Your Toughest Job

[Jeff Urban]

The ones that separate the Men from the boys. For me it was an RCA CTC169.

The symptom was display shifted on the screen and no sound. I don't remember or sure which but IIRC shifted to the right was a known fault, but this was shifted to the left. And it was intermittent.

It would never act up when I had the chassis stood up on edge for testing. Three people with good eyes already combed the board for bad connections and even resoldered everything, no good.

I am not even that much a digital guy, but...

This fault was caused by the 503KHz crystal for the sweep circuit. Check out this failure mode;

When the unit is turned on the uProcessor had always been running to receive the remote command. However the EEPROM ran off a sweep derived source. There was only a small window of time for the EEPROM to load its data into the uProcessor and the 503KHZ crystal had not failed completely but enough to make the oscillator slow to start - sometimes. I other words the data missed the bus. (I think that is pretty damn good metaphor...)

It was not easy but it was a contract job. you cannot just give back the money,it would be like if your house burnt down and the insurance company say "Well here's your premiums back". It doesn't work that way. It would not have been a nosebleed for the company, more like a decapitation.

Literally, "Do not leave until this is done, we will even get you beer at midnight just DO NOT GIVE UP, DO IT".

So who else has had "one of those" ? And this was not lucking out on it, I had to know much more about the theory of operation than was included in even the training manual. I actually though maybe they were going to try to take my car keys ! It was that critical to them. But if I decide to work for you I have loyalty. There are only a few things I won't do, like sabotage or wholesale ripping people off. I fired a guy from my shop for sabotaging a unit. Guy walked in and told him what he wanted and the guy did it. We do not do that and he was out the door immediately. He became a bricklayer...

So, anyone have anything like that " I got others, I bet Allison also does, maybe a few others. I would like to hear about them. you are being a desk potato (or in Quaylese potatoe) on the internet, what else you gonna do ?

I am not a democrat or a liberal but Dan Quayle was not an exemplary...whatever. Anything really. Should not even be trusted to teach grade school spelling and VP of the country. Man Bush 2 was really fucking stupid, couldn't find oil in Texas. Are you kidding me ?

Anyway, back to the topic. Your hardest job, brainwise.

 

[Phil Allison]

Jeff Urban wrote:

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

In other words the data missed the bus.

(I think that is pretty damn good metaphor...)

** ROTFL - so do I.

So, anyone have anything like that " I got others, I bet Allison also
does, maybe a few others. I would like to hear about them.

The Yamaha R1000 digital reverb unit was a nasty case.

The complaint was that it made a loud noise, like white noise, after warming up for some time. Soon verified with hot air blown onto the main PCB but not able to be localised.

Found that a soldering iron tip placed on the middle of the 40 pin processor reliably induced the fault - so ordered a new one and fitted it with a mating socket. Made not the tiniest difference.

A call to the importers struck oil when the tech went through the service bulletins in his filing cabinet. One of them was headed " Big Noise ".

The fix was to change 6 memory buffers to high speed CMOS.

All R1000s in the initial batch had the same issue, loss of speed when running hot caused memory data clashes and random data produces white noise when decoded to analogue.


...... Phil

 

[John Robertson]

On 2020/02/12 10:30 p.m., Phil Allison wrote:

Jeff Urban wrote:

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

In other words the data missed the bus.

(I think that is pretty damn good metaphor...)


** ROTFL - so do I.


So, anyone have anything like that " I got others, I bet Allison also
does, maybe a few others. I would like to hear about them.


The Yamaha R1000 digital reverb unit was a nasty case.

The complaint was that it made a loud noise, like white noise, after warming up for some time. Soon verified with hot air blown onto the main PCB but not able to be localised.

Found that a soldering iron tip placed on the middle of the 40 pin processor reliably induced the fault - so ordered a new one and fitted it with a mating socket. Made not the tiniest difference.

A call to the importers struck oil when the tech went through the service bulletins in his filing cabinet. One of them was headed " Big Noise ".

The fix was to change 6 memory buffers to high speed CMOS.

All R1000s in the initial batch had the same issue, loss of speed when running hot caused memory data clashes and random data produces white noise when decoded to analogue.


