I am aware of the risk of shock because they can hold a charge for a while until they are discharged, but are there other things that I need to be cautious of? Can't they leak and ruin things, just like a battery? What else should we look out for related to the power supplies? Ashley

John, Thanks for the vivid visual images! I suppose I've been lucky so far. Actually, most all of the stuff I've acquired has recently been powered on by someone else before it came into my hands, so if it was going to go BANG!, it probably would have done so before it came into my possession. I'd be interested in hearing any horror stories relating to exploding capacitors and more info on the proper caution that needs to be exercised when messing with old power supplies. Ashley

On Wed, 2005-03-16 at 17:53 -0500, der Mouse wrote: Nor me, but a recent tale... I tried bringing that Sun 2/120 up the other day, and one of the big filter caps on the output side let go with an impressive hissing and *masses* of smoke. Funny thing was, it had been brought up after many years dormant on light load first (I didn't want to risk any of the boards in there!) and allowed to sit for a bit like that. Tested voltage levels and they were OK, so shut down and fitted all the boards. Back up again, all voltage levels OK. It ran like that for maybe 30 minutes, and only then did one of the big caps fail. Luckily they had safety valves otherwise it would have been a rather impressive explosion. Thing is, a variac wouldn't have helped at all - everything checked out. Things were up to temperature well before the cap blew, and there wasn't anything we did to put the PSU under any stress at the point when it broke (*unless* something died and shorted in one of the boards). So cautionary tale is not to stick your eyes near *any* big old caps in running equipment I suppose, no matter how good they are! I've replaced the dead cap (and its twin just to be sure) - the blown one failed as a dead short. Ran for about 30 mins on a dummy load and it was giving out useful voltages. This weekend I'll get to see if any boards in the machine were cooked by the failure, but I'm hopeful that the destroyed cap was the only casualty... cheers Jules

John Lawson wrote: Only for linear supplies! Don't do that with switchers. It's probably worthwhile to disconnect electrolytic capacitors and reform them separately first. There are tons of web pages about that, but the basic principle is to apply current-limited DC to it and slowly let it charge. If you apply rated voltage without current limiting, and it needs reformation, that's likely to cause it to fail catastrophically as John described. For a very overengineered approach to capacitor reformation which we used on the PDP-1 project, see the draft of the restoration procedure, the fifth document in this directory: http://pdp-1.org/Documents/RestorationProject/ For software that automates that procedure using an HP or Agilent programmable power supply, see: http://www.brouhaha.com/~eric/software/wrec/ Eric

I do not own a variac, so I will need to get one of these to properly reform the TU56 capacitors. I see several different varieties on eBay. What would I need exactly? Ashley

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=50968&item=3… Thanks for all the good info and the lessons on how to use this stuff. I looked at the one on ebay and it appears that the Watt meter does not work. Will that be a problem? Also, I did not see any place where I could plug a supply directly into this variac. I don't know if I'll need to do that. Based on all the info that you and others have provided, it looks like an option is to remove the capacitors from the unit and hook them up directly to the variac. Or is that not the right thing to do? Excuse me if I'm asking dumb questions, it's because I'm learning as I go. Ashley

From: "Ashley Carder" <wacarder at usit.net> need As I understand it, the right tool for an electrolytic that has been removed attempting to reform them in place, by reducing the AC line voltage to the power supply that they are a part of. Vince

Your explanations of the two methods all seems to make sense to me. But then again, I'm too new at this to know if it's good or bad! I've always just plugged the stuff in and it hasn't exploded (yet!). But I'm a dummy, just tryng to graduate to something else beyond dummy. I have a long ways to go to become an expert. There probably aren't enough years left in my lifespan to allow me to become an expert, but I may be able to get beyond dummy some time during my lifetime. Ashley

http://www.vintage-radio.com/projects/capacitor-reformer.shtml Above is link to a circuit for a capacitor reformer with details for use in the states on 110v. Geoff.

I now have Two nice machines (Tek scope, VT100) that blow fuses and have been dancing aound the issue to get one for a few weeks now. I'm not really an electronics guy but I've been led to believe that a variac is the single best thing to get to help remedy fuse blows. Any strong yes or no votes towards this decision? John A.

