13 Jan
2006
13 Jan
'06
11:23 p.m.
All:
I'm playing around with my relatively new IMSAI and I was
pondering the following. The unit is constructed with the front panel power
switch unconnected and the corresponding terminals on the power supply board
jumpered together. There is a mains switch on the rear panel which takes the
place of the front panel switch.
While for safety reasons I cannot disagree with the approach -
it's plain stupid to put uninsulated mains voltage in a low-voltage area
(and unpassable by UL evaluators I'm sure) - it detracts a bit from its use
because you have to reach behind the unit to turn it on.
Has anyone come up with an elegant solution to being able to use
the front panel power switch while keeping it safe? I was toying with some
sort of low-voltage circuit (small LV transformer and relay, with the front
panel switch in the loop on the coil side). Another idea I had was to hot
glue a dielectric insulator board (that gray cardboard kind of stuff) over
the parts of the front panel that would be exposed.
Ideas? Too much time on my hands?
Rich
Rich Cini
Collector of classic computers
Build Master for the Altair32 Emulation Project
Web site: <http://highgate.comm.sfu.ca/~rcini/classiccmp/>
http://highgate.comm.sfu.ca/~rcini/classiccmp/
/************************************************************/
14 Jan
14 Jan
12:20 a.m.
I don't think there is anything intrinsically unsafe
about the switch itself, which is rated for line
voltage and likely UL approved. The idiocy of the
IMSAI design is that the switch is connected via PC
traces to solder pads at the top edge of the board
where the wiring to the power supply attaches -- live
line voltage exposed at a location that just invites
contact while one is otherwise careful to avoid the
power-supply internals. I've seen people put a few
strips of electrician's tape over the pads to insulate
them, though a fold of fishpaper and hot-glue might be
less likely to fall off with age (??). Another piece
of bad design is the bare metal fuse clip that is so
close to the edge of the power supply PC board that I
would be concerned about the possibility of contact
with the case cover, e.g., if the case was
sufficiently deformed. Another fishpaper might be a
good idea there.
--Bill
1:15 a.m.
Interestingly, there is no fuse holder at that location (bridged with a
jumper). Instead, there is a standard 1-1/4" glass fuse holder in the back
panel. Certain spots of the power supply board I've glued-over with
insulating plastic (salvaged from the HV section of various power supplies
over the years).
Very bad design indeed. The pads at the top are very close to the metal cage
and the HV traces on the power supply board are uninsulated.
-----Original Message-----
From: cctalk-bounces at classiccmp.org [mailto:cctalk-bounces at classiccmp.org]
On Behalf Of William Maddox
Sent: Saturday, January 14, 2006 12:21 AM
To: General Discussion: On-Topic and Off-Topic Posts
Subject: Re: AC power on front panels
I don't think there is anything intrinsically unsafe
about the switch itself, which is rated for line
voltage and likely UL approved. The idiocy of the
IMSAI design is that the switch is connected via PC
traces to solder pads at the top edge of the board
where the wiring to the power supply attaches -- live
line voltage exposed at a location that just invites
contact while one is otherwise careful to avoid the
power-supply internals. I've seen people put a few
strips of electrician's tape over the pads to insulate
them, though a fold of fishpaper and hot-glue might be
less likely to fall off with age (??). Another piece
of bad design is the bare metal fuse clip that is so
close to the edge of the power supply PC board that I
would be concerned about the possibility of contact
with the case cover, e.g., if the case was
sufficiently deformed. Another fishpaper might be a
good idea there.
--Bill
1:58 a.m.
New subject: dangerous voltages inside
On Sat, 14 Jan 2006 01:15:07 -0500
"Richard A. Cini" <rcini at optonline.net> wrote:
Interestingly, there is no fuse holder at that
location (bridged with a
jumper). Instead, there is a standard 1-1/4" glass fuse holder in the back
panel. Certain spots of the power supply board I've glued-over with
insulating plastic (salvaged from the HV section of various power supplies
over the years).
Very bad design indeed. The pads at the top are very close to the metal cage
and the HV traces on the power supply board are uninsulated.
