References:
Organization: D and D Data
Followup-To:

On 2012-02-23, Michael A. Terrell wrote:

"DoN. Nichols" wrote:

On 2012-02-22, Michael A. Terrell wrote:

I'm talking about the all mechanical cash registers that were the
main products of NCR. The had several large buildings in the Dayton
area that were machine shops cranking out all the parts to feed their
assembly plants. They started building them over 100 years ago.

O.K. I thought that we were still talking about the NLS digital
voltmeter readouts. :-)


Sorry, I thought 'NCR' was enough information. Of course I grew up
near there and knew some employees.

I was expecting NLS, and saw a TLA which started with 'N', so I
stopped reading. :-)

[ ... ]

We had a large DataProducts drum printer on the SATE for the PRC-77.
It would shake the entire computer booth when it printed a line of
asteriks to signal a failure. I was there one day when the engineer
from Data general was doing a core dump to the printer. it didn't have
a refolder, so the paper was sraying off the ceiling. ;-)

I can believe it.

Years later, I had to scrap a tractor trailer load of those printers,
along with a trailer full of VAX based CAD systems that had come off
lease. The owner wanted cerified destruction to keep them off the
market. It made me sick to trash the uncased NTSC studio grade
monitors. but the contrast said they could not be resold. A lot of small
TV stations would have paid $1000 each for them, for spares.

Ouch! That truly hurts. To what level did they have to be
destroyed? Individual board level in the VAXen? I can understand
trashing the software, and perhaps even the hard disk packs, but at the
generic component level, that is painful. Those could have been used to
repair many non-competing VAXen.

I wonder why they make the polarized DC versions? Perhaps
brighter? I didn't see a clue that the AC ones cost any more -- though
they *should*. :-)


Probably for the anal retentive types? Also, it would allow for
simple bicolor indicators, by simply inverting the polarity if two are
wired in parallel.

As long as they could tolerate the full reverse voltage. I
didn't look up the specs on the individual lamps.

but colored lamps would eliminate the need for
those tiny colored silicone rubber covers.

The ones which tended to tear when you tried to pull them over
the rim of the bulb sleeve, or when you pressed on them to eject a bad
bulb. :-) Who was it -- Honeywell -- who made those switch assemblies.
I know that MicroSwitch made the snap-on switch elements.


Master Specialties. I may still have a new switch or two left.

http://www.marineairsupply.com/catalogs/10_Series.pdf

Yep -- those are the later ones I remember. The earlier ones
were made by Honeywell, and had little tabs which stuck out either side
at slightly different depths, which engaged gray divider/barrier
modules, so a stack of them would couple together and snap into the
panel all at once. On those, the fronts fully removed, and a special
tool would reach in and unlatch the lamps and remove them for
replacement.

I think that I have a few of each style -- somewhere around
here. :-)

I first saw the 327/328 lamps on a tour of the VOA Bethany facility
while it was being upgraded from the original design. It was one tech's
job to wander around the facility to test & replace bad lamps. There
were over 3,000 in use, and they were still installing new equipment.

With a pushbutton on each panel to light all bulbs on that
panel? There were such on the flight simulator instructor's control
panel which a company I worked for for a while made.


They used the switch itself, with the console in diagnostics mode.
He would just push a switch, then pull it out with what looked like an
IC extractor and replace a bad lamp, and test again.

O.K. This company mounted a T-bar relay every so often on the
back of the panels to switch *all* lamps to test voltage at once.

As for the Master Specialties switch/lamp assemblies, you didn't
need an IC extractor, as long as you had fingernails on the thumb and
index finger. Just grab notches in the two ends of the clear front
plate, pull and twist.

BTW I've trimmed the accumulated "References: " header. I've had
problems in the past when it got past a certain length, though I
think that my modification to my preferred editor, jove, has
fixed that. :-)

Enjoy,
DoN.