...... Phil

One of my more memorable fixes (we all have some of those) was trying to
sort out why a proven board design would not reliably read an opto
switch for a game that had had 1,000s sold. This was one of the first
optos used in pinball games on a game called Space Station and I knew
that the original factory computer was perfectly happy with the opto,
but a replacement MPU (exactly the same schematic and layout as the
original) was not. The opto would read erratically in self test -
essentially it flickered. DV supply was clean, MPU clock was good, the
new MPU board used what appeared to be good parts. Poked at it for a
little while with a 'scope and all the levels looked good.

After a while I started looking at the chips this aftermarket
manufacturer used for IO. The original factory TTL was LS, and it turned
out this designer figured that HCT was the way to go to replace it. All
well and good, except that they inserted HC instead of HCT for the IO
chips...

I wrote them a note suggesting they not do that any more.

Today I spotted another of this same company's boards in for service
(different generation MPU) and they are STILL mixing HC and HCT, with
this board made about ten years after the previous one...

Sigh. The board looks good, but the complaint is that it has some
erratic problems from time to time... I wonder which chips I'll change
first?

John :-#)#

--
(Please post followups or tech inquiries to the USENET newsgroup)
John's Jukes Ltd.
MOVED to #7 - 3979 Marine Way, Burnaby, BC, Canada V5J 5E3
(604)872-5757 (Pinballs, Jukes, Video Games)
www.flippers.com
"Old pinballers never die, they just flip out."

 

[Fox's Mercantile]

On 2/12/20 6:27 PM, Jeff Urban wrote:
> Anyway, back to the topic. Your hardest job, brainwise.

Being one the guys that had to make shit work in the Aerospace
industry was always full of fun and excitement.

Emitter Coupled Logic was the "New Big Thing(tm)" at the time.
We had a project that required a huge amount of data, and it
had to be correct.

It wasn't.

Worse than being random, it, after analysis, appeared to be
pattern sensitive.

We were using a Textronix DAS9600 as a logic analyzer. Not
fast enough. Call Textronix. They send a "Super duper" front
end snapshot capture. Oooh, this is fucking cool! It has a
sample rate of 5 pico seconds. Our customer says, "Buy it!"

I fill out a purchase order. Submit it. And wait for the
confirmation from Textronix. Two days later, I have it, and
a delivery date of six months. I notify our customer. They
said, "I'll fix that."

The next day, I'm at my desk and I get a call from Tektronix.
It's the CEO. "Tell your customer, you'll have that in six
weeks."

Of course, I had to ask him how that happened. "The President
called me last night and told me to make it happen."

Oh my.

Well, true to his word, it showed up in six weeks. Now the
fun begins, trying to find the problem. I was proud of myself.
I found it within a week. "Good tools = good work."

There was a problem with the memory from Fairchild. And it
WAS pattern sensitive. A call to them got us all new memory.

There was more to it than that, but that's all you get from
me.


--
"I am a river to my people."
Jeff-1.0
WA6FWi
http:foxsmercantile.com

 

[Michael Terrell]

On Thursday, February 13, 2020 at 1:30:46 AM UTC-5, Phil Allison wrote:

Jeff Urban wrote:

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

In other words the data missed the bus.

(I think that is pretty damn good metaphor...)


** ROTFL - so do I.


So, anyone have anything like that " I got others, I bet Allison also
does, maybe a few others. I would like to hear about them.


The Yamaha R1000 digital reverb unit was a nasty case.

The complaint was that it made a loud noise, like white noise, after warming up for some time. Soon verified with hot air blown onto the main PCB but not able to be localised.

Found that a soldering iron tip placed on the middle of the 40 pin processor reliably induced the fault - so ordered a new one and fitted it with a mating socket. Made not the tiniest difference.

A call to the importers struck oil when the tech went through the service bulletins in his filing cabinet. One of them was headed " Big Noise ".

The fix was to change 6 memory buffers to high speed CMOS.

All R1000s in the initial batch had the same issue, loss of speed when running hot caused memory data clashes and random data produces white noise when decoded to analogue.

