Bob’s Mailbox

Dear Bob:

                The “Tube Sound” letter in the Sept. 19, 1994 issue asks about a simple way to get “tube” sound from solid-state amps. The Oct. 3, 1994 issue of Electronic Engineering Times has an article on tube emulators which mentions one such device from Déjà vu Audio, Berkeley Springs, Va. The article goes on to discuss other “over drive” factors. While some key factors are discussed, such as overload distortion characteristics, other tube “signatures” are omitted, such as the tube amplifier’s better stability into nasty reactive loads (like electrostatic speakers), high peak to rms power ratios (due to high-voltage rails of tube circuits), and lower damping factors (due to output transformer coupling and lower negative feedback).

                My specialty is speaker engineering and I find that a number of characteristics attributed to amplifiers are often the interaction between the output stage and the crossover network and/or the voice coil of the speaker. The differences between amplifiers or cables that are perceived in listening evaluations may not be directly due to the cables themselves, but caused by interactions with time-varying phenomena.

                When these components are changed midway into the playback of a CD, the shift in sound characteristic may actually be the effect of the speaker voice coil heating up and the resultant impedance rise. Passive crossover network -3-dB points can shift one-half octave or more when due to temperature coefficients. Often, a system’s approach will reveal that what is being heard is an interaction between components. The differences perceived are real, but cannot simply be measured component by component. They can only be documented by looking at the whole system.


Menlo Scientific Ltd.

Berkeley, Calif.

Thanks for your notes on odd thermal effects in speakers and crossover networks. Maybe we need liquid-cooled crossovers? However, those “high-voltage” rails of vacuum-tube circuits aren’t really very high when properly viewed through the output transformer!—RAP


Dear Electronic Design Readers:

                I know many of you have purchased, integrated, and operated test equipment that failed to perform in accordance with the manufacturer’s specifications. Like most test engineers, I have found dozens (maybe hundreds) of problems while integrating software with “off-the-shelf” test equipment. Many of you have also spent long hours on the phone with the manufacturer trying to find workarounds—workarounds which would not have been necessary if the equipment complied with published specifications. Perhaps you have spent several hours tracking down and classifying a problem only to find the manufacturer has a deaf ear. Do you need an alternative? Are you ready to try something new?

                I am compiling a test-equipment problem database. If you will send me written descriptions of your test-equipment problems, I will attempt to categorize these problems reports. The data will be used to encourage test-equipment manufacturers to address these problems.

                We should all take time to inform vendors when we have a problem and we should also take the time to warn fellow test-equipment users when vendors are unresponsive to our complaints. Please take the times to do something positive for yourself and your fellow equipment users.

                I want to make this easy and flexible. You can send your complaint(s) to a mail address, a Compuserve address, or an Internet address. Since I could be swamped with complaints, I can not promise individual replies, but I will try to let the user community know the results of the study. Please supply the following information when responding:

                Manufacturer’s name

                Model number of test equipment

                Description of the problem

                Approximate date of the problem

                Results of contact with vendor

                Problem significance (will you still buy from this vendor?)

                Complainer’s name (will not be released to test-instrument vendor without consent)

                Complainer’s address (will not be released to test-instrument vendor without consent)



7227 E. Chadowes

Wichita, Kans. 67206

Compuserve: 74722,2622


[email protected]

This seems like a good idea for constructive criticism, rather than just griping. I recommend that everybody do this.—RAP


Dear Bob:

                First, let me say how much I enjoy your column. I have your book, too, and love it. I like the way your mind works! You see right through to the essence of the true nature of things, whether in semiconductor circuits, mechanical devices, or whatever. And we both know that a little humor serves as an essential lubricant in life.

                In a previous column you mentioned Tom Milligan, who was then at Philbrick, telling his people that when fully results were obtained when a circuit was under development, they should Record the Amount of Funny. This sounds like it just might be my good friend Tom Milligan, whom I once worked with in broadcasting eons ago in Charlotte, North Carolina. He then went to MIT, and I went to work for the government, and we lost track of each other. If you have his address, I would love to have it, so I can contact him.


Springfield, Va.

No, we have been looking for Tom for several years. Does anybody know his whereabouts? –RAP


Dear Bob:

                After my return from a long trip, I can today respond to TWO of your historical (and enjoyable) refreshers:

1.       Teledeltos paper was indeed limited to two-dimensional analysis, but I believe I once used a semilog resistance paper. To obtain 3D solutions for TWT focusing fields, I used an electrolytic tank when I worked (loosely speaking) at the Air Force Lab in Dayton, Ohio. Its probe even automatically followed equipotential lines, and the electrolyte was quite harmless—a bucket of sea water will do fine.

2.       I built a V-F converter in the early 1950s, but I never knew it. My lab at the Technical University of Vienna (the real one) got a grant got building a device to measure core iron losses. Those on the market needed several pounds of test samples, while I did it with as many grams. Essentially, I had to integrate the product of current and voltage over one full ac period—a hysteresis microwattmeter.

The V-F solution came to me while listening to a then young pianist named Gulda playing Beethoven (I believe it was the Hammerklavier sonata). Make constant-width pulses, I said, whose amplitude varies with the voltage, while the frequency followed the current (converted to a voltage in a grid resistor). The gadget worked well enough to get the lab its grant, and me not only a doctorate, but a patent in 1953.

Just before I left for the above-mentioned Dayton lab, Siemens wanted to prove I infringed on their electroMECHANICAL wattmeter! I convinced them to buy my patent for about $500, which was an extra month’s income on my new job. (Those were the days when a buck still got some respect.)


Boonton, N.J.

That’s interesting to hear about your old V-F integrators! And thanks for the comments on the Teledeltos paper: I got a lot of comments on 2D and 3D analogs! –RAP 

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.