Repair sagas welcome here, or new designs only ?

[Milacron]

I ask as I use European and Japanese CNC machine tools and measuring instruments, and when something electronic goes wrong, the manufacturer often refuses to provide any schematics..... and a replacement module is outrageously expensive (like over $10,000 in some cases...on a machine worth $15,000).   So, I end up on quite the "treasure hunt" for the bad component on a multi thousand dollar board sometimes and have tales to tell in that regard.  My last repair involved desoldering/resoldering over 400 contacts before I finally found the bad IC.  The desoldering was probably as difficult as it gets too...thru hole DIP's, very tight spaces, required the smallest nozzle that Hakko offers... and I wanted to remove each 20 pin DIP intact for testing (1990 vintage static RAM's)

Two instruments that have been very helpful to me are the Huntron Tracker and the BK IC tester, which I've not heard Dave mention.  I could elaborate more on those if anyone cares. 

Plus we could get into a nice round of hissy fits over Stabilant...which I have found extremely useful....but it is some controversial stuff... 

East Coast, USA

[Conrad Hoffman]

Hey Milacron, I think I know you from somewhere! I'd like to be of help, but my suggestions would include things like the Huntron Tracker! I'm a test equipment junkie and my basic philosophy is to find some way, any way, to figure out what's going on without disturbing components. I get pretty disgusted with myself if I end up unsoldering a lot of parts to find a problem. Most of what I know goes back to ohms law and I'm not above measuring microvolt drops on circuit traces to see what chip might be drawing more current than others. In circuit impedance checks are good, as is discovering slight edge and amplitude differences that suggest a bad driver or such. Thermal searches with a tiny thermocouple probe can be useful. Ultimately though, I don't think all machine control circuits are as well designed for service and troubleshooting as they should be (OK, what is?) and that's why you get paid the big buck for fixing this stuff! You do, don't you? BTW, I'm very new to this board and since it's a ways away, you may find the times of maximum activity are a bit unusual.

[Milacron]

  my basic philosophy is to find some way, any way, to figure out what's going on without disturbing components. I get pretty disgusted with myself if I end up unsoldering a lot of parts to find a problem.

I agree, that's one of the nice things about the Huntron Tracker..if nothing else it's useful for indication what is probably not wrong with a board !  Also, I have an ESR meter for testing capacitors on the board, but even there you "have to hold your mouth right" to know if the reading is for real or bogus. 

The latest desoldering marathon was only after exhausting all other practical possibilities.  I say "practical" since in this case, thermal probe with board under power would have been a monumental undertaking due to all the test wires (40+) that would have been necessary to connect the board remotely just to get to components under power.  The 20 pin DIP's were replaced with sockets so at least the "new" static rams will never have to be desoldered in the future.

---------------

On another subject.... is there no way to include location as part of profile shown on each post ?  And even if logged in, we still have to type the "captcha" letters and the "mathmatics" question just to post or edit ? (what a PITA)



 

[vk6zgo]

I for one,am interested in hearing about people's repair adventures.
I don't subscribe to the"Eeek! Repairing modern Electronics is way too difficult for little moi!",school of thought.

It is damn annoying that manufactures don't provide schematics----the usual excuse is Intellectual Property!
Most of the time they are using public domain circuitry,so the King definitely has no clothes!

Many companies seem to assume that you live in the same country as them,so you can pick up replacement PCBs easily,but if you live on the other side of the world,it is much more difficult!
It is even worse,if the original manufacturer either doesn't support the unit anymore,is defunct,or,as in your case, wants a "King's Ransom" for the board.
In these cases,there is no other option ,but to fix the thing in-house!

We had a Huntron Tracker at one of the TV Stations I was at.Only one Tech ever used it much,& he swore by it.
As far as I could see,you needed another good board to compare with,& if I had that,I could do the same things with an Oscilloscope & DMM,so I never spent the time needed to learn its use.(probably my loss!)

You shouldn't have to mess around like that to post or edit.I misread your comment & had to edit
this--I just hit "modify"& there I was.I'll now hit "Save" & I'll be back on the forum!
You can put your location in your profile,by Clicking on "profile" & digging around a bit,till you get the right page to do it.(I had trouble doing this,too,as it sometimes comes up wrong)

VK6ZGO

[Frangible]

I wouldn't mind reading a few Repair sagas.  They're just like debugging sagas!

[ElektroQuark]

Yep, another interested here.

[Zad]

Many students spend a year before uni "travelling" (i.e. going abroad and getting drunk for a year). I spent mine repairing electronics! 12 months spent repairing a wide variety of consumer electronics probably taught me more about real world electronics than the next 4 years at uni did. The sort of diagnosis skills you learn are just not taught in a formal environment.

[Milacron]

We had a Huntron Tracker at one of the TV Stations I was at.Only one Tech ever used it much,& he swore by it.
As far as I could see,you needed another good board to compare with,& if I had that,I could do the same things with an Oscilloscope & DMM,so I never spent the time needed to learn its use.(probably my loss!)

 

Yes, you need a good board to compare signatures on the screen, otherwise might as well use Oscope and DMM.  Huntron T is good for finding bad solder joints and bad resistors, caps, diodes, etc.... anything that has gone defective that shows with board not under power.  But if defect only shows with board under power, such as with most IC's,  the Tracker is of no use.  And if you can't find anything wrong with board out of control and you can't access the board back in the control under power, then things get really tricky. 

I've heard that folks that really know what they are doing, could connect the good board to a (large, upper end, very expensive) logic analyser, which would record what should be going on and then connect bad board to same instrument so it could pinpoint the defective component.  But seems to me, even that would be quite a PITA in some cases due to the odd board connector and 40+ wires to therefore patch in.   Anyone know more about this ?

[Conrad Hoffman]

My experience with logic analyzers is limited and ancient, but for a while they were incredibly cheap on the surplus market, at least around here. There are/were a variety of types and it seemed that as logic families advanced and then processors became the norm, the old analyzers got dumped for new ones with more and different features (and speed). I can only imagine one being useful if you had schematics, or for identifying signals around a single possibly bad chip, and it seems like a scope with 4 channels might be more useful for that. Without a schematic, where would you connect all those wretched probes?

My thing is analog, and my guess is the troubleshooting methods for that are substantially different from digital. Still, the root cause of most problem is usually mechanical. My guess is bad connections, cracked traces, open wires and various short circuits outnumber failed ICs by a good margin. #1 failed item on my bench would be capacitors, but I don't know how many you find in machine tools, other than the power supply and various bypasses. That's not to say ICs don't fail; goodness knows I've replaced a lot of 'em.

The worst thing I've ever fixed was an old GR impedance comparator. It basically worked but as the frequency was increased there was a phase shift that couldn't be calibrated out. I sunk about 40 hours into the thing and replaced a variety of components to no avail. I'm not even sure how I found the problem, but right at the input grid of a tube, a trace was cracked. The capacitance across the crack was sufficient to pass signal and make the instrument work. But, the crack also introduced phase shift- remember that the input impedance of a tube can be rather high, so even with a cracked trace everything looked more or less normal. One touch of the soldering iron and it worked perfectly. Time to find > 40 hours, time to fix > 40 seconds!