Plugging them into a variac won't address the primary issue of What's Wrong With This Widget? It can help you find something that's getting real hot if you can switch the unit on at about 60-70 volts and ramp it up slowly, if that doesn't blow the fuse. But I'd check the obvious stuff first, and then troubleshoot the power supply, before trying the above. Check the blatantly obvious first: 1. Right fuse? 1a. Right current? 1b. Right type (slow blow vs. fast blow?) 2. Shorted big filter caps or diodes in the power supply? 3. Lost screws or other metal parts loose inside? Don't laugh. I've seen it. Item two is easy enough to find, but only in a relatively simple power supply. If the PS is linear, not a switcher, I'd try disconnecting it from everything else and troubleshooting it in isolation. If it's a switcher, it's much to difficult (for me) to repair. Switchers have small transformers (proportionate to the current they produce, and employ a high frequency oscillator. You shouldn't run a switcher without a load. You may let all the smoke out. For linear power supplies, try to isolate the power supply circuitry from the rest of the unit. Find the transformer, see if the input and output voltages (AC) are reasonable. Should be, transformer failure mode in 99.999% of cases is no juice at all. Can you find diodes? Or a bridge rectifier? It is any good? A shorted diode or bridge will kill the fuse. You might have to unsolder one end of the diode to test it. Clip something metalic onto the wire lead near the diode so the heat from desoldering it doesn't do further damage. If you don't have any clamp-on heatsinks, loop a rubber band around the handles of a small needle nose pliers and use that. Classic linear power supples may have a grounded center tap on the transformer, or not, perhaps two diodes, more likely four or a bridge. Look for a capacitor or several on the output of the rectification stage. Depending on what sort of regulation circuitry follows, they may/may not have big ugly electrolytics on the output of the rectifiers. If you can unsolder or unscrew one lead on the capacitors, if present, try metering them for shorts or really low impedance that doesn't float as the ohms battery in your meter charges the cap. If you can find the outputs of the power supply, and you've been living right and they are all socketed, and this is a linear, and you can plug it in and run it unplugged from the rest of the unit, and have proper voltage on all the outputs of the PS, then obviously something is smoked further into the unit. Bad luck. Look for evidence of heat damage (fried components, burned printed circuit board, etc.). At 12:26 PM 3/17/2005 -0500, you wrote: [Computing] I have been told that _Introduction to Objectivist Epistemology_ was required reading at the Xerox PARC lab where OOP was invented, but this may be merely an urban legend. -- wilcoxb at cs.colorado.edu (Bryce Wilcox) --... ...-- -.. . -. ----. --.- --.- -... tpeters at nospam.mixcom.com (remove "nospam") N9QQB (amateur radio) "HEY YOU" (loud shouting) WEB ADDRESS http//www.mixweb.com/tpeters 43? 7' 17.2" N by 88? 6' 28.9" W, Elevation 815', Grid Square EN53wc WAN/LAN/Telcom Analyst, Tech Writer, MCP, Cisco Certified CCNA

On Thu, 2005-03-17 at 11:00 -0500, Ashley Carder wrote: If I'm not mistaken the whole reforming capacitors discussion was for electrolitic capacitors (ie DC). The motor run capacitors are AC. I'm not sure the reforming process would be the same (or if it is even possible). Anyone want to comment? -- TTFN - Guy

On Thu, 17 Mar 2005 10:20:10 -0800 "vrs" <vrs at msn.com> wrote: I doubt if a motor capacitor is made up out of a back-to-back electrolytic pair. Usually, motor capacitors are some form of 'oil filled' capacitor or modern variant, and are non-polarized. Usually not of a high enough value to need to be an electrolytic design. I have the 'perfect instrument' for testing and restoring capacitors. It's a US Navy LCR bridge that has built in charging power supply. So you can install a capacitor onto the 'unknown' terminals, charge it up to as high as 400 volts DC, while measuring the leakage current. You can then measure it's capacitance and quality while charged to said voltage. It's intended for tube-era parts, but will work fine for lower voltages. The best device to check out and 'form up' electrolytics of lower voltages (IMHO) is a regular constant-voltage/constant-current bench supply. You set it up to the working voltage for the capacitor, then turn the current limiting very low and attach the capacitor. If the capacitor is 'good' it should charge up to the supplie's set voltage and stop drawing current. If it draws current for awhile, you're not going to 'rush' current into it and cause damage. Important to remember when charging capacitors up is that they then have a fairly high 'instantaneous' current capability. If you charge, for instance, a 4700 uF capacitor up to 40 volts, care should be taken not to short it out. Discharge it slowly through a resistance, say 10K ohms or so. You might 'melt' whatever shorting item you use if you just crowbar it, and the hot metal can burn. Fooling around in the lab, we used to hang big 'computer grade' capacitors (I had a 470,000 uF one) across regular 'low current' bench supplies to 'arc weld.' You can melt copper wire with a part like that 'trickle' charging up to 40 volts on an ordinary supply. -Scott

