Test Ideas: Produce AC test signals
Use this circuit as a DC-biased AC source when you need both low-distortion and power driving capability.
By Tiger Zhou and Robert Dobkin, Linear Technology -- Test & Measurement World, 12/1/2008 2:00:00 AM
AC testing of electronic systems often requires a low-distortion signal source to excite a DUT (device under test). Instead of using a signal generator and a power amplifier to produce a low-distortion AC signal, you can build a power oscillator with just one IC and a few discrete components.
|
Have a test idea you'd like to share? Publish it here and receive $150. Send us your ideas. Read other Test Ideas. Read Design Ideas published by EDN. |
Figure 1 shows a power oscillator that uses a twin-T filter network and a high-power LDO (low dropout) regulator to produce a DC-biased AC signal. The LDO regulator’s op amp and output transistor form a voltage follower. A precision 10-μA current source and the RSET resistor set the circuit’s DC-level output.
 Figure 1. A power oscillator uses a twin-T filter in its feedback loop to set output frequency. |
Oscillation begins when positive feedback is present in the loop. The oscillation will be stable when the loop gain is unity. When the loop gain is greater than unity, the oscillation amplitude will keep increasing until the components saturate. When the loop gain is less than unity, the oscillation will diminish. A feedback potentiometer lets you adjust the loop gain so it is close to unity, which maintains a stable oscillation.
A notch filter in the feedback loop, made from two T-type filters in parallel (one low-pass, one high-pass), sets the oscillator’s output frequency. The LDO regulator then amplifies the signal and drives the load.
The twin-T notch filter in the feedback loop between the OUT pin and SET pin attenuates frequencies above and below the filter’s center frequency (f0), which is the oscillator’s output frequency. Using the equation below, you can calculate values of resistors and capacitors to get f0. For f0 = 400 Hz, R = 8.45 Ωk and C = 47 nF.
 Figure 2. Loop gain changes with the value of K from Figure 1. |
At f0, the twin-T network’s gain is maximized near unity. The maximum gain, though, changes with the value of K in Figure 1. When K = 2, the oscillator’s maximum gain is 1.0 (Figure 2). When K = 5, the gain increases to 1.1, and the gain decreases even more if K should become greater than 5. Thus, you should select a K value from 3 to 5 for a larger than unity gain, then attenuate the feedback signal with the potentiometer to achieve oscillator stability.
While the potentiometer will work, it will not provide automatic gain control; you can get that if you replace the potentiometer with a light bulb or a MOSFET. (View schematics for both implementations at the end of this file).
Because a light bulb heats when turned on, its resistance will increase with the oscillation amplitude. Thus, the bulb’s change in resistance tunes the feedback loop’s gain to maintain the oscillation. In the MOSFET implementation, a Zener diode detects the oscillator’s peak voltage. The MOSFET’s resistance decreases as oscillation amplitude increases, which decreases the oscillator’s loop gain and maintains the oscillation.
Figure 3 shows the oscillator’s signal output and harmonics using the light bulb in the feedback loop. With f0 of the twin-T oscillator at 400 Hz, the circuit’s THD (total harmonic distortion) is 0.1%. We also measured the circuit’s performance at f0 = 8 kHz. At that frequency, THD jumped to 7%.
 Figure 3. Using a light bulb in the oscillator's feedback loop produces a 400-Hz sine wave with low distortion for the second harmonic. |
The most significant harmonic contribution comes from the second harmonic (lower trace), which is less than 4 mVpk-pk in the light-bulb implementation. View the oscillator’s output and harmonics using the MOSFET implementation at the end of this file. With the MOSFET, the circuit’s THD is 1% with a 40-mVpk-pk second harmonic. The MOSFET’s resistance changes more quickly and thus produces greater harmonic content than the light bulb.
When designing and using an oscillator, you must also consider the oscillator’s dynamic response. The startup is one of the few criteria representing the oscillator’s stability. The startup of a well-designed oscillator should have minimum overshoot and have no swinging of its output amplitude. Unlike many oscillators, both the light bulb and the MOSFET implementations maintain a constant peak-to-peak amplitude.
The twin-T oscillator can drive any type of load—inductive, capacitive, or resistive—with a drive capability up to 1.1 A, which is the limit of the LT3080 LDO. Your circuit’s load will limit the maximum frequency that you can get from the oscillator while maintaining low distortion.
When testing the oscillators, we observed a very small amount of overshoot at startup. The MOSFET variation stabilizes faster than the light-bulb version because the latter has a long thermal constant caused by heating.
| The authors would like to thank Tony Bonte, Mitchell Lee, Jim Williams, and Todd Owen for fruitful discussions. |
Web-exclusive figures:
 A light bulb increases resistance as it heats, providing negative feedback. |
 A MOSFET in the feedback loop responds more quickly than a light bulb, but with higher distortion. |
 A MOSFET increases resistance more quickly than a light bulb at the cost of greater harmonic content. |
 With a light bulb in the feedback loop (a), the power oscillator takes longer to stabilize than with the MOSFET in the feedback loop (b). |
No related content found.
- 0 rated items found.
Datasheets.com Electronic Parts & Inventory Search
185 million searchable parts
- Part Number
- Description
- Inventory
- Products
- Manufacturers
View All Webcasts
