# Oscilloscope Amplitude Calibration

Discussion in 'General Electronics Chat' started by ELECTRONERD, Oct 14, 2009.

1. ### ELECTRONERD Thread Starter Senior Member

May 26, 2009
1,146
16
Hey Everyone,

I have an old Tek 7904 oscilloscope, and I need to calibrate the amplitude on it so I get an accurate reading. What I'm wondering is for my meter; when I select the setting to AC voltage measurement, do I get the RMS value or the peak value? I'm hoping that it's the peak value, since that is what my oscilloscope reads. Just nead a verification!

Thanks!

2. ### beenthere Retired Moderator

Apr 20, 2004
15,815
283
Peak, of course. The o'scope should have a calibrator output to check the probe compensation and vertical deflection, too.

3. ### Audioguru New Member

Dec 20, 2007
9,411
896
Your Multimeter reads the "average" of a 50Hz or 60Hz AC voltage. It is calibrated for reading the RMS value of the low frequency sine-wave.

I made a Cmos oscillator with an accurate 5.0V supply for calibrating the low frequency and higher frequency p-p input of my 'scope.

4. ### ELECTRONERD Thread Starter Senior Member

May 26, 2009
1,146
16
That's what I was hoping, thanks. I'll check into those suggestions!

5. ### ELECTRONERD Thread Starter Senior Member

May 26, 2009
1,146
16
Interesting, I didn't know that it read the average, but in low frequency response it will read the RMS value. There must be a certain point in frequency when it changes, I should be able to find it in the manual, correct? I was using an approximate 6kHz sine wave, measured by the frequency selection in my meter. I would be interested in the CMOS oscillator you made for calibrating your o'scope. Would be so kind as to provide it a schematic diagram?

I suppose most modern o'scopes don't need to be calibrated manually, due to their supreme features. I did notice, however, that my scope has a self calibration technique and I think I can put the output of that calibrator (which is inertly in my o'scope) to the input of my o'scope and adjust the amplitude and frequency for the most accurate reading. Is this true? Unfortunately, I haven't been able to get my hands on the manual for my o'scope so I'm unsure. If anyone does have a link or PDF file for my Tektronix 7904 oscilloscope?

Thanks!

6. ### Audioguru New Member

Dec 20, 2007
9,411
896
Your meter is made to average the input AC voltage then is calibrated for a low frequency sine-wave. The capacitance of its input voltage divider attenuates frequencies above about 200Hz.

7. ### ELECTRONERD Thread Starter Senior Member

May 26, 2009
1,146
16
I understand. Therefore, I would need a circuit like you had to calibrate the o'scope for precise readings both on frequency and amplitude. Would you please provide the schematic for your circuit? Or, any other calibrator with the same function? If you can't do a minor favor like that, would it be possible for me to use the "Calibrate" setting on my o'scope to adjust the amplitude and frequency of my o'scope?

8. ### ELECTRONERD Thread Starter Senior Member

May 26, 2009
1,146
16
Why didn't you respond Audioguru? In any case I assume you don't want to share your CMOS circuit schematic? That's all right, I can find a schematic for another one, or design one myself.

Good news! For the "calibrator" function on my o'scope, I produced a 1kHz square wave. Based on the x axis, one complete cycle was 1000μS, which is 1kHz. Therefore, the frequency seems to be accurate enough. Now, all I have to do is get my amplitude accurate enough. According to beenthere, I can just measure the peak of my input sine wave and adjust it from there.

Thanks for the help everyone! If I want to come back regarding that CMOS circuit, I'll post here in the future.

Last edited: Oct 14, 2009
9. ### russ_hensel Distinguished Member

Jan 11, 2009
820
47
Actually the average of most AC is 0. But you probably mean the average of the absolute value. Just to be picky.

10. ### ELECTRONERD Thread Starter Senior Member

May 26, 2009
1,146
16
What!? How can the average of most AC be zero? Can you please explain that?

I wasn't talking about average at all, only the peak value. The only similarity I can think of is the RMS value but even then it doesn't make sense when you say average.

11. ### The Electrician AAC Fanatic!

Oct 9, 2007
2,300
335
Think about what the voltage coming out of a wall socket looks like. It's supposed to be a sine wave, and if it were perfectly distortion free, it would spend as much time on the positive half cycles as on the negative half cycles. That means that its average value would be zero.

In fact, if you set your meter to measure DC, what it measures is average voltage. Try measuring the line voltage with your meter set to measure DC, with an appropriate voltage range, and it should say essentially zero. Sometimes there will be a few millivolts of DC at a wall socket due to half-wave loads such as hair dryers set to low heat (in your house, or in the neighbors).

12. ### ELECTRONERD Thread Starter Senior Member

May 26, 2009
1,146
16
Thanks The Electrician,

I visualized a sine wave with a 3V peak. Therefore the average, $\frac{[3V+(-3V)]}{2} = 0$ is 0.

Jul 17, 2007
22,183
1,728
14. ### The Electrician AAC Fanatic!

Oct 9, 2007
2,300
335
What kind of meter is it? Is it a digital meter?

Who is the manufacturer? Model number? Do you know if it's a "true RMS" meter?

