Hi,

I am very ew to electronics and need some help hijacking the signals from a cheap kitchen scale I bought. I realise that the signal from the scale is analog, and intercepting the signal to the lcd would be easier. Problem is, I want to display what is show on the lcd, on my pc, as I want to manipulate the readings ie average etc). How do I interpret the signals from the pcb. Do I use th RS232 port? The lcd pcb has 16 connectors connecting to the lcd, and the characters are fairly big(1 line of chars).

I have googled for info on the lcd model no. M1094-10TR, but found nothing. I would like to log the input from the lcd connectors.

Many thanks,
Dion.

 

A raw LCD will have an apx 40-100Hz backplane signal as well as a signal lead for each visible segment and symbol on the LCD. Segments/symbols are driven out of phase to the backplane when ON and in-phase when OFF. A LCD with a built in driver, which is probably what you have, is another story all together and you'd really need the data sheet on the LCD for that.
So what you really need is a microcontroller to interpret this data, convert it to ASCII, and send it to your PC via RS232 or USB. You could also do it via the parallel port and discreet logic, but that would be much more difficult.

__________________
Inside every little problem, is a big problem trying to get out.

 

How about taking some really good pics of the PCB and posting them? We might be able to make some more informed suggestions then. You may very well be able find a better signal on the scale board than the LCD will provide.

__________________

 

Thanks for the reply guys. I will get hold of a decent camera and send pics. I like the idea of hijacking the signal directly from the scale outputs. I just need to find info on what I will be receiving as input from the 4 wires leading from the scale, as well as how the scale actually works.

dion.

 

Most now a days use a 4 wire load cell resistance bridge. Two of the wires are used to 'excite' the bridge, typically using 10volts or 5 volts, but in your case they probably use less as the battery voltage is sure to be much lower. This excitation voltage must be regulated. The other two wires are used to measure the differential voltage generated from the weight on the load cell. It's just millivolts in value and needs amplification with an op amp.

Lefty

__________________
Measurement changes behavior

 

You could use a webcam, and some OCR software functions to get the values onto the computer.

 

Sorry for the long delay. Here is the image.

Attached Thumbnails

 

 

If you're lucky the amplifier for the weight sensor signal is not underneath that blob. You'd have to probe around with an oscilloscope or a multimeter while pressing up and down on the scale to see if you can find a trace that carries the analog signal, which you can feed into your own ADC and log it that way. Looks like the row of pins above the blob in the picture is the LCD connection if you still want to go that route, but that picture is so blurred and you didn't take one of the other side of the board it's hard to tell anything.

__________________

 

Use the macro function on your camera (Usually a flower icon) so we can see the PCB better. Do the 4 wires on the left go to the load cell? What is the part number on the 8pin SMD IC?

__________________
Inside every little problem, is a big problem trying to get out.

 

The 4 wires on the left do attach to the load cell. I have attached 2 more pics. Any ideas about an ADC?

Thanks,
Dion

Attached Thumbnails

   

 

IF I recall it correctly a simple high input resistance op amp (Jfet inputs) should be able to directly read the millivolt signals coming off of the load cells and be able to amplify them.

After that a simple single channel analog signal monitoring unit will send the right information to any off the shelf data logging or monitoring type software you use.
Some analog data or signal collection devices are capable of direct measurement from standard load cells too!
Look for commercial or industrial data logging systems for more information. They are very common in manufacturing. You may end up with a complete plug in and go set up!

__________________
"Issue a general safety warning. Then look the other way and allow stupidity the chance to eliminate itself." -- tcmtech
"Those who can, Will. Those who can't, will achieve positions of power over those who can and then promptly stop them." -- tcmtech
"Your impossibility may just be my day to day routine." -- tcmtech

 

It looks like the sensor wires ( S+ & S- ) go straight into the epoxy blob IC. You could measure the voltage between S+ & S- for different weights and then use a differential amplifier to reference the voltage to ground for your ADC. It looks like the calibration table is stored in U2 so the output of the load cell may, or may not, be linear. How much resolution do you need?

__________________
Inside every little problem, is a big problem trying to get out.

 

I am a little out of my depth here. My strength is in app dev, so please be a little patient. If I do not respond fairly quickly, I am more than likely researching the concepts you guys use. What exactly do you mean by resolution in this context? The accuracy of the reading?

Thanks.

 

>> It looks like the sensor wires ( S+ & S- ) go straight into the epoxy blob IC...
If you mean the 4 wires from the load cell, they do not go directly into the epoxy blob IC, but are soldered onto the pcb on the left(look at the first of the last images attached previously).

 

Yes, but the wires enter the PCB, go through 2 zero Ω jumpers and into the epoxy blob. There are some bypass capacitors in parallel to the lines (And maybe some resistors- can't tell from the image), probably shunting noise to ground, but the traces on the PCB appear to go straight from the wires to the blob. ie: There is no OpAmp between the load cell and the blob.
By resolution, I meant how much precision do you need. ie: a 10bit ADC will give you 1024 discreet steps.

__________________
Inside every little problem, is a big problem trying to get out.