so, Dave sez "get a differential probe!", it will save your life - maybe - but most likely will save your CRO and device under test.
I selected this modestly spec'ed one based on price and my needs.
I wanted good isolation of both differential probe inputs from my CRO, this is good for Cat 3 600V between either and ground.
I also wanted some decent amount of attenuation so as to view high voltage without blowing my CRO inputs.
The DP-25 has 200X, 50X and 20X attenuation settings, an over range light warning and it requires an external 9V 35mA supply.
When connected to your CRO and you select 1V/div and these attenuation settings will then be 200V/div, 50V/div and 20V/div
respectively.
These D.P. are made in Taiwan, seemingly OK prices for what you get. See
http://www.triosmartcal.com.au/oscilloscope-differential-probes/2303-pintek-dp-40lv-10mv-650v-differential-probe-40mhz.htmlor
http://www.globalmediapro.com/dp/A01LT0/Pintek-DP-25-Differential-Probe-25MHz-1000V/ (whoever
they are)
or ebay, where I bought mine.
You get a DP-25, two reasonably good quality banana style cables, big alligator clips, long narrow nose type clips, a short BNC-BNC cable and a 240VAC 9VDC plug pack supply.
I was surprised to see the power supply uses 9V +ve on the outside conductor, not inside as is usual.
I must confess to powering it with reversed polarity when I first used it and it survived. No smoke!
So that's good idiot proof design in the DC supply end....
The banana plug input is important to me as I planned to use the excellent quality Fluke test leads and clips, probes etc that come in their meter test lead kit in place of the crap you usually get supplied with cheap gear.
Pintek surprised me with the higher than expected quality of the supplied leads.
So what's it look like and what is inside?
It appears to generate it's own isolated power via a SMPS, to power the differential attenuator/amp circuit.
This uses a 2N3958 JFET (the TO-71 can package) and a LM6365 opamp.
Shielding is from a metalized coating applied to the inside of the ABS plastic casing.
The input attenuation and isolation are shown in the below photos. I can't comment on the design and build quality, maybe
others here can tell if it's OK or dodgy or whatnot.
I show a 800V p/p signal from a tiny HV supply from a photocopier light. This voltage is too high for the DP-25 and it shows "over range", but does not clip. It seems it will clip when the input exceeds approx 8 divisions (1 V/div on CRO) for all three attenuation settings.
So the maximum input before clipping seems to be +/- 800V. The manual clearly states +/- 500V as the maximum.
Finally I show the DP-25 compared with a normal old 10X probe, connected to the CRO's calibration square wave.
There is a fair bit of noise unfortunately, but I suppose that is to be expected. We are comparing noise from the JFET, opamp, unshielded probe leads etc., with a shielded piece of wire (10X 100MHz probe)
Test conditions: Ch1 DP-25 output, 10mV/div, 20X attenuation, Ch2 10X probe, 20mv/div, 500mV square wave cal signal from CRO.
So, there you go. The $250 differential CRO probe from Pintek.