Buying and using an oscilloscope

[ppppenguin]

£68 seems a bit high - similar instruments have sold at the BVWS Harpenden meetings for £20 to £30 - but it's not outrageous.

The kinks in the sinewave are likely to be from your generator, not the scope. If your generator is one that does sine, square and triangle waves then the sinewaves are usually distorted at the peaks.

Yes, you can use any convenient square wave to adjust a probe, but you need to be confident that it is a decent square wave. Easily done by connecting it directly to the scope without a probe.

[Bonky]

Thanks Jeffrey.

[Refugee]

On the Y amplifier controls have you checked that the variable pot is set to cal?
It will make a big difference to the gain.

[Mark Hennessy]

The inbuilt sq wave generator for calibrating the probe didn't work (and still doesn't). The output is meant to be 0.2V but was only 0.1. I've adjusted the trimmer on the pcb to its max (but only 0.125V).  Has anyone any advice?

So it is working, but not giving the expected output voltage? There's a world of difference between the two...

I'm sorry to say this, but by twiddling that trimmer, you have made a fundamental mistake here. You are new to this, so this is forgiveable, but it's really important that you take this on board and learn from it.

Before twiddling that trimmer, you made a huge assumption. You assumed that the vertical sensitivity of the 'scope was correct. That was an invalid assumption. You know nothing about the pedigree of that 'scope, so you have no reason to believe that any part of it is correctly calibrated. You placed your trust in something that wasn't deserving of it.

This is a big deal. You need to trust your instruments, and that trust must be earned. In other words, assume nothing!

If that seems like I'm telling you off, I'm not really - well, perhaps only enough to get your attention! But please, take the lesson from this - no mistake is wasted if you learn from it. Don't become the "Phantom Twiddler". Adjusting pre-sets is the last thing to do, not the first. Especially in a bit of test gear.

And of course, once you've found out what the problem is, you'll have to go back and readjust the calibrator preset. But how will you do this with the required 1% accuracy? Think about that one...

You haven't described why you believe the calibrator is low. Perhaps you do have good reason to suspect it, but you haven't explained that - therefore I'm assuming nothing.

So is it the chicken or the egg? Or both?

An important check - is the error the same on both channels?

If so, then it might be the calibrator that is genuinely low, as the two channels are fairly separate internally, and it's unlikely that both channels will have the incorrect gain by the same amount. However, that's not conclusive, because actually, the two channels do share a common amplifier after the beam switching. But nonetheless, it's a check that should be done.

You have a multimeter that you trust? And you can lay your hands on a battery? If so, you have the means to check if the problem is with the generator or the input channel of the 'scope. Remember, 'scopes measure DC as well as AC

Ref's suggestion of checking the variable gain controls is a good one - a common mistake that we all make from time to time. Some 'scopes light a red LED when the control is not in its calibrated position, but this one does not.

Let's think more deeply about this...

The calibrator circuit is a CMOS IC running from 5V. The 5V rail is likely to be OK because if not, there would be many other problems with the 'scope. However, it's easy to check, and would be the first thing I'd do if I knew the calibrator was the culprit. Next, the 4011 CMOS IC making the signal is generally reliable in service, but when chips fail, they tend to fail completely. If you were unlucky, it's possible that a failure of 1 of the 4 NAND gates (either a or b) could cause the output signal to half in this particular circuit. It's easy to check - just make sure that you have a 5V square wave on pin 3 and pin 4 of IC2001. For anyone not looking at the service manual, each of these two square waves goes through a 15k resistor to a summing point which has 270 ohms and the 100 ohm preset going to ground, and this is the output...

So with this particular circuit, there is a possible failure mode that could result in the signal being half the size you expect. That's relatively unusual. Whereas when it comes to a gain error in a vertical channel, that's ordinarily much more likely. There is a lot more circuitry, switch contacts and many more pre-sets to play with (DON'T!), so on purely statistical grounds, that's where you tend to point the finger of suspicion in the first instance. As I say, a battery and a multimeter is all you need - literally a 30 second job.

Final point. Despite the name, the calibrator isn't there to calibrate the 'scope. The voltage accuracy of the calibrator might be quite good (1% in this case), but often they aren't as good as that, and they are intended to provide a quick check but nothing more rigorous - calibrating the vertical channels of a 'scope requires something with more precision (the spec is 3%, so as a rule of thumb, a calibrator ought to be accurate to around 0.3%). The primary purpose of it is to check the HF compensation of the probe. That much you might already have gathered, but it's worth writing down for anyone else in a similar position.

I hope all this helps. It will need a couple of re-reads, but if nothing else, try to take the general guidance - try to understand the thought processes that I'm explaining. Yes, specifics are important, but one of the most important things about becoming an engineer is to develop the right mindset - that's what enables us to take on new challenges and problems.

[Craig]

FWIW on Tektronix scopes the calibrator has a jumper that puts the circuit into DC mode by disconnecting the feedback on the 1kHz oscillator. That enables you to use a DVM to accurately set the level. Then when the jumper is moved to a "parked" position the oscillator starts up again - but now the square wave amplitude is accurate.

[Bonky]

Thanks everyone, especially Mark for his detailed response(s) (and admonishments).

Being a newbie does reflect my ignorance I know but I'm not entirely incompetent either. Both my degrees are in Biology rather than electronics/physics (the maths was a bit above me at the time). It may appear I've taken a scattergun approach to this and others but I'd like to think I have been a bit more focussed than that.

