22 Jul
2016
22 Jul
'16
12:53 a.m.
Is there anyone with documents on the Nova 3 front panel, and what
drives it? It has some number of custom DG chips, which hopefully are
good if I want to try to fire it up to play, but am interested in that
on good authority there are 28v incandescent lamps.
A friend has an Eclipse front panel with nearly identical bezel, which
has LED's and a number of differences in the logic (different connector
to the system, for instance). So it is probably all run on 5v.
I have not had time to figure out the driver circuit for any of the
lamps to see what that may turn up, and wanted to know whether it was
28v lamps before I buy 40 of them. (the thing has only 2 out of a lot
of lamps).
Thanks
Jim
22 Jul
22 Jul
3:25 a.m.
From: jim stephens: Thursday, July 21, 2016 9:53 PM
I have not had time to figure out the driver circuit
for any of the
lamps to see what that may turn up, and wanted to know whether it was
28v lamps before I buy 40 of them. (the thing has only 2 out of a lot
of lamps).
I mentioned http://www.foxdata.com/blog/tag/nova-312/ to you
earlier, which suggested the 28V bulbs. I also found
http://www.chookfest.net/nova3/ledmod.html
which makes it clear that the voltage is likely a fair bit over 12V.
The DG Components guide only lists part numbers for the
28V bulbs, 24V bulbs, 5V, and 6V bulbs (at least that I found
that fit with the wire base and the form factor). I think installing
28V bulbs in 24V circuits would be fine (they'd last longer).
Alas, I have no authoritative information on the exact correct
bulb.
Vince
3:46 a.m.
On 7/22/2016 12:25 AM, Vincent Slyngstad wrote:
From: jim stephens: Thursday, July 21, 2016 9:53 PM
That's not a problem. I was also interested in the info if I ever get
around to using it for a simh front panel blink'n version. Unless a Nova
3 drops in my lap. (and still would want info).
Thanks for the tips. I see that the foxdata site lamps don't appear
quite as bright as the chookfest ones, which is interesting. I'm
thinking now if I do anything it will be the LED route, so will evaluate
that, as I agree opening the thing back up will suck.
Jay did your DG stuff on the back dock move? just curious. PM if you like.
thanks
Jim
I have not had time to figure out the driver
circuit for any of the
lamps to see what that may turn up, and wanted to know whether it was
28v lamps before I buy 40 of them. (the thing has only 2 out of a
lot of lamps).
I mentioned http://www.foxdata.com/blog/tag/nova-312/ to you earlier,
which suggested the 28V bulbs. I also found
http://www.chookfest.net/nova3/ledmod.html
which makes it clear that the voltage is likely a fair bit over 12V.
The DG Components guide only lists part numbers for the 28V bulbs, 24V
bulbs, 5V, and 6V bulbs (at least that I found that fit with the wire
base and the form factor). I think installing 28V bulbs in 24V
circuits would be fine (they'd last longer).
Alas, I have no authoritative information on the exact correct bulb.
Vince
11 a.m.
Jim wrote...
-------
Jay did your DG stuff on the back dock move? just curious
-------
Not yet. One listmember expressed interest but nothing firm. Two racks of DGblue just
waiting for someone to cart them off ;) One is an S/200 and it uses bulbs. My S/130 uses
LEDS I seem to recall, but haven't had the panel open in a good while.
My DG Eclipse S/130 project is at a long impasse for now.
But once the S/130 project is done I'll be diving into a sizeable pile of nova
800/1200 stuff and I'll have DGtonnage to make go away then....
I do not have a Nova3, but I will see if I have any docs that speak to the bulbs used
there.
J
1:39 p.m.
From: jim stephens: Friday, July 22, 2016 12:46 AM
On 7/22/2016 12:25 AM, Vincent Slyngstad wrote:
Those chookfest bulbs are doing just what I'd expect from 12V bulbs
in a 28V circuit -- way too bright, and then a drastically shortened
lifespan. (The right bulb can last for years; the wrong one for
minutes.)