--
Remove oil spill source from e-mail
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---

"DoN. Nichols" wrote:

Michael A. Terrell wrote:

"DoN. Nichols" wrote:

Michael A. Terrell wrote:

I'm talking about the all mechanical cash registers that were the
main products of NCR. The had several large buildings in the Dayton
area that were machine shops cranking out all the parts to feed their
assembly plants. They started building them over 100 years ago.

O.K. I thought that we were still talking about the NLS digital
voltmeter readouts. :-)


Sorry, I thought 'NCR' was enough information. Of course I grew up
near there and knew some employees.

I was expecting NLS, and saw a TLA which started with 'N', so I
stopped reading. :-)


TLAs happen! ;-)


We had a large DataProducts drum printer on the SATE for the PRC-77.
It would shake the entire computer booth when it printed a line of
asteriks to signal a failure. I was there one day when the engineer
from Data general was doing a core dump to the printer. it didn't have
a refolder, so the paper was sraying off the ceiling. ;-)

I can believe it.

Years later, I had to scrap a tractor trailer load of those printers,
along with a trailer full of VAX based CAD systems that had come off
lease. The owner wanted cerified destruction to keep them off the
market. It made me sick to trash the uncased NTSC studio grade
monitors. but the contrast said they could not be resold. A lot of small
TV stations would have paid $1000 each for them, for spares.

Ouch! That truly hurts. To what level did they have to be
destroyed? Individual board level in the VAXen? I can understand
trashing the software, and perhaps even the hard disk packs, but at the
generic component level, that is painful. Those could have been used to
repair many non-competing VAXen.


I tried to sell the VAX boards to several service companies, but they
were only interested in PDP-11 parts. They ended up at a refiner to
recover the precious metals.


Probably for the anal retentive types? Also, it would allow for
simple bicolor indicators, by simply inverting the polarity if two are
wired in parallel.

As long as they could tolerate the full reverse voltage. I
didn't look up the specs on the individual lamps.


Two LEDs, in antiparallel work OK. That's where we get Bicolor LEDS
with two leads. Now they make four lead RGB LEDs.


Master Specialties. I may still have a new switch or two left.

http://www.marineairsupply.com/catalogs/10_Series.pdf

Yep -- those are the later ones I remember. The earlier ones
were made by Honeywell, and had little tabs which stuck out either side
at slightly different depths, which engaged gray divider/barrier
modules, so a stack of them would couple together and snap into the
panel all at once. On those, the fronts fully removed, and a special
tool would reach in and unlatch the lamps and remove them for
replacement.

I think that I have a few of each style -- somewhere around
here. :-)

They used the switch itself, with the console in diagnostics mode.
He would just push a switch, then pull it out with what looked like an
IC extractor and replace a bad lamp, and test again.

O.K. This company mounted a T-bar relay every so often on the
back of the panels to switch *all* lamps to test voltage at once.

As for the Master Specialties switch/lamp assemblies, you didn't
need an IC extractor, as long as you had fingernails on the thumb and
index finger. Just grab notches in the two ends of the clear front
plate, pull and twist.


The tool helped prevent sore fingers when you were pulling them out
all day long.


BTW I've trimmed the accumulated "References: " header. I've had
problems in the past when it got past a certain length, though I
think that my modification to my preferred editor, jove, has
fixed that. :-)


--
You can't have a sense of humor, if you have no sense.

On 2012-02-24, Michael A. Terrell wrote:

"DoN. Nichols" wrote:

[ ... ]

Ouch! That truly hurts. To what level did they have to be
destroyed? Individual board level in the VAXen? I can understand
trashing the software, and perhaps even the hard disk packs, but at the
generic component level, that is painful. Those could have been used to
repair many non-competing VAXen.


I tried to sell the VAX boards to several service companies, but they
were only interested in PDP-11 parts. They ended up at a refiner to
recover the precious metals.

O.K. So pretty much destroyed. It was not late enough for
there to be a demand for VAX boards.

Probably for the anal retentive types? Also, it would allow for
simple bicolor indicators, by simply inverting the polarity if two are
wired in parallel.

