Short: Solves prob with high voltage discharges The following article was written originally by Dick Landry of New Brighton, MN and appeared in the Oct 89 issue of Amiga Transactor on pgs 11,12. I have copied it for posting since several people have asked for it; not having access to the magazine. Any spelling errors are mine, all other material is not. nuff said. Here it is. -------------------------------------------------------------------------- A potential hardware bug that may result in a high voltage discharge (snap) is prevalent in several early models of RGB monitors supplied for C128 and Amiga Computers. Since the advent of the Amiga and C128 computers, a small but significant number of us have experienced a problem with a sudden snap, crackle, or popping sound accompanied by a momentary loss of picture on the monitor. This event woul usually be infrequent to start with, but the snapping would become more frequent with time. On one Amiga 2000 system, it developed immediately, and was so bad it would crash the system. I suggested that the owner return the system and have it replaced. I haven't seen the owner since, and I suspect he returned the whole Amiga system. In the first year that the Amiga was on the market, I purchased my Amiga 1000 and the associated monitor at a Chapter 11 sale. The monitor occasionally cracked an electrical snap. About a year ago, it gave a snap that evoked a guru (system crash). The snap seemed to go away for a few months. Later, when my computer system was used for a demonstration at a club meeting the electrical snap returned. Inquiries did not provide any answers. I had the monitor checked over. We found taking the cover off caused the snapping to stop. A muffin fan was attached, but soon after, the computer developed an occasional lockup when I attempted to save or print a file. The frequency of the lockup increased, and in short order, the computer could be used for less than an hour before it locked up. I observed the lockup occurred after there was a snap in the monitor. Intermittent problems are the hardest to correct, but the problem had to be isolated and solved. Frank Gerard of ECS, an authorized Commodore service shop, worked on the problem with several different monitors and concluded that it was related to the high-voltage wire going from the high- voltage transformer to the anode cap on the top of the picture tube. He sprayed a plastic insulating coating around the high-voltage transformer where the red wire emanated. This seemed to correct the problem in one or two monitors, but not in mine nor several others. Bill Hanley, an engineer for the local public TV station, also brought in monitors with snapping problems. After a discussion with Frank, Bill checked several bulletin boards, and found information about a trace being established on the monitor mother board by arcing. Frank studied the bottom side of the mother board and found evidence of carbon traces created by shorts from a heat shield tab to ground on several snapping monitors. There is a heat sind for a voltage regulator on the side of the monitor near the high-voltage transformer. This heat sink is a L-shaped metal plate about four inches high and fige inches long. It angles at the rear corner for about an inch along the back of the monitor. The heat sink is attached to the mother board of the monitor by two screws and a twist metal tab that extends through the mother board. The metal twist tab is very close to the ground foil on the bottom side of the mother board. When twisted, it comes very close to the ground foil. And it doesn't take much voltage to bridge such a narrow gap. The pathway is on the bottom side of the mother board and out of site, so it seemed an improbable source for the problem. We concluded that the large heat sink seems to act as a large capicator near that high voltage wire. When voltage builds up high enough in the heat sink, it discharges with a small spark to the ground foil. Consistent discharges inside a dirty monitor will help build a trace path to ground increasing the frequency and size of the voltage discharge. If the discharge is large enough, the high voltage on the ground trace will be reflected back to and through other monitor components and back to the computer. This problem seems to be prevalent on Commodore/Amiga models 2002, 1902, and 1080. The Amiga 1084 monitor and other RGB Commodore monitors have a smaller heat shield and seem to have a better high-voltage cable path that is farther from possible conducting paths and seems to prevent the problem from occurring in those monitor models. Four steps were taken to solve the problem (please remember, even powered down, monitors can retain potentials of over 20,000 volts. Tech- nical experience is a must. Don't say we didn't warn you! -Ed.): 1. Clip the shorting metal twist tab from the heat sink. The two screws can hold the heat sink adequately. Carefully scrape any carbon traces on the mother board created by arcing from the tab slot to the ground trace. 2. Provide extra electrical insulation around the red HV wire. Split heat shrink tubing to wrap around the HV wire and use electrical tape to completely encase the wire with extra layers of insulation. 3. Spray Koloid Clear Acrylic plastic around the wire and the high- voltage regulator to reduce leaking from the high-voltage sources. 4. Use TV Corona Dope to plastic coat the area around the twist tab slot and the ground trace. Frank completes the repair at a cost of $30 (US), but other service centers say they can't do the job for that price. A reasonable charge might fall into the $50(US) range. In two months of heavy use, no noticeable snapping has occurred. --------------------------------------------------------------------------- This is an additional posting by Ken Wallewein on problems with the 1084 monitor. > I have the problem with my 1084 monitor where it snaps occasionally with a > jump in the picture. I read the article about this problem in the > Transactor, and they suggested several things including using a > high-voltage insulating spray. They also mentioned a spot on the bottom of > the logic board which suffers from arcing. I opened up my 1084 to fix this > and discovered that there is a large flat sheet of metal soldered to the > bottom - some sort of shield - which makes it impossible to get at the spot > they were talking about in the article. None of the suggestions they gave > could be applied to my 1084 so I'm out of luck. Has anyone else had a > similar experience? > ________________________ > Bob Fillmore, Systems Software & Communications BITNET: FILLMORE@EMRCAN > Computer Services Centre, BIX: bfillmore > Energy, Mines, & Resources Canada Voice: (613) 992-2832 > 588 Booth St., Ottawa, Ontario, Canada K1A 0E4 FAX: (613) 996-2953 I'd read that article too, as well as various postings that have appeared here. My monitor had been snapping quite a lot, and I'd been quite unsuccessful in determining the location of the spark, even though I could hear it. The article pointed me in the right direction, but I didn't like the solution -- it didn't seem to address the root cause of the problem, which was that the heat sink simply wasn't electrically connected to anything! It picked up a static charge from being in close proximity to all that high voltage, but there was no way for the charge to bleed off. Eventually the charge got high enough to arc, and then the whole process started over. I verified this with a FET VOM (meter), which has sufficient sensitivity that making contact to the metal place after a few minutes of operation caused the needle to jump. Leaving the meter connected stopped the sparking and snapping altogether, even though the current was too low to register. The meter has an internal resistance of 10 megohms or so; it obviously didn't take much conductivity to drain off the static charge. I could have replaced the meter's action with a high-value resistor. However, after using the meter to confirm that the plate had no other electrical connections, I decided to simply ground it with a short piece of wire. Everything has worked properly since. That's it, folks. Ground the plate with a piece of wire and the problem is solved. ----- Another piece of hacking I did on the monitor I don't recommend so highly, but I'll pass it on anyway. By tweaking (read: frobbing) a ferrite slug and a variable resistor, I was able to get the screen to shrink by about an inch both horizontally and vertically (sorry, I can't tell you which ones; I just tweaked until I got what I wanted). I then used 'morerows' to increase the logical size of the screen by nearly 100 horizontally and 40 vertically. It's like having a bigger monitor! ----- Let me warn anyone else who wants to try these hacks; I took a chance. I don't have the schematics, and I don't know if they will cause long-term damage. But I sure like the results. /kenw Ken Wallewein kenw@noah.arc.ab.ca (403)297-2660