Re: 50g RS-232

On 04 Aug 2006 17:28:50 +0300, in msg <op4pwslgm5.fsf@xxxxxxxxxxxxxxx>, Antti
Louko <alo@xxxxxx> wrote:

"James M. Prange" <jmprange@xxxxxxxx> writes:

For the 50g, the full (6V nominal) battery power of the calculator
is available at the port, and that seems to me to be the most
obvious and convenient way to power a level-shifter for it; of
course, that's not the only way.

I would use an interface IC made for just this purpose like Sipex
SP232AC. They come in different packages: old-fashioned DIP16 or more
compact surface mount packages. SP232AC needs a four 0.1uF capacitors
and one filtering capacitor (10uF). The maximum supply voltage is 6V
so it is better to put one or two diodes in the supply to drop the
voltage a little bit. With this IC you get correct RS-232 voltage
levels.

I've used the MAX232 (or one of its many clones, such as the SP232AC) in many
designs. If wall power is available, it's *definitely* the way to go, and I've
used it in dozens of products.

But those parts are current hogs when it comes to a battery operated device!

I was suggesting a way of doing level conversion that allows you to get more
than a few days out of a set of batteries!

The math is simple (as I reach for an HP-41C since my trusty HP-42S is at work):

The idle current specs on the MAX232 parts range from 8mA to 15mA, depending
upon the manufacturer (the SP232AC requires 15mA). That's when it's just sitting
there plugged into the HP-50g, but not plugged into a device. When plugged into
a device it can go up another 2mA, and will remain constant regardless of
communications.

I don't know the size of the batteries used in the HP-50g, but I'll assume AAA's
rated at 2000mA hours, and that 10mA's is being used by the MAX232 part.

This will discharge the batteries to 0V in 2000 / 10 = 200 hours (even assuming
the calculator is turned off - if the 6V line from the HP-50g remains active!),
or about 8 days. Regardless of whether communications are even taking place!

However the calculator is going to quit working when the voltage drops below 4V
or so, not 0V. You'll only have a few days of battery life when using a MAX232
type driver chip.

The design I posted uses *much* less power. In the idle state, with no
communications taking place, whether or not an RS-232 device is plugged in, the
circuit uses the power of one CMOS input (going into the HP-50g), and the
leakage current across the transistors. The typical leakage of a CMOS input is
..1uA, and the transistor's spec's claim .05uA leakage each, so .2uA or .0002mA.

This will discharge the batteries in 1141 years! The shelf life of the
batteries will obviously kill them before this circuit will.

The circuit uses more power when active communication is taking place. This is
determined mostly by the 5K resistor on the 2N3906 which is required for RS-232
compatibility. (The 10k on the 2N3904 was selected for TTL inputs, but can be
changed to 100k for CMOS, pretty much removing it from the calculations).

The RX level will be typically -8 volts (being driven by a MAX232 chip on the
RS-232 device! ;-), so about 2.8mA will flow from the battery during a RS-232
transmitted space condition. (Given random data you could divide this by two
since half the bits will be marks, but we'll go worst case).

During *constant* RS-232 transmissions of worst case data, the batteries will
discharge to 0V in 29 days. Which is really only achievable if you put the
HP-50g in a hard loop writing '00's to the output. Typical data logging
intervals and data would in the order of 10's to 100's of years.

You decide which method you prefer, but when a cable becomes available for the
HP-50g, and if it uses a MAX232 type chip, I'm ripping it out and replacing it
with the transistor level conversion circuit I posted!

Before anyone implements this, we still need to know what is the
allowed input voltage range for the RX line. As the serial connector
doesn't provide calculatorts regulated logic supply (3.3V, I guess),
it is difficult to regulate the output of the interface IC to 3.3V
level.

You just need a resistor and a zener diode (and even the zener diode may be
optional).

-Zonn
.

Relevant Pages

  • Re: micro hydro load controller
    ... used to supply emergency power to alarm systems, ... is designed to charge some 12V batteries from the 240V mains, ... of application where the input voltage may be variable. ...
    (sci.electronics.basics)
  • Re: 50g RS-232
    ... device's TX voltage level and output impedance, ... The positive voltage used is the 6V from the HP-50g, which is above the +3V dead ... I only use it on battery powered applications since it uses virtually no power ... The scheme I posted would not work without the 6V available from the HP-50g. ...
    (comp.sys.hp48)
  • Re: Im in over my head, would someone kinly throw me a life ring?
    ... Is it possible to tap a lower DC voltage and use an off the shelf 12vdc ... The source of the DC power is ... 6-12v batteries so the voltage could be anywhere from 72vDC ... Some times a client will ask me for a mouse, so I find him a mouse, ship ...
    (sci.electronics.basics)
  • Re: efficient low-voltage ac/dc conversion for bicycle battery charger
    ... > I want to construct a device that takes the AC power generated by my ... > batteries may at the same time power a load through a low-drop ... This switches the voltage up to the 6V range, ... switcher designed to run from low startup voltages. ...
    (sci.electronics.design)
  • Re: whine from dying Ultralife lithium 9volts in a Lectro UM 400
    ... Well there is no switching power supply in the 195D's. ... If the external input voltage is low enough, ... That doesn't explain the whine from failing intrnal batteries. ... When my NP1 starts going down sometimes I will hear it and switch to ...
    (rec.arts.movies.production.sound)
Loading