mains-derived. I guess they decided that as the EHT was only 2kV overall, it wasn't worth going down the HF PSU route

I remember these scopes being pretty reasonable for what they were when we used them at university, especially given how low the EHT was. The 60MHz version (HM604 IIRC) was much brighter/sharper though - that was 12kV, with the expected separate HF PSU.

From the point of view of the beginner, my favourite thing about the Hameg 'scopes was the green "triggered" LED. As we know, triggering is often what causes most confusion for people new to the 'scope, and this goes a surprisingly long way to help. And it's effectively free, as you've got some form of trigger detection monostable in there for the auto "BBL" timing, so I've always felt it's a shame it's not more widespread. Also from the beginners POV, I really liked the triangular "overscan" LEDs that showed when the trace was shifted off the screen, but those obviously require a bit of extra circuitry, so not all models got them. Secretly, I just like LEDs