I'm not a big fan of the surgery involved in LED conversion, though
I understand why folks do it. Particularly if they, like Emil, have used
the wrong bulbs and found themselves replacing them all the time.
Somewhere, I have Oshino's write-up about bulb rated voltage and
bulb operating voltage, but I do remember the lifespan varies as
some power of the ratio, and it makes a huge difference. (There
are also formulae for derating brightness, etc.)
Vince
I mentioned
http://www.foxdata.com/blog/tag/nova-312/ to you earlier,
which suggested the 28V bulbs. I also found
http://www.chookfest.net/nova3/ledmod.html
which makes it clear that the voltage is likely a fair bit over 12V.
Thanks for the tips. I see that the foxdata site lamps don't appear
quite as bright as the chookfest ones, which is interesting. I'm
thinking now if I do anything it will be the LED route, so will evaluate
that, as I agree opening the thing back up will suck.
2 p.m.
From: Vincent Slyngstad: Friday, July 22, 2016 10:39 AM
Somewhere, I have Oshino's write-up about bulb
rated voltage and
bulb operating voltage, but I do remember the lifespan varies as
some power of the ratio, and it makes a huge difference. (There
are also formulae for derating brightness, etc.)
I found it -- the lifespan varies as the 12th power of the ratio.
The current varies as the 0.55 power, and the brightness as the
3.5th power (of the reciprocal).
So, a 12V bulb in a 28V circuit can expect 1/26000 the lifetime,
and will take about 1.6X the current to burn about 20X brighter
than normal.
A 28V bulb in a 24V circuit would last about 6.3X it's rated lifetime,
take 92% of the rated current, and about 58% as bright.
Vince
2:18 p.m.
On 2016-Jul-22, at 11:00 AM, Vincent Slyngstad wrote:
From: Vincent Slyngstad: Friday, July 22, 2016 10:39
AM
Nice to see the data & math on that.
Somewhere, I have Oshino's write-up about
bulb rated voltage and bulb operating voltage, but I do remember the lifespan varies as
some power of the ratio, and it makes a huge difference. (There are also formulae for
derating brightness, etc.)
I found it -- the lifespan varies as the 12th power of the ratio.
The current varies as the 0.55 power, and the brightness as the 3.5th power (of the
reciprocal).
So, a 12V bulb in a 28V circuit can expect 1/26000 the lifetime, and will take about 1.6X
the current to burn about 20X brighter than normal.
A 28V bulb in a 24V circuit would last about 6.3X it's rated lifetime, take 92% of
the rated current, and about 58% as bright. 
2:27 p.m.
On Fri, 22 Jul 2016, Vincent Slyngstad wrote:
. . . and that presumably clarifies the "Livermore Firehouse Bulb".
Somewhere, I
have Oshino's write-up about bulb rated voltage and bulb
operating voltage, but I do remember the lifespan varies as some power of
the ratio, and it makes a huge difference. (There are also formulae for
derating brightness, etc.)
I found it -- the lifespan varies as the 12th power of
the ratio.
The current varies as the 0.55 power, and the brightness as the 3.5th power
(of the reciprocal).
So, a 12V bulb in a 28V circuit can expect 1/26000 the lifetime, and will
take about 1.6X the current to burn about 20X brighter than normal.
A 28V bulb in a 24V circuit would last about 6.3X it's rated lifetime, take
92% of the rated current, and about 58% as bright.
2:16 p.m.
On 2016-Jul-22, at 10:39 AM, Vincent Slyngstad wrote:
From: jim stephens: Friday, July 22, 2016 12:46 AM
Not to mention that the extra heat from a bulb run over-voltage can damage plastic front
panels and light shrouds.
Even at the rated voltage, different bulb models can have a wide range of specified
lifetimes,
On 7/22/2016 12:25 AM, Vincent Slyngstad wrote:
Those chookfest bulbs are doing just what I'd expect from 12V bulbs in a 28V circuit
-- way too bright, and then a drastically shortened lifespan. (The right bulb can last
for years; the wrong one for minutes.)
