# Oscilloscope Probe Noise and Safety

I have a Tektronix 531A tube oscilloscope from the late 50s. I recently bought some 20$probes and noticed a lot of noise on the 0.005 volt/cm setting when that spring loaded hook attachment was clipped on. When I bring my hand in closer to the plastic of the probe the ripple on the scope goes crazy and when I'm actually holding the plastic, it goes off screen on the 0.005 volt/cm setting. When the spring loaded hook attachment is removed the noise level is normal and when I touch the scope's chassis with the same hand the probe is in, it returns to normal levels as well. My question for this part would be if this was to be expected from a 20$ probe and if there is any way I can get less noise with this clip. One thing that I should've mentioned at the beginning was that my house is old and has no mains ground in any of the wall sockets. I mean the hole is there but it was never wired when the new standards came in. I have done my best to ground the machine from its ground clip in back to my radiator, which has solved a lot of noise problems already on the higher settings like 0.05 volts/cm. Now its just the noise at the 0.005 volts/cm as mentioned above.

This brings me to my second and last question. Is it safe to take measurements with a "floating" oscilloscope that has no mains reference/grounding? Will people be shocked if they're next to that radiator now? I have read that if your machine is grounded, it is safe to measure battery powered equipment, isolation transformer stuff and basically anything else that doesn't have a ground connection where the circuit can be completed.

But what if your scope is not grounded? (Haven't found much helpful information on this for such an old scope as mine although I have seen some for the newer scopes) Is it still safe to measure battery powered and isolated stuff or is that now dangerous?

P.S. I only do this for a hobby. I have basic understanding of electricity and use my tools to repair tube powered equipment where voltages are in the 500+- range. Is there any dangers I should be aware of and any fail safe techniques that I could use to save my oscilloscope because its the only one I've got.

• Without an earth ground any appliance that is 'leaky' may shock you. Controlling noise in sensitive equipment will prove difficult. The correct solution is to have an electrician run grounds wires to all outlets. That is also a very expensive solution.
– user105652
Jun 30, 2016 at 19:53
• Just what is your house wired with? Knob and tube? Old Romex without grounds? BX? Conduit? Jun 30, 2016 at 22:24
• I went down in my basement and I opened an access hatch in the ceiling down there and found the Romex type wire running through the floor supports. I also found a plug behind some stuff that looks exactly like this one: [link]c03.apogee.net/contentplayer/templates/foe/sg2op.jpg Jul 1, 2016 at 19:39

5mV per div is a really small scale, unless you've got really good grounding and a low impedance ground connection (that clip-on fly lead isn't a good ground at higher frequencies) you'll see noise as the loop area in the fly lead basically acts like an AM antenna (which are just wire coils too) and it picks up all kinds of crud from the airways, hence all the noise you're seeing on the scope.

A floating scope isn't necessarily a hazard provided that there isn't a low impedance path from mains to come into contact with something or someone. It's when there's unacceptably high leakage in the power supply that you'll have a problem as it'll allow enough mains to get through that it might give someone a zap (you could try doing a current measurement from scope ground to earth - like a tap or something, I'd be weary if it reads more than a few uA). Almost every laptop and phone charger in existence is floating and will have no direct connection to mains earth so floating supplies aren't a hazard on their own, provided that they're built properly.

Battery powered or isolated stuff is safe to measure with just about any scope as there's no direct path to ground, it's when you have non-isolated stuff (and a non-floating scope) that you can run into problems, I've seen the odd mildly exploded scope because someone was measuring the output of a mains rectifier and connected scope ground to rectifier ground (which will actually be -1/2 mains voltage below ground, hence the explosions). Having said that, if you've got mains leakage then your battery/isolated item will now be floating around mains voltage so be careful.

500V is probably getting into the realm where I wouldn't trust most cheap scope probes to handle the voltage well, there are plenty of HV probes which offer an extra level of safety through higher levels of isolation. One of the best things you can do is to stand on a nice rubber mat and keep one hand in your pocket when probing the circuit, that more or less eliminates any direct current paths through your body. Current from one hand to the other is the most dangerous as the current will be passing through your heart.

• I just measured the voltage from the radiator to the GND plug on my machine and I'm getting 96 volts ac. I was shocked by this current while attaching the wire to the radiator and it felt very unpleasant. You can also see sparks when the wire is touched against the radiator. I couldn't get a reading on the amperage because both my multi-meters have had their amp settings fried (by me, by accident of course) but I believe that it is significantly higher than a couple of micro-amps. Possibly in the low milli-amp range. Does this change anything? Jul 1, 2016 at 19:57
• a couple of mA at ~90Vac is not very safe, while it might not quite be enough to give you a heart attack, that's still enough to potentially cause some arrhythmia, I'd look into getting the earth connections fixed or having a dedicated ground point installed in you lab. A couple of mA at 90V is more than enough to totally fry most mosfet/jfet based electronics unless you're really careful.
– Sam
Jul 1, 2016 at 23:35
• Thanks for all the help Tom. Are you saying that if I try to measure transistors with my scope that they will fry because I have this 96 volts floating around the chassis of my scope? Jul 3, 2016 at 17:30
• If what you're probing is well isolated then you should be alright, but stray capacitances can still couple in unwanted voltages. If you're working with BJTs then you should be ok as they've effectively got a diode in parallel with the base terminal so it's near impossible to feed in more than a volt or so into the base, HV on the collector may cause breakdown (where it acts like a zener from collector to emitter) but a lot of BJTs handle that pretty well too. Mosfets and CMOS logic is what you need to be careful of, it doesn't take much to obliterate the gate oxide layer in those things.
– Sam
Jul 3, 2016 at 22:48