As long as they could tolerate the full reverse voltage. I
didn't look up the specs on the individual lamps.


Two LEDs, in antiparallel work OK. That's where we get Bicolor LEDS
with two leads. Now they make four lead RGB LEDs.

Depends. If the current limiting resistor is outside the LED
package, yes. Each LED prevents much reverse voltage from being applied
to the other. However, if the LEDs have the resistor (or other current
limiter) packaged inside the lamp assembly, then it gets the full
reverse voltage (unless they also included either a series diode to
block reverse voltage, or a parallel diode to shunt the reverse
voltage.)

[ ... ]

As for the Master Specialties switch/lamp assemblies, you didn't
need an IC extractor, as long as you had fingernails on the thumb and
index finger. Just grab notches in the two ends of the clear front
plate, pull and twist.


The tool helped prevent sore fingers when you were pulling them out
all day long.

O.K.

Enjoy,
DoN.

--
Remove oil spill source from e-mail
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---

"DoN. Nichols" wrote:

wrote:

"DoN. Nichols" wrote:


O.K. So pretty much destroyed. It was not late enough for
there to be a demand for VAX boards.


I traded one about 60 memory boards for the PDP-11 for a half dozen
National Semiconductor VAX memory boards. I needed the 256 KB RAM.


Two LEDs, in antiparallel work OK. That's where we get Bicolor LEDS
with two leads. Now they make four lead RGB LEDs.

Depends. If the current limiting resistor is outside the LED
package, yes. Each LED prevents much reverse voltage from being applied
to the other. However, if the LEDs have the resistor (or other current
limiter) packaged inside the lamp assembly, then it gets the full
reverse voltage (unless they also included either a series diode to
block reverse voltage, or a parallel diode to shunt the reverse
voltage.)


The opposite polarity diode will keep the voltage low enough to
protect the one not in use. In some applications they feed an AC signal
to the LED for a third color.

http://www.vishay.com/leds/bi-color/

http://www.vishay.com/docs/83056/tluv5300.pdf


--
You can't have a sense of humor, if you have no sense.

On 2012-02-25, Michael A. Terrell wrote:

"DoN. Nichols" wrote:

wrote:

"DoN. Nichols" wrote:

Two LEDs, in antiparallel work OK. That's where we get Bicolor LEDS
with two leads. Now they make four lead RGB LEDs.

Depends. If the current limiting resistor is outside the LED
package, yes. Each LED prevents much reverse voltage from being applied
to the other. However, if the LEDs have the resistor (or other current
limiter) packaged inside the lamp assembly, then it gets the full
reverse voltage (unless they also included either a series diode to
block reverse voltage, or a parallel diode to shunt the reverse
voltage.)


The opposite polarity diode will keep the voltage low enough to
protect the one not in use. In some applications they feed an AC signal
to the LED for a third color.

http://www.vishay.com/leds/bi-color/

http://www.vishay.com/docs/83056/tluv5300.pdf

You're assuming two things:

1) The two LEDs are connected in parallel (with one reversed).

2) The current limiting element is external to the package.

I do not believe that (2) is true with the LEDs in the 327 lamp
package which I found earlier, and note from your PDF the following for
the bare LED:

Absolute maximum ratings:

Reverse Voltage Per Diode VsubR 6 V

So -- two of those connected as described, with individual current
limiting resistors (to operate at the 28 VDC of the 327 lamp) would
successively subject each to over four times the max Vr

And the forward current per diode is 100 mA at about 2.2 V,
which is a maximum time of about 50 uS.

As for the 327 base LEDs, I've re-found and I see no bi-color
lamps, and no discussion of maximum allowed current -- just a voltage
rating of 24 to 28 V, which says that the current limiting feature is
built into the base of the LED lamp. And for the DC only ones (either
polarity) I can't find a spec to tell me whether it will accept the same
voltage in reverse.

Enjoy,
DoN.

