Does anyone use a Faults database ?

[blindboybenton]

Good afternoon or good morning

Over the years iv'e always had thoughts of designing and using a faults database
Never seem to got round to it.
The easy bit is designing it, The hard bit would be making sure you enter all the info
I use Linux only and the customer database is in Open office Base

Working alone from a spare room in my house means time is important and anything that saves my time
must be good.
I did years ago do a spares database. It was more based on what semi conductors i have
I would write down whatever i took on a clip board and once a week i would remove and add parts
But after about 3 months it all stopped and sort of fell apart

I'm moving my workshop to the local guitarshop in a few weeks , I already do there repairs, with the access to the workshop
restricted to their opening hours. so no working on in the evening or sundays etc

Does anyone here use a faults database ?

Have a good sunday
BBB

[nosaj]

Good afternoon or good morning

Over the years iv'e always had thoughts of designing and using a faults database
Never seem to got round to it.
The easy bit is designing it, The hard bit would be making sure you enter all the info
I use Linux only and the customer database is in Open office Base

Working alone from a spare room in my house means time is important and anything that saves my time
must be good.
I did years ago do a spares database. It was more based on what semi conductors i have
I would write down whatever i took on a clip board and once a week i would remove and add parts
But after about 3 months it all stopped and sort of fell apart

I'm moving my workshop to the local guitarshop in a few weeks , I already do there repairs, with the access to the workshop
restricted to their opening hours. so no working on in the evening or sundays etc

Does anyone here use a faults database ?

Have a good sunday
BBB

With electronic components being the way that they are. I would think a faults database might lead you on a wild goose chase versus a well established troubleshooting procedure. And for me the only way to get there seems to be repetition and repetition on working on things. The more I work on things eventually the more that sticks in my brain on how things work.
Me I'm a work in progress one day one day I keep telling myself.

nosaj

[Enzo]

I would never use it. On the face of it, it assumes that faults are generic, ie when this symptom occurs, it must be this fault. And that is often not the case. Let's say "trem doesn;t work" is the fault. There surely is not one thing to look at for that. We could come up with a troubleshooting tree for that specific symptom, but then all that is is application of standard troubleshooting techniques to the particular circuit.

Consider the very common Fender trem. The one with the bug, rather than the bias wiggler. Now we have to wonder if the LFO is working or if the interface is shot. The LFO could have a tube bad or weak, the caps could be leaky, a resistor open, the jack bad, a bad pot, a broken connection ANYWHERE, etc. If the LFO works, we could have a bad bug (even that is bad bulb or bad photocell), a bad pot, a bad connection. There are typically five caps: three feedback caps, one cathode bypass, and one coupler to the output. We can't assume one of those is the problem all the time. We could decide just to replace all five caps up front, but then what is the point of a list? The resistors could go open, or way off value, but then how far off do they have to go before it matters? If a 1 meg resistor is up to 2 meg, mostly all that does is slow down the trem speed.

So what would my list say? it could say check the following parts, and then just list them all. Not helpful.

Now the other Fender trems. The LFO part is about the same, but the interface is wholly different. How to organize this, should I have a list for some model AB763, and AA864, and etc.? Or just general for that model, hoping all variants are the same?

Then we have Marshall trems, and AMpeg trems, and Gibson trems and so on. And once we have all the tube amps listed, we have solid state trem drive in endless array. But the bottom line is the trems are all very similar, so better to just understand how trems work and learn to troubleshoot than hoping a short cut pays off.

What if we have "Low preamp output". Pick a model, say PV 5150. We could guess a plate resistor is wonky, but which one? WOuld it always be the same one? I'd have to agree with anyone that in a Fender HR DeVille, it would most likely be the phase inverter plate resistor than any other plate resistor in it. But then instead of a list offering check plate resistors, I would have already just gone down the row of tubes and checked pins 1 and 6 for plate voltage and 3 and 8 for cathode voltages.