Occasional batches of embedded controllers for a Telemetry receiver, where the MPU oscillatior wouldn't start and run. Asked the long time employees, it had been a problem since it was first released to production. Their solution was to keep changing the PU and crystal until it worked. I looked at the schematic and spotted the problem. They had designed it for a PMOS IC, which went obsolete before production started. So they went into production with the CMOS version, but they didn't change the capacitors. The original used a pair of 20 pF, the CMOS needed 200 pF for that crystal. Engineering insisted that I was wrong, because it had been in production for ten years. They refused to issue the ECO, until a large customer wanted to know where their equipment was. I told the Corporate VP that they had been sitting in shipping for a month, waiting for that ECO so they could be shipped. He turned pale. He had started as a production test tech, and he was familiar with the problem. "I'll be back in five minutes, with your ECO."

So much for Design Engineering's policy of no ECOs being written from the Production floor, or on designs that were over two years old.

 

[amdx]

I went into a consumer electronics repair shop to apply for repair
position, I had a couple dozen certificates from Manufacturer training
classes with me. The owner told me he just just hired a guy or he would
have hired me. I wasn't home an hour and the guy called me back and
said, I noticed several of your certificates were from Sony. I have this
Sony projector TV that has been a dog, if you want to take a look at it,
I will pay you.
So I drove the 1/2 mile from my house and got the manual out. The
symptom is no light out of any of the three tubes. There was no high
voltage, so I'm poking around with the scope and the owner says "Hey,
you got it going". I didn't know it had lit up, being on the backside, I
didn't see it, but I didn't say that.
It turned out a previous tech had changed two caps in that circuit, but
he replaced them with a lower value. Just the capacitance of the scope
probe was enough to bring the circuit to life. Probe connected it works,
disconnecting the probe it didn't work.
He hired me, even though it was part time, I already had a full time job.

Mikek

 

[Randy Day]

In article <b4854b08-3ce9-43a5-8c5e-44a1d8e867d3@googlegroups.com>,
jurb6006@gmail.com says...

[snip]

> Anyway, back to the topic. Your hardest job, brainwise.

Big Name retailer, mall anchor store, large company
customer. Symptom: Jewelry department POS terminal (full
UPS, on 24/7) keeps shutting down, staff have to wait 20
minutes each morning for restart.

Went on for weeks/months, different techs had tried
various components, no change, lots of hair pulling,
no one is happy.

So I get the service call one afternoon, working away.
Replaced the parts suggested in the ticket, unit runs
fine. As I finish up, the store is closing, staff are
turning out the lights. All of a sudden the UPS
starts beeping - WTF?

Turns out, that one terminal was wired to a breaker
that got turned off every night - the poor UPS kept it
going for a few hours after closing, so no one saw a
problem...

 

[Michael Terrell]

On Wednesday, February 12, 2020 at 7:27:04 PM UTC-5, Jeff Urban wrote:

The ones that separate the Men from the boys. For me it was an RCA CTC169..

The symptom was display shifted on the screen and no sound. I don't remember or sure which but IIRC shifted to the right was a known fault, but this was shifted to the left. And it was intermittent.

It would never act up when I had the chassis stood up on edge for testing.. Three people with good eyes already combed the board for bad connections and even resoldered everything, no good.

I am not even that much a digital guy, but...

This fault was caused by the 503KHz crystal for the sweep circuit. Check out this failure mode;

When the unit is turned on the uProcessor had always been running to receive the remote command. However the EEPROM ran off a sweep derived source. There was only a small window of time for the EEPROM to load its data into the uProcessor and the 503KHZ crystal had not failed completely but enough to make the oscillator slow to start - sometimes. I other words the data missed the bus. (I think that is pretty damn good metaphor...)

It was not easy but it was a contract job. you cannot just give back the money,it would be like if your house burnt down and the insurance company say "Well here's your premiums back". It doesn't work that way. It would not have been a nosebleed for the company, more like a decapitation.

Literally, "Do not leave until this is done, we will even get you beer at midnight just DO NOT GIVE UP, DO IT".