--
Remove oil spill source from e-mail
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---

"DoN. Nichols" wrote:

On 2012-02-25, Michael A. Terrell ? wrote:
?
? "DoN. Nichols" wrote:
??
?? wrote:
?? ?
?? ? "DoN. Nichols" wrote:

?? ? Two LEDs, in antiparallel work OK. That's where we get Bicolor LEDS
?? ? with two leads. Now they make four lead RGB LEDs.
??
?? Depends. If the current limiting resistor is outside the LED
?? package, yes. Each LED prevents much reverse voltage from being applied
?? to the other. However, if the LEDs have the resistor (or other current
?? limiter) packaged inside the lamp assembly, then it gets the full
?? reverse voltage (unless they also included either a series diode to
?? block reverse voltage, or a parallel diode to shunt the reverse
?? voltage.)
?
?
? The opposite polarity diode will keep the voltage low enough to
? protect the one not in use. In some applications they feed an AC signal
? to the LED for a third color.
?
? http://www.vishay.com/leds/bi-color/
?
? http://www.vishay.com/docs/83056/tluv5300.pdf

You're assuming two things:

1) The two LEDs are connected in parallel (with one reversed).

2) The current limiting element is external to the package.

I do not believe that (2) is true with the LEDs in the 327 lamp
package which I found earlier, and note from your PDF the following for
the bare LED:

Absolute maximum ratings:

Reverse Voltage Per Diode VsubR 6 V

So -- two of those connected as described, with individual current
limiting resistors (to operate at the 28 VDC of the 327 lamp) would
successively subject each to over four times the max Vr


You want a single resistor, not two.


And the forward current per diode is 100 mA at about 2.2 V,
which is a maximum time of about 50 uS.


That is for pulsed or multiplexed operation, like a moving LED sign.
Typical current is much lower, for a long life. The light output is
nonlinear with current. After some point the current increases faster
than the output, which causes the LED to run hot and reduce it's useful
life.


As for the 327 base LEDs, I've re-found and I see no bi-color
lamps, and no discussion of maximum allowed current -- just a voltage
rating of 24 to 28 V, which says that the current limiting feature is
built into the base of the LED lamp. And for the DC only ones (either
polarity) I can't find a spec to tell me whether it will accept the same
voltage in reverse.


You would need to wire the lamp sockets for bicolor operation, so an
external resistor per pair of LEDs is no big deal.

If there was enough space in the base you could use an 0402 SMD
resistor inside the base for a bicolor, for the 6 volt lamp. It sounds
like that lamp is intended for TTL logic drive of under 5 V and the AC
version has a higher resistance for the 6..3 typical voltage. There
could be a problem if they were fed from a higher voltage, with a
dropping resistor.

A pair of SMD 5.1 volt zenners anti-parallel and in parallel with the
LEDS would protect them. It would only ad a few cents to the
manufacturing costs, but the space is limited in a base that small. I
finally convinced the antique radio crowd to use a pair of anti-parallel
9.1V zenners across the dial lamp to protect the tapped filament in
transformerless radios. it was no big deal to lose the rectifier when
the pilot light burnt out, when they cost about 50 cents and the pilot
light was 20 cents. They cost quite a bit more today, and they refuse to
convert the radios to used white or yellow LEDs.

That style of switches really made equipment look nice though.


--
You can't have a sense of humor, if you have no sense.

On 2012-02-26, Michael A. Terrell wrote:

"DoN. Nichols" wrote:

[ ... ]

? The opposite polarity diode will keep the voltage low enough to
? protect the one not in use. In some applications they feed an AC signal
? to the LED for a third color.
?
? http://www.vishay.com/leds/bi-color/
?
? http://www.vishay.com/docs/83056/tluv5300.pdf

You're assuming two things:

1) The two LEDs are connected in parallel (with one reversed).

2) The current limiting element is external to the package.

I do not believe that (2) is true with the LEDs in the 327 lamp
package which I found earlier, and note from your PDF the following for
the bare LED:

Absolute maximum ratings:

Reverse Voltage Per Diode VsubR 6 V

So -- two of those connected as described, with individual current
limiting resistors (to operate at the 28 VDC of the 327 lamp) would
successively subject each to over four times the max Vr


You want a single resistor, not two.

