monolithic capacitors

[guitarman0]

Does anyone have any suggestions for an easy method of testing monolithic capacitors in circuit? I'm just looking for a quick and easy way of saying good or bad.

[Dave Dixon]

I'll jump in. Be prepared for a lot of different suggestions, as so many of us here are experienced in many different types of circuitry. In many applications, you can tell that a capacitor is "probably" okay by probing it one direction with an ohmmeter - watching the reading count up or down - then reversing the leads and seeing it do the same.
In many cases there will be a resistor, or resistive device in circuit, that will cause you to just get a resistance reading using this technique. If you have a 1K resistor in parallel with the cap, you will most likely get a 1K reading, which won't help much. Obviously detecting a short across the cap doesn't always mean it is the capacitor that is shorted. That is when it is usually necessary to start pulling up legs or pins for further troubleshooting.
I'm sure there are tons of suggestions coming your way. Hope this helps,
Dave
BTW Troubleshooting audio equipment, one technique we used was to take a know good capacitor of the same value, and parallel it with a suspected open capacitor in the circuit. If the device returned to "normal" operation, it was time to pull the component for testing, and probably replacement.<p>[ September 01, 2005: Message edited by: Dave Dixon ]</p>

[Dean Huster]

Caps are pretty hard to test in-circuit. Most of the smaller values are nearly perfect devices for all practical purposes: zero leakage, infinite resistance, zero ESR, zero inductance, etc. As a nearly perfect open circuit, it's nearly impossible to test in-circuit because nearly anything at all in parallel will affect the "perfect" test instrument reading.<p>One nice thing to consider is that most non-electrolytic caps just don't go bad unless they're of the ancient paper dielectric, and those got leakier than all heck or sometimes shorted. Usually, the most commonly defective caps will be the electrolytics, so concentrate on them more. There will be exceptions, of course ....<p>In general, by the time you've isolated the problem down to a specific circuit, it shouldn't be that big a deal to lift one end of a suspected cap and check it. Now, if what's happening is that you're trying to find a better way than "Easter egging" the trouble, it isn't by in-circuit cap checks of all caps -- it's by logical troubleshooting to narrow the problem down. "Easter egging" is the term we used in the military to describe a troubleshooting method by which you went through from beginning to end, checking every component rather than considering symptoms and isolating the problem to the most logical circuit, etc.<p>Dean

[Robert Reed]

And Dean-Don't forget "shotgunning", where the trouble was isolated to a section of the circuit, and then all those components were replaced.
Both of these methods were the true mark of the inept troubleshooter.
G'itman-
Of all the non 'lytic capacitors I have replaced fell into two categories (1) high voltage circuits from the tube days where when things went bad it happened in a hurry, and (2) in more recent low powered circuits that ran very "close to the wire" low volt rated capacitors (<10V).
If one thing has improved over the years, it has been component reliability. I can hardly beleive the accuracy oif 5% resistors and 10% capacitors I have checked in the last 10 years--most have checked within 1or 2% (however stability is another matter), even large value 'lytics are very close to the money. Why are you so concerned about checking monos in particular? Generally, when circuits fail, the pecking order goes something like this--Semis- 'lytics-power components- everything else.

[L. Daniel Rosa]

Military "easter egging" and "shotgun" troubleshooting? At first glance they look like methods of frustration management- paint a grenade pastel or just put the bead on the DUT and blow your troubles away...

[Dean Huster]

Off-topic for the cap checks, on-topic for the move in the troubleshooting techniques direction:<p>When I was stationed with the U.S. Naval Security Group Activity at Hanza, Okinawa, we had some Naval electronics equipment to maintain down at the nearby Army Security Agency. Understand that (at least in 1973), Navy Communications Technicians - Maintenance, Electronics Technicians, Fire Control Technicians and Radar Technicians spent anywhere from 9 to 14 months in electronics training before they were ever sent to the fleet to actually work on equipment. Of course, they were also enlisting for six years.<p>Anyway, our techs were amazed down at the ASA facility when they had a Teletype machine go bad. Some low-echelon army tech would come out and, according to a preordained list, replace each unit of the Teletype machine until the machine again worked properly. It might be a printer malfunction, but they'd start with the power supply, then the keyboard, then the .... well, whatever. They never varied from that ordered list. The Navy guy would look over at the obvious symptoms and tell the tech, "Hey, you've got a bad skanookanizer," or whatever. The Army tech would finally get down his list to finally replacing the skanookanizer and the machine would be operational. The bad skanookanizer was taken to the next echelon of repair where another tech would troubleshoot the thing and figure out what was wrong with the skanookanizer. The next echelon would actually affect the repair on the skanookanizer. Another echelon would test the skanookanizer to make sure it was working properly and then put it on a repair cart for the next exciting move in the world of electronic "troubleshooting".<p>I was in metrology -- test equipment repair and calibration. Unfortunately, we had the training but not all the equipment needed to calibrate our signal generators, at least according to the folks at Pomona. Anyway, we had to have our test equipment calibrated and contracted with the Army to do it. They'd rumble, rattle and bounce out on-site in two typical green AWD deuce-and-a-halfs, hook them together into one big room and open up all their rack-mounted calibration equipment. We sent a Hewlett-Packard 606A HF signal generator out there for calibration and it came back all screwed up with a non-sinusoidal waveform on a couple of the bands. Well, it went out that way, but I knew that a calibration would take care of the problem. I sent it back. The Army said that it was defective and needed repair. I said that all it needed was tweaked up. I tried to show them in their hallowed Army T.O. where the problem would be fixed. Alas, the Army T.O. did not have them do any of the tweaks inside the RF cage where the hp manual has you make a minimum of 24 tweaks to set the frequency and adjust the tank, and the Army techs were not allowed to vary outside that T.O. by regulation. Adjusting the tank is what takes care of the non-sinusoidal problem and I knew that after calibrating dozens of these things. So, right there in front of them, and to their horror, the squid opened up the RF cage, hooked up some test equipment and demonstrated to the dogfaces something that day. The Army guys decided that although they weren't allowed to vary an inch from the Army T.O., they would use the pages I copied from the hp manual whenever they got a 606 in again, keeping them hidden away from their C.O.<p>Dean

