RAW666

's profile
image
Pr. Engr.

Biography has not been added


RAW666

's contributions
  • 10.17.2017
  • A lifetime designing PCBs: Switching to eCAD
  • Charles, Thanks for the memories. My initial contact with PCB CAD (as a customer) was with early really bad autorouted designs in the late 1970s. As you say, the cost of CAD designs was well over twice that of taped designs. For several years, the in-house PCB design process went from taping to having the PCB designers having to draft the routing with pencil and paper and then those drawings were placed on a drafting table sized digitizer and another person entered the drawing into the computer. The argument for this huge increase in cost was the claim that the digitized data was easier to change later. However, the first time I tried to get a small change made to a digitized design resulted in full corruption of the whole design. The CAD gradually has improved but I still morn the loss of taping technology for small to medium sized designs.
  • 10.24.2017
  • An overlay transistor cautionary tale
  • John, When transistors are "second sourced" using just registered JEDEC parameters, many different transistors can result. One decades long problem in the company I worked for concerned what was originally a Motorola complementary pair (2N6341 and 2N6438). These were what were called "triple diffused" and had very good high frequency performance. However, other companies were able to pass the registered parameters with transistors that came nowhere near the performance of the Motorola design. We had to use military qualified (JANTX) parts so the real Motorola parts became scarce and then unobtainable. We were able to struggle on by screening for the available parts that would not burn out immediately but it still made the amplifier continue to have a very high failure rate. In later years, things got even worse when the only company supplying JANTX2N6438s began building them with two paralleled die and no emitter ballasting. The company obviously was able to pass the JEDEC parameters but also obviously the transistors fail more frequently because one die hogs most of the current and overheats. Also, I remember well the 2N3866 and 2N3375 from my video amplifier design days.
  • 01.31.2017
  • Adrift on the meandering path of product development: DC-probe case study
  • Very interesting blog Thomas, Thanks for promoting ExpressPCB and FrontPanelExpress. I made use of both extensively for test equipment projects during my last 20 years of employment. I was always impressed by the bug free software and extremely user friendly online interface of both services. I found the ExpressPCB software fully bug free from the very beginning - unlike some other early online services. I used the FrontPanelExpress service not only for control panels but also for mounting fixtures and for transparent plastic alignment overlays.
  • 08.10.2012
  • The shocking truth about FETs
  • Small signal FETs in TO72 packages (such as Signetics SD210s) used to be shipped with either a metal clip or a conductive elastic band around the four leads. The clip or band had to be in place until the part was assembled into the board. I remember one large lot of parts that by military contract had to be sent out to an outside testing contractor. The testing contractor removed the protective clips and sent the parts back in plastic bags. Needless to say, most of them were destroyed. Failures after assembly were rare as long as the clips remained in place until after assembly. Most of this was back when ESD procedures were rare.
  • 03.17.2016
  • Could test and measurement crack Farook's iPhone?
  • It is interesting to speculate about how to hack an iphone but don't forget that the FBI is not really expecting to find any useful information on Farook's employer supplied cell phone. He destroyed his personal cell phone before the attack. The FBI wanted to obtain a legal ruling that would eventually prohibit any ordinary person from using strong encryption on a commercial device. When Justice Scalia died, the chances of obtaining such a ruling in the near future became remote and the FBI dropped the effort.
  • 02.29.2016
  • Leap day fun stuff
  • The pong concept has been popular for a long time. The "round pong" clip reminded me of a test pattern that I wrote for a test set used for testing the JHMCS helmet display (used on several US fighter jets).
  • 06.28.2013
  • Detect charged bodies with electronic electroscope
  • That seems to suggest that an actual electronic electroscope would have some physical structure like a leaf 'scope but a low bias current op amp and a voltage meter at the "leaf" end. The sensing structure would have to be reset to "zero" every few minutes as the op amp bias current will eventually charge the sensing structure. There are several CMOS op amps that claim bias currents well below 0.1 picoamps.
  • 03.27.2015
  • Ode to the Oscilloscope
  • Rather than having a cart, I usually had to make do with a USM-24 or another MIL scope that I have forgotten the designation. I recall that those MIL scopes had an elegant means to get around the lack of a linear horizontal sweep. They incorporated burst generators that could modulate the trace. Thus you counted dots on the trace rather than lining up with a graticule to measure time. Actually more accurate in some ways than a linear sweep. This feature never seemed to make its way into commercial scopes. Perhaps my delayed mode is also stuck in 1968. My time in the USAF was 1964 - 1968, all working on the old EC-121 AWACs. Roger Weise
  • 11.29.2014
  • Classic oscilloscope front panels show how it's done
  • I think that the ease of use of the Tek 400 series and 7000 series front panels likely will never be replicated. The mechanical design of the time base selector was terrific. Although the 500 series front panels were not as refined as later designs, those 500 series machines made good space heaters. My Tek 585 could keep a small unheated room comfortable in the winter. The Tek 556 with its separate power supply box may have been good to heat a house! I wish that I now had room for a Tek 547 since I consider it the best. It was the last of the high resolution CRTs and thus one could see very small details of a signal that are blurred out on later low res scopes.
  • 12.02.2014
  • Make high-quality double-sided PCBs – at home
  • Very good question! All of the boards that I can remember where we used the Polaroid coating tended to be rather low impedance so a moderate amount of leakage would not have been noticed. If I ever run across an old applicator I will try to remember to revive it with some alcohol and do a leakage test.