Re: Wall Wart for a Gecko
Posted by
caudlet
on 2002-08-21 16:20:14 UTC
The primary reason you would want to have 5V for the inputs is
because the circuit that drives the inputs (in most cases the PC
printer ports) runs on 5V. It is a lot more selective about what
voltage it sees when its pins go high. More that about one diode drop
(.6V) over the supply limit on some chips will reverse bias the
output pins (current flows backward...a BAD thing on IC's).
All that being said the optoisolator input is normally an LED that
turns on with a current flow. The LED needs just enough voltage
across it to overcome the forward diode junction voltage (2 junctions
in series = 1.2V) to start conducting current. Once it starts
conducting the increase in current versus applied voltage happens
pretty much straight up on a graph...in real life you get the
flashbulb effect: One quick flash and its over!. Since in a series
circuit the same amount of current flows through each component you
simply add a resistor in series with either leg of the LED and bingo
the current is limited to what Mr. OHM predicts. The Gecko's have
this resistor built in and calculated to give optimal performance at
5 volts.
The really good things about current devices is that they are much
more noise immune and you can easily increase their SNR (signal to
noise ratio) by raising the drive voltage and reducing the current at
the other end. If you were to double the series voltage on the input
of the isolator then you would have to force twice the voltage in
signal to get the same current flow. Think about how much noise it
would take to false trigger a circuit that needed 12 or even 15 volts
before it turned on or better yet, how much further away from true
ground you could make the LED turn off.
IF you were using open collector drivers (see earlier posts about
parallel port buffer circuits) with a parallel port to Gecko
interface than 6V or probably even 8V would be no big deal since the
current ratings on most optoisolators is pretty wide. Since the
current requirements are modest (15ma per Gecko) you can
easily "steal" the +5 off of the computer bus. Oddly the parallel
printer port itself has lots of ground pins and no +5. Other ports
do and you can get the voltage there. The only thing that would cause
me to pause on using the PC power is if the motor drives were located
more than a few feet from the PC. The possibilty of interference
increases ; especially with a plasma cutter when you hang cables out
in the air.
Probably more than you EVER wanted to know about this subject but you
are not dealing with normal people here! We in this hobby are
fooling with electricity, electronics, PC's, and metalworking ; all
things that cry out for precision (which turns a lot of us on) so we
tend to go on about tiny details when the answer is: Use 4.8 to 5.1V
@ 100ma from a filtered regulated source.
because the circuit that drives the inputs (in most cases the PC
printer ports) runs on 5V. It is a lot more selective about what
voltage it sees when its pins go high. More that about one diode drop
(.6V) over the supply limit on some chips will reverse bias the
output pins (current flows backward...a BAD thing on IC's).
All that being said the optoisolator input is normally an LED that
turns on with a current flow. The LED needs just enough voltage
across it to overcome the forward diode junction voltage (2 junctions
in series = 1.2V) to start conducting current. Once it starts
conducting the increase in current versus applied voltage happens
pretty much straight up on a graph...in real life you get the
flashbulb effect: One quick flash and its over!. Since in a series
circuit the same amount of current flows through each component you
simply add a resistor in series with either leg of the LED and bingo
the current is limited to what Mr. OHM predicts. The Gecko's have
this resistor built in and calculated to give optimal performance at
5 volts.
The really good things about current devices is that they are much
more noise immune and you can easily increase their SNR (signal to
noise ratio) by raising the drive voltage and reducing the current at
the other end. If you were to double the series voltage on the input
of the isolator then you would have to force twice the voltage in
signal to get the same current flow. Think about how much noise it
would take to false trigger a circuit that needed 12 or even 15 volts
before it turned on or better yet, how much further away from true
ground you could make the LED turn off.
IF you were using open collector drivers (see earlier posts about
parallel port buffer circuits) with a parallel port to Gecko
interface than 6V or probably even 8V would be no big deal since the
current ratings on most optoisolators is pretty wide. Since the
current requirements are modest (15ma per Gecko) you can
easily "steal" the +5 off of the computer bus. Oddly the parallel
printer port itself has lots of ground pins and no +5. Other ports
do and you can get the voltage there. The only thing that would cause
me to pause on using the PC power is if the motor drives were located
more than a few feet from the PC. The possibilty of interference
increases ; especially with a plasma cutter when you hang cables out
in the air.
Probably more than you EVER wanted to know about this subject but you
are not dealing with normal people here! We in this hobby are
fooling with electricity, electronics, PC's, and metalworking ; all
things that cry out for precision (which turns a lot of us on) so we
tend to go on about tiny details when the answer is: Use 4.8 to 5.1V
@ 100ma from a filtered regulated source.
Discussion Thread
ksprayberry428
2002-08-21 12:56:32 UTC
Wall Wart for a Gecko
JanRwl@A...
2002-08-21 13:53:45 UTC
Re: [CAD_CAM_EDM_DRO] Wall Wart for a Gecko
Tom Benedict
2002-08-21 14:13:37 UTC
Re: [CAD_CAM_EDM_DRO] Wall Wart for a Gecko
Alan Marconett KM6VV
2002-08-21 14:39:41 UTC
Re: [CAD_CAM_EDM_DRO] Wall Wart for a Gecko
Alan Marconett KM6VV
2002-08-21 15:13:25 UTC
Re: [CAD_CAM_EDM_DRO] Wall Wart for a Gecko
studleylee
2002-08-21 15:19:35 UTC
Re: Wall Wart for a Gecko
caudlet
2002-08-21 16:20:14 UTC
Re: Wall Wart for a Gecko
Steve Blackmore
2002-08-21 17:48:07 UTC
Re: [CAD_CAM_EDM_DRO] Wall Wart for a Gecko
Tom Benedict
2002-08-21 19:39:48 UTC
Re: [CAD_CAM_EDM_DRO] Re: Wall Wart for a Gecko
ksprayberry428
2002-08-22 04:33:14 UTC
Re: Wall Wart for a Gecko
cncnovice
2002-08-22 04:44:58 UTC
Re: Wall Wart for a Gecko
Carol & Jerry Jankura
2002-08-22 05:36:45 UTC
RE: [CAD_CAM_EDM_DRO] Re: Wall Wart for a Gecko
Tim Goldstein
2002-08-22 09:51:55 UTC
Re: [CAD_CAM_EDM_DRO] Wall Wart for a Gecko
mariss92705
2002-08-22 09:53:22 UTC
Re: Wall Wart for a Gecko
mariss92705
2002-08-22 10:12:07 UTC
Re: Wall Wart for a Gecko
Alan Marconett KM6VV
2002-08-22 11:04:31 UTC
Re: Wall Wart for a Gecko
caudlet
2002-08-22 13:49:17 UTC
Re: Spitting on the fence
JanRwl@A...
2002-08-22 15:13:18 UTC
Re: [CAD_CAM_EDM_DRO] Re: Wall Wart for a Gecko
Keith Bowers
2002-08-22 15:54:58 UTC
Re: [CAD_CAM_EDM_DRO] Re: Wall Wart for a Gecko
mayfieldtm
2002-08-22 16:45:13 UTC
Re: Spitting on the fence
William Scalione
2002-08-22 18:40:51 UTC
Re: [CAD_CAM_EDM_DRO] Re: Wall Wart for a Gecko
slo_rat
2002-08-24 21:51:44 UTC
Re: Wall Wart for a Gecko
mrjensan
2002-08-25 11:53:08 UTC
Re: Wall Wart for a Gecko