But -- the limiting resistor (or perhaps a FET wired as a
constant current source) is *already* present inside the 327 format LED,
since it is designed to run from a 27V supply. This is what I have been
trying to point out this whole time.

If you got some designed with no limiting resistor -- that would
be different -- but would be asking for trouble in a system with a lot
of 28V rated 327 style LED assemblies, and a few of the color reversing
unlimited ones. What are the chances of someone swapping them without
knowing the difference?

[ ... ]

You would need to wire the lamp sockets for bicolor operation, so an
external resistor per pair of LEDs is no big deal.

*If* the LEDs did not already have built-in limiting resistors
or alternate circuits.

If there was enough space in the base you could use an 0402 SMD
resistor inside the base for a bicolor, for the 6 volt lamp. It sounds
like that lamp is intended for TTL logic drive of under 5 V and the AC
version has a higher resistance for the 6..3 typical voltage. There
could be a problem if they were fed from a higher voltage, with a
dropping resistor.

You are assuming two colors of LED in a single 327 style lamp
base? But this maker does not offer such, so we are stuck u sing the
ones which they already make -- or contracting for a special run.

Enjoy,
DoN.

--
Remove oil spill source from e-mail
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---

"DoN. Nichols" wrote:

On 2012-02-26, Michael A. Terrell ? wrote:
?
? "DoN. Nichols" wrote:

[ ... ]

?? ? The opposite polarity diode will keep the voltage low enough to
?? ? protect the one not in use. In some applications they feed an AC signal
?? ? to the LED for a third color.
?? ?
?? ? http://www.vishay.com/leds/bi-color/
?? ?
?? ? http://www.vishay.com/docs/83056/tluv5300.pdf
??
?? You're assuming two things:
??
?? 1) The two LEDs are connected in parallel (with one reversed).
??
?? 2) The current limiting element is external to the package.
??
?? I do not believe that (2) is true with the LEDs in the 327 lamp
?? package which I found earlier, and note from your PDF the following for
?? the bare LED:
??
?? Absolute maximum ratings:
??
?? Reverse Voltage Per Diode VsubR 6 V
??
?? So -- two of those connected as described, with individual current
?? limiting resistors (to operate at the 28 VDC of the 327 lamp) would
?? successively subject each to over four times the max Vr
?
?
? You want a single resistor, not two.

But -- the limiting resistor (or perhaps a FET wired as a
constant current source) is *already* present inside the 327 format LED,
since it is designed to run from a 27V supply. This is what I have been
trying to point out this whole time.

If you got some designed with no limiting resistor -- that would
be different -- but would be asking for trouble in a system with a lot
of 28V rated 327 style LED assemblies, and a few of the color reversing
unlimited ones. What are the chances of someone swapping them without
knowing the difference?

[ ... ]

? You would need to wire the lamp sockets for bicolor operation, so an
? external resistor per pair of LEDs is no big deal.

*If* the LEDs did not already have built-in limiting resistors
or alternate circuits.

? If there was enough space in the base you could use an 0402 SMD
? resistor inside the base for a bicolor, for the 6 volt lamp. It sounds
? like that lamp is intended for TTL logic drive of under 5 V and the AC
? version has a higher resistance for the 6..3 typical voltage. There
? could be a problem if they were fed from a higher voltage, with a
? dropping resistor.

You are assuming two colors of LED in a single 327 style lamp
base? But this maker does not offer such, so we are stuck u sing the
ones which they already make -- or contracting for a special run.


Then you would fall back on the old method of using lamps of separate
colors, on separate circuits. Or add current steering diodes to protect
the LEDS. to use the more modern bicolor methods. With the price of the
switches & lamps, there isn't a lot of call for either these days. They
even make pushbuttons with built in displays that can be programmed with
whatever message you want on the buttons.


--
You can't have a sense of humor, if you have no sense.