[guitarman0]

Thanks for all of the suggestions on the monolithic caps. Just a bit of explanation...the company I work for works on PBX phone systems for hotels and motels. I work in the electronics lab doing board level repairs on the boards, power supplies etc... The problem I'm having is one of the PBX manufacturers we support has a station card that has 12 ports per card. Each port contains 8 monolithic capacitors. A common failure item is for a mono cap to short causing the entire board to go down. Of course this one cap being shorted causes ALL of the mono's on the board to show short. Pulling a leg on each cap individually untill the shorted one is found is what we are currently doing, but it a time consuming, tedious task with that many caps to go through..and of course, the bad cap is ALWAYS going to be one of the last you come to. I was just hoping someone knew of a shortcut I hadn't heard of to test these little suckers.<p>By the way, my initial training was in the Army. I was an Avionics Special Repair tech. Worked on IFF transponders, weather radar, target aquisition, etc... for Hueys, Cobras, Apaches.

[Robert Reed]

WOW
Are you talking 8 monos per downed port to check or all 96 ( 8 X 12 ports) that have to be checked. If so I can see where you have a problem.

[rshayes]

One method that I have seen used to locate shorts on the power busses of printed circuit boards is to inject a current that flows through the short circuit and then try to locate the path that the current takes. If direct current is used, the path can be located by using a sensitive voltmeter to locate the point with the highest voltage drop across the power plane. This may require measuring voltage drops in the 100 microvolt range, but this is not impossible. <p>The same technique could probably be used with a low level signal in the kilohertz range and a high gain amplifier. An audio oscillator with an step down transformer could supply a low voltage and low impedance test signal. The current path could be detected with a pickup coil insteaad of a probe. This might not get you to a specific capacitor, but it should get you within about half an inch or so.<p>A current that is too high might blow the short open. Unfortunately, this might result in a temporary cure, since the short could reoccur in the future.<p>If this is a recurring problem, it would certainly be worth the time to set up some kind of systematic test rather than continuing to disconnect 96 parts per board.

[guitarman0]

Well, it would be all 96 if the shorted cap turned out to be the last cap in the last port. Luckily THAT hasn't happened yet, but I have had to go through sometimes 8 to 10 ports (80 caps) before finding the bad cap. <p>The audio oscillator might not be a bad idea. I'll have to check that out. Might at least be able to narrow it down to a specific port, if nothing else, and that would be a big big help.

[Robert Reed]

Gitman
Would it be possible to post a partial schematic of the area in concern and that circuit portions intended purpose. May be looking at this a different way, one of us could come up with a solution.

[positronicle]

--Edited by Positronicle--

[guitarman0]

Well, that is another little problem we have. We are working without schematics on these boards. They are not available to us. So, we are basically doing the old hunt 'n peck deal, educated guess as to what is what in the circuit and what may be wrong depending on the symptom, circuit tracing, visual inspection, etc... Makes for interesting days at times.

[Robert Reed]

Well now you have opened up a whole new can of worms-sorry, too many for me to sort out.

[thesprocket]

<blockquote><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><hr>Originally posted by Dean Huster:
Off-topic for the cap checks, on-topic for the move in the troubleshooting techniques direction:
...Anyway, our techs were amazed down at the ASA facility when they had a Teletype machine go bad. Some low-echelon army tech would come out and, according to a preordained list, replace each unit of the Teletype machine until the machine again worked properly. It might be a printer malfunction, but they'd start with the power supply, then the keyboard, then the .... well, whatever. They never varied from that ordered list. The Navy guy would look over at the obvious symptoms and tell the tech, "Hey, you've got a bad skanookanizer," or whatever. The Army tech would finally get down his list to finally replacing the skanookanizer and the machine would be operational. The bad skanookanizer was taken to the next echelon of repair where another tech would troubleshoot the thing and figure out what was wrong with the skanookanizer. ...Dean<hr></blockquote><p>Some of us dogfaces were enlisting for 6 years too, with 14 months training as well. We tended to fix all of our skanookanizer's in the field, lest they be sent off to return sometime after we rotated back to the states or what not. <p>Funny that you got onto the subject of instrument calibration, as that was our worst nightmare. as we couldn't forge/aquire the certification stickers we had to let the calabration shop have the instruments on schedual. As in your situation, they often came back worse than when they left.