The worst high voltage design I have ever seen was a Fluke
Digital Voltmeter that I used to have. I think it was an 8300.
It used Nixie displays, and the high voltage for the Nixie tubes
ran on a trace that just weaved it's way across the circuit board
(with feedthroughs, etc,) without isolation or anything about it
to differentiate it from other traces. The board was not
solder-masked, so the metal was bare and exposed. Adjacent
tracks were low-level signal lines.
Granted, it was a high-end precision DMM so nobody was supposed
to go inside except the high-priest calibration dude, but
still....
Does anybody else here appreciate old Fluke gear? My newest
piece is an 8060 DMM, which is late-70's 4-1/2 digit meter. What
I really like are the Fluke Differential voltmeters. I currently
have a nice 881AB. The John Fluke Company made their name with
Differential Voltmeters and high-end lab-grade meters. (I'm
biased against anything in a Fluke DMM without a 4-digit model
no., the 8060 is the best handheld they ever made, the newer
'autorange' lines have always been a disappointment to me.)
2:34 a.m.
New subject: dangerous voltages inside
I don't know how practical it would be, but might it be possible to actuate
a remote AC switch by using an insulated mechanical linkage?
Otherwise, I'd be sore tempted to just add a small step-down transformer
and a relay.
The MITS Altair's like that too. Damned dangerous.
Cheers,
Chuck
5:22 p.m.
New subject: dangerous voltages inside
The worst high voltage design I have ever seen was a
Fluke
Digital Voltmeter that I used to have. I think it was an 8300.
It used Nixie displays, and the high voltage for the Nixie tubes
ran on a trace that just weaved it's way across the circuit board
(with feedthroughs, etc,) without isolation or anything about it
to differentiate it from other traces. The board was not
solder-masked, so the metal was bare and exposed. Adjacent
tracks were low-level signal lines.
Well, the supply yo nixie tubes is likely to be a lot less deadly than
mains. I am thinking of the HP9815 calculator. That has a Panaplex
display, the HV for that comes from the PSU/printer board to the
keyboard/display interface via one wire of a 24 pin ribbon cable. All the
other pins are TTL level signals. And HV on exposed components on both
boarrds.
The worst I've seen (as it made me leap) was the PSU from a Tektronix
terminal (I forget he model number, but it was a colour, raster-scan
one). This was a little SMPSU in a metal can. Unfortunately, there was no
bleeded resistor on the mains smoothing capacitors. If the startup
resistor went open-circuit, the capcitors had no way to discharge. And
the way into the supply was to undo a couple of screws and slide the PCB
out. If you didn't realise the capcitors were likely to still be charged,
you would most likely accidentally touch the mains smoothing capacitors
with painful results.
Does anybody else here appreciate old Fluke gear? My
newest
Yes. OK, I've got a new-ish Fluke too (original 85) that I use all the
time too.
I have one of their data loggers, a 2240 or something like that. It's
controlled by a pair of 4040s, one for the main control, one to drive the
paper tape output. I have only one 10 channel input board (but I do have
the temperature-compensated input connector block on it), the
high-resolution ADC (of course this is isolated from ground, all the
digital inputs/outputs to it are opto-isolated), the internal printer,
and so on. It's very well made, and a joy to use.
-tony
5:59 p.m.
New subject: dangerous voltages inside
Tony Duell wrote:
The worst I've seen (as it made me leap) was the
PSU from a Tektronix
terminal (I forget he model number, but it was a colour, raster-scan
one). This was a little SMPSU in a metal can. Unfortunately, there was no
bleeded resistor on the mains smoothing capacitors. If the startup
resistor went open-circuit, the capcitors had no way to discharge. And
the way into the supply was to undo a couple of screws and slide the PCB
out. If you didn't realise the capcitors were likely to still be charged,
you would most likely accidentally touch the mains smoothing capacitors
with painful results.
I've done exactly that with an SMPSU from a SCSI tape unit.
I grabbed the PCB after the unit had failed, been taken out
of service, and left unplugged for a day or so. The main
smoothing capacitor was still charged up (and remember we
have 240V mains here) and discharged across the second and
third fingers of my left hand. I yelled quite a bit. Then,
I found that I had "+" and "-" burned into my fingers from
the capacitor terminals.
