14 Jan
2015
14 Jan
'15
1:34 p.m.
Yep, I'm still working on "restoration" of the Microdata Reality.
I say "restoration", because due to another restoration project that
actually has a deadline, I only have time to clean up the microdata and
repair just obvious visual brokenness, not actually start testing/repairing
it electronically. That will come after the "deadline project" J
As most know, schematics and detailed documentation are nonexistent for this
machine.
I've done the rack, the disc drive, the tape drive, and the power supply. On
to the last item, the front panel.
Virtually all of the LED's on the front panel board had so much mouse pee
that the leads on the LEDS actually corroded through, so I need to replace
them all.
Without schematics, can anyone suggest how I might go about determining a
suitable replacement LED from an electrical perspective?
Pictures at www.ezwind.net/microdata/restoration
Best,
J
14 Jan
14 Jan
1:42 p.m.
Virtually all of the LED's on the front panel board had so much mouse pee
that the leads on the LEDS actually corroded through, so I need to replace
them all.
Without schematics, can anyone suggest how I might go about determining a
suitable replacement LED from an electrical perspective?
Long answer :
Trace out the circuit for one LED and its driver.
Most of the time such LEDs have a series resistor. The +ve side of the combination goes
to a supply line, the -ve side to the collector of a driver transistor, the emitter of
which is
grounded (this may be an open-collector gate like a 7406 or 7407). From that you can
estimate the voltage across the LED+resistor (= supply voltage - saturation voltage
of the driver transistor (say 0.5V)). Then the current through the thing is
(that voltage - forward voltage of LED)/series resistor. Substitute the forward voltage
of
some possible replacement LEDs and see if the current is sensible (compare to the
maximum current on that LED's data sheet).
Short answer :
Go and get some standard (not high brightness, high efficiency, or anything fancy
like that) red 0.2" (5mm, whatever) LEDs. Solder them in. I will bet it will work
fine.
-tony
1:49 p.m.
Without schematics, can anyone suggest how I might go about determining a
suitable replacement LED from an electrical perspective?
Or, if you can find even one working LED on the original panel, remove it and
connected it (correct way round) with a 330 Ohm or 470 Ohm (or so) resistor
in series to an adjustable bench supply. Turn up the supply until the LED is
glowing at what you consider to be normal brightness (this will be 5-10V I would
guess). Now measure the voltage across the LED itself and the current through
it (or deduce the latter from the voltage across the series resistor and the known
value of that resistor). Try to find an LED that matches those values for forward
voltage and operating current.
Or as I said, just stick in standard red LEDs. I've replaced LEDs many times without
thinking what I am fitting.
-tony
2:03 p.m.
Tony wrote...
Or, if you can find even one working LED on the original panel, remove it
and connected {snip}
That is probably the way I'll go. Thanks (as always) Tony!!!
J
3:24 p.m.
Jay,
there are no issues with the LED's that I know of other than they do
have a load resistor in series with them, so they need no internal
dropping resistor to work across +5.
they have long leads as well, as there is a 1" (or there bouts)
stiffening board to keep the leads from shorting out, and to position
them in a line to poke thru the front panel. There are other solutions
to allow for long leads, but they just used a long circuit board.
Jim
On 1/14/2015 2:03 PM, Jay West wrote:
Tony wrote...
Or, if you can find even one working LED on the original panel, remove it
and connected {snip}
That is probably the way I'll go. Thanks (as always) Tony!!!
J
3:41 p.m.
One thing to be careful about is that modern LEDs are *much* more
effecient than old LEDs.
I recently replaced a LED on my IMSAI front panel. The new LED stuck
out like a sore thumb, it was insanely bright.
From memory, I replaced the resistor with about a 1k2.
Doug
On 15/01/2015 10:24 AM, jwsmobile wrote:
Jay,
there are no issues with the LED's that I know of other than they do
have a load resistor in series with them, so they need no internal
dropping resistor to work across +5.
they have long leads as well, as there is a 1" (or there bouts)
stiffening board to keep the leads from shorting out, and to position
them in a line to poke thru the front panel. There are other
solutions to allow for long leads, but they just used a long circuit
board.
Jim
On 1/14/2015 2:03 PM, Jay West wrote:
Tony wrote...
Or, if you can find even one working LED on the original panel,
remove it
and connected {snip}
That is probably the way I'll go. Thanks (as always) Tony!!!
J
4:18 p.m.
JS wrote....
----
they have long leads as well, as there is a 1" (or there bouts) stiffening
board to keep the leads from shorting out, and to position them in a line to
poke thru the front panel. There are other solutions to allow for long
leads, but they just used a long circuit board.
----
Yep. A picture of what you describe (stiffening board and long leads):
http://www.ezwind.net/microdata/restoration/frontpanel/IMG_0299.JPG
J
15 Jan
15 Jan
8:59 a.m.
On Wed, 14 Jan 2015, Jay West wrote:
JS wrote....
----
they have long leads as well, as there is a 1" (or there bouts) stiffening
board to keep the leads from shorting out, and to position them in a line to
poke thru the front panel. There are other solutions to allow for long
leads, but they just used a long circuit board.
----
Yep. A picture of what you describe (stiffening board and long leads):
http://www.ezwind.net/microdata/restoration/frontpanel/IMG_0299.JPG
I had wondered if you had finished the Microdata... I can't remember for
certain now if I ever followed up, but I did check my stash of old
Allen-Bradley resistor networks and I couldn't find any that were the same
value and bus configuration as the ones the mouse pee ruined on your
board.
14 Jan
14 Jan
2:25 p.m.
-----Original Message-----
From: cctalk [mailto:cctalk-bounces at classiccmp.org] On Behalf Of tony
duell
Sent: 14 January 2015 21:50
To: General Discussion: On-Topic and Off-Topic Posts
Subject: RE: front panel LED help sought
and
Without schematics, can anyone suggest how I might go about
determining a suitable replacement LED from an electrical perspective?
Or, if you can find even one working LED on the original panel, remove it connected it (correct way round) with a 330 Ohm or 470
Ohm (or so)
resistor in
series to an adjustable bench supply. Turn up the
supply until the LED is
glowing
at what you consider to be normal brightness (this
will be 5-10V I would
guess).
Now measure the voltage across the LED itself and the
current through it
(or
deduce the latter from the voltage across the series
resistor and the
known
value of that resistor). Try to find an LED that
matches those values for
forward
voltage and operating current.
Or as I said, just stick in standard red LEDs. I've replaced LEDs many
times
without thinking what I am fitting.
"without thinking" is that an impostor pretending to be Tony? :-) :-)
Regards
Rob
10:08 p.m.
Or as I said, just stick in standard red LEDs.
I've replaced LEDs many
times
without thinking what I am fitting.
"without thinking" is that an impostor pretending to be Tony? :-) :-)
3582
days inactive
3583
days old
9 comments
6 participants
participants (6)
-
ard＠p850ug1.demon.co.uk -
doug＠doughq.com -
jwest＠classiccmp.org -
jws＠jwsss.com -
robert.jarratt＠ntlworld.com -
tothwolf＠concentric.net