Golden Rules of Troubleshooting

First analysis instrument to use on removing the case cover - your nose

a) Assume nothing. b) The nose knows. c) The complexity of the problem is in inverse proportion to the symptoms displayed. d) That one problem has been found and fixed does not imply that all problems have been either found or fixed. f) First cause may be intermittent. What is observed may be secondary.

Peter Wieck Melrose Park, PA

smell test is good.

check power, fuses, probe around for expected voltages. For example if there's any sort of linear regulator check output, input, check voltages on ICs. I deal with old stuff so there's always some 7400 series logic so

+5 is expected to be all over the place.

Look for burned parts, puffy caps, probe anything in areas with heat-discolored circit boards, check and resolder and suspect joints by parts that warms up. Diode check is a great feature on DMMS.

If you're dealing with suspect dried up caps, just solder a new one on the solder side of the board and see if that helps at all. I like to add parts on the bottom of the board, get things working then position parts where they belong if all good.

If these things appear OK, then you have to narrow the problem down and dig around into what the circuit should do vs. what's happening.

For intermittent problems tap the hell of boards, cable assemblies etc. Check power supply voltages with the device cold AND after it warms up.

Just came across a machine that would crash fairly often, The problem appears to be a mix of loose sensor and power connectors, a power supply that drops voltage as it warms up, plus a dead cooling fan in the electronics section. Just to keep things interesting, nothing warms up that much with the service panels removed.

There is problem with "twizzling" for intermittant soder joint failure, if it is a PbF board. Its too easy to temporarily "fix" a failed joint. I start with light touching stuff with a bird feather, while switched on. Then firmer twizzling. Then hot air or freezer spray. Then if I auapect a PbF failure ,an adapted engraver tool with a shaped nylon bolt in place of the engraving tip, to rattle the board, listening on headphones , if an audio amp.

I'm pleased to say I don't really deal with PbF stuff. How often do you find bad joints that looked OK (well, as good as PbF looks) and did have weird cracking or separating from the part lead?

Do you do repairs with real solder or that fake tin stuff?

Well, usually that is a very good rule, and saves time chasing what you THINK the problem must be, when in fact it is something novel you haven't seen before.

But, then, I have repaired a few thing recently, where past experience told me what the most likely failure was, and that was the correct thing to fix.

Jon

** Troubleshooting is like a maze, there are many starting points and blind alleys to negotiate on your the way to the goal.

Unfamiliar equipment requires a more cautious approach than familiar items.

Always observe the fault carefully *before* starting disassembly and find out what functions work and what do not.

Owner's descriptions are often misleading or fail to mention important issues, but you still need to find out what prompted that description.

Bench tests need to simulate actual use or worst case use, but beware of simulating abuse.

When stuck on a problem, take a break and start over a day or so later.

.... Phil

Make a block diagram. It may only be a mental picture, but it's an important step toward tracking a symptom down.

I taught a class of technicians once, from a local factory... the block-diagram-exercise got them talking amongst themselves more productively than any other lesson.

This thread, to me, represents the best of what newsgroups like this are for.

1. The aforementioned "assume nothing". (probably should have been ALL CAPS)

1. Correlate stated or observed malfunction with: a.) Visual inspection and yes, often smell checks. b.) Power supply checks. c.) Again, check for mechanically intermittent connections.

2. Don't assume too much from DC voltage checks with a multimeter. A 'scope can often give you the best "eyes" into what the circuit is actually doing.

Nevertheless, a good multimeter can give really great information interpreting voltage drops, finding leaky or shorted nodes, bias issues, etc.

Too many people just start off re-capping for example, assuming this fixes almost everything. I'm sick to death of hearing this from customers who have been on the internet.

Another thread here, I see...

Mark Z.

"When stuck on a problem, take a break and start over a day or so later." I cannot tell you how many problems I solved while doing that. Sometimes getting just getting away from the problem allows clearer thought.

Posting a query to some "social media" may not ellicit the solution but some suggesting will often set you thinking in a different train. Its so easy to decide in your own mind that the problem must be in some particular narrow area and you totally overlook a completely valid alternative area.

Most important, assume nothing. As soon as you think you know what is wron g, it is impossible not to ignore contrary evidence.

I work with some engineers who are good designers because they keep their b rain fixed on the goal and work through every obstacle. They THINK they ar e also good troubleshooters but they are wrong, obstinately frustratingly w rong, for the same reason. They KNOW what is wrong too early, and nothing shakes their belief.

Second. When anybody tells you the symptoms, listen. When they tell you w hy, STOP! Don't listen. It will probably be theoretically impossible, and then you end up doubting the symptoms. Sorry but this is especially a pro blem with techs and mechanics. They observe correctly but theorize contrar y to the laws of physics, so you need to stop listening when they get to th at point.

Capacitors?

I must confess we do replace almost all electrolytic caps that are more than about 30 years old in our vintage game monitors and amplifiers (video/pinball/jukebox) as this saves a tremendous amount of troubleshooting time. It is true that perhaps 60% of the caps replaced are fine, but the rest are marginal at best and make all sorts of errors creep in that would take hours to find otherwise.

While we are in there we check resistors on collectors/emitters/plates/cathodes as they often drift out of value. However those we only change if off more than 10%

John :-#)#

--
(Please post followups or tech inquiries to the USENET newsgroup) 
John's  Jukes Ltd. 2343 Main St., Vancouver, BC, Canada V5T 3C9 
(604)872-5757 or Fax 872-2010 (Pinballs, Jukes, Video Games) 
                      www.flippers.com 
        "Old pinballers never die, they just flip out."

