Help needed: simple, *very* low power DC voltage increaser

[Lisandro Pin]

I have a problem with an old PC keyboard i really love; it's very
picky on the input voltage (+5VDC), and most modern motherboards,
having to deal with main power and standby power tend to use diodes /
resistors on the +Vout of the PS/2 connector, resulting in lower
voltages (my motherboard measures at 4.8, and yes, that's not good
enough).

Does anyone have any idea on how to get 1 extra volt from the +5VDC
one? Keep in mind these are low current power outputs (fused, i think
at 500mA but i wouldn't feel safe going over 300mA, knowing a keyboard
needs 100-170mA by itself). Batteries and extra wiring are out of the
question; the idea is to be able to modify any troublesome keyboard so
it works everywhere, hassle free.

I thought of DC inverters, but where i live (Argentina) the IC kind
are hard to come by; so if anyone knows a discrete design of this sort
(nothing fancy), i'd be most thankful.

 

[Robert Baer]

I have a problem with an old PC keyboard i really love; it's very
picky on the input voltage (+5VDC), and most modern motherboards,
having to deal with main power and standby power tend to use diodes /
resistors on the +Vout of the PS/2 connector, resulting in lower
voltages (my motherboard measures at 4.8, and yes, that's not good
enough).

Does anyone have any idea on how to get 1 extra volt from the +5VDC
one? Keep in mind these are low current power outputs (fused, i think
at 500mA but i wouldn't feel safe going over 300mA, knowing a keyboard
needs 100-170mA by itself). Batteries and extra wiring are out of the
question; the idea is to be able to modify any troublesome keyboard so
it works everywhere, hassle free.

I thought of DC inverters, but where i live (Argentina) the IC kind
are hard to come by; so if anyone knows a discrete design of this sort
(nothing fancy), i'd be most thankful.

If the +5V supply actually goes thru a diode to the connector, then i
am very suprised, as all motherboards that i have looked at use the same
1980's style 8-bit microcontroller with TTL technology that canot take
low voltages like you mention.
I would double-check the motherboard first, and then see if was easy
to bypass the stupid diode (if there) with a jumper across it.
Failing that, the +5V could be supplied by a USB port, and if one had
a driver for a USB keyboard, then an adaptor may be possible, as an
adaptor from USB to PS/2 (for mice) appear to require no circuitry.
Worst case would be two connectors; USB for power and PS/2 for signal,
could "Y" into the single connector for the keyboard (P/S-2 or older
5-pin DIN).

 

[Externet]

Hi Lisandro.
Look for the regulator on the power supply that provides the 5 V, and
insert a diode to the ground reference lead of the regulation
circuitry.
That will raise the voltage by 0,7V. For a lesser increase, you can
use diodes with 0,5 V or 0,2 V forward drop.
Example for a 7805 (5V) regulator that probably is NOT your case :
Pin 1 to source voltage;
Pin 2 to anode of a 0,7V diode and its cathode to ground. (instead of
direcly grounded)
Pin 3 is now +5,7V !!! output.
Miguel

 

[Lisandro Pin]

If the +5V supply actually goes thru a diode to the connector, then i

am very suprised, as all motherboards that i have looked at use the same
1980's style 8-bit microcontroller with TTL technology that canot take
low voltages like you mention.
I would double-check the motherboard first, and then see if was easy
to bypass the stupid diode (if there) with a jumper across it.

I have seen it a lot of times, you wouldn't beleive how many modern
motherboards are wired this way. Most have the +5V to the PS/2
connector via a small diode, and the "sleep" +5V line through a
resistor. This is fine for most modern keyboard controller chips,
which will work with voltages as low as 3V, but for older ones it's
not enough. In fact, most modern PC components take specifications
lightly (non-working AGP cards, anyone?), but older keyboards, like
the venerable IBM model M, are quite sensitive to the input voltage.
Hacking the motherboard is possible (for me), but i'm looking for a
general solution.

Failing that, the +5V could be supplied by a USB port, and if one had
a driver for a USB keyboard, then an adaptor may be possible, as an
adaptor from USB to PS/2 (for mice) appear to require no circuitry.
Worst case would be two connectors; USB for power and PS/2 for signal,
could "Y" into the single connector for the keyboard (P/S-2 or older
5-pin DIN).

I actually thought of using either the USB or some other +5V line
inside the PC, but i wanted something... "cleaner" . The idea is to
have a small board that can be fitted inside the keyboard and will
pump the voltage up just a notch (1v to have 5,6 - 5,8v would be more
than enough). It doesn't have to be super clean DC or complicated,
just a simple inverter.
I have came across a voltage doubler using a 555 and a couple of
caps / resistors; this and a voltage divider on the output could do
the trick, but i have to simulate it first in order to be sure it
doesnt go above 300-400mA from the PS/2 power connector. The thing is
i'm really not experienced with inverter designs; so i was hoping
someone had a brighter idea than me or did something similar in the
past.

 

[Lisandro Pin]

Tried looking in my motherboard; it uses a regulator for some +5v
signals internally but these are shared with a lot of other circuitery


I came across this circuit today:
http://www.reconnsworld.com/power_voltdoubler.html

It's a very crude voltage doubler using a 555. If you take this
circuit, reduce C2 to, say, 15uF and raise C3 to 470 or 1000uF to
fight the increased ripple you get a 0.8 - 1.x increase in the input
voltage, depending on the load. A small 5,5v zenner at the output is
all that's needed to have yourself a very passable "1 volt up"
circuit, or so it seems. This would allow the keyboard to work with
both "bad" motherboards and regular ones at a modest increase in
consumption (it's about 80% efficient), which is exactly what i need.

I'll build this circuit tomorrow to see how it performs.

 

[Glenn Gundlach]

I have a problem with an old PC keyboard i really love; it's very
picky on the input voltage (+5VDC), and most modern motherboards,
having to deal with main power and standby power tend to use diodes /
resistors on the +Vout of the PS/2 connector, resulting in lower
voltages (my motherboard measures at 4.8, and yes, that's not good
enough).

Does anyone have any idea on how to get 1 extra volt from the +5VDC
one? Keep in mind these are low current power outputs (fused, i think
at 500mA but i wouldn't feel safe going over 300mA, knowing a keyboard
needs 100-170mA by itself). Batteries and extra wiring are out of the
question; the idea is to be able to modify any troublesome keyboard so
it works everywhere, hassle free.

I thought of DC inverters, but where i live (Argentina) the IC kind
are hard to come by; so if anyone knows a discrete design of this sort
(nothing fancy), i'd be most thankful.

Wow, I guess what I'm typing this on shouldn't be working. The
keyboard is a 1988 Maxi-switch keyboard with macros and an AT type DIN
to PS2 adaptor. The motherboard is a Gigabyte GA-7N400PRO2 (current
model) with an Athlon XP 3200.

I really like this keyboard, too.
GG

 

[Lisandro Pin]

Good the circuit works just fine (boosted power line to +5,7v).
Bad it didn't help anything. It's funny, because the keyboard
seems to run fine if i feed it from a +5v line off a molex conector...


Time to keep looking i guess.

 

[Lisandro Pin]

Problem solved; it was voltage related but in an unexpected way. Check
comp.sys.ibm.pc.hardware.misc for details.

Thank you for you help!

 

[JeffM]

http://groups.google.com/groups?q=group:comp.sys.ibm.pc.hardware.misc+author:Lisandro-Pin