--
John Honniball
coredump at gifford.co.uk
17 Jan
17 Jan
3:10 p.m.
New subject: dangerous voltages inside
I've done exactly that with an SMPSU from a SCSI
tape unit.
I grabbed the PCB after the unit had failed, been taken out
of service, and left unplugged for a day or so. The main
smoothing capacitor was still charged up (and remember we
have 240V mains here) and discharged across the second and
That wouldn't make much difference actually. Most of these SMPSUs have 2
capacitors in series. They each charge to around 160V, no matter whether
the mainse is 115V or 230V. For the former, the input rectifier acts as a
voltage doubler, chearging each capacitor to the peak of the AC input on
alternate half cycles. for the latter, the input cirucit is a
conventional bridge rectifier.
third fingers of my left hand. I yelled quite a bit.
Then,
I found that I had "+" and "-" burned into my fingers from
the capacitor terminals.
Ocuh.
I can understand cheap PC-related PSUs missing out the bleeder ressitors
to xave all of 5p, but I was suprised Tektronix did it. Oh well...
The PSU in my HP9845B is relatively friendly in this respect. It has
bleeder resistors. It also has well-insulated holes in the top cover
through which you can insert voltmeter probes to check the residual
voltage on the smoothing capacitors before taking the PSU module out of
the machine. The procedure given on the top of the supply is to isolate
the mains, wait 5 minutes, insert the discharger tool (which I assume is
just another bleeder resistor, or maybe a shorting link!) and then to
check the thing is discharged with a meter. Then yoy take it out.
-tony
14 Jan
14 Jan
6:12 p.m.
New subject: dangerous voltages inside
----- Original Message -----
From: "Scott Stevens" <chenmel at earthlink.net>
snip-------
Does anybody else here appreciate old Fluke gear?
Certainly do. I have an 803B here I use fairly often.
DaveB, NZ
--
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.1.371 / Virus Database: 267.14.17/229 - Release Date: 13/01/2006
16 Jan
16 Jan
6:05 a.m.
New subject: Fluke Re: dangerous voltages inside
At 01:58 AM 1/14/06 -0500, Scott wrote:
Does anybody else here appreciate old Fluke gear?
I do. Some of it is VERY good. FWIW a lot of old Fluke Nixie tube meters
have been showing up around here lately. I've been grabbing them for the
Nixie tube stuff.
My newest
piece is an 8060 DMM, which is late-70's 4-1/2
digit meter. What
I really like are the Fluke Differential voltmeters. I currently
have a nice 881AB. The John Fluke Company made their name with
Differential Voltmeters and high-end lab-grade meters. (I'm
biased against anything in a Fluke DMM without a 4-digit model
no., the 8060 is the best handheld they ever made, the newer
'autorange' lines have always been a disappointment to me.)
I have a Fluke model 87 DMM that I use constantly. I recently bought a
HP LogicDart but I've hardly touched it. I use the 87 when I need a meter
and a HP 54510 DSO when I need a scope. (I bought the scope out of a
non-working junk pile and fixed it and I love the thing.)
10:37 p.m.
New subject: Fluke Re: dangerous voltages inside
If its a good measurement, its a Fluke!
----- Original Message -----
From: "Joe R." <rigdonj at cfl.rr.com>
To: "General Discussion: On-Topic and Off-Topic Posts"
<cctalk at classiccmp.org>
Sent: Monday, January 16, 2006 11:05 AM
Subject: Fluke Re: dangerous voltages inside
At 01:58 AM 1/14/06 -0500, Scott wrote:
Does anybody else here appreciate old Fluke gear?
I do. Some of it is VERY good. FWIW a lot of old Fluke Nixie tube meters
have been showing up around here lately. I've been grabbing them for the
Nixie tube stuff.