It all depends on the cost and how much later down time you want to take a chance on.

At work we had a 200 HP motor drive control that quit and we called in a factory repair man. He determined that 2 large diodes ( arund 600 volt and

100 amp or so) were bad. AS this was a 3 phase unit , there were 3 diodes. I told him to replace the 3 rd one as it may have had some stress on it and may fail later. He said he would but it was $ 50 for the diode. I told him that at what we were paying him and the ammount of production it would cost that $ 50 was nothing worth even thinking about.

That often works. If possiable after working on a problem for a while, as someone else to take a look at it. Don't even tell him what you have done or checked.

I found that it is beter for me when asked for help to not even listen to what others have checked. It does pay to listen to see what parts they may have changed out. Then check those parts to make sure they are good and installed correctly.

I have done two things that rate as my worst boners. First was a problem with a SCR temperature controler. Thought I had repaired it, but when turning on the power it would blow a fuse. Did that 3 times and I just could not understand it. The next fuse I installed, I checked it and it was bad. Turned out when I checked the parts bin that half of the new ones were bad.

Another one was when I replaced a part that came with about 20 feet of 5 wire cable of about # 20 wire. I used the old wire to pull the new wire through the conduit. Some how I got distracted for a while and then when I connected the wires, I had forgotten to cut off the old wire. That ment I connected one end of the old wire back up and the other end was not connected to anything. I asked someone else to see what it was that I had done wrong as I knew it must be something simple that I over looked. He took one look in the junction box and asked why there was so much wire in it. Then found out what I had did wrong.

Back in the old days, McIntosh used to have clinics where owners could have their equipment restored up to specifications for free. I remember watchi ng them start by routinely replacing all of the cathode bypass electrolytic s; they would then check and replace the tubes, as neceeary. They rarely h ad to go farther (but would if required).

There is no correct approach. There are only approaches that work under specific circumstances. Like the universal solution to all problems, and the answer to all questions (42), the correct approach are only useful after you found the solution.

1. That which you know to be functional, beyond any need of checking, is usually the problem.
2. Solutions found after midnight are best forgotten as they are likely to be the product of temporary insanity.
3. Things go better if you have a schematic, gallon of coffee, a working unit to compare, and a parts unit to cannibalize.
4. While it is theoretically possible to work on two problems at one time, it is much easier to fix problems in the order they appear. I like to fix the little things first.
5. Nothing is considered fixed until after it is cleaned. If you can't fix it, at least make it look like it's been fixed.
6. Approximately half the disassembly videos on YouTube are wrong, useless, dangerous, or out of focus.
7. For a good collection of counterfeit parts, just buy them on eBay.
8. Your most valuable tool is a visual inspection, often with a magnifying glass. This helps identify exploded, incinerated, or overheated components. It also shows where the previous repairs were performed, a sure sign of problems to come.
9. Always ignore the symptoms and theories of the customer. They're often misleading and will usually waste your time chasing down false theories. Test the unit yourself to confirm the symptoms. I often find a very different set of problems or that the buzzwords in the original description were misused.
10. Fix everything that you find, even if the customer hasn't complained about it. If it's wrong, even slightly, they'll be back to complain and you might end up fixing it for free.
11. Take photographs of before and after. Treat is as if it were evidence in a lawsuit or insurance investigation. I also take a photo of the insides of the customers laptop, and use it for wallpaper on their desktop. The reactions are "interesting" and help fortify my exorbitant rates. The photos also help when I can't recall how to put it back together.
12. Never do what cannot be undone. For computahs, make an image backup of the hard disk both before and after the repair. Image backup programs are now very fast and Terabyte USB 3.0 hard disk drives are cheap.
13. No repair is complete without a culprit. The first step to doing a repair is to blame someone or something. This is an important part of the repair as most customers are worried that they may have done something wrong (such as buying a piece of junk). By blaming someone else, this eases their guilty conscience.
14. Never let the customer see you reading the instruction manual. They will immediately suspect that you don't know what you're doing.
15. Beware of spare parts left over after reassembly.

More later. I'm late (as usual).

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

Does it, though? I mean, you could save even more time by just running all the caps through an ESR meter test and just replacing those that fail. Given how fragile some of those old PCB traces can be, I'd only want to replace the caps that really way out of spec.

Um, we use temperature controlled soldering irons and the traces on most of our arcade grade games stand several soldering cycles...

The Bob Parker design ESR meter (which we've been selling since 1999) doesn't handle small value caps (under 10UFD) very well, and on the monitors a bunch are 4.7ufd or less.

There are only about fifteen caps in the average monitor, and it takes about an hour to pull the chassis, replace the caps, do the ring/LOPT test, check the fuse, reinstall and start the burn-in process. Then it goes away for five to ten years.

Amps can be worse, particularly early stereo tube amps where all the parts are connected on standoffs - those can take three to five hours to recap. They don't come back for ten to twenty years after a recap and resistor check (tubes too of course).

John :-#)#

--
(Please post followups or tech inquiries to the USENET newsgroup) 
John's  Jukes Ltd. 2343 Main St., Vancouver, BC, Canada V5T 3C9 
(604)872-5757 or Fax 872-2010 (Pinballs, Jukes, Video Games) 
                      www.flippers.com 
        "Old pinballers never die, they just flip out."