On Thu, 17 Mar 2005 21:26:33 -0800 "vrs" <vrs at msn.com> wrote: The capacitors could be oil-filled, hence the 'goo' they'd squirt. Said oil might be nasty and carcinogenic, mind you. (PCBs) But 100uF is rather high for it to be a non-electrolytic.

Very true. There's a reason railgun power supplies tend to be capacitor banks which are slowly charged and very quickly discharged. (Railguns, at least conventional railguns, demand very high current for a very short time - mega-amps for milliseconds.) Right. There's a reason high-power computer supplies have labels showing slashed-circle "no!" signs over watches and rings: they are entirely capable of turning a metallic object like a ring or watch red-hot right on your hand. If you're really unlucky, it can simultaneously spot-weld it to whatever it touched and fail to overload trip (.1 ohm contact resistance at 5 volts draws only 250 watts, which is not that much to an even moderately beefy computer power supply). I once deliberately tripped a mains breaker by shorting hot to ground with a screwdriver. It made an impressive crack! sound, and even more impressive was that the screwdriver shaft was vaporized half through - probably a couple of grams of steel gone. Not something I'm going to repeat; it gave me a good deal more respect for such things. /~\ The ASCII der Mouse \ / Ribbon Campaign X Against HTML mouse at rodents.montreal.qc.ca / \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B

Motor start capacitors tend to be far more beefy* than filter caps, so I would doubt any of them would ever need reforming, unless they are REALLY pushing the age envelope (as in 100 years). I have never heard of nor have seen a motor start cap blow due to "deforming" (although I have seen some blow due to bad seals at the lugs, making either moisture get in or dielectric get out. That is a different failure mode, however.), and I have been around quite a lot of old motor start caps. If the caps looks good (look at the seals), it very likely is good. *Due to the harsh electrical environment they tend to be subjected to. William Donzelli aw288 at osfn.org

Ashley wrote: When the electrolytic capacitor is first manufactured, it has to go through a "forming" process to grow a thin layer of oxide on one electrode. This oxide layer is the dielectric. It is formed by application of current-limited DC. If the capacitor is unused for an extended period of time, the oxide layer breaks down. At that point if a DC voltage is applied that is not limited to a relatively low current, it will punch through at the point that the oxide is the thinnest, resulting in permanent damage. Provided that the capacitor has not failed either by leaking, the electrolyte drying out, or punch-through, applying current-limited DC will reform the oxide layer, restoring the capacitor to a useful state. Eric

Jay West wrote: replacements is about $70 each! For the four motor >run caps anyways. Just what are the caps? ... digs thru surplus catalogs... :) Ben alias woodelf

Vince said: My TU56 arrived and I spent some time looking over it. It is missing a G742 module, but has a "homemade" card in its place. It is also missing two M941 jumper/extender cards. Otherwise it appears to be complete. I need a couple of empty take-up reels. Does anyone have any of these or the G742 or M941s? I don't want to waste decent DECtapes just to get empty reels. It is quite dusty and has some cat hairs on it. I'm thinking I should dis-assemble some of it and get rid of the dust that has adhered to most things. What's the best way to clean the backplane to restore it to what should be a functional, clean condition? I looked at the capacitors and they don't appear to have leaked (yet). I also discovered that it likely has not been powered on since 1989. Would it be prudent to go ahead and replace the capacitors, or should I use the "power up slowly" method described by others so the the capacitors can be re-formed? The tape spindles spin by hand and don't appear to be too tight. I will take pictures of the work-in-process and put on my web site at http://www.woffordwitch.com. I'll add a "Current Projects" link on the left-hand nav menu and add TU56/TC11 under there. I'd like to keep a log of everything. Perhaps others can use this information in the future. Ashley