My newest
piece is an 8060 DMM, which is late-70's 4-1/2
digit meter. What
I really like are the Fluke Differential voltmeters. I currently
have a nice 881AB. The John Fluke Company made their name with
Differential Voltmeters and high-end lab-grade meters. (I'm
biased against anything in a Fluke DMM without a 4-digit model
no., the 8060 is the best handheld they ever made, the newer
'autorange' lines have always been a disappointment to me.)
I have a Fluke model 87 DMM that I use constantly. I recently bought a
HP LogicDart but I've hardly touched it. I use the 87 when I need a meter
and a HP 54510 DSO when I need a scope. (I bought the scope out of a
non-working junk pile and fixed it and I love the thing.)
14 Jan
14 Jan
8:14 p.m.
Another piece of bad design is the bare metal fuse
clip that is so close to the edge of the power supply
PC board that I would be concerned about the possibility
of contact with the case cover, e.g., if the case was
sufficiently deformed.
A good solution to this, that was told to me by a man who
assembled new IMSAIs "back in the day", is to use the bottom
3/4 inch of a Tic-Tac candy box. You get the fuse covered
in plastic but can still see it. It fits tight and will stay
in place. He did this on all of the units he built (and he
said he built a lot), so if you happen to come across a unit
set up like that, he may well have built it.
15 Jan
15 Jan
2:04 p.m.
A good solution to this, that was told to me by a man
who
assembled new IMSAIs "back in the day", is to use the bottom
3/4 inch of a Tic-Tac candy box. You get the fuse covered
in plastic but can still see it. It fits tight and will stay
I'ev seen open-type fuseholders with official transparaent plastic
covers. Voth for 20mm and 1~1/4" (is that what you'd call a 3AG fuse)
fuses, and both for chassis mounting and PCB mounting.
I normally use them for any fuse that's carrying significant voltage [1]
or current. It's a lot cheaper than the results of accidentally touching
it or shorting it to something else.
[1] Dor the pedants, a fuse (or anything else) can'r carry a voltage.
What I mean is that there is a significant voltage between the fuse
terminals and the local ground.
-tony
14 Jan
14 Jan
1:54 a.m.
Has anyone come up with an elegant
solution to being able to use
the front panel power switch while keeping it safe? I was toying with some
sort of low-voltage circuit (small LV transformer and relay, with the front
panel switch in the loop on the coil side). Another idea I had was to hot
glue a dielectric insulator board (that gray cardboard kind of stuff) over
the parts of the front panel that would be exposed.
The bad thing about the IMSAI design is the bare connections to mains at
the top of the panel - normally the remainder of the panel is fairly safe
behind the acrylic front - even with the cover off. You can cover the pads
at the top of the board with a few strips of good electrical tape.
If you really don't want mains power at the front panel at all, then a relay
is paobably the best solution - lots of room to mount it (and the xfmr to
power it), and you don't have to modify the front panel.
--
dave04a (at) Dave Dunfield
dunfield (dot) Firmware development services & tools: www.dunfield.com
com Collector of vintage computing equipment:
http://www.parse.com/~ddunfield/museum/index.html
6:09 a.m.
If you really don't want mains power at the front
panel at all, then a
relay
is paobably the best solution - lots of room to mount it (and the xfmr to
power it), and you don't have to modify the front panel.
Surely a transformer would need to be powered all the time?
Why not use a nicad to pull the relay in, and charge it up when the unit
is on? If the nicad was flat, you could just have an ordinary pushbutton
to bypass the relay to get you going.
Gordon.
1:45 p.m.
On 1/14/2006 at 11:09 AM gordonjcp at gjcp.net wrote:
Surely a transformer would need to be powered all the
time?
So what? Your central heating system and doorbell have one. Much, if not
most, modern electronic equipment, such as PCs, televisions, etc. all have
some tiny draw from the AC mains even when powered off. If all you
needed was a very small DC supply you could do away with the transformer
and perhaps use a couple of capacitors in series with each side of the
line, followed by a rectifier and a zener--it wouldn't provide quite the
isolation from the line that would be ideal, but it wouldn't need a
transformer.
Cheers,
Chuck
2:59 p.m.
