Global Sources
EE Times-Asia
Stay in touch with EE Times Asia
EE Times-Asia > Amplifiers/Converters

FAQs: Amplifiers (Part 3)

Posted: 17 Dec 2008 ?? ?Print Version ?Bookmark and Share

Keywords:FAQ? amplifier?

What is a closed loop buffer?
A closed loop buffer is an amplifier with high input impedance and low output impedance, with a fixed gain of +1. It is typically used for isolation, increased output drive, capacitive load drive, etc. No gain setting resistors are required.

What is a open loop gain, voltage feedback op amp?
Open loop gain, voltage feedback Op amp: The open loop gain is specified at DC and is defined as the ratio of an output voltage change to an input voltage change. It is also referred to as differential voltage gain (no feedback or input networks added). When specified using a sinusoidal waveform, it varies in magnitude and phase relative to frequency.

What is an amplifier's intercept point?
Intercept Point is the fundamental output power where the specified distortion term (2nd, 3rd, or 3rd order intermodulation) is equal in power to this fundamental value power.

What is an amplifier's specified supply range?
The specified supply range describes the power supply voltages required to power the op amp.

My amplifier design works OK on a single 5V supply but if I try to put 4 volts on the input, the output will not go above about 3.6 volts. What's wrong?
In the device datasheet, look for the specification labeled Input Voltage Range or Input Common Mode Voltage Range. This specification tells how close to the upper or lower supply voltage the amplifier can operate. Most amplifiers cannot operate with inputs closer than one to two volts to the supply rail voltage. Some op amps can go to the negative supply rail but not the positive rail. If you need the input to go very close [within 20 to 200mv] to the supply rail, select a Rail-to-Rail Input amplifier, or one that allows the input to desired supply rail. If the output must also go very close to the positive rail, select a Rail-to-Rail Input/Output (RRIO) amplifier.

I selected a Rail-to-Rail Input/Output amplifier but the output does not go all the way to the negative rail or all the way to the positive rail. What am I doing wrong?
The term "Rail-to-Rail" is misleading. To be completely accurate, it should be "almost Rail-to-Rail" or "very nearly Rail-to-Rail." The output for most R-R amps is from 20 to 200mv from either supply rail, almost never all the way to the rail. And as more load current is required, the output will "pull" further away from the supply rail voltage. Most amplifiers will provide their maximum output voltage swing with a load of 100k? or greater. The Electrical Characteristics table and the Characteristic Curves in the product datasheet will specify the output voltage swings that can be expected.

What's the difference between GBW, unity gain bandwidth, gain bandwidth product, and the -3dB frequency?
For many op amps, the open-loop gain drops with frequency at a steady rate, -20dB/decade. At any point on this downward slope, the product of the gain and the frequency at that point is constant, and is known as the gain bandwidth product or GBW. If the op amp has been stabilized to operate at unity gain, then the unity gain bandwidth, or the frequency at which the open loop gain is unity (gain of one) is usually equal to the gain bandwidth product. This is shown on the "Open Loop Gain and Phase" plots as the frequency where the gain crosses through 0dB.

Some op amps do not have a steady GBW, especially those not stabilized to operate at unity gain. In these cases, the GBW will be different from (usually higher than) the unity gain bandwidth.

The -3dB frequency is a measurement of the bandwidth of a closed-loop application of the op amp. The -3dB frequency is the point where the overall closed-loop gain drops by 3dB. The frequency at which the closed-loop application gain is unity can be calculated using BW=GBW/Av. Both the -3dB frequency and the unity gain bandwidth of the application depend on freedback gain setting, output swing, load, and circuit layout.

What is the typical input capacitance of an op amp?
The input capacitance is typically about 2-3 pf. About half of the capacitance is in the chip and half is in the package.

How do I protect the amplifier inputs from possible voltages above/below the supplies?
You must either clamp the input, or limit the current into the device, or ideally, both. The simplest way to limit this current is with a resistor selected such that the maximum voltage applied to the circuit input creates a current on the input pin that is less than the maximum pin current rating. A 1K to 100K resistor in series with the input pin is usually enough. The series resistor of an inverting configuration usually serves this function fairly well. However, a non-inverting amplifier may require this input protection resistor because the signal is usually applied directly to the non-inverting input pin. For low-impedance circuits that cannot contain a large resistance, a pair of clamping diodes between the negative rail, the input, and the positive rail, along with a small series resistance, will protect the device. For high-impedance circuits, use a larger resistor and/or low-leakage diodes.

How should a very low frequency (
The traditional differentiator uses a series Rs-Cs input and a parallel Rf-Cf feedback around an op amp. But there is no simple solution, no "one size fits all" solution. Try more Rs or Cf to minimize the noise. The only reason the output of the differentiator is noisy is because there is a lot of gain, and the INPUT is noisy. Adding more Cf or Rs reduces the gain. Miracles are not expectedBesides, just because the differentiator's output is noisy, does not mean it is doing harm. It is just magnifying the noisy signal that is there, as well as amplifying the signals! If you are closing a loop, the differentiator's output noise may be HELPING make the loop quiet and stable. If the differentiator's output is rather noisy, and if that is because the input is too noisy, try to figure out what is really creating the input noise.

What's a good way to minimize 1/f noise while amplifying low-level DC signals?
For a high signal-to-noise ratio, the circuit must be well-engineered. This includes choosing the best amplifier for (a) the bandwidth of interest, and (b) the impedance level of the input signals. Choosing a low voltage noise amplifier will not be helpful if the input source has rather high impedance, and the amplifier has high current noise. An amplifier with extremely low 1/f noise can be made using the LM394 transistor pair, better than most integrated op amps.