15. ### ELECTRONERD Thread Starter Senior Member

May 26, 2009
1,146
16
The meter is a Velleman DVM890F.

16. ### The Electrician AAC Fanatic!

Oct 9, 2007
2,300
335
I looked up the specs for your DM890F here:

http://www.vellemanusa.com/us/enu/product/list/?id=342990

It appears that it is not a "true RMS" responding meter. I checked several of the handheld meters, and none of the ones I checked mentioned "true RMS". The only one I saw that was "true RMS" was the DVM645BIU bench meter.

I'm surprised by this because there are a number of ICs for implementing DVMs which are true RMS responding. I wonder why Velleman didn't use them.

Anyway, meters which aren't true RMS responding measure AC voltages by rectifying (convert to absolute value) the incoming AC and then measuring the average value of that, and finally multiplying by PI/SQRT(8).

The PI/SQRT(8) factor is the ratio of RMS to rectified average of a distortionless sine wave. If the grid waveform is distorted, which it always is these days, then such a meter will give a reading slightly in error.

But, even if your meter were a "true RMS" meter, that wouldn't help you determine the peak value of a grid frequency waveform because you wouldn't know how much the waveform was distorted. The peak value of an undistorted sine wave is 1.414 times the RMS value.

The output of the "cal" function on oscilloscopes is usually a square wave at 1 kHz with a DC component; zero volts low and 5 volts high (some scopes are 2.5 volts high). Happily, the RMS value of such a square wave is the same as the rectified average and non-"true RMS" meters will read the same as "true RMS" meters (as long as the meter is not set to read "true RMS" AC+DC). This means that you can measure the cal square wave with a DVM and determine its amplitude with the accuracy of a DVM.

The only problem I see is that your meter has a specified upper frequency limit of 400 Hz on AC, but it might still be ok at several kHz; I say several because to measure a square wave accurately the meter must respond to the significant harmonics of the square wave.

If you have an audio square wave generator, set it to 50 Hz and measure with your scope and with your meter; note the meter reading. Adjust the frequency up to 1 kHz and use the scope to make sure the peak value is the same; measure with your meter and see how far off it is from the reading you got at 100 Hz. If it's still fairly accurate, you can use it to measure the scope's cal voltage. If fact, if you note the factor by which the meter is reading low (or possibly high), you can use that factor to adjust your reading of the scope's 1 kHz cal voltage.

17. ### Audioguru New Member

Dec 20, 2007
9,411
896
My 'scope calibrator circuit is a simple ordinary Cmos oscillator made with a CD4069 hex inverters IC. Its output is buffered by 4 of the inverters connected in parallel and it has a regulated 5V supply so its output is exactly 5.0V peak-to-peak.
Here is an ordinary Cmos oscillator circuit:

File size:
23.2 KB
Views:
116
18. ### ELECTRONERD Thread Starter Senior Member

May 26, 2009
1,146
16
Hey The Electrician,

Sorry It took me a while to get back to you.

Firstly, I'm assuming the equation you gave "PI/SQRT(8)", can be represented as this?:

$\frac{\pi}{sqrt{8}}$

I did take the output from my calibrator which was a 1kHz square wave (I measured exactly 1kHz, so the frequency is accurate enough), but I didn't get a 0V to 2.5V or 5V square wave, I'm assuming your implying the peak? Instead, I simply got a square wave that didn't start from zero, but it had a negative 5V (if i remember the correct voltage) and a positive 5V; which a standard square wave should have.

I do have an audio signal generator that can produce both sine waves and square waves, but the square waves are distorted and completely unusable. I'll have to fix that in my spare time eventually. Perhaps I could use a sine wave instead and yet still have the frequency as 50Hz?

It's truly unsatisfying knowing my meter can't find the true RMS value. When I purchased the DVM, I didn't know much about electronics at the time and so it appeared to be a grandoise purchase. However, I now consider it to be a profligate purchase. I share your confusion on why Velleman didn't implement those ICs used to find the true RMS value. I might have to purchase a more proficient meter in the future; that is, when I save de' mullah.

19. ### The Electrician AAC Fanatic!

Oct 9, 2007
2,300
335
That's correct.

That's ok, too. That square wave will read 5 volts on either a "true RMS" meter, or an average responding AC meter.

The problem is that a meter needs to respond to some of the harmonics of a square wave to get an accurate result, perhaps up to the 10th. So, if you want to be able to accurately measure a 50 Hz square wave, your meter needs to accurately measure a sine wave up to 500 Hz.

Your scope can respond to DC, so just use a variable source of DC, which you can accurately measure with your meter, and see if the scope gives the right value.

Audioguru's CMOS oscillator is a very good way to go. The square wave output amplitude is very close to the value of the DC voltage supplied to the CMOS chip if its output is lightly loaded, as it will be by the scope.

20. ### Wendy Moderator

Mar 24, 2008
20,772
2,540
To calibrate the scope I'd build a CMOS 555 oscillator, then force it to a high and a low, measuring the DC voltages. This will give you a P-P reference. Then feed the oscillator into the scope.

I don't think a meter is a good way to directly calibrate a scope myself.