----- Original Message -----
From: "Chuck Guzis" <cclist at sydex.com>
To: <cctalk at classiccmp.org>
Sent: Sunday, January 15, 2006 7:45 AM
Subject: Re: AC power on front panels
On 1/14/2006 at 11:09 AM gordonjcp at gjcp.net wrote:
For a practical implementation of this sort of arrangement see the
circuit here-
http://homepages.ihug.co.nz/~tractorb/collins/Collins%2075S%20mains%20switc…
I threw this together to get round an issue with under-rated wafer
switch contacts being used for mains switching.
Regards
DaveB, NZ
--
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.1.371 / Virus Database: 267.14.17/229 - Release Date: 13/01/2006
Surely a transformer would need to be powered all
the time?
So what? Your central heating system and doorbell have one. Much,
if not
most, modern electronic equipment, such as PCs, televisions, etc.
all have
some tiny draw from the AC mains even when powered off. If all
you
needed was a very small DC supply you could do away with the
transformer
and perhaps use a couple of capacitors in series with each side of
the
line, followed by a rectifier and a zener--it wouldn't provide quite
the
isolation from the line that would be ideal, but it wouldn't need a
transformer.
Cheers,
Chuck
4:44 p.m.
On 1/15/2006 at 8:59 AM Dave Brown wrote:
For a practical implementation of this sort of
arrangement see the
circuit here-
http://homepages.ihug.co.nz/~tractorb/collins/Collins%2075S%20mains%20switc
h%20.pdf
Very nice job, particularly using the grounded/earthed neutral as the
return. Here in the USA, one side of the 120vac is tied to the third prong
of the mains receptacle in the distribution panel, so this would also be a
way to avoid tying into an incorrectly-wired receptacle (more common than
you'd think!).
Back in the bad old days of glass-enclosed active devices, a dodge for
powering a single 6.3v 0.3a filament from the mains without a filament
transformer was to use a 1.0 uf nonpolar cap as a dropping device. In that
application, if the cap failed, the filament of the tube/valve would simply
burn out. Perhaps a very small (7w) incandescent bulb might be used in
your circuit in place of a 100 ma fluse. Would a common NE-2 neon also
work as protection? It might be nice to use as a "this equipment is
connected to the mains" indicator.
Cheers,
Chuck
12:02 p.m.
On 1/15/2006 at 8:59 AM Dave Brown wrote:
SNIP
For a practical implementation of this sort of
arrangement see the
circuit here-
http://homepages.ihug.co.nz/~tractorb/collins/Collins%2075S%20mains%20switc
h%20.pdf
Very nice job, particularly using the grounded/earthed neutral as the
return. Here in the USA, one side of the 120vac is tied to the third prong
of the mains receptacle in the distribution panel, so this would also be a
way to avoid tying into an incorrectly-wired receptacle (more common than
you'd think!).
Cheers,
Chuck
Chuck is correct, this is major reason for using standby transformers
(AC or SMPS) to provide low current power for power controlling the
main power.
To other one who did this modifcation on a radio equipment and very
low current control of the SSR relay, find a cheap LED small digital
alarm clock, uses 1" cube transformer and is wired. Majority of
these are not on the PCB board and ready to use, cut a small area of
PCB to get diode(s), cap for DC output where transformer wires are
soldered on. Usually 100mA, 16 to 12VAC. Add a fuse for the mains
side of transformer and another fuse after it.
Cheers, Wizard
5:25 p.m.
If you really don't want mains power at the
front panel at all, then a
relay
is paobably the best solution - lots of room to mount it (and the xfmr to
power it), and you don't have to modify the front panel.
Surely a transformer would need to be powered all the time?
Why not use a nicad to pull the relay in, and charge it up when the unit
is on? If the nicad was flat, you could just have an ordinary pushbutton
to bypass the relay to get you going.
Having had experience of NiCd powered power switches in machines like the
Torch XX and the Whitechapel MG1, I would not recomend this. NiCds always
seem to be discheargeed when you need them.
-tony
15 Jan
15 Jan
5:20 a.m.
So? DEC did that in their power controllers. I am
quite sure you have
plenty of trnasformers powered all the time anyway -- the doorbell, the
clock/timer in the cooker/microwave oven, the VCR, ATX PSUs, etc.... OK,
some of those are SMPSUs running in standby mode, but the point is that
there's plenty of circuity energised all the time.
