Date: Mon,May 2 2005 8:01
From: "Glenn Ashmore"

The power to run the meter has to be isolated from the circuit being measured.
They need either a battery or an isolated dc/dc converter to power them.

After some search it looks like the typical 200mV digital meter modules
take about 1 to 2 mA.
Even if I use a couple of them, for voltage and some currents (to
trolling motor, other loads), I need max. about 10 mA. For this low
current, a transformer to isolate the DC seems to be overkill.

I am considering to take a simple 555 and insert a capacitor not only
in the output signal line, but also the return (no common 0V anymore)
line. On the "secondary" side would be a simple bridge rectifier, a 9V
Zener diode, and a capacitor to smooth the ripples.

Did anybody try something like that already? Other simple solutions?

TIA,

Stefan

"stefanhanoi" wrote in
oups.com:

Other simple solutions?

Why not use one of the ICs that is already designed for this purpose? Sure
reduces your parts count to nearly zero, and the output is nicely regulated
by the IC's circuitry.

http://www.mouser.com/?handler=data....Ne=500&N=57960

cheap....as low as 48 cents. Mouser will sell you just one if you're
willing to pay the shipping....

Larry,
those DC/DC charge pumps obviously produce an output voltage that is
higher than the input voltage, or negative. But the output still has a
common ground (0V) with the input voltage. For the panel meters we need
a supply voltage that is within the range of the available voltage (12
to 24V available, 9V needed) (no need for those DC/DC charge pumps),
but that needs to be isolated/free flowing from the main/"primary"
power supply so that we can measure it, i.e. no common 0V/GND. I don't
see how the ICs that you show can do that. On the other side, I wonder
whether there are not really specialized ICs that do that.

Thanks,
Stefan

There are special chips for the purpose, inverting dc/dc isolators. They
are expensive; $10 a piece as I recall.

I had a lot of trouble with this whole setup; you need a separate isolator
for each panelmeter [volts, amps] I had a meter for each of several
alternators.

a loose connection on a heavy user [DC air compressor powered by a starter
motor] caused a spike that blew 2 isolators and panel meters. I added more
protection, reverse diodes and clamps; a few months later, again some
distant fault blew a couple of my panelmeters, even though I had bought more
expensive ones this time that were supposed to be internally protected as
well..

I'm not sure that the idea is at all practical in reality, these meters seem
just too delicate for our world.

I'd love to hear from people who have successfully installed a bunch of
panel meters on a boat or similar system. I have a couple set up to read
temperature as well, one for the engine room in general and one for the
house batteries [you need to adjust the charging voltage for battery
temperature].

regards, Mark Holden



"stefanhanoi" wrote in message
oups.com...
Date: Mon,May 2 2005 8:01
From: "Glenn Ashmore"

The power to run the meter has to be isolated from the circuit being
measured.
They need either a battery or an isolated dc/dc converter to power them.

After some search it looks like the typical 200mV digital meter modules
take about 1 to 2 mA.
Even if I use a couple of them, for voltage and some currents (to
trolling motor, other loads), I need max. about 10 mA. For this low
current, a transformer to isolate the DC seems to be overkill.

I am considering to take a simple 555 and insert a capacitor not only
in the output signal line, but also the return (no common 0V anymore)
line. On the "secondary" side would be a simple bridge rectifier, a 9V
Zener diode, and a capacitor to smooth the ripples.

Did anybody try something like that already? Other simple solutions?

TIA,

Stefan

On Sun, 29 May 2005 13:30:07 GMT, "Mark Holden"
wrote:

I'd love to hear from people who have successfully installed a bunch of
panel meters on a boat or similar system.

First post here and probably no help because of the common rail or
some other part that I've missed but:

Years ago I designed, built and raced an electric 'motorcycle' (more
like stretched moped). It had 2 x 12V car batteries wired in series
and I used a single panel meter (and additional components) to
display:

Voltage on battery 1
Voltage on battery 2
Total voltage
Total current
Speed.

The voltage was just a matter of calibrating the meters (with a
potential divider) to read a max of 199.9 volts.

The current was the voltage drop across the longest main battery cable
(to 199.9A).

Speed was done magnetically by sensing teeth on a drive gear within
the donor mopeds final drive unit then using a frequency to voltage
converter chip to display mph (to 199.9). [1]

The voltage was fairly easy to calibrate using a couple of good
(calibrated) DMM's and the current with an accurate high power bench
PSU. The 'speed' was resolved imperially and mathematically down to
'pulses / mph' and then an oscillator used to set / calibrate the
speedo. There were no affordable GPS units available in those days. I
didn't need all that level of accuracy but it was nice to have.

That was later tested against a Police RADAR gun. Mine showed 45.7 mph
and theirs said 46!

I can't remember how I got round the common link between powering the
unit and the sense inputs. Knowing it only had to last an hour or so
probably using a PP3 battery and an on/off switch but I guess all but
the Battery 2 voltage could have been done using the same common
ground point

I don't suppose it helps but has reminded me it might be handy if /
when I find a suitable electric outboard for my 12' Porta-Bote.

T i m

[1] I assume I could even make use of the speed function using a
marine log / transducer?

p.s. I came back to my machine after the lunch break to find a '1st'
rosette stuck to it? It turns out the IEEE had some judges there that
day and my award was for 'Technical innovation (apparently for my
electronics stuff). That was a nice surprise!

On Sat, 25 Apr 2009 21:24:25 +0100, T i m wrote:

On Sun, 29 May 2005 13:30:07 GMT, "Mark Holden"
wrote:

Bad form etc but I just looked again and realised that message was 4
years old! (doh).

Note to self; check the display order when subscribing to a new group
and before replying. ;-(

T i m