View Single Post
  #14   Report Post  
Posted to sci.electronics.repair
John Robertson John Robertson is offline
external usenet poster
 
Posts: 907
Default Diffferent techniques in troubleshooting

On 12/31/2015 12:14 PM, John Heath wrote:
On Thursday, December 31, 2015 at 12:14:59 PM UTC-5, Jeff Liebermann wrote:
On Wed, 30 Dec 2015 16:10:40 -0800 (PST), John Heath
wrote:

I would like to know if others have found this to be true.


It's true, but only if you have a production line environment. During
the late 1960's, I worked for a company that did warranty service on
mostly consumer tape recorders. We would get the rejects and returns
from the distributor. We had 4 people working the line. One to unbox
and machine and organize the paperwork. One do to nothing but
diagnostics. Me to tear it apart and do parts replacement, more
testing, and reassembly. One more to do more paperwork, a final
check, and reboxing. Anything that couldn't be fixed in 10-15 mins
was put aside for later troubleshooting. Anything with more than 5
faults was deemed not worth repairing. It was quite efficient. As I
vaguely recall, we could do about 4 to 10 machines per hour.

What was interesting was the general lack of test equipment. As you
noted, it takes far too much time to setup and probe. For audio,
putting a finger on the base of a transistor and listening for hum
worked well. To make it quick an easy, we had cardboard templates,
with holes punched for all the common injection points with notes on
what to expect.

Head alignment and channel balancing was a different story. For that,
we had a separate test fixture near final check. It had an audio
generator and oscilloscope, but with a good ear, the scope wasn't
really necessary. Unfortunately, this doubled the cycle time adding
an additional 4 to 10 minutes.

After you've worked on several hundred of exactly the same model tape
recorder, you see patterns of failure. For example, if some component
was inserted backwards on the first 10 or so units, it's a fair guess
that the rest had the same problem. So, pre-emptive replacement was
possible and worked well.

Fast forward 30 years and I'm now doing marine radios and data
transceivers. Over about a 10 year period, I was either involved or
watched at least 5 different product test and QA systems. The one
that worked best was visual inspection. I brought some of the
assemblers into test to do the inspection. They couldn't read a
resistor color code, but they could spot anything that had changed
quite easily. The worst was a complex series of bench tests required
by the customer, which involved a large pile of test equipment. I did
a quick time-and-motion study and found that 30% of the test time was
in setup, changing setup, and teardown. This would have been a great
candidate for computer driven ATE (automagic test equipment) except
that the PC of fashion at the time was an Apple III and RS Model II.
We tried, and failed. So, we continued to use a pile of test
equipment for test and troubleshooting. After about 3 years of this,
some of the rotary switches on the HP equipment started to fail.

Somewhat later, I had changed company and now had a product with
sufficient quantity to justify ATE. The PC of fashion was the HP 9816
with an HP-IB(IEEE488) bus. It controlled a rack full of HP RF test
equipment. I don't have time to go into detail on the setup. When it
worked, it was great and could test, align(tune), print measurements,
print reports, and live test amazingly quickly. When it didn't work,
everything came to a grinding halt. I recall having 4 techs at idle
while someone furiously drove to Fry's to buy a replacement part. I
wanted to build a manually operated backup test line but management
vetoed the idea. Getting all the equipment calibrated at the same
time was tricky, but reduced downtime. ATE would have been a marginal
failure had I not been able to "borrow" test equipment from the
engineering dept.

However, that was production test, not troubleshooting. When the
manual or ATE test line declared something worthy of troubleshooting,
it was NOT done while there were units waiting to be tested. If the
products were cheap enough, they could just be "remanufactured" which
means tear them apart down to the board level, run them back through
board test, and build a "new" unit. This worked well with radios that
had plugin cards. Eventually, there was a backlog of boards to be
troubleshot. Usually between production runs and contracts, someone
would decide that it's time to clean up the backlog. However, instead
of having the test techs do the troubleshooting, the engineer who
designed the radio would be sentence to overtime diagnosing the
boards. Yeah, that was usually me. Unless the problem was trivial, I
would only tag the likely parts, borrow someone that could do the
soldering, retest the board, and send it on to production test. This
worked well when we needed 25 identical board immediately, and was a
total waste of time for 5 or less boards that took too long to setup.

Today, much of what I see is ATE run by techs that only know how to
push the buttons. Troubleshooting is done partly with BITE (built in
test equipment). Parts replacement and soldering is outsourced to a
contractor, specialty shop, or refurbisher. In short, there's nobody
on the production line that knows which end of the soldering iron to
grab, much less how the product actually works. The rule is that
production must continue unimpeded by any form of troubleshooting or
repair. If there is any troubleshooting, it's to fix the production
line and ATE, not the actual product.

Might as well say something about bugs and fixes. With todays fairly
short product lives (typically about 6 months), it's not uncommon for
a company to have 3 or 4 generations of replacement products in
development at the same time. If something is found lacking or
defective in the current product, it is cheaper and easier to simply
wait for the next model to appear in production, and replace it with a
later version. What this means is that NOTHING gets fixed in
production, even when it's a known problem, easy fix, or major
enhancement. If you see products with known design flaws that
continue to be manufactured with the flaws, you now know why.
http://www.designnews.com/archives.asp?section_id=1367

Of course, there will still be independent repair shops (like mine)
that fix things one at a time. They're a dying breed mostly because
of the declining cost of modern electronics. As the retail price of
new hardware declines, the amount of money that a customer is willing
to pay for a repair also declines. As a rule-of-thumb, most customers
will buy a new replacment instead of repairing something when the
repair cost exceeds 50% of the replacement cost. So, if someone drags
in an out of warranty laptop that cost them $500 about 3 years ago, my
maximum bill (parts+labor) cannot exceed $250. Assuming no parts were
replaced, that 3.3 hrs at $75/hr. The number of things I can fix in
under 3.3 hrs is rapidly shrinking. For example, I have an APC XS1300
UPS on the bench that can be purchased new for about $125. In order
to replace one lousy bulging electrolytic capacitor, I've already
burned 2 hrs and I'm not done. Can I bill the customer $150 to fix a
$125 UPS? Nope.

So, the question isn't as much how to troubleshoot, but whether it's
worth you time troubleshooting in the first place.

Happy end of the tax year.



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


Sounds like you touched a lot of bases in the service life. I took a similar path ending up self employed for about the last 25 years. It has changed so much from the old days walking around with a tube caddy cleaning TV tuners. I now walk around with a laptop pinging IP addresses. How did that happen? I think the writing is on the wall. How many shoe repair and watch repair shops do you see. The electronic repair tech could meet the same fate. Just the other day I saw a flat screen monitor sitting in a garbage bin. To see the rain falling on it was too much so I took it inside. When I plugged it in everything worked and it had a VGA plus DVI input! What was that doing in the garbage. My only guess is someone wanted a wider flatter monitor so he or she tossed it out. I am typing on that monitor as we speak. How can one make a living in service with this going on. It is so bad that I have to work on 50,000 dollar LED displays boards. They will still fix those at that price. Just abou

t everything else is disposable. However I have no regrets as electronics has been an existing ride with technology changing so fast that there is never a dull moment.


There is always arcade game repair. There is enough interest in that
that most cities need a few people who do house calls.

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."