The main problem with lists is that it makes you spend time going through possible things in hopes your problem winds up being exactly the same as some other guy's problem. When I'd prefer to just track down the actual problem in the first place. Yes there are various makes and models which have a few certain symptoms that tend to be consistent. But how do you know from your list that the amp you are working on has a problem like that versus some other issue?

A parts data base is an inventory, that can be useful to tell you what is in stock, if you con't remember. I have a service document data base for my paper files. But that is just a list of physical assets, notanswers to problems.

[R.G.]

I have used faults databases before, on complex computer systems. They worked OK-ish. Got the problem isolation down to one or two replaceable assemblies, or cards.

In a guitar amp, so many problems are continuous spectrums, not true binary faults, that it's hard to say that something is or isn't a fault. Take bias, for instance. Is the bias right in this here amp I have in my hand? Umm.... maaaaaaaybe. What's the B+? Which output tubes? Crossover distortion %? How much crossover can you see on the scope?

Humans are truly amazing pattern matchers. We all do use fault databases, they're just inside our heads. It's amazing how large a % of amp faults can be diagnosed by one of the "databases" that read this forum from only a few lines typed on the screen. The build up over years of having seen X amp do Y funny stuff all goes into the fault database we carry. We just call the database "experience". In many instances, it is difficult to beat a human for recognizing patterns. In the few instances where machines have done this (such as, for playing chess) it's a combination of the computer being able to model massive numbers of possibilities and some "heuristics" (which means, a human told me that if this, then this, where the two are not necessarily logically connected) that let the computer do the job.

There is at least one database every tube amp owner should fill in. The owner should get a set of known-bad tubes - open, low emission, noisy, microphonic and so on, and go plug them into his amp, one at a time, then listen. That would cover the vast mass of problems in amps for most tube amp owners. Beyond that, get a more experienced "database" to look at it.

[blindboybenton]

Thanks for all the replies. If I was to do it. It wasn't going to be on things like .. changing tubes or plate resistors etc. It was just a thought. I spent years doing CRT TVs and we used a database of faults and some fault books. Was always helpful. As for the spares database. It was very useful once you know what semi's you had. Save you hunting . I had found that I thought I ain't got an 2sa xxxxx and when I ordered it . And put any extra ones in the draw I already had it. I'm an organised person. Hate to buy stuff I've already got. The only thing that triggered this post was I had a Ashdown mag300 in with a low output. I spent 30mins or so hunting the fault down in the preamp and when I found the fault I realised I had had the same fault 1 yr earlier .. I looked back though the customer database and couldn't find any info on the part and it just said op amp changed. BBB

[mozz]

Just keep case histories if your memory is not the greatest. If the unit has not been touched by anyone else, i can fix it. If it has been "fixed" by someone who is not a repair person, i spend more time looking to see what's what. My scope is my best tool, next to a volt meter and ohm meter.

[52 Bill]

Like a lot of others, I use this forum as my database.

I seem to remember that a couple of posters here that ran large service centers used them. For them it helped for warranty repairs that would show batch manufacturing problems, like a run of amps with bad tube sockets, etc.

That sort of thing may helpful if you do a lot of that kind of work, but understanding the circuit and noting the symptoms of the unit that is in front of you is probably more likely to lead you directly to the fault.

It's not unlike some of the posters here that point out problems that they have had with similar things. Sometimes it is related and sometimes it isn't and to the novice poster, it can sometimes lead them down the wrong path. That's why I like Enzo's fix the problem first and then upgrade approach. You don't waste time looking in the wrong direction and you don't have more problems than you started with.

[nickb]

To be of any statistical value, you need to have volume. I have written a web-based engine linked to mySQL for fault tracking in the past. It was used for specific production PCBs to understand failure modes. Given enough samples a pattern starts to emerge.

For a small workshop I think it would take more time to maintain, enter and retrieve data than would be saved on the few occasions that it might have been of benefit. Keeping good customer records reaps much greater rewards.

[Enzo]

RG's Geofex.com site has a nice tube amp debug routine.

GEO's Guitar Amp Pages