So who else has had "one of those" ? And this was not lucking out on it, I had to know much more about the theory of operation than was included in even the training manual. I actually though maybe they were going to try to take my car keys ! It was that critical to them. But if I decide to work for you I have loyalty. There are only a few things I won't do, like sabotage or wholesale ripping people off. I fired a guy from my shop for sabotaging a unit. Guy walked in and told him what he wanted and the guy did it. We do not do that and he was out the door immediately. He became a bricklayer....

So, anyone have anything like that " I got others, I bet Allison also does, maybe a few others. I would like to hear about them. you are being a desk potato (or in Quaylese potatoe) on the internet, what else you gonna do ?

I am not a democrat or a liberal but Dan Quayle was not an exemplary...whatever. Anything really. Should not even be trusted to teach grade school spelling and VP of the country. Man Bush 2 was really fucking stupid, couldn't find oil in Texas. Are you kidding me ?

Anyway, back to the topic. Your hardest job, brainwise.

Metrodata NTSC, Motorola Exorcisor Bus based six channel character Generator. at a CATV headend. It locked up for about 10 minutes. Just long enough not to get a scope there and the covers off the rack mounted chassis. It worked about six months, and did it again, and again at shorter intervals until it locked up and stayed that way. There were no spare boards for the pair of systems, and they were essential to the system's operation.

One system had a SMS disk storage subsytem with a pair of 8" floppies, but the failing system only had 32KB of RAM on a MC6800 8bit processor. No operating system, and information was stored as 'pages' so the addressing was selected by a block of RAM. The two systems and the dis system had cost us a little over $60,000, in 1980. Then the company that built the two systems promptly went bankrupt and shut down.

There I was, standing on a ladder, leaning into the chassis with a logic probe. One of the address lines on the Exorcisor bus was shorted to ground.. One of the address decoders had failed. a 29 cent TTL IC that took over two years to finally die.

 

[Phil Allison]

Jeff Urban wrote:

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

Anyway, back to the topic. Your hardest job, brainwise.

** The Gestetner machine:

Back in the early 80s, I took a job in a city office that was set up as a small design and repair business.

The boss designed and protoyped high end audio and carried out elaborate modifications to commercial, domestic audio gear.

His name was "Allen Wright" and as I was soon to discover, I had walked into a den of Scientologists. Little did I know I was the ONLY non Scientologist or "wog' as they like to call such people.

I did the real work, fixing domestic hi-fi and guitar (mostly tube) amplifiers all day. Allan did whatever he did.

Various bods would pay Allan a visit, some of them were high ranking Scientologist from their "Church" up the road a bit. A real trap for the terminally gullible.

Anyhow, one day a large ugly machine arrived on the spare work bench. I remember the name was "Gestetner" and it was a stencil cutter.

It refused to work and had been rejected by the local service agent as being beyond repair - more on that later.

Such machines are partly electronic with a high voltage/ high frequency supply creating an arc discharge that cut the special paper stencils so multiple copies could be made of a typed document. The damn thing had suddenly died and would not cut.

Allan spent hours on it, day and night cos it was for his "Church" and his reputation was all wrapped up in the job.

I overhead a conversation between the Grand Poo Bah and him about the agents refusal to service the beast - both reckoned it was because the manager HATED Scientologists. And why not?

Having a few minutes up my sleeve one afternoon and with Allan out of the premises I took a look at the contraption. There was a pile of stencil paper near it and I fingered through intending to take a sheet.

I noticed that not all looked alike - so grabbed my analogue multi-meter and probed the carbonised surface. Some of them conducted nicely and others were insulating. Faulty paper.

When the boss returned I showed him what I found and astonished he set it up with a good sheet and off the dam thing went- just perfect.

Soon the Grand Poo Bah arrived to see this modern miracle and Allan was beaming. He explained what had happened and how the problem was bad paper all along. Although sitting 5 yards away, my name was never mentioned.

Just a tad grumpy about Allan taking full credit for my discovery, I later approached him with a request for some financial compensation - as I had saved the "Church" a great deal of money in not having to purchase a modern photo copier.

I was told to write a bill for my time.

I complained is was less than 30 minutes.

"Then a bill for 30 minutes is what you get" - he replied.