Can op-amps be purchased with a selected offset voltage?
Not unless you plan to pay a premium price and order 100,000 units. Small scale special testing of this sort is very expensive. If you want parts with consistent offsets, you will have to screen them first. Our advice is to redesign your circuit to be less sensitive to offsets. This will also prevent trouble in the future if devices need to be replaced in the field, as they can throw in any "off-the-shelf" device, instead of a "selected" device. There are ways to add/force an external offset.

Why do some amplifiers oscillate with capacitive loads?
Oscillations caused by capacitive loads are a result of interactions between the op-amp's output impedance and the capacitive load. The output impedance and the capacitive load form an R-C pole in the output stage, causing additional phase lag in the feedback signal. CMOS amplifiers have a higher output impedance that causes the pole to come near, or below, the op-amp's unity-gain frequency.

The additional phase lag of the pole will erode the phase margin of the op-amp, causing the total amplifier phase lag to increase to more than 180 degrees at the unity-gain frequency, resulting in oscillation due to a total feedback phase shift of greater than 180 deg at unity gain. CMOS parts can have an output impedance between 100 and 500, causing the pole frequencies to be relatively low. Similar speed bipolar op-amps have output impedances in the 1 to 100 range, resulting in pole frequencies much higher than the CMOS op-amp, keeping the pole away from the unity-gain frequency of the part.

The capacitive load drive of CMOS parts can be increased with the use of output resistors and external "feed-forward" capacitors placed on the output. The datasheets for CMOS op-amps have a section on how to compensate for capacitive loads.

Do you need to carefully bypass the supply pins of a high speed (>200MHz) op amp if you are only using it at 1MHz?
Absolutely! If it's not bypassed, it may ring at frequencies between the signal and the amplifier's bandwidth, causing unexpected errors. Ringing at 11Mhz would bum you out! If you have to use a 200 MHz op-amp, it is only fair to put in pretty good bypassing!

What is the difference between an amplifier's output current and short circuit current?
"Short Circuit" current is the current that the device will deliver if the output is directly tied to one of the supply lines. Depending on the device's design, this represents the current limiting of the output stage. However, short circuiting current does not represent the true output driving capability of the output stage. Due to output stage resistance, the maximum output current is determined by the output voltage swing under load. The output will have higher swings with lighter loads, and less swing with heavier loads. The "Output Swing vs. Load" or "Vout vs. Iout" graphs in the device datasheet should be consulted to determine if the op-amp can safely drive the load to the correct levels. Don't forget to account for feedback resistor load, as it can be significant with high-speed or micropower circuits.

If the output of the op amp is stuck at a voltage close to one of the supply rails (i.e. the output is railed), what's the cause?
There are many ways to "rail" an op amp. The hard part is keeping it from railing. If the input exceeds the input voltage range, the output will generally go to one supply voltage rail. If the output would theoretically go beyond the actual supply voltage, given a hypothetically higher voltage supply, then again the op amp will rail. If the feedback around the amplifier is either nonexistent or in the wrong polarity, again the op amp will rail. Also, if the positive input is higher than the negative input, the op amp will rail.

The op amp application should be analyzed to be sure the input voltage(s) and gain are appropriate for the supply voltage used, so that in normal operation the input voltage is within the operating ratings, and the output voltage is within its normal bounds.

Why do application schematics involving op amps seldom show supply connections?
Power supply connections are often omitted to simplify the application schematic. This was done historically when op amps were first being used, and the tradition is perpetuated itself, mainly to keep the application schematic from looking overly complicated. Drawing the supply connections to all op amps can be especially confusing with multiple (e.g. dual or quad) amplifiers in one package. In such cases, only one supply pin per package needs to be connected, and the supply is routed internally to all the op amps in the package.

Even if there is no supply connection shown on the application schematic, a power supply must be connect to the amplifier for it to operate.

What is the difference between a single supply op-amp and a dual supply op-amp?
Generally, "single supply op amp" means the op-amp has an input common mode range that includes V- (Gnd). However, since there is no "ground" pin on an op-amp, there may no difference in the actual circuitry, layout, or behavior of op-amp itself that allows it to operate on a single supply vs a dual or split supply. The only difference one might find upon a thorough examination is in the op-amp data sheet. An op-amp that is designated as a dual supply op-amp usually references the output load with respect to ground, whereas a single supply op-amp usually references the output load with respect to the midpoint voltage of the single supply. Although op-amps that are designated as single supply op-amps generally operate at lower voltages, this is not a requirement. So whether you operate your op-amp from a single 5 volt supply and ground, or from +2.5 and -2.5 volts, makes no difference to the op-amp. All the op-amp cares about are the relative voltages of those supplies with respect to each other, and with respect to the input and output voltages.

How do you check stability of an op amp circuit?
To check the stability of a control loop such as an op amp circuit, apply a pulsed load and observe the output voltage change. The pulsed load can be a pulse or step change in load current, connected to the op amp circuit output by a series R-C network (such as 10k/0.01? F). The greater the ringing or oscillation, the less the stability of the circuit. This process is often referred to as "banging" the output.

For part 1, click here

For part 2, click here

Source: National Semiconductor

Article Comments - FAQs: Amplifiers (Part 3)
*? You can enter [0] more charecters.
*Verify code:


Visit Asia Webinars to learn about the latest in technology and get practical design tips.

Back to Top