Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Digital Fat Calliper Sensor

Status
Not open for further replies.

vivian_lys90

New Member
Hi everyone,

PC212531.JPG
Anyone know what kind of sensor is used in the picture attached?? This is a circuit for digital body fat calliper.


Thanks.
 
Looks to me as though this is a mostly mechanical device with a straight digital input - presumably to a microprocessor.

There are 8 big pads and each big pad has 8 smaller pads next to it. I would guess that every 8th smaller pad is connected.

So all you need is a micro-controller with two 8-bit ports free for reading calliper position, and some simple code to make it work.

EDIT: ... except that there are only an 8 pin connector and a 4 pin connector to the board - so there must be some kind of multiplexing going on.
... would need to examine the sensor board and the control more carefully to get the full picture.
 
Last edited:
Hi, thanks for the reply.
Actually I want to rebuild the digital fat calliper and edit its coding. Here attached is the other part of the circuit of the digital fat caliper. Anyone of you have ideas what is the components that i need to rebuild it?
2012-11-01-216.jpg
PC212530.JPG
PC212536.JPG
 
The top picture is just the pincers for the fat fold and the top side of the PCB shown previously and in the middle photo.

The diodes in the middle image are arranged to suggest that input from the 8 large ranges is by Charlieplexing (i.e, the position of 8 button switches can be read by only 4 wires). That is the way multiplexing suggested by hexreader above seems to have been done. See the example and demonstration about 60% down this page: **broken link removed**

The bottom picture shows the chip-on-board (COB)(i.e., the microcontroller that runs everything is under the blob of black epoxy) and some passive components, mostly small capacitors. As mentioned in an earlier thread, I suspect reading and changing the program in the COB will be impossible without, even if you are lucky, a Herculean effort.

John
 
Last edited:
Good morning Vivian,

I vaguely remember the other thread. As I recall, one option was to manually record results, but you wanted a more automated system. At the time, we did not know how the sensor worked. With the new information here, we can say that digital information is supplied to the MCU. That is, an A/D converter is not needed.

Rather than try to figure out what is gong on inside the COB, I think it would be easier to start anew. You could use a 16-pin or 18-pin PIC to take the data from the sensor's 12 wires and calculate the slider's (moving contact) position. Then using geometry or a calculated table, determine the distance between the pincer tips. Finally, transmit the data to your data logger and to a serial LCD display. It is probably easier to buy a new display, rather than hack the one on the device you already have. You would need to study the current display more to decide whether it can be hacked.

John
 
Hi John,
Thanks for your information. Your information is very useful and it help me a lot! For now, the sensor and the PIC problem is solved. So the other problem is the coding problem. Is that possible for me to retrieve the programming code from the existing circuit?

Vivian:eek:
 
Hi John,
Is that possible for me to retrieve the programming code from the existing circuit?

Vivian:eek:

Some people shudder at the concept of "impossible." Microcontrollers can be decapped (opened) and the connections microscopically probed to recover coding. That is what I meant by Herculean task.

From a standpoint of practicality, I would say it is impossible to recover that coding. It would be easier to write your own code, even if you have to learn how to code first.

Where to begin? Have you done any coding? If not, and since this is all digital, I would suggest using Assembly. There are plenty of tutorials out there that I can recommend, but first, what is your experience level in coding?

John
 
The hardware for Charlieplexing is already there. That is the one piece of the device you opened that you will certainly be using. The rest is relatively simple coding.

John

Edit:

Here is a list of some tutorials for Assembly language. Some of them have accompanying C tutorials as well.

1) http://www.amqrp.org/elmer160/lessons/
2) **broken link removed**
3) http://winpicprog.co.uk/pic_tutorial.htm
4) http://www.gooligum.com.au/tutorials.html
5) **broken link removed**

For code examples, math routines, radix conversions, serial communication, etc:


#3 is by Nigel Goodwin on this site and has been translated into C by Ian Rogers. I personally used Elmer (#1) and Tracey (#2) early on, but now usually go to #4 for questions and PicList for finding ready-to-use routines.
 
Last edited:
Hi Vivian,

From your other thread, you state that you want to transfer the readings to a PC. You also mention storing the readings - do you mean storing them on the device? The coding for the calliper connected directly to the PC is very simple. The version with storage is also simple, but requires a little more thought.

Do you have an preference for microcontroller to use? If you don't have any equipment, perhaps using an Arduino is the simplest approach as it contains everything you need to program (it has a bootloader) and test it. You could also just use an ATMEGA microcontroller and a cheap programmer, e.g. USBISP **broken link removed**
 
Hi everyone,
Thanks for the reply. I have a question, can I found the output signal just looking on the printed circuit board?

Are you asking about the signal from the pincers to the MCU or from the MCU to the digital display?

The former signals have been discussed. They are carried in the 12 wires (4 +8).

The signals to the display may be harder to access. It is possible they are through the board near or even under that black blob of epoxy. The display may be attached to the board in such a way that trying to access its signals may destroy it. If it is a graphical display, there is also a likelihood that communication to the MCU is two-way. Since only one black blob is visible in the pictures you post, it is possible that COB does it all and is also the display driver.

If you are thinking of just tapping into the display signals and sending them to a PC, that might be doable if the display takes serial data and does not have two-way communication to the MCU. But even in that simple case, you will need to do some programming, and I believe it would be easier to just start with the signals from the pincers (goniometer) and manipulate them.

Without seeing more disassembly of the device and display, it is hard to say how the devices are connected, but I think it is pretty unlikely to be the simple case (one-way serial) described above. That opinion is based on taking apart other consumer goods.

John
 
I suspect any potentiometer that can rotate at least 60° or so will work, but then you will need to digitize its output, such as with a constant current source and an analog-to-digital converter (ADC). Many microcontrollers (perhaps almost all except the very basic ones) have ADC inputs. The charlieplexed output is already digital.

John
 
Hi John,

Thanks for your information. Do you know about Multi plate electrode stator? Beside, I can't found any charlieplexing in the market, do you know where can I get it? Thanks for your help. :)


Vivian
 
Out of context, I did not recognize the term "multi plate electrode stator." Googled it and came up with some designs using multiple piezoelectric devices for micromanipulation and multiplate variable capacitors. You will see examples of the latter in the tuning section of some radios. Something like that could be used to measure angle. Here is a book I found on Google about that application. I think using a variable resistor will be easier.

As for charlieplexing, I don't know that it is something you can buy as a plug-in component. It is a way to design a multiplexed circuit, as described in the link given in post #5. Your device apparently already has that circuitry. If you are still planning to capture the data for automatic logging, what you have in your device is the way to go. It is basically just a PCB board and some diodes. Of the alternatives you have mentioned, my second choice would be the variable resistor.

John

Here's the link to the book:
https://books.google.com/books?id=T...ce=gbs_ge_summary_r&cad=0#v=onepage&q&f=false
 
Last edited:
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top