Skip navigation
Electronic Design


Pay Attention To What You're Printing
I was very disappointed at the standard of the article, "Shed Some Pounds With This AC/DC Transformerless Power Supply" (Nov. 22, 1999, p. 109). I will leave aside the safety implications of a nonisolated power supply, as there were several warnings given in the article. I'll just highlight some technical errors:

  1. The use of a single-phase rectifier. The input power factor of this circuit will be very poor, typically <0.5. So for the 20-A line fuse, the maximum output power will be limited. The ripple current in the dc link capacitor C1 will be very high. No ratings were given for D1, which calculations show to need a substantial part due to the very high peak currents and high rms-to-average ratio. To be CE marked, the input current harmonics need to meet EN standards. The single-phase circuit will not, except at powers less than a few tens of watts.
  2. Use of 180-V clamp diode D8. Because the 115-V ac supply has 10% tolerance, the maximum peak is 179 V. The 180-V diode will typically have a 5% tolerance, so this is 171 V min. Thus, at high supply the diode will burn out. Furthermore, there is only the 20-A line fuse to limit power in D8 and this part is rated at about 5 W = 27 mA at 180 V.
  3. Output ripple and size of smoothing capacitor. It is stated that for lower ripple, use a larger smoothing capacitor, e.g., 10,000 µF rather than 1000 µF. I would suggest that the ripple at line frequency in the output is due to poor control-circuit frequency response. Adding more capacitance is not the answer. This will in fact make the input power factor even worse.

What worries me is that other articles, which I read to educate myself about areas in which I am not expert, could have as many errors.
P. Barrass
Power Design Manager
Control Techniques, U.K.

Give Credit Where It Belongs
I am a reader of your publication as well as one of the founders of Entridia Corp., a company that was mentioned in the article "Of Holodecks And Transporters: The Promise Of Unlimited Bandwidth" (Jan. 10, p. 56).

My dismay arose when I noticed that our copyrighted diagram of network infrastructure segmentation (Fig. 3) was not attributed to either Entridia or Jeremy Bunting of Thomas Weisel Partners (co-copyright holders). I trust that credit for all future artwork placed in Electronic Design will be duly noted and printed.
Paramesh Gopi
Cofounder of Entridia Corp.

Don't Sign On The Dotted Line
I enjoyed your column in the Feb. 7 Electronic Design ("Revel In Your New Ideas Despite Negative Feedback," p. 164). (Decimal-keeping slide rule? I'd love to see one. I like to baffle my younger colleagues by using the slide rule in my office from time to time...)

You said, "If you invent in the field of your employer, he or she will own the invention..." Quite often, it's worse than that. A lot of employers have standard agreements that say that everything you invent when an employee, whether OR NOT in their field, whether or not on your own time, belongs to the employer.

As I recall, that was true at one company where I worked. Interestingly enough, it was not at the next one. So when I joined my next employer, a small startup, I asked for a change from their standard agreement so it covered only work in their field, rather than all work. At least they were agreeable on that. They didn't even give me a hard time.

This is something people may want to watch out for. I suspect a lot of engineers don't read the fine print when they show up for the first day of work, and probably didn't think to ask for a copy of the employee agreement before they sent their acceptance. (I know I didn't until the third time.)
Paul Koning
Xedia Corp.

In the March 6 issue, the wrong caption was run with the figure in Timeless Ideas For Design (page 134). The correct version is shown here.

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.