# How compensate for frequency derating when measuring voltage O-scope

Discussion in 'Radio Circuits, Repair & Performance' started by VE4AJM, Apr 30, 2020.

1. ### VE4AJMHam MemberQRZ Page

How do you compensate for the derating with frequency change? I am using a techtronics scope to measure voltages at 10-30 megs and I see from the manual that voltage derates up to 5 fold from 1 to 9 MHz and continues thereafter. How do you take any meaningful measurement of voltage without an accurate reference voltage at near the same frequency?

I need to measure the voltage generated by a heterodyne oscillator circuit that puts out 5 different selectable frequencies depending on the band selected. The frequencies range from 10-35 MHz, and the voltage range from 6-11 volts. When I measure them, they all come up at nearly half what is specified, but I am thinking based on the manual for the Techtronics probe, the measurements will of course be low.

Last edited: Apr 30, 2020
2. ### SM0AOMHam MemberQRZ Page

There is a general recommendation that to make anywhere meaningful measurements of voltage levels at radio frequencies, you need a probe+oscilloscope bandwidth at the -3 dB points of at least twice and preferably three times the fundamental frequency of the waveform that is measured.

Oscilloscopes are characterised in bandwidth, rise time or both.
Normally the - 3 dB bandwidth is stated, under which the amplitude of the applied signal is shown constant with frequency, dropping off to 70% when the upper limit is reached.

Due to the low-pass characteristic of the oscilloscope vertical amplifier, all waveforms will be shown as sinusoidal when the upper limit frequency is approached or exceeded.
Usually, nothing can be said of the amplitudes of signals that exceed the - 3 dB point with a factor two or more.

Some oscilloscopes had the shape of their amplitude versus frequency curve shown in their manuals.

This would enable you to at least estimate the level of a fundamental component above the vertical amplifier bandwidth.

73/
Karl-Arne
SM0AOM

PY2RAF, KB0MNM, KD0CAC and 2 others like this.

Specifically what oscilloscope are you using?

Your reference to the manual doesn't seem right. For instance a 100 MHz oscilloscope should only roll off the amplitude measurement by ~30% (0.707) at 100 MHz. If you really have a five fold roll off in amplitude at 9 MHz then that's a very low bandwidth scope.

N0TZU likes this.
4. ### VE4AJMHam MemberQRZ Page

It was pointed out to met that the derating with frequency most liking refers to the maximum voltage, and not the measurement. Sorry for the really dumb question.

The scope is a tektronics 100 MHz scope, and I looked up the manual online, and clearly didn't understand the figure / graphic that referred to voltage derating.

Last edited: Apr 30, 2020
WA7ARK, N0TZU and K7TRF like this.

That makes a lot more sense. You should have no problem making accurate voltage measurements (well accurate to within screen reading abilities) for your Het oscillators with a 100 MHz scope.

N0TZU likes this.
6. ### KL7SGPlatinum SubscriberPlatinum SubscriberQRZ Page

Unless you need to see the wave form, I would just get an RF probe and plug it into your voltmeter.
They are inexpensive.
I just checked ebay and they can be had for about \$70.00. The fluke 85RF is good to 500 Mhz.

N8VIL, KB0MNM, KM4YTW and 1 other person like this.
7. ### VK6ZGOHam MemberQRZ Page

Yeah, that usually refers to the characteristics of test probes, rather than the bare Oscilloscope.
Most of the probes have decent ratings at DC & low frequency ac, but it falls off above that.

8. ### VE4AJMHam MemberQRZ Page

I've been doing some more reading on the subject of oscilloscopes and bandwidth, and it seems there is something to the suggestion that oscilloscopes don't measure amplitude correctly throughout their bandwidth. Apparently the standard by which scopes are judged is the db falloff at 50% of bandwidth. So, while a scope may display frequencies up to 100 MHz, at 50mhz and above there is certain rated fall off in accuracy of amplitude.

According to tektronix, bandwidth is defined as the frequency at which the signal amplitude drops to -3db in measuring accuracy. So that is up to 30% inaccuracy at the top end of the frequency range.

https://www.mouser.com/pdfdocs/Tektronix12_things_to_consider1.pdf

So, a scope needs to be 5X the bandwidth of the maximum signal you intend to measure, to be accurate to within 2%.

9. ### W9WQAHam MemberQRZ Page

or, if you are just looking for presence of signal, gain in a stage, or change when you adjust tuning, you might just not be too concerned about super accuracy and
move on. its an indicator...

10. ### VE4AJMHam MemberQRZ Page

I agree, in many circumstances the accuracy of the voltage measurement in absolute terms isn't relevant.

I've been struggling with a potentially low output on a heterodyne oscillator circuit, so reasonably accurate amplitude measurements are critical to resolving this. It seems my measurements are indicating it is around 30% lower than spec, so I have been questioning whether my measuring technique is valid or not. Also, it makes me wonder what equipment was used to come up with the specs. Was it measured accurately?

I have to admit, that what SMOAOM wrote in his post, went right over my head.

Last edited: Jun 17, 2020