...... Phil

 

[John-Del]

On Wednesday, February 12, 2020 at 7:27:04 PM UTC-5, Jeff Urban wrote:

> Anyway, back to the topic. Your hardest job, brainwise.

I wish I could recall most some of them, but I think I buried them in my subconscious where they would do no harm..

One that pops in my head was an old 60" Hitachi analogue projection TV, the ones that weighed about 400lbs.

At first, it seemed simple: blown horizontal output. This unit used separate horiz output and high voltage output transistors. After replacing the output, it fired right up, although the raster appeared misconverged more than it should, but the new transistor was running cool at least.

I connected the cable to the back of the TV to get a better look at the raster and there was a spark at the cable fitting and the TV went dead. New transistor blown. There was a leakage between the AC input and the cold ground where the cable was. I started checking those safety caps that bridged hot to cold ground, then pulled the SMPS transformer to see if it was shorted primary to secondary, but nothing was found. It was in a basement and I didn't want to haul that beast up and out to the shop, so I went back for the service manual to fight it in the basement.

To make a reaaaaaaaaaaaly long story short, it turns out it was a bad deflection yoke on one of the three projection CRTs causing the ground fault. Yep, the vertical circuit was on the cold side of the chassis and the horiz deflection and HV was on the hot side. There was arcing across the gap between the vertical and horizontal output windings of one of the yokes.

 

[Arie de Muynck]

In this case the toughness was more in logistics than electronics...

End of 1977 I designed all electronics (console and motor driver box)
for the 6 mirrors above the stage of the Genesis "...And then there were
three..." world tour. It controlled 12 big 24VDC motors with
potentiometer feedback. Final drivers were triacs, phase control, from a
2kW 34Vac transformer output. The driver box was built in a 19" rack in
a flight case, must have weighed 60 kg. Since I was the only one that
could service it, the road manager had my private number.

Apart from phone calls in the middle of the night like "Come here
immediately" - "Where is here" - "Dallas, Texas!" (I live in the
Netherlands) the system was remarkably robust. A 5 minute conversation
with the local technician usually solved the problem.

The toughest problem was a call "We are in Paris, the system is
completely mad, everything is oscillating". I got on a plane, paid a
cabdriver extra to drive through the meadows between the cows to get
from the blocked highway to a secondary road, and made it to the Palais
des Sports an hour before the show. They would not let me touch it, too
close to the show starting. They played with fixed mirror positions.

Next morning, with the tech, I checked the system. Indeed the motors
moved too fast, would not position stable, stayed oscillating, one was
dead. Time to start measuring. Hmm, peak motor voltages are high? DC
controller rails voltages OK. The big transformer is humming a bit, and
OUCH when I touch it very hot. WTF? Measure secondary - 50% too high.
Measure primary - 70% too high. OK, turn off immediately. Maybe they
connected the 220V single phase input to 380V? Follow the mains cable to
under the stage to a big humming and very hot box. The tech first goes
pale then red - it is the 3kW 115V to 220V autotransformer they had used
in the USA part of the tour...
Remove that box, replace one fuse, and the show can go on again. All the
electronics had survived.

Arie de Muijnck

 

[amdx]

O

 

[Jon Elson]

On Wed, 12 Feb 2020 16:27:01 -0800, Jeff Urban wrote:

> The ones that separate the Men from the boys.
OK, one crazy one was a Datum mag tape controller on a PDP-11 computer.
These had one big standard board and a smaller plug-in board to customize
it to the specific computer. The logic was just a whole bunch of TTL
random logic, no microcode. It was totally dead, all registers read back
as zero when you wrote to them. So, I started looking at the global
reset. It was not being reset by the CPU, but the buffered side on the
board was held in reset (low). I changed a few chips, and that didn't
fix it. So, I then measured resistance to ground with power off, and it
was REALLY low, like 2 Ohms. I started cutting traces, and finally got
down to a 2" stretch of PCB trace that was shorted to ground. There was
NO grounded trace within 1/4" of that trace that was at ground. I peeled
the trace off the board, there was no speck of copper visible anywhere.
I finally put a piece of wire wrap wire across where I'd peeled off the
trace, being sure the insulation was intact. The unit worked fine until
the end of its life.

