On Monday 01 April 2019 11:31:08 Jon Elson wrote: > On 04/01/2019 06:24 AM, Gene Heskett wrote: > > It once took me around 3 years to find an engineering > > goofup by Chyron. > > Well, here's one crazy fault that I finally debugged. It > was a rack-mount mag tape controller for a PDP-11, made by > Datum. Basically one huge 15" square board in an aluminum 3U > or so cabinet. > It was stuck in reset. The reset/ line ran all over the > board, but only to inputs, the driver was on the system > bus. And, reset/ was stuck low, and even when powered off, > it was just a couple Ohms to ground. I started cutting the > reset/ trace in spots to isolate it, and pulled the last 2 > chips, but it was still grounded. Finally, I had it down to > a 2" strip of copper trace, and this was a 2-layer board, > and there was nothing on the other side, and no ground trace > nearby. I finally peeled the trace off the board, and you > could not see ANYTHING odd there! Replaced the chips, > soldered a piece of wire-wrap wire (with the insulation LEFT > ON in the middle) and it worked fine for years. > > That was certainly one of the odder faults I ever had to > debug! Not an engineering goof, just a weird defect. > > A piece of gear I worked on had a photomultiplier that > sensed a green light from a flame to indicate the sulfur > content of the air. This also provided a flame on/off > signal. The range of the signal from flame on to flame off > was about .6 V. They set up two transistors, one lit a > flame-out light, the other held the gas solenoid open. So, > either one transistor should be on, or the other one should > be on. But, the .6 V difference was almost the entire > signal swing, so the circuit was really flaky. I eventually > figured out how to do it with ONE transistor, as the lamp > drew so little current it could not possibly hold the > solenoid open. I put the lamp across the transistor that > switched the solenoid valve.
I'd have done the same, except I'd have added a flyback diode across the solenoid to absorb the shutoff pulse from the solenoid. Perhaps with a small r in series to absorb more of the stored energy in the R. > > Since this was in a safety system, I was a bit hesitant to > be so bold with this change, but, my circuit WORKED, the > original required frequent tweaking, maybe as room > temperature changed or like voltage changed. > Actually, it seems like you did exactly what a good designer should have done in the first place. I always ask myself will I lose any sleep worrying about it. If the answer is no, it gets done. I will fix a blown circuit according to the schematic maybe twice. The 3rd time it fails, I'll redo the design to remove that failure. Case in point, a balancing relay with platinum points that cost a couple hundred a set was used to run the motor of a 10kw powerstat in an RCA TT-25 transmitter for automatic line voltage control. But the thing was mounted on a transmitter wall that had a constant vibration from the belt driven 10 hp cooling blower. Contact life was about 6 months and you could see the fire between the contacts. I thought about it, looked up a starting and extinguishing voltage of an NE-2, and put a network of 5 NE-2's from each contact to a 2200 ohm 5 watt fire proof R connected to the common point. It was sorta entertaining to watch the NE-2's flicker, but couldn't see any arcing at the contacts. It regulated closer than it ever did new, and for all I know, that was in the early 60's at KOTA-tv in RCSD, it could have been running yet on the same contacts when they shut it off for the final time at midnight June 30 2008 when the digital switch was made. I was there till about '70 when I went to KXNE-tv-19 for Nebraska ETV and learned about high power klystrons. Thats a whole nother story. Lets just say that they confirm E=MC2 as their major distortion source. And tuning adjustment mistakes can destroy a $125,000 klystron in 10 milliseconds. Never had to explain that, only tube I lost in 8 years was due to a heineman breaker single phasing its cooling pump. We designed a 50 kilovolt vacuum relay to detect that. On test, it ran the filter caps, used to 20kv, up to almost 40kv, but they survived. They were a LOT cheaper than klystrons in any event. Cheers, Gene Heskett -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) Genes Web page <http://geneslinuxbox.net:6309/gene> _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
