Igaging scale current draw

[matteou]

Hi, first of all thanks Yuriy for a wonderful work!

I've made the dro based on arduino with Igaging Scales.
I would like to supply the scales with a battery but I've some problems.
A new scale draw about 15uA (50uA with original display on).
When I connect the scale to arduino, the current jumps to 600uA (or even more, sometimes 10-20mA).
The scale still works but, also with original display, it draws always 600uA or more.
I've tried with 5 scales...


No errors in the circuit (very simple, just 2 resistor for voltage divider for the clock line).
Anybody have got same problem? Any solution?
Thanks

[fixerdave]

...
A new scale draw about 15uA (50uA with original display on).
When I connect the scale to arduino, the current jumps to 600uA (or even more, sometimes 10-20mA).
... Any solution?

Just a guess but it could be that the Arduino code is reading the scales a lot more often than the original iGaging heads.  You could slow down the Arduino code.  Not really much point in reading more than a couple times a second. 

But... why not just power the scales from the 3.3vdc output of the Arduino? It's spec'd to source 50ma.

David...

[matteou]

It isn't a problem related to the number of readings per second... If I plug the scale to the original display, the current draw remains the same (>1mA). 
It seems that the arduino input port has low input impedance and overloads the output-stage of the scale (permanently damaging the scale).

I've tried to put an opamp (unity gain stable, single supply, rail-to-rail) and it works, the (new) scale draws only 15uA (the old scales continue to draw a lot of mA).

[fixerdave]

It isn't a problem related to the number of readings per second... If I plug the scale to the original display, the current draw remains the same (>1mA). 
It seems that the arduino input port has low input impedance and overloads the output-stage of the scale (permanently damaging the scale).

I've tried to put an opamp (unity gain stable, single supply, rail-to-rail) and it works, the (new) scale draws only 15uA (the old scales continue to draw a lot of mA).

The schematic I worked from shows a 10k pull-down resistor on the data line from the sensor.  The clock line to the sensor is, as you say, through a voltage divider to reduce the 5v output to 3.3v.  

You make no mention of the pull-down resistor on the data line.  Are you using one?  If so, is it sufficiently large to not draw much load?  If you're not using any, perhaps the output of the sensors is not capable of bleeding off and the voltage is rising beyond the 3.3v limit.  The Arduinos have a high-impedance input but also have an optional internal pull-up resistor.  I've no idea what the circuit looks like inside the sensor, but I expect that 10K pull-down on the Arduino end is there for some reason.  Again, just guessing, and just answering because no one else is...

Where are you measuring current?  At the usb cable to the sensor?  I've a spare scale sensor here, and an original iGaging head.  It has been used in-circuit with the Arduino.  I just rigged up a usb breakout such that I can measure current "in the cable".  With my Fluke, I'm getting 0.04mA while the original iGaging head is showing a moving reading from the scale.  That's pretty near the bottom of what the meter can do, so don't look for absolute accuracy, but it's not 10mA.  Again, that sensor has been used in testing with my Arduino circuit that includes the pull-downs.

David...