(Due Friday, March 28th)
Peter J. Kindlmann
Spring 2003
(Posted on 2/19/03, last edited 2/19/03)
Note the distinction between the minimum necessary, and the extra credit opportunity. Doing the extra credit portion, measurement and calculations, is worth about 50% extra.
Carefully build the audio amplifier in accord with the instructions of the kit. See Ed Jackson about any soldering or other assembly questions. Some components must be inserted with the right polarity--reversing them will damage them. We may be short on replacement parts. Please follow the instructions carefully, ask if you're not sure!
Setup:
Power the amplifier from the bench power supply set to 18V (take care to get the polarity right!), but set the current limit to about 300mA so as to keep the output transistors from over-dissipating (getting too hot) by some combinations of crossover adjustment, frequency and amplitude of operation. See Ed Jackson or Yanxiang Liu if you have questions about this. (A key thing to remember about the proper use of a previously set up to be current-limited: turn off the supply before you connect it to the circuit, then turn it on.)
Do the setup for crossover distortion correction with signal generator and oscilloscope. Note that it is essential to have the speaker (or, if you or your lab colleagues get tired of the sound, an equivalent resistance of adequate power rating, e.g. 7.5 ohms at 5 Watts--either on top of the parts cabinet in C048 or else Ed Jackson can direct you) connected during all functional tests. The "bootstrapping" of the Q2 collector resistor R4 depends on the presence of a load, and with it the basic operation of the circuit. Under quiescent conditions (i.e. powered up but not amplifying anything), the DC voltage at the junction of R6 and R7 should be about half the supply voltage.
In making any of the temporary circuit changes below for purposes of measurement, always be sure to do so with power off.
Bench measurements:
(a) Gain (normal closed-loop, i.e.with R5 feedback in place): With amplifier output well short of clipping, measure input (signal at Q1 base) and output (speaker/load voltage) amplitudes with the oscilloscope at 1kHz, and determine the gain. You should always use the x10 scope probes, unless you can not afford the loss of gain. Be sure to read and understand the discussion of passive oscilloscope probes at <http://www.eng.yale.edu/EE-Labs/morse/probe/probe.htm>. This measurement gives you the gain in the presence of the feedback provided by R5, which is also part of the biasing of Q1.
There is a further caution to observe in making these gain measurements, that of grounding the scope probes. To avoid a so-called "ground loop" which can seriously affect the input (low-level) amplitude measurement, ground only the scope probe associated with the amplifier input, and do so at the amplifier input. Leave ungrounded the scope probe looking at the speaker output. I'll be happy to explain the details in class and/or the lab.
(b) Frequency Response: With an output amplitude setting well below maximum, say around 2Vpeak-to-peak (can you venture a guess why I suggest this reduced amplitude?) determine the low and high "3 db frequencies" (the frequencies at which the gain has fallen by 3db, or to 0.707 of its mid-band value).
(c) Peak Output Swing (normal): Briefly turn the input volume way up at some mid-frequency (say 1kHz), and note the output amplitude at the onset of clipping of the positive and negative sine-wave peaks.
(d) Gain and frequency response (open loop, i.e. with feedback from R5 disabled): This is a bit trickier, because the R5 feedback also determines the quiescent operating conditions, something we do not want to disturb. Turn the power off. Locate the terminals of resistor R3 and solder a shorting wire across them. Then power up the circuit again, and carefully raise the input level until a modest output is reached, as in the frequency rersponse measurement above. Again determine the now open-loop gain and 3db frequency response points.
(e) (extra credit) Disabling bootstrapping: With power turned off again, unsolder the end of R4 that connects to the speaker terminal and lift it out of the board. Using a short jumper wire, connect that end of R4 to 'ground' (the negative battery terminal or other speaker terminal) instead. Power up the circuit again and slowly raise the input level to 2V peak-to-peak again. (This is still R3 shorted to disable signal feedback.) Determine the circuit's gain again and its 3db points. Now make a mid-frequency large amplitude test, observe the positive and negative clipping behavior and any distortion as the amplitude is raised toward that limit.
In your report writeup include: