Intermittent fault finding

[Welsh_sparky_21]

Is there anything more frustrating than an intermittent fault

 

[buzzlightyear]

Is there anything more frustrating than an intermittent fault

Explaine.

 

[timhoward]

Is there anything more frustrating than an intermittent fault

Maybe only if you spend hours chasing your tail based on false information and there's no fault at all?

 

[ImpededLoop]

Explaine.

He's probably been to a job with a dodgy ring final and a knackered RCD

 

[buzzlightyear]

He's probably been to a job with a dodgy ring final and a knackered RCD

I hope he's not gone to the one I had.

 

[buzzlightyear]

Had two fault call out with in a week they must catching.

 

[James]

fixed one this morning.
tumble drier sometimes works, sometimes not.
appliance engineer had replaced the start button and the door switch.
fault was still there and still intermittent.

after 5 mins of fault finding had to try really hard not to look smug as I pushed the plug ALL THE WAY in to the socket.

that is a fantastic win for before 9am

 

[brianmoooore]

Finding what's causing intermittent or obscure faults is the highlight of the job, especially if others have tried and failed before you.

 

[timhoward]

I had a very different sort of win today, but not after some frustration, and it's nothing really to do with usual electrics.
A frail neighbour / friend who has Parkinson's disease called me to ask if I could do a diagnostic scan on his car.
I said sure but what's the story?
The fault was that the headlights only work on main beam. He'd been to a small garage and main dealer and been told by both it wasn't anything simple. The dealer wanted him to spend £120 on a diagnostics scan. That was booked in for tomorrow. He was getting cold feet and wondered if I'd take a look. I was sceptical a scanner would tell us anything useful, but played along.
So I take the scanner over there, and it draws complete blank, there are no faults stored or pending on any module. Then I check fuses. Then I dismantle enough to test the light switch. Then I check main beam relay. All fine.
Eventually I think it's time to distrust all previous information, and check the obvious so I put a meter on the bulb connections, and flipping heck there's been 14v there all along! All it needed was two new bulbs.
It made my blood boil a bit that two different places were trying to prey on vulnerable chap. I've started a letter of complaint to the main dealer as while I can cope with them being smarmy and pretentious this was downright incompetent at best or dishonest at worst.

 

[brianmoooore]

I take it this car doesn't have lamp failure warning? Unlikely that both lamps would fail at the same time, so he's probably been driving a one eyed monster around for some time, until the second one failed.

 

[buzzlightyear]

diagnostic scan

Which make of scan I'm looking for a good car one.

 

[timhoward]

I take it this car doesn't have lamp failure warning? Unlikely that both lamps would fail at the same time, so he's probably been driving a one eyed monster around for some time, until the second one failed.

I wondered about that, but didn't want to get more involved as I was already late home for tea.

Which make of scan I'm looking for a good car one.

Unless you want to spend many hundreds on a pro tool it's best to get something specific for the manufacturer, so this case it was VCDS running on a laptop for V.A.G group cars. I've got the Vauxhall one too, can't remember what that's called!
Best to join a forum for the relevant make and ask what's best.

 

[Lucien Nunes]

I see intermittents more or less daily. Knowing how to approach and provoke them is an important part of my skillset. Yesterday's was on a valve 1/4" tape recorder from the 1960s. It had come in a year or two ago for a service with a report that there was no playback audio. It did need a service but I found nothing that would have completely stopped the playback. It came in again back end of last year with the report of no playback, but worked perfectly on arrival. The record / playback switching is often the most troublesome part of tape recorder electronics so I went over this carefully and tried all reasonable means to provoke a no-playback response over a period of weeks but the machine worked consistently. Tapping, shaking, poking various likely parts, flexing PCBs, freezer spray, hairdryer, nothing made any difference. I switched it on yesterday morning for one last test before returning it to the customer and presto! No playback.

The total absence of hum and hiss indicated that the fault was near the output stage but before I got as far as a buzz test it was obvious that the last valve in the signal path was not lit. A close inspection of the PCB revealed a slightly suspect joint on pin 5 (one of the heater pins). When heated with the iron, the solder fell away like water off a duck's back, i.e. it had never wetted the valve holder pin during manufacture; it had only ever been touching mechanically. The problem had taken over 50 years to show up and had resisted my attempts to provoke it. But the customer was right after all.

 

[littlespark]

I think @Lucien Nunes has been taking English lessons from @buzzlightyear

I had to read that at least twice before I understood what he was tapping, shaking and poking!

 

[Lucien Nunes]

I should perhaps have used semicolons instead of commas to form the list. Sorry.

Back in my live audio days I would get to know the gear from some of the hire companies quite well. There was a front-of-house processing rack that would repeatedly turn up with one equaliser that had an intermittent fault. Twice I stuck a note on it pointing out the fault was intermittent and it could not be assumed to be OK to send out just because it seemed OK in the warehouse. Eventually I spoke to the warehouse manager, who admitted they kept sending it to me because they couldn't find the fault but hoped I would. Apparently I had gained enough of a reputation for fixing gear in the field that they thought I stood a better chance than their own techs.

 

[Lucien Nunes]

A theatre lighting dimmer rack that we installed would occasionally blow one or more of its three 2A control fuses (the control module takes all three phases to get a mains timing reference and will power itself from any or all of them.) There was nothing apparently wrong with the SMPSU so it seemed likely that there were transients on the supply that were being clamped by the control module's VDRs. After a few visits to replace fuses, Richard twigged that there was 4-pole isolation in the 3-phase supply. He tested the isolator and found that the neutral was breaking first and making last, causing a momentary neutral loss and hence overvoltage impulse on a random phase at each switching cycle. Replacing the isolator stopped the fuses blowing.

 

[Lucien Nunes]

The action of the Compton organ at Southampton Guildhall generates a good selection of intermittents. This mammoth instrument contains about 7,000 electromagnets, over 50,000 sets of contacts and over 100 miles of 26SWG cotton-covered wiring joining it all up. All of the electrical gear, much of which is installed in a dedicated relay room, is original 1937 equipment, almost unique today. The main problem is that it doesn't get used enough, so all those contacts sit there from one month to the next, gathering dust and tarnish (they are mostly silver) and take a few cycles of use to get cleaned and back to reliable operation.

When we attend for a service visit, it takes me the best part of half an hour to exercise everything and take notes of what is not coming back to life. I'll make a list of a few dozen problems that are all likely within a particular location - console, relay room etc - and usually find half of them unrepeatable or genuinely cleared by the time I get there. Once I've dealt with those, I'll return to the console for another round to find half of the half that had cleared are back. We never actually get them all, but on an instrument of that size, they tend to go unnoticed in normal use.

What does not go unnoticed is any failure of the setter action that stores and recalls stop combinations on the thumb pistons. This is a fully electromechanical binary memory of some 4-5 kilobits, that works using around 700 solenoids to manipulate grids of spring contacts into the correct slots in busbars. It is remarkably reliable but if you get one loud stop that fails to go off when a quiet registration is recalled, it is a real problem.

Typical problems that need dealing with on the electrical side are:
Dirty contacts
Dirty contacts
Inadequate tension of contact springs
Dirty contacts
Mechanically sticky parts associated with electromagnets and contacts
Flux corrosion to lead-out wires of solenoids
Dirty contacts
etc.

Later when I am on the other computer I will post some pics of two of the most challenging sources of intermittents in this instrument, which are the console main cable connectors. When you see them, you will understand!

FWIW I am posting all this stuff to amuse myself while sitting in hospital connected to the chemo infusion pumps. I must do some proper work.

 

[timhoward]

The action of the Compton organ at Southampton Guildhall generates a good selection of intermittents. This mammoth instrument contains about 7,000 electromagnets, over 50,000 sets of contacts and over 100 miles of 26SWG cotton-covered wiring joining it all up. All of the electrical gear, much of which is installed in a dedicated relay room, is original 1937 equipment, almost unique today. The main problem is that it doesn't get used enough, so all those contacts sit there from one month to the next, gathering dust and tarnish (they are mostly silver) and take a few cycles of use to get cleaned and back to reliable operation.

I hear you. (I'm an organist btw). I get called in to play for carol services on instruments that barely get played the rest of the year, and real ingenuity can be called for to work around the faults. Tracker (mechanical) action instruments are normally ok, but there's a couple of electro-pneumatics that are long running adversaries of mine. Each year I write a polite essay in the tuning book, and the following year I read the witty responses to last years faults. One year I had to avoid about 6 notes for the whole evening, substituting with right foot and octave coupler.
There's one considerably smaller EP action in a church near me has contacts that have been hopelessly bent out of shape by someone dropping the keyboard stack of the manual above on them. I've recently spent most of a morning gently re-shaping them. About 3 have snapped off leading to a fun session lying on my back on the pedalboard soldering in mid-air directly above my face!
I think these ones are phosphor bronze and I need to find a source of contact wire sometime...

BTW I too have been on the receiving end of those infusion pumps. Are they the Siemens ones, and have you learnt to silence the 5 minute remaining alert yourself yet?

 

[timhoward]

Just for @Lucien Nunes benefit - favourite quotes from the tuning book:
Organist: "Evidence of pedal felt falling off".
Organ tuner: "Evidence removed"

Organist: "Oboe Tenor C almost out of tune".
Organ tuner: "Oboe Tenor C almost tuned"

 

[littlespark]

One year I had to avoid about 6 notes for the whole evening,

That would have sounded like Les Dawson at his best