I'm not a big fan of the surgery involved in LED conversion, though I understand why
folks do it. Particularly if they, like Emil, have used the wrong bulbs and found
themselves replacing them all the time.
Somewhere, I have Oshino's write-up about bulb rated voltage and bulb operating
voltage, but I do remember the lifespan varies as some power of the ratio, and it makes a
huge difference. (There are also formulae for derating brightness, etc.) I mentioned
http://www.foxdata.com/blog/tag/nova-312/ to you earlier, which suggested the 28V bulbs.
I also found
http://www.chookfest.net/nova3/ledmod.html
which makes it clear that the voltage is likely a fair bit over 12V.
Thanks for
the tips. I see that the foxdata site lamps don't appear quite as bright as the
chookfest ones, which is interesting. I'm thinking now if I do anything it will be
the LED route, so will evaluate that, as I agree opening the thing back up will suck.
from a few thousand hours to 25,000 or 40,000 or more
hours.
Although the common trick with computer front panels seems to have been to just run bulbs
under-rated,
e.g. a 6V bulb in a circuit with Vcc=5V, with some further voltage drop in the drive
circuit the bulb sees well under 5V.
4:29 p.m.
On 7/22/2016 10:39 AM, Vincent Slyngstad wrote:
I'm not a big fan of the surgery involved in LED
conversion, though I
understand why folks do it. Particularly if they, like Emil, have
used the wrong bulbs and found themselves replacing them all the time.
I
wouldn't do the surgery, but here's the situation. Someone already
used this panel for the source of lamps, or they were all burned out.
There are only two bulbs left to remove.
The surgery wasn't bad, it doesn't look like there are lifted etch pads,
so that part is good. So this one is a good candidate for the
conversion as any. I'm not a stickler for historic accuracy, and if
anyone gets it to use as a running unit, they will hopefully never have
to open the panel again.
Thanks for the analysis too, that is good to know, as i still have
1600's with 5v bulbs to deal with. Luckily the 1600 uses socketed parts
and the main issue with them is pulling the panels two 50 pin ribbon
connectors off. The connectors on the ends of that get worn out.
The Nova has an edge card connector (as does my buddies Eclipse panel)
so that isn't a problem with this design. Between the two problems I'll
take the problem with the connector over the problem with soldering
lamps any time.
thanks again
Jim
http://jimsoldtoys.blogspot.com/2016/07/nova-3-front-panel.html
5:03 p.m.
Jim;
Thanks for the link... I went up the tree and saw the pics of that ADDS Envoy terminal.
I'm exceedingly green with envy. Congrats on an awesome piece of kit!
J
2:20 p.m.
DG generally used 28V, 0.040 Ma (nominal), fragile wire-lead
incandescent bulbs for the Nova/SuperNova/Nova2/Nova3 front panels as
well as the early Eclipses. The S/130 was DG's first LED-based front
panel and was much-appreciated by Field Service.
More followup off-list...
Bruce
On 7/21/2016 10:53 PM, jim stephens wrote:
Is there anyone with documents on the Nova 3 front
panel, and what
drives it? It has some number of custom DG chips, which hopefully are
good if I want to try to fire it up to play, but am interested in that
on good authority there are 28v incandescent lamps.
A friend has an Eclipse front panel with nearly identical bezel, which
has LED's and a number of differences in the logic (different connector
to the system, for instance). So it is probably all run on 5v.
I have not had time to figure out the driver circuit for any of the
lamps to see what that may turn up, and wanted to know whether it was
28v lamps before I buy 40 of them. (the thing has only 2 out of a lot
of lamps).
Thanks
Jim
2872
days inactive
2872
days old
11 comments
6 participants
participants (6)
-
Bruce＠Wild-Hare.com -
cisin＠xenosoft.com -
hilpert＠cs.ubc.ca -
jwest＠classiccmp.org -
jwsmail＠jwsss.com -
v.slyngstad＠frontier.com