Not really - I make a point of avoiding things that draw power when
they're not in use. I want to keep as much stuff as I can when I move
back up north, and the house there is going to be off-grid.
Gordon.
5:14 p.m.
So? DEC did that in their power controllers. I am
quite sure you have
plenty of trnasformers powered all the time anyway -- the doorbell, the
clock/timer in the cooker/microwave oven, the VCR, ATX PSUs, etc.... OK,
some of those are SMPSUs running in standby mode, but the point is that
there's plenty of circuity energised all the time.
Not really - I make a point of avoiding things that draw power when
they're not in use. I want to keep as much stuff as I can when I move
back up north, and the house there is going to be off-grid. from the mains.
-tony
16 Jan
16 Jan
4:01 a.m.
So? DEC did that in their power controllers. I am
quite sure you have
plenty of trnasformers powered all the time anyway -- the doorbell,
the
clock/timer in the cooker/microwave oven, the
VCR, ATX PSUs, etc....
OK,
some of those are SMPSUs running in standby mode,
but the point is
that
there's plenty of circuity energised all the
time.
Not really - I make a point of avoiding things that draw power when
they're not in use. I want to keep as much stuff as I can when I move
back up north, and the house there is going to be off-grid.
6:13 a.m.
At 11:09 AM 1/14/06 +0000, you wrote:
I would be VERY hesitant to put a NiCad in any computer system that I
wanted to preserve. They have a nasty habit of leaking without warning and
the electrolyte is extremely corrosive.
Joe
If you really don't want mains power at the
front panel at all, then a
relay
is paobably the best solution - lots of room to mount it (and the xfmr to
power it), and you don't have to modify the front panel.
Surely a transformer would need to be powered all the time?
Why not use a nicad to pull the relay in, and charge it up when the unit
is on? If the nicad was flat, you could just have an ordinary pushbutton
to bypass the relay to get you going.
Gordon.
14 Jan
14 Jan
2:53 p.m.
All:
I'm playing around with my relatively new IMSAI and I was
pondering the following. The unit is constructed with the front panel power
switch unconnected and the corresponding terminals on the power supply board
jumpered together. There is a mains switch on the rear panel which takes the
place of the front panel switch.
While for safety reasons I cannot disagree with the approach -
it's plain stupid to put uninsulated mains voltage in a low-voltage area
I would agree.... That is asking to either come into contact with some
signal line (and blow chips all over the place) or you (and blow you all
over the place).
(and unpassable by UL evaluators I'm sure) - it
detracts a bit from its use
because you have to reach behind the unit to turn it on.
Has anyone come up with an elegant solution to being able to use
the front panel power switch while keeping it safe? I was toying with some
sort of low-voltage circuit (small LV transformer and relay, with the front
That's how DEC did it in some of the PDPs. A little transformer energised
all the time with a rectifier and smoothing capacitor which suppled the
voltage to operate a relay. DEC, of course, added circuity to allow for
remote power on, emergency shutdown (if the machine overheated), etc. We
had a discussion of the DEC power bus a week or so back.
panel switch in the loop on the coil side). Another
idea I had was to hot
glue a dielectric insulator board (that gray cardboard kind of stuff) over
the parts of the front panel that would be exposed.
This has the advantage of being closer to the original design (in that
the machine is electrically unchanged).
But I would go for the relay solution if it was my machine. It would be
safer I think (both to you and the machine).
-tony
6795
days inactive
6798
days old
24 comments
13 participants
participants (13)
-
ard＠p850ug1.demon.co.uk -
bshannon＠tiac.net -
cclist＠sydex.com -
chenmel＠earthlink.net -
coredump＠gifford.co.uk -
dave04a＠dunfield.com -
gordonjcp＠gjcp.net -
jpero＠sympatico.ca -
rcini＠optonline.net -
rigdonj＠cfl.rr.com -
tractorb＠ihug.co.nz -
wh.sudbrink＠verizon.net -
wmaddox＠pacbell.net