Anyway, the calibration.

Are we talking about the same thing here? On this 'scope on the bottom left of the panel is a small copper protuberance on which one hangs the end of the x10 probe. Pressing the 'cal' button is meant to produce several traces on the screen and one uses a trimmer tool on the probe body to get the traces as they should be (the manual is very good here). Only the focus and intensity controls work at this setting; nothing else does.

I do NOT get ANY traces on the screen. Testing the output on the copper protuberance is meant to show 0.2V; it shows half this, hence my diving into the set to alter the trim pot hoping to see if an increase in voltage produced a trace; it didn't. The circuit for this in-built sig gen seems relatively simple and as there is a cheap-as-chips IC there, that was going to be my first call in the investigation. Unfortunately I'm on holiday for 3+ weeks and did not take the 'scope with me (my wife baulked at the inclusion of a Brompton...she's have gone ballistic if I'd have packed a 'scope as well!).

Obviously I'll investigate further on my return.

I am NOT using the internal trim pots to alter the X/Y settings.

BTW. I've just won an auction for several dilapidated radios which I'm going to use as guinea pigs.

Thanks again,

Richard.

[Terry]

Pressing the 'cal' button is meant to produce several traces on the screen and one uses a trimmer tool on the probe body to get the traces as they should be (the manual is very good here). Only the focus and intensity controls work at this setting; nothing else does.

What 'cal' button?

Nothing else works?

Not according to the manual!

Probe adjustment

The TIMEBASE switch should be in the 0.2ms/cm position. The input coupling is set to DC. If the attenuator sensitivity is set to 5mV/cm (variable control to C), the display will have a height of 4cm when a x10 probe is being compensated.

That doesn't agree with what you wrote! However:

Component Tester

By simply pressing a single pushbutton, the HM203 can be switched into test mode without actually affecting the oscilloscope measuring set-up.

Now that sounds more like what you are describing!

   

There is NO 'cal' button!

As for your half-amplitude calibration waveform, that is correct! The waveform is a 50/50% square wave so it only sits at 0.2V for 50% of the time. Your meter is averaging out the signal it sees so, correctly, displays 0.1V, which is 50% of 0.2V.

If viewed correctly on the screen it would measure 0.2V peak.

[pwdrive]

This might have been covered before but here goes anyways, Cal and Cal, first one...Cal 0.2 volts, you've already found that function on the scope, the bit you hang the probe on for probe cal (calibration) the other is  also Cal but is shown as C on your scope, there are three of them, one on each of the Y amplitude controls and one on the horizontal time base control, all three being the variable controls not the selector switches.

Taking the Y amplitude cal first, there's two of them one for each channel, channel 1, turn the control anti clockwise so the white pointer is in line with the C shown on the V/cm scale, channel 2, do the same. That now means that the V/cm switch settings are calibrated to the vertical divisions on the screen, eg: if the V/cm switch is set to 2 volts and the amplitude of the waveform on the screen occupies one complete division then 2 volts peak to peak is what you have if a X1 probe is being used.

The same setting procedure applies to the horizontal time base, turn the control anticlockwise so the white pointer points to the C on the ms/cm scale. That now means that the ms&us/cm switch settings are calibrated to the horizontal divisions on the screen, eg: if the ms&us/cm switch is set to 5ms/cm and the time period of the waveform on the screen occupies one complete horizontal division (X mag. off) then a waveform of 5ms time period is what you have, if you want to know the frequency of that then use  f=1/T where f is in Hertz and T is in seconds...eg: 1/0.005 = 200 Hertz.

Hope that helps.

""I do NOT get ANY traces on the screen. Testing the output on the copper protuberance is meant to show 0.2V; it shows half this""

What are you measuring it with?

This post might cross, any errors then guilty as charged.

Lawrence.

[Bonky]

Humble apologies. Too much sea air and out last night celebrating what would have been my father's 100th birthday obviously took a toll on my mental faculties as I cut and pasted: " Only the focus and intensity controls work at this setting; nothing else does."... which was, of course, to do with the component tester - which is now working - and not the probe calibration.

""I do NOT get ANY traces on the screen. Testing the output on the copper protuberance is meant to show 0.2V; it shows half this""

What are you measuring it with?

A standard 'Maplin's like' multimeter - which does seem accurate (it's not a Fluke however)..Thanks Lawrence for the tips...but I cannot do anything for 3+ weeks as Sidmouth is a long way from Sheffield. How frustrating!

"As for your half-amplitude calibration waveform, that is correct! The waveform is a 50/50% square wave so it only sits at 0.2V for 50% of the time. Your meter is averaging out the signal it sees so, correctly, displays 0.1V, which is 50% of 0.2V.

If viewed correctly on the screen it would measure 0.2V peak."

There are no traces on the screen when the probe is attached to the 'protuberance'. I can get a similar trace to that mentioned in the manual, however, by using an external sig gen set for a sq wave.

I'll experiment when I get back.

Thanks again for your combined patience and expertise.

Richard.

[Murphyv310]

Hi Richard.
Slightly off topic. Mark has evaluated the Aneng AN8002 and the 8008. These are incredible meters for the money and very accurate, ideal to compliment your scope. I've both and use them most of the time.
Good luck with your scope, playing around with them, feeding different signals will give you the feel of using a scope. Good luck.