I Knew Those Diodes When
Tunnel diodes probably seemed like pretty obscure and rare devices over most of the electronics industry, but they were so commonplace when I worked for Tektronix in their Oklahoma City Service Center \["40 Years Ago," Nov. 8, 1999, p. 40\]. Tunnel diodes, with their sub-nanosecond switching speeds, were the center of nearly every triggering circuit used in the 465, 475, 485, and 7Bxx series of plug-ins. It was to the point that I was usually surprised if someone didn't know all about the little guys. I'd bet that Tektronix was one of General Electric's biggest tunnel-diode customers. I was so proud of myself when I was finally taught what to look for when that little gold-plated device was put on a curve tracer.
Your column and Bob Pease's column make Electronic Design a real treat to receive.
Poplar Bluff Technical
The Way It Really Happened
I agree with your comments, except the analogy of IBM offering CPM in addition to DOS \["OFDM Links Two Controversies," Dec. 6, 1999, p. 22\]. The fact is IBM did offer purchasers their choice of PCDOS or CPM—the initial announcement mentioned both. At that time, I worked for Burroughs Corp. in Coral Springs, Fla., and many of my neighbors worked for IBM in Boca Raton, Fla. Their comments were that the PC was targeted for CPM, but initially they couldn't come to suitable terms with (aging memory dropout—I don't remember his name).
At the same time, they were negotiating with Bill Gates for his BASIC interpreter and happened to mention their CPM problem. He immediately promised an alternative based on CPM. The port of the clone to IBM's hardware just turned out better than the port of CPM.
Div. Chief Technical Advisor
Don't Try To Sell Me On 8-VSB
While I think your opinion might be a valid one to you, and no doubt was somewhat researched when you wrote it, time has not confirmed the facts \[Dec. 6, 1999, p. 22\].
That announcement about the VSB/QAM receiver you quoted has turned into one of the industry's most famous examples of vaporware. No working demonstrations have been performed for anyone, anyplace as of the date of this message \[Dec. 14, 1999\].
Plus, Sinclair has every right IMO to attempt to do two things, based on the fact that they may also have enough money to actually be a thorn in the FCC's side over this. I personally hope such is the case.
The first thing is of course to make them actually think about it using the most recent data available—hopefully obtained by their own people attending a series of tests as unbiased data takers. If they plead no budget, then that is a problem for Congress to address, and who has been emasculating the FCC budget in recent years?
Second is the sincere hope that the math will prove Sinclair is correct, and that this will give the commission the data (and the intestinal fortitude) they need to undo their mistake.
By mandating that 8-VSB is the standard of the land, the FCC is making one of its more colossal blunders. COFDM also has had some advances, most notably a reduction in the range advantage often given to 8-VSB based on the headroom required in transmitters to handle differing signal formats. Whether this can be completely ameliorated or not remains to be seen.
But if one gives any thought at all to the amount of CPU power required to train a set of filter taps to receive a ghosted 8-VSB signal in the presence of 10 or more ghost images, he or she will quickly understand that offloading the FCC's mistake into a drainage of the consumers' wallet to pay for that CPU power is going to be a huge mistake at the box office/cash register. Sets containing that much computer power aren't going to sell competitively, ever.
Where I live and work as the CE of a CBS affiliate in north central W. Va., we would gladly double our tx power if that's what it takes to do COFDM. With COFDM, we stand a chance of being seen in the bottoms of the valleys here, which is where the majority lives. Roads go there easily and all that.
This Design Idea Can Lead To Failure
Readers who attempt to use this circuit may well be in for a rude surprise in the form of seemingly random and sometimes spectacular failures \[Dec. 17, 1999, p. 114\]. This is caused by excessive current flow through the rectifier, should power be applied to the circuit at, or near, the positive peak of the input voltage. The discharged capacitor represents an impedance approaching 0 Ω, which leaves the rectifier to carry the current. This circuit was experimented with by various contract manufacturers with the same results. A fixed-series current-limiting resistor is required in all such cases to prevent those failures.
Thanks for your input. You are right. The capacitor I used was an aluminum electrolytic with a built-in ESR great enough to avoid your stated problem. Certainly, if one used an oversized input capacitor or a high-voltage ceramic capacitor, watch out! As this application is aimed at those requiring a low-cost, off-line solution, Murphy helps us here. The price of the capacitor goes inversely with its ESR.