How to test an O-Scope?

[newguy]

Hi,

I recently purchased a used oscilloscope (arriving today, if the UPS gods are kind to me), and I'm curious what if anything I can do to test it out over the holiday weekend to see if it is calibrated correctly, working, etc.

The obvious ones that I thought of (once for each channels):

* Hook it up to the square wave output calibration thing on the scope itself, to make sure it shows a clean trace

* Hook it up to a known voltage source (checked w/a DMM) to make sure that what it shows is calibrated consistently

* Download a signal generator to my laptop and run some square, sine, triangle waves off the soundcard

... anything else? I have the XYZs of oscilloscopes book from the Tek site, but am basically a complete novice, so please bear with me if this is a really dumb question.

Hopefully they ship the probe it is supposed to come with!

[Sch3mat1c]

What kind, Tek 465/75?

Voltage standard is easy. Frequency is harder. Sound card is crystal controlled so it shouldn't be that bad for calibration.

The specs on my 475 say the squarewave test loop thing on it is only 10% or so accurate. It also has soft corners due to the way it's produced. You should really test it with a squarewave which has very sharp corners, and goes up to a few MHz. If you see any peaks or ringing on the wave, the vertical amplifier isn't tuned right, and has an RF peak somewhere. It should be nice and square. Note that in the MHz range, hookups are critical and you may be seeing ringing of the wires instead!

Tim

[newguy]

Thanks Tim!

I ended up going with the Tek 2221, hopefully not a mistake ... had read good things about the 465 and 475 as well.

For some reason I've seen the 2221 listed as both 60Mhz and 100Mhz. The auction described it as a 100Mhz unit, not sure if it depends on manufacture date? (Or was the guy just blowing smoke?)

[Bricolo]

Anybody knows how to calibrate a Tek P6109 probe? They have 2 pots, I know hot to calibrate a probe that has 1 knob, but 2...
And I prefer not to try before knowing what I'm doing

[jackinnj]

there's not much you can do to "test" a 2210 if you don't have a lot of very high cost equipment -- and at any rate you won't be using it to make ultra-precise frequency or voltage measurements, it's the stuff in between -- like watching the tops of waveforms clip off, comparing the phase between a pair of signals --

[newguy]

there's not much you can do to "test" a 2210 if you don't have a lot of very high cost equipment -- and at any rate you won't be using it to make ultra-precise frequency or voltage measurements, it's the stuff in between -- like watching the tops of waveforms clip off, comparing the phase between a pair of signals --

Fair enough, I guess it was kind of a silly question...just wanted to make sure that I looked it over for any obvious failures out of the box. Friends of friends have had trouble with used scopes.

Good news is the scope arrived, with a switchable 1x 10x probe, trace is bright and both channels pick up the cal signal just fine (slightly soft edges, consistent with Tim's point above). It is indeed a 100Mhz scope.

Only started to play with the digital storage / measurement features, which seem really, really neat.

Just brought home a laptop from the office, so I'll see how square and triangular waves look tonight. Of course, I have no reference to compare it against ... is it safe to assume that any square wave quirks (ringing at the edges, tilted tops, rounded "corners") are due to the laptop? I'm assuming so.

Happy holidays!

[jackinnj]

on the front of the scope is a small jack in which you insert the scope probe -- this should have a clean square wave out -- you adjust the trimmer on the scope probe until the square wave is symmetrical.

occasionally you will see "Time - Mark" generators for sale on EBay. I have a mil-surplus Ballantine -- they have extremely quick rise-time so you can check the timing on the scope to nano-seconds.

the 221X series don't like to be banged around, by the way. It's a lab scope, not a service-call scope. If the waveform is badly skewed -- like an old TV set out of adjustment -- you may need to replace the CRT.

[Sch3mat1c]

From a soundcard output, you might find various noise (picked up from the noisy environment it's in) and limited frequency response. For instance, to filter DAC and sampling noise, SCs have a low-pass filter. But say you play a 10kHz square wave - this has its first harmonic at 30kHz and second at 50kHz. That's only two out of at least 10 needed harmonics being passed, and will be an ugly thing to look at. Down by 1kHz it should be better, since you have what, the 43rd or 45th order harmonic being cut off. Amplitude and linearity should be okay unless it's a really crappy card.

Tim

[BC108man]

Here are a couple of other things you could look for:

Channel Matching - Use the digital storage feature to freeze the Cal signal onscreen, then feed it into the other channel and view it live there. Check that they both have the same amplitude and waveshape.

Noise - With the input grounded, and the gain turned all the way up, you should see no more than 1 or 2 minor Divisions of noise on the trace.

As others have said, the best test waveform is a square wave. It's actual frequency doesn't matter (even 1KHz is fine) but it must have REALLY fast edges. They contain extremely high frequency components, and are the best way to evaluate a scope front-end. Try to borrow an Agilent or LeCroy Pulse Generator, if you can get your hands on one. Otherwise, the scopes cal signal will at least show you if there's anything seriously wrong.

Have fun. And was it a 2210 or 2221? They're pretty different beasts.

Cheers, BC

[newguy]

the 221X series don't like to be banged around, by the way. It's a lab scope, not a service-call scope. If the waveform is badly skewed -- like an old TV set out of adjustment -- you may need to replace the CRT.

I think I saw something about this on one of the earlier threads - related to the mesh on the screen, right? The picture looks crisp and clean, so it looks like I lucked out with respect to the CRT.

The waveform off the internal calibrator looks really clean, and it triggers very smoothly off that signal, so the basics seem to be working well.

I hooked it up to my soundcard, and had somewhat mixed results:

Square waves (even at 1 Khz) were really ugly, but I'm 99% sure that this is a result of the soundcard / software and not the scope. Rise time was poor, and it was easy to see the "humps" at the ends of each pulse indicating the missing harmonics.

I also looked at sine waves at a variety of frequencies. They were super noisy, but again, I'm 99% sure that this is a result of the soundcard / software - putting the scope into storage mode and activating the "average" function yielded a clean trace, and I was able to use the cursor to get back to approximately the right frequency by identifying the peak and trough of the wave.

Now to the part that concerns me a bit...there were times that the scope seemed to get "confused" on where to trigger. It would work great 9 times out of 10, but sometimes if I changed the time division it would split the trace into two overlapping waves. Should I be concerned about this? I could always get it to trigger correctly by tweaking the trigger voltage or switching between the two trigger modes (apologies, I forget the terms, something like "P-P" and "Norm") ... I suspect that I've just got it in the wrong setting, but I was confused on why this would happen.

BC108man: clever suggestions, thanks! I'll give those a try tonight and see what I can find.

The scope is a 2221A ... just found out that I have the manual for it on CD-R as well!

PS: I just noticed how long this post is ... thanks to anyone who reads this far!