Voltage Divider for Oscilloscopes

[StevenPituch]

Hi All,
I've done my research, but haven't quite answered this one for myself.

I just ordered a Hantek DSO5102P Oscilloscope.  I have been trying to educate myself on safe voltages for oscilloscopes.  It seems 400 Volts is a maximum that is often quoted.  That is about the DC voltage on the plates of the power tubes on my AB763-2 amp.  I believe using a 10x probe I should still use a one-to-one voltage divider to safely measure this high a voltage.  I will being measuring the AC signal but suggestions still say to consider the 400 Volts DC.

Can I just use two 1 megOhm 1/2 Watt resistors to make a voltage divider and just measure half the existing AC voltage?  If so, I wonder if it would be OK to leave them in the circuit for future measurements.

I know I can just use a voltmeter to measure the DC, but I am interested in seeing the signal wave forms through the amp to educate myself on distortion.

Regards,
Steve Pituch, W2MY

[shooter]

look this over and see if it makes sense

https://how-to.fandom.com/wiki/How_to_make_a_100X_oscilloscope_probe

I didn't prof it 
I did make a 10 R string version, no cap
eventually broke down n got a tek 100X, with cap 

[jjasilli]

"The “old school’” approach is to put a large capacitor in series with the probe to reject the dc component so that the [AC] signal could be centered on screen and zoomed-in on for analysis."  See:  https://www.electronicdesign.com/technologies/test-measurement/article/21801607/11-myths-about-oscilloscope-probes. Item 7

[shooter]

fwiw

I quit measuring AC at the PA tube plates in favor of the speaker.  I measure "left" of G1's coupling cap, then move to speaker for everything but hard down amps.
I get a "closer to reality" measurement at speaker, the math is easier there also.

[pdf64]

When overdriven, the Vac on power tube plates can exceed 2kV. A 100:1 high voltage probe is required.

[DummyLoad]

https://www.digikey.com/product-detail/en/cal-test-electronics/GE3421/BKGE3421-ND/6041716


good to 4KV - will work fine for tube amp measurements.


--pete

[StevenPituch]

Thanks everyone for the clarification.  I ordered a 100x probe.   I’m guessing the procedure should be to work slowly, measure the dc voltages at the grids with a voltmeter to assure they’re within the capability of the probe, and then use the scope on the grids.   Also avoiding the power tube plates and probing instead at the speaker side of the transformer.  Makes sense when I studied the schematic. 
The important thing is to verify before using the scope.

[DummyLoad]

the 100x probe would allow to scope the plates, depending on the input scaling that your scope supports, and the operating voltage of the device (your amps B+). however still, your scope has a 400V absolute input limit spec, so then a scope with a 400Vpk limit would have: 400V x 100 = 40KV. 

most old(er) tektronix and HP scopes have a 50V/div scale and that will allow up to 5KV/div DC+pk. the probe i linked to is rated for use up to 4KV. most measurements you would make will be with AC coupling.

aside from the scopes' input rating, the other important spec to be aware of is the MAXIMUM operating voltage rating of your probe: that is the maximum voltage of the body and the divider components will withstand. it is not wise to exceed those ratings.   

--pete

[sluckey]

most old(er) tektronix and HP scopes have a 50V/div scale

No. that should be 5v/div scale.

[PRR]

Even a 100x probe may have a 400V compensation cap, and be unsuitable for poking power plates.

For most purposes, an audio transformer is VERY tightly coupled. Look at the secondary and multiply by turns ratio to picture the primary.

Or since it is medium Z and medium bandwidth, roll your own. 100k:1k is 100:1 and likely to be flat far beyond the audio band without capacitance compensation. Yes it it a heavy load on a preamp plate, but not at a power plate.

[DummyLoad]

most old(er) tektronix and HP scopes have a 50V scale

No. that should be 5v scale.

depends on how old slucky: my 500 series stuff has 20V, the low BW one 10V. at 1x. 

[shooter]

WOW, I'm in love!
the bench scope in the Navy went to 20v/div but it was 3 times as wide as that, we called her bench bertha 
we made calibrated 100ft cables because NOBODY wanted to move her when the "portables" were out for cal 

[pdf64]

...since it is medium Z and medium bandwidth, roll your own. 100k:1k is 100:1 and likely to be flat far beyond the audio band without capacitance compensation. Yes it it a heavy load on a preamp plate, but not at a power plate.

Given the high voltages developed at the OT primary and the <=500V limit of many common resistor types, it may be a good idea to use several resistors in series for the 'series dropper' resistive element (eg the 100k bit) of such a potential divider.

[DummyLoad]

WOW, I'm in love!
the bench scope in the Navy went to 20v but it was 3 times as wide as that, we called her bench bertha 
we made calibrated 100ft cables because NOBODY wanted to move her when the "portables" were out for cal 

i had the chance to buy a pallet of scopes with several of that model. most were 545 with dual input plug-ins. i still regret passing on the offer, funds weren't scarce at the time either. i was still renting so i didn't want to piss off the landlord with a pallet of "junk" in the already full of junk garage which was basically VAX1170 in 3 racks, a pair of RM05 disk drives, several disk paks. 
 
--pete

[DummyLoad]

Even a 100x probe may have a 400V compensation cap, and be unsuitable for poking power plates.

the one i linked to is RATED to use to 4KV - is that an issue with 400-500V B+?

--pete

[PRR]

> ...use several resistors in series
> the one i linked to is RATED to use to 4KV

Points made.

I still think 99% of what you need to know can be seen on the 27V secondary instead of the 900V primary.

[DummyLoad]

> ...use several resistors in series
> the one i linked to is RATED to use to 4KV

Points made.

I still think 99% of what you need to know can be seen on the 27V secondary instead of the 900V primary.

100% agree. if i had that tek310 or a tek 545a with that probe i liked to, i wouldn't be as apprehensive about looking at plate swing.

an old friend here had the 545a with the tek 20KV 1000:1 probe - it was pretty cool to see plate swing on a scope.

--pete

[sluckey]

Ah, you talking about the real old Tek scopes. I cut my teeth on the 545A. A real tube scope! Loved the scope even though the trigger stability was usually a pain in the ass when checking rise/fall time on a 1µS pulse. We had six on carts at the first radar site I worked at back in '68. We all knew which one to grab when we needed to look at the fast stuff. I had forgotten the vertical went to 20V/cm on those scopes.

[pdf64]

...I still think 99% of what you need to know can be seen on the 27V secondary instead of the 900V primary.

Totally agree, but actually seeing the enormous magnitude of the voltage spikes on the power tube/s is a salutary lesson 

[shooter]

a pain in the ass when checking rise/fall time on a 1µS pulse

YUP, we called it "averaging" 

even in the early 2000's calibrating rise_time on switch mode DC power supplies was near impossible even with Gig/sec digital scopes (that weighted 2 pounds :)

I cannot even fathom a cart full of 545s, what'd you use diesel motor carts/