That was one of the stranger ones.

Jon

 

[John Robertson]

On 2020/02/19 12:15 p.m., Jon Elson wrote:

On Wed, 12 Feb 2020 16:27:01 -0800, Jeff Urban wrote:

The ones that separate the Men from the boys.
OK, one crazy one was a Datum mag tape controller on a PDP-11 computer.
These had one big standard board and a smaller plug-in board to customize
it to the specific computer. The logic was just a whole bunch of TTL
random logic, no microcode. It was totally dead, all registers read back
as zero when you wrote to them. So, I started looking at the global
reset. It was not being reset by the CPU, but the buffered side on the
board was held in reset (low). I changed a few chips, and that didn't
fix it. So, I then measured resistance to ground with power off, and it
was REALLY low, like 2 Ohms. I started cutting traces, and finally got
down to a 2" stretch of PCB trace that was shorted to ground. There was
NO grounded trace within 1/4" of that trace that was at ground. I peeled
the trace off the board, there was no speck of copper visible anywhere.
I finally put a piece of wire wrap wire across where I'd peeled off the
trace, being sure the insulation was intact. The unit worked fine until
the end of its life.

That was one of the stranger ones.

Jon

The ShortSqueek by Global Specialties is handy for finding shorts on
circuit boards...

https://flippers.com/pdfs/GlobalSpecialities_ShortSqueek_Model_SQ-1.pdf

It gets a bit of a workout here in my shop every now and then when we
have a shorted trace.

John ;-#)#

 

[Chuck]

On Wed, 12 Feb 2020 16:27:01 -0800 (PST), Jeff Urban
<jurb6006@gmail.com> wrote:

The ones that separate the Men from the boys. For me it was an RCA CTC169.

The symptom was display shifted on the screen and no sound. I don't remember or sure which but IIRC shifted to the right was a known fault, but this was shifted to the left. And it was intermittent.

It would never act up when I had the chassis stood up on edge for testing. Three people with good eyes already combed the board for bad connections and even resoldered everything, no good.

I am not even that much a digital guy, but...

This fault was caused by the 503KHz crystal for the sweep circuit. Check out this failure mode;

When the unit is turned on the uProcessor had always been running to receive the remote command. However the EEPROM ran off a sweep derived source. There was only a small window of time for the EEPROM to load its data into the uProcessor and the 503KHZ crystal had not failed completely but enough to make the oscillator slow to start - sometimes. I other words the data missed the bus. (I think that is pretty damn good metaphor...)

It was not easy but it was a contract job. you cannot just give back the money,it would be like if your house burnt down and the insurance company say "Well here's your premiums back". It doesn't work that way. It would not have been a nosebleed for the company, more like a decapitation.

Literally, "Do not leave until this is done, we will even get you beer at midnight just DO NOT GIVE UP, DO IT".

So who else has had "one of those" ? And this was not lucking out on it, I had to know much more about the theory of operation than was included in even the training manual. I actually though maybe they were going to try to take my car keys ! It was that critical to them. But if I decide to work for you I have loyalty. There are only a few things I won't do, like sabotage or wholesale ripping people off. I fired a guy from my shop for sabotaging a unit. Guy walked in and told him what he wanted and the guy did it. We do not do that and he was out the door immediately. He became a bricklayer...

So, anyone have anything like that " I got others, I bet Allison also does, maybe a few others. I would like to hear about them. you are being a desk potato (or in Quaylese potatoe) on the internet, what else you gonna do ?

I am not a democrat or a liberal but Dan Quayle was not an exemplary...whatever. Anything really. Should not even be trusted to teach grade school spelling and VP of the country. Man Bush 2 was really fucking stupid, couldn't find oil in Texas. Are you kidding me ?

Anyway, back to the topic. Your hardest job, brainwise.

In the early 80s,our shop had an19" Hitachi tv that had a ghosting
problem. I spent hours in the video circuit but found no defective
components. I finally hooked the chassis to a crt on another Hitachi
tv and the ghosting was gone. Replacing the crt repaired the set. At
the time we were the largest Hitachi dealer in the U.S. but only saw
the problem two other times. All were 19" Hitachis.