Dear Editor (Aug. 1975)

Transient IM

Dear Editor:

Thank you for publishing the Leach article on "Transient IM Distortion in Power Amplifiers" (Audio, Feb., 1975). This is the type of analysis that is very much needed in the design of audio equipment, to take it out of the "arts and crafts" category and give it a firm scientific basis.

From 1948 until about 1968 (when I bought my first transistor amplifier, a Dynaco Stereo 120), my standard of comparison was a vacuum tube amplifier that I designed and built and wrote up for Audio (Engineering) in 1948. All stages were push-pull. The output, from a pair of Western Electric 300-Bs, was 30 watts, at about 2% IM distortion, and the power curve was within 1 dB from 20 Hz to 22 kHz.

There was no feedback of any kind.

My friends used to bring in other amplifiers for A-B tests, especially some of the early transistor jobs, and this old 300-B amplifier always won without any argument, until about 1968.

The best praise I have for Leach's article is to say that I wish I had written it!

Curtiss R. Schafer; Sandy Hook, Ct.

More on TIM

Dear Sir:

In his otherwise excellent article on transient IM distortion (TIM) in the February Audio, Prof. Leach states that operational amplifiers like the 741 are prone to TIM because of their low 2 to 5 Hertz open loop bandwidth. In audio design, as in servo and other control systems, the term "open-loop" gain means gain around the entire forward and feedback paths. But in op amp terminology, "open-loop" response means the response of the device without external feedback. The compensation controlling the bandwidth is usually lead type, rather than lag which (as he notes) reduces TIM instead of increasing it.

The slew rate of the 741 is about 0.5 volt per microsecond-roughly 5000 times faster than it would be if lag compensation were used. This fast slew rate insures that any "holes" resulting from transients will be so brief as to be inaudible. I have critically compared 741s with high grade tube circuits and could hear no difference (closed-loop gain was limited to 20 dB max.). There may be ways of using op amps that could lead to problems, but local feedback around each op amp will prevent this.

-Donald E. Phillips; Cedar Rapids, Iowa

The Author Replies:

The 741 op amp is internally lag compensated by a 30 pF capacitor.

This is necessary to prevent oscillation when external feedback is added.

Normally, when the op amp is used with external negative feedback, the bandwidth of the closed-loop amplifier is reduced by placing a capacitor in parallel with the feedback resistor from the output terminal to the inverting input terminal. This capacitor increases the feedback at higher frequencies, thus reducing the bandwidth of the closed-loop amplifier. If the feedback loop is broken, the capacitor has in effect increased the bandwidth around the complete forward and feedback paths, and this is, technically, lead compensation. However, in the case of the 741, lead compensation in the feedback loop is not necessary for stability since the op amp is already internally lag compensated for this purpose.

Although lead compensation in the feedback loop of an op amp will reduce its tendency to produce TIM when used as a voltage amplifier on its own, we must examine what would occur when this op amp is included in the forward path of a power amplifier with negative feedback of its own.

Since the bandwidth of the op amp has been reduced by "local" lead compensation in its feedback loop, the open-loop bandwidth of the power amplifier as a whole has been reduced. This can lead to TIM in other stages of the amplifier, although this depends heavily on the particular design.

Nearly all state-of-the-art power amplifiers today employ an op amp input stage, whether discrete components or an IC. I prefer the discrete designs until some IC maker markets a unit in which the amplifier designer can adjust the internal frequency compensation (these exist, but are not in production), and more importantly, the local negative feedback in each internal stage. This way the op amp will be capable of being tailored to meet the specific objectives of the amplifier designer. Otherwise, I think most IC op amps are better suited for d.c. and low-frequency instrumentation amplifiers. However, some of the IC amplifiers designed for r.f. applications may be ideally suited for audio purposes. I have not investigated this in any detail.

Whether IC op amps are suited for mixer and low-level applications where they will not be inside the feedback loop of another amplifier is something I have not investigated. I would refer those interested in this application to Hoge, W.J.J., "Tubes Versus Transistors: A Further Comment," J.A.E.S., June 1974, p. 338. Mr. Hoge gives the results of an actual listening test performed on commercially available mixer amplifiers. His conclusion was that mixer amplifiers which have open-loop bandwidths smaller than the audible frequency spectrum sound inferior.

In my article on TIM, I concentrated on describing its worse possible manifestation in power amplifiers. There are more subtle effects which I did not describe relating to how the THD of an amplifier varies with frequency at a constant power level. If the amplifier's THD increases rapidly with increasing frequency, a subtle form of TIM can be triggered by wideband audio signals with excessive high frequency levels, even though the input stages do not clip. I believe this form of TIM is the one most likely to be produced by a commercially available amplifier.

-W. Marshall; Leach School of Elec. Eng. Geo. Ins. of Tech.

FTC Violators

Dear Sir:

Although the FTC has passed the "honest power in advertising" rule, it has had no effect whatever in our area. Ads still flood the newspapers here claiming 200-watts for 8-track tape players. I phoned several of these advertisers and none of them was aware of the rule; most of them refused to believe me. And one man said, "Even if there is such a ruling, they'll never bother with us-we're too small."

I'm sure this problem exists all over the country in smaller cities. To what department of the FTC should one send clippings of offending ads, along with a letter asking the FTC to inform the violators that they are wrong?

-Bruce Cullom Sound Town, Inc., Texarkana, Tx.

The rule most certainly applies to 8 track players as well as receivers and amplifiers. Inquiries about the rule should be sent to:

Carthon E: Aldhizer Div. of Special Statutes Federal Trade Commission Washington, D.C. 20580

Tel: (202) 963-7124.

In addition, the EIA has published a batch of data on the ruling, and should be glad to send materials to stores in your area. Address Jack Way man, Electronic Industries Assn., 2001 Eye St. N. W., Washington, D. C. 20006.

Telephone (202) 296-5550.

Electo-O-Phonie Enthusiast

Dear Sir:

Since I am an owner of the Elekt-O-Phonie FU-100's immediate predecessor, the F/Mc2, I read Professor Lirpa's review in Audio (April, page 54) with great interest. Almost before I finished reading I ran to the nearest Five and Ten Cent store where I bought an FU-100 for my car. It works fine on 12 volts d.c. My only problem came when I fitted the Klipschorns (I refuse to listen to any other speakers) into my Volkswagen. It was worth the trouble, though. Can you imagine my satisfaction as I cruise down the Boston Post Road on my way to work every morning at 140 mph, listening to all eight channels simultaneously?

-Alan M. August N. Providence, RA.

Dear Sir:

My hat is off to Professor I. Lirpa for his enlightening equipment profile of the Elekt-O-Phonie Model FU-100 Octaphonic Receiver. One question plagues me about this obviously sophisticated piece of gear. What does the "FU" stand for? Frequencies Unthinkable, maybe?

-Jack Stevens; WXCL Radio Peoria, Ill.

(Source: Audio magazine.)

= = = =