Measure DC offset, Hum, DBs gain, Oscillation

[walters]

How do i measure DC offset?

Where is DC offset mostly at on a analog or digital Circuit?

How do i measure DB gain?

How do i measure DB?

How do i measure Oscillation?

How do i measure Hum?

 

[zachtheterrible]

Mostly with an oscilloscope.

 

[walters]

Yea but how do i do it with a oscilloscope what do i do and how do i
do this stuff and measurements?

 

[zachtheterrible]

https://www.electro-tech-online.com/threads/i-want-to-build-a-mic-amplifier.240/

Take a look at that article. It gives explanations as to how to measure basic but important things on your oscilloscope, like frequency and voltage measurements.[/quote]

 

[walters]

Thanks Zach

but it doesn't show how to measure DC offset,DB gain,oscillation,hum

 

[zachtheterrible]

Well, oscillations are measured in frequency, and it shows how to do frequency.

Hum is measured in voltage usually.

DB gain gets measured by comparing the input voltage and the output voltage, or power, or current.

Dunno about DC offset.

Someone will come along shortly who knows more than me :lol:

 

[Morgen]

DC offset can be measured with a DC voltmeter or an oscilloscope. Remember that it isn't unusual, as in the case of op amp inputs, that a DC offset is in the microvolt range.

 

[walters]

Thanks guys

How do i measure overloading the input?

why does it overload the input?

 

[eblc1388]

Remember that it isn't unusual, as in the case of op amp inputs, that a DC offset is in the microvolt range.

Most cheap, non-precision opamps have input offset voltages in the millivolt (mV) range, that's 1000 times higher than microvolt range.

 

[Nigel Goodwin]

Thanks guys

How do i measure overloading the input?

You don't really, you measure the output, then adjust the input level until the output starts clipping - you're VERY unlikely to overload the actual input, more likely the later stages instead - and if you DO overload the inpout stage, the later stages will be MASSIVELY overloaded.

why does it overload the input?

Take a 5 gallon bucket, fill it with 5 gallons of water - no problem!. Now take the same bucket and fill it with ten gallons of water!, you then have very wet feet - same principle really!.

 

[Morgen]

Remember that it isn't unusual, as in the case of op amp inputs, that a DC offset is in the microvolt range.

Most cheap, non-precision opamps have input offset voltages in the millivolt (mV) range, that's 1000 times higher than microvolt range.

The general-purpose uA741 datasheet shows a typical input Vos of about 2mV, whereas the precision OPA27 shows a typical input Vos of about 20uV. A 100x difference, but I get your point.
My point was only that DC offset can be a very small quantity not as easily measured as say, finding the voltage of a dry cell.

 

[eblc1388]

My point was only that DC offset can be a very small quantity not as easily measured as say, finding the voltage of a dry cell.

Point taken. Most often one needs to use the gain of the Opamp itself to amplify the offset voltage by 1000x to get a reasonable reading on a DVM.

 

[walters]

Thanks alot guys for your help

How do i Track down Oscillation in a circuit when im troubleshooting?

When im using the oscilloscople and i see a oscillation frequency
riding on the AC waveform what should i do to try to track down
this oscillation and where its coming from?

 

[Nigel Goodwin]

Thanks alot guys for your help

How do i Track down Oscillation in a circuit when im troubleshooting?

When im using the oscilloscople and i see a oscillation frequency
riding on the AC waveform what should i do to try to track down
this oscillation and where its coming from?

There's no 'magic' answers, you track it down by understanding how the circuit works and what might cause it not to work correctly. This applies to pretty well all fault finding.

I would suggest you spend sometime playing with your scope and trying to understand how circuits work - you can't expect to get 5 years education in a few posts on a forum!.

As with everything, start with the basics and work up, make sure you have a good understanding of discrete components before moving on to IC's.

 

[Morgen]

Oscillation requires gain and positive feedback, so that portion of the circuit which provides those should be your focus. It would be helpful to use your scope to determine the frequency of oscillation. That information could assist in pinpointing where the fault lies. I must agree with David and Nigel, though. You seem to be overreaching yourself.
It's not that overreaching is inherently wrong. That is how we learn. But take steps too big and you'll trip up more often than learn.

 

[walters]

Thanks guys for your help

1.) Oscillation requires gain and positive feedback,

2.) It would be helpful to use your scope to determine the frequency of oscillation

This some good information i can give me some insight to know how
to start thanks

 

[awright]

"How do I measure..."

"How do i measure DC offset?"

Configure your op amp with a convenient, relatively high gain, say, 1000. Short the input. Measure the DC output voltage with a DVM. Divide by 1000 (or whatever gain you configured your op amp with). That is your input offset voltage.

"Where is DC offset mostly at on a analog or digital Circuit?"

Not clear on what your question is here. Offset voltage at the input stage of a circuit most strongly influences the total offset seen at the output because all subsequent gain stages amplify the input offset. Offset voltages arising after the input stage are amplified less and have, therefore, less influence on the output offset voltage.

"How do i measure DB gain?"

(A) If you have a voltmeter that offers dB measurements, you measure the input and output voltages with a logarithmic voltmeter and subtract (arithmetically) the input dB from the output dB reading. Since you are measuring GAIN in dB, not LEVEL in dB, the reference level of the dB meter is immaterial (e.g., dBm, dBv, dBarbitrary, will all give you the same answer for GAIN). Lots of DVMs offer relative measurements in dB these days. With these, you measure the input voltage and set that as the REFERENCE. Then the output voltage is indicated as gain in dB.

(B) If you do not have a meter that reads dB directly, you measure Vin and Vout. Gain in dB is 20[log(Vout/Vin)]

Decibels are logarithmic units of ratios. "Level" is the term referring to signal magnitudes (not gains) in decibels relative to some known reference value. Thus, dB(micro)V [I don't know how to enter a Mu for micro] is 20 times the logarithm (base ten) of the ratio of the measured voltage relative to 1 microvolt, i.e., 20[log(V/10^-6)]. Decibels are very confusing until you understand their basis, then they become very natural and convenient to use and greatly simplify many types of measurements.

The "20" in decibel calculations arises from the fact that the decibel is basically a power ratio indicator. The "Bel" (obsolescent) is the log(base 10) of a power ratio and was named after Alex. G. Bell. Since it was an inconveniently small number, requiring most values to be expressed to a couple of decimal points (since a power ratio of ten is one Bel), the decibel (1/10th of a Bel) became more commonly used. If you were to measure input and output signals in units of power (Watts), the gain would be calculated from Gain = 10[log(power out/power in)]. Since it is generally much more convenient to measure signals in volts, rather than Watts (how many of you own wattmeters?), we have to use volts squared to get values relative to power. But to square a value expressed in logarithms, you multiply by 2. Thus, decibels for ratios of signals measured in volts are multiplied by 2, i.e., we multiply the log of the ratio by 20, rather than 10.

More than you really wanted to know?

"How do i measure DB?"

See above.

"How do i measure Oscillation?"

As earlier posters have said, view the signal on an oscilloscope, being sure that the 'scope bandwidth includes the possible oscillation frequency. You can also infer that oscillation is occurring if you see an output signal level higher than you would expect from the known gain of the circuit and the input signal level (after you eliminate DC as the cause of excessive signals).

"How do i measure Hum?"

Same way you measure any signal (carefully). You want to be sure you are not including non-"hum" signals, such as audio signals, noise, DC offset, radio interference, etc. The best way is to use a spectrum analyzer so you can look for signals at "hum" frequencies (in the U.S., harmonics of 60 Hz.). If you don't have a spectrum analyzer, use a narrow bandpass filter to select and measure signals at 60 Hz, 120 Hz, etc. If you don't have a variable bandpass filter, build a set of fixed filters at harmonics of 60 Hz. See app notes from National, Analog Devices, Linear Technology, etc. for circuit details.

If you don't have a 'scope, you can differentiate between hum and the broadband noise that can get through a bandpass filter by varying the bandwidth of the filter. If the measured signal level increases by 3 dB when you double the bandwidth, you are measuring broadband noise. If the measured signal level does not change with changes in bandwidth, you are measuring hum. This is because doubling the filter bandwidth allows double the amount of broadband noise energy to pass through the filter, while doubling the filter bandwidth has no affect on the hum component. Of course, you have to know that the filter gain at the hum frequency does not change with changes in bandwidth.

Have fun.

awright

 

[walters]

NICE thats really good information thanks

 

[awright]

Measuring DC Offset

My earlier response on measuring "DC offset" was too simplified, as I assumed that you were referring to input offset VOLTAGE (and I was responding at 2 am).

The input offset of an amplifier is actually fairly complex, as it always consists of two distinct phenomena, offset VOLTAGE and offset CURRENT. Which one dominates in a particular application is a function of the source resistance seen by the amplifier. I am not really well qualified to discuss these in detail, but, to wade in over my head....

Offset VOLTAGE is defined as the input voltage required to result in zero offset at the output of the amplifier being evaluated. It is evaluated with a very low resistance shunting the input so that offset CURRENT is shunted to ground with minimal effect on the output. This is because the input offset CURRENT, normally being quite a small value, will not develop a significant voltage across a low source resistance.

To clarify my earlier post, if you set up the DUT (Device Under Test) with a nominal gain of 1000 (using 100 ohm input and 100Kohm feedback resistance) and ground the non-inverting input terminal and the input to the input resistor, the gain of the circuit to an offset VOLTAGE between the input terminals is that of a follower-with-gain, or 1001. Therefore, the more accurate calculation is to divide the measured output offset by 1001, not 1000.

Input offset CURRENT is the difference between the input bias currents at the two input terminals. The input transistors of even the best bipolar input op amps require some bias current to function, since bipolar transistors are current operated devices. Modern processing and super-beta transistors have greatly reduced the input bias currents and improved the inherent balance between the two input transistors (by careful thermal layout, laser trimming, improved processing, etc.), but there is still some bias required and some lack of perfection in balancing. This leads to input offset currents that can cause problems in high source impedance applications. Because input currents are extremely small for good FETs, FET input amps normally have very small input offset currents.

Offset CURRENT is evaluated with a very high resistance at the inputs and is a little more difficult and tricky to measure than offset VOLTAGE. Basically, you set the DUT up as a straight follower with extremely well matched (0.1% or better), very high resistance (10 to 1000 Mohm), source (from ground to the non-inverting input terminal) and feedback (from the output to the inverting input terminal) resistors, measure the output offset voltage, and calculate the equivalent input current for the resistor values used (Ioffset = Vout/R). Shunt both resistors with very low leakage 0.01uF capacitors to limit noise. Obviously, this measurement setup is highly susceptible to noise and leakage currents and should be carried out in a screened enclosure using clean teflon terminals.

Both offset voltage and offset current vary with temperature and should be evaluated at several accurately known temperatures.

awright

 

[David Bridgen]

Re: "How do I measure..."

(B) If you do not have a meter that reads dB directly, you measure Vin and Vout. Gain in dB is 20[log(Vout/Vin)]

So long as the impedances across which the voltages are measured are equal.
https://www.